Regulation of the Expression of Aminoacyl-tRNA Synthetases and Translation Factors

Putzer H, Laalami S.

Publication Details

Aminoacyl-tRNA synthetases and translation factors are key enzymes required for pro tein biosynthesis. Escherichia coli and Bacillus subtilis often use different strategies to regulate the expression of the genes encoding these enzymes. Synthesis of several E. coli aminoacyl-tRNA synthetases is controlled by different mechanisms acting at the transcriptional or translational level. By contrast, in B. subtilis, expression of the majority of these proteins is regulated by a common, yet specific transcriptional antitermination mechanism. However, all of these controls share a common effector, the tRNA. In the case of the E. coli translation factors, the primary enzyme function is often exploited for autoregulating their own synthesis at the translational level. Here we will focus on the gene organization and the multiple types of gene regulation governing prokaryotic aminoacyl-tRNA synthetase and translation factor expression.


Decoding the genetic message is the major and most energy consuming process in the cell. In addition to ribosomes, mRNA and tRNA, translation requires amino acids, nucleotides and specialized proteins. The latter include aminoacyl-tRNA synthetases and translation factors transiently associated with the ribosomes. These proteins catalyze sequential steps during translation, starting with the charging of tRNA, ribosome-dependent polypeptide synthesis, final release of the protein and ribosome recycling. Here, we will consider our current knowledge of the gene organization and the expression of aminoacyl-tRNA synthetases and translation factors in prokaryotes. We focus essentially on two bacterial systems, the Gram negative bacterium Escherichia coli and Bacillus subtilis, the best-studied Gram positive organism. Translation factors are only considered for E. coli. Structural and mechanistic aspects of these enzymes are treated in accompanying chapters.

Aminoacyl-tRNA Synthetases

Aminoacyl-tRNA synthetases (aaRS) play a central role in protein biosynthesis by catalyzing the attachment of a given amino acid to the 3' end of its cognate tRNA. They do this by forming an energy-rich aminoacyl-adenylate intermediate of the cognate amino acid, which serves to transfer the amino acid to the tRNA. The intrinsic proofreading capacities of the aaRS and their balanced expression contribute greatly to the accuracy of translation of the genetic code. In addition to their crucial role in protein biosynthesis, aaRS are involved in a number of regulatory processes via their product, the charged tRNA. The control of the expression of amino acid biosynthetic operons by the level of tRNA aminoacylation in vivo is well documented. Another phenomenon, the pleiotropic stringent response, operates when tRNA aminoacylation is diminished in wild-type (relA) strains and inhibits rRNA and tRNA synthesis as well as the synthesis of some other macromolecules involved in translation.1 As a by-product of the amino acid activation step aaRS can form diadenosine 5'-5'''-P1, P4-tetraphosphate (AppppA), considered as a pleiotropically acting alarmone that has been associated with oxidative stress or timing of cell division.2 Eucaryotes exploit yet other functions of aaRS, e.g., some aaRS are essential factors in certain splicing activities.3 It is obvious that control of the cellular levels of the aaRS is important for any organism.

E. coli Aminoacyl-tRNA Synthetase Genes

General Regulatory Phenomena

In E. coli a single aaRS is found for each amino acid, with the exception of lysyl-tRNA synthetase for which two species have been characterized. The cellular abundance of the different aaRS is quite similar;4 at a doubling time of 40 minutes, the cell contains 10 to 20 times less molecules of each enzyme than ribosomes. This amounts to between1300 and 2600 molecules per cell. In contrast, the concentrations of different charged tRNA isoacceptor families vary more than ten-fold, from about 700 (tRNAGln) to 8000 (tRNAVal) at a doubling time of 60 min.5 Under the same conditions the individual aminoacyl-tRNA/synthetase ratios vary by a factor of 15 and the turnover rate of the aminoacyl-tRNAs is between two and eight per second. Surprising differences in the in vivo activity of individual aaRS result from these measurements; while glutamyl-tRNA synthetase charges only two tRNAs per second, threonyl-tRNA synthetase (ThrS) charges as many as 48 per second. These values are much higher than what can be obtained in vitro with purified enzymes, up to 240-fold higher in the case of ThrS.5

A quantitative analysis of O'Farrel 2D gels in which 18 out of the 21 synthetases were identified shows that in most cases their cellular concentration increases with growth rate.6 Overall, the level of enzyme (normalized to total cellular proteins) increases two- to threefold for each fivefold increase in growth rate. This increase is less than that observed with ribosomes but corresponds to that of elongation factor EF-Tu and the initiation factors. Growth rate-dependent regulation affects aaRS expression rather rapidly; new steady-state levels are reached within 2 to 3 minutes after a upshift. There is no general rule as to whether growth rate control of the aaRS is transcriptionally or translationally regulated. Recent DNA array data (LaRossa, personal communication) show a decrease of mRNA levels for several synthetases with increasing growth rate while the earlier proteome data clearly found increased protein concentrations under similar conditions. Even an increased mRNA level is not necessarily an indication of transcriptional regulation and may reflect translation enhanced mRNA stability.

Superimposed on growth rate-dependent regulation is a regulatory response of individual synthetase genes to limiting amounts of the cognate amino acid. At least half of the aaRS are synthesized more rapidly under these conditions.7 Derepression can be transient (hisS, leuS, metG, proS, serS, thrS and valS) or permanent (argS, ileS and pheST). In all cases, addition of the amino acid after starvation reverses the effect. However, this leaves open the question of whether it is the free amino acid or the cognate tRNA, which is involved in regulation. For E. coli metG, or hisS from the closely related organism Salmonella typhimurium, it appears clear that regulation depends on the level of aminoacylation of the corresponding tRNA. The specific control mechanisms governing pheST and thrS expression (see below) also respond directly to the tRNA aminoacylation level. In contrast, there is no clear evidence so far that aaRS genes are coregulated with those of their cognate tRNAs. In both enteric and Gram-positive organisms amino acid deprivation provokes an immediate adjustment of cellular metabolism, the stringent response, by inhibiting the synthesis of rRNA and tRNAs. The extent of aminoacylation of tRNAs plays a central role in this regulatory response probably by controlling the level of guanosine-5'-diphosphate-3'-diphosphate (ppGpp) synthesized. However, the stringent response has only a minor effect, if any, on the expression of at least 10 of the 21 aaRS.8 At present, no mutation is known which affects the control of all or even a small number of aaRS genes.

Genetic Organization in E. coli

The molecular weights and subunit structures are surprisingly variable among the different aaRS (see chapter 2 by D. Söll and M. Ibba). With the exception of the threonyl- and phenylalanyl-tRNA synthetase genes (thrS and pheST) which are in close vicinity, all aaRS genes are scattered throughout the chromosome (Fig. 1). The orientation of transcription of different aaRS genes is not correlated with the direction of replication. There is also no correlation between the positions of the aaRS genes and tRNA genes except for the gene gltX encoding glutamyl-tRNA synthetase (see below). All aaRS genes are expressed from σ70 promoters with weak consensus sequences, including lysU, which is heat-shock induced. In most cases the transcription initiation sites are known.9

Figure 1. E.

Figure 1

E. coli aminoacyl-tRNA synthetase genes. The aaRS genes are shown in boldface type together with adjacent genes. Positions are given in min and are calculated values from the complete genome sequence (Colibri database). Arrows indicate the directions (more...)

The glycyl- and phenylalanyl-tRNA synthetases are the only synthetases composed of two different subunits. Both are tetramers of the α2β2 structural type. The genes for the two different subunits are grouped on the chromosome and are expressed as an operon in which the promoter-proximal gene codes for the small subunit. Finally, there are a few aaRS that are co-expressed with other proteins, generally also involved in translation (see below).

Specific Control Mechanisms

Many aaRS genes are, often only transiently, derepressed between two and fourfold following starvation for the cognate amino acid. This is much less than what is observed for the corresponding biosynthetic operons under the same conditions (generally 10-fold and more). In the following section we will describe several studies at the molecular level. They shed considerable light on the ingenuity of the cell to regulate the expression of individual aaRS genes at all possible steps of genetic expression.

Autoregulation by Transcriptional Repression (alaS)

Alanyl-tRNA synthetase (AlaS), encoded by the alaS gene, is a tetramer of the α4 type. In an in vitro system, the synthetase is able to repress its own transcription when present in micromolar quantities.10 On the other hand, when present in amounts corresponding to the intracellular concentration, autogenous repression is only observed in the presence of alanine. Experiments analyzing protection against DNase I digestion show that AlaS can actually bind to the -10 region of its promoter, suggesting that transcription initiation of alaS might be controlled by the alanine concentration in the cell. Unfortunately, the derepression of AlaS synthesis expected upon alanine starvation was never shown to occur, and no in vivo data have been presented to confirm and further investigate the model.

Regulation by Transcriptional Attenuation (pheST)

Phenylalanyl-tRNA synthetase (PheST) is a tetrameric enzyme of the α2β2 type. The pheS and pheT genes, encoding the small and the large subunits, respectively, are cotranscribed, with pheS being the first gene in the operon. The major promoter initiates transcription 368 bp upstream of the pheS initiation codon. However, about 30% of the pheST transcripts originate from the upstream thrS-infC-rpmI-rplT operon. The pheST genes are derepressed 2.5-fold under conditions in which the cellular concentration of phenylalanyl-charged tRNA is decreased. This was achieved, for example, by using a host strain carrying a mutated PheST with a high Km for phenylalanine. This strain is bradytrophic (a leaky auxotroph) for phenylalanine. The key elements of pheST regulation are a Rho-independent transcription terminator, located just in front of pheS, preceded by an open reading frame coding for a 14-residue peptide (the leader peptide) containing five phenylalanine residues, of which three are consecutive. A number of in vitro and in vivo experiments, including a detailed mutational analysis and the use of transcriptional and translational fusions of the pheST leader to lacZ, established the regulatory model presented in Figure 2.8,11,12 This regulatory mechanism, described in the legend of Figure 2, depends on differential translation of the leader peptide in the presence or absence of phenylalanine. This translation determines the local position of the ribosome, which in turn influences the formation of alternative RNA secondary structures, among them, the terminator. The default state for this mechanism, that is in the absence of any leader peptide translation, is termination (Fig. 2A). The pheST operon is different from similarly controlled amino acid biosynthetic operons, since its expression is essential for bacterial growth; absence of its expression cannot be compensated for by supplementation of the growth medium. Thus, some readthrough of the terminator must occur even in the repressed state. There is also a necessity for tight coupling between transcription and leader mRNA translation (Fig. 2C). Only when the ribosome immediately follows the RNA polymerase can its influence on RNA folding have an immediate effect on the RNA polymerase. Accordingly, increasing the distance between the leader mRNA stop codon and segment 2 leads to derepression; the ribosome, although having reached the stop codon due to the availability of phenylalanine, cannot hinder the formation of 2/3 antiterminator structure.

Figure 2. Alternative structures of the pheST leader mRNA.

Figure 2

Alternative structures of the pheST leader mRNA. The initiation codon, the five phenylalanine codons of the leader peptide and the U stretch of the terminator are written out, the stop codon is boxed. Important regions 1, 2, 3, and 4 are indicated. A) (more...)

The pheST operon is also derepressed in strains carrying a mutant allele of the miaA gene whose product modifies tRNAPhe and tRNATrp, among other tRNAs. The absence of the isopentenyl modification of the adenine adjacent to the anticodon of these tRNAs alters their translational properties. A concomitant slow-down in leader peptide translation favors formation of the antiterminator structure (Fig. 2C) and is thus the likely cause for the observed derepression.11

Autoregulation at the Translational Step (thrS)

Threonyl-tRNA synthetase (ThrS) is a homodimeric enzyme whose transcription initiates 162 bp upstream of its structural gene. Transcripts generally extend at least beyond the first downstream gene, infC, which is separated from thrS by only three nucleotides. ThrS expression is negatively feedback regulated at the translational level. When ThrS is overproduced from a plasmid, expression of translational thrS-lacZ fusions (producing a fused ThrS-LacZ polypeptide) is repressed, but that of transcriptional fusions (translation of thrS and lacZ is initiated at distinct sites) is not.13 In vitro, addition of exogenous ThrS to a cell-free protein synthesizing system decreases the initial rate of ThrS synthesis by 90% without affecting thrS mRNA synthesis.14 Constitutive mutations in the thrS leader, i.e., mutations that abolish negative feedback control, mostly occur in a GU sequence located at positions -31 and -32 relative to the first nucleotide £è the initiation codon (Fig. 3A). Not by chance, this dinucleotide also occurs as the second and third bases of the anticodons of all tRNAThr isoacceptors (Fig. 3B) and represents the major recognition site for the synthetase. It is also present in the loops of domains 2 and 4 which makes them structurally similar to the anticodon stem of tRNAThr. Both domains are part of the translational operator of thrS, which is confined to the first 110 bases upstream of the thrS structural gene15 (Fig. 3A). Threonyl-tRNA synthetase can bind to its operator in vitro with an affinity close to that of the tRNAThr-ThrS complex. In doing so, domains 2 and 4 are specifically protected from chemical modification or ribonuclease digestion. In addition, ribosome binding to a thrS leader transcript is completely inhibited by the addition of ThrS in physiological concentrations but a 10-fold excess of tRNAThr over synthetase completely relieves this inhibitory effect. Together with similar results obtained in vivo, this shows that tRNAThr acts as an antirepressor and that the intracellular concentration of uncharged tRNAThr modulates the repressor activity of ThrS. Since the tRNAThr concentration in the cell increases with growth rate, specific and growth rate-dependent regulation may be achieved via a single mechanism in the case of thrS. Surprisingly, operator mutations, which increase thrS expression also, increase the steady-state mRNA concentration, without altering mRNA stability. This is best explained if one assumes that thrS mRNA molecules that escape repression are fully translated and stable while those where the ribosome is blocked from initiating translation are immediately degraded.16 The operator-ThrS interaction is now extremely well characterized, practically down to the atomic level. Domain 2, which contains the anticodon-like arm, is the major recognition site of the synthetase. The crystal structure of a ThrS-domain 2 complex clearly indicates that domain 2 and the anticodon domain of tRNAThr are recognized by the same C-terminal site of the synthetase (A.C. Dock-Brégeon, pers. comm.). This is a perfect example of molecular mimicry where the structure of an RNA molecule has evolved to fit a binding site on a protein that normally interacts with a different RNA. Final proof for the authentic functional similarity comes from experiments based on tRNA identity rules. In both E. coli tRNAThr and tRNAMet, the anticodon is the major determinant of aminoacylation specificity. The replacement of CGU in the anticodon-like sequence of the thrS domain 2 by CAU is sufficient to abolish control by ThrS and establish control by methionyl-tRNA synthetase in vivo and in vitro.17,18 Domain 3 is a single-stranded linker region between domains 2 and 4 but also important for efficient translation of thrS.19 In fact, the 3' single-stranded end of domain 3 is part of a split ribosomal binding site whose two regions are brought into close proximity by the hairpin structure of domain 2 (Fig. 3A). The RNA domains involved in ribosome and ThrS binding are thus not strictly overlapping, but interspaced. Domain 4 behaves similarly to domain 2 in all aspects but with a lower efficiency. ThrS is a homodimeric enzyme and binds two tRNAThr molecules. Domains 2 and 4 bind to ThrS with the same stoichiometry as tRNAThr but only one intact operator binds per dimeric enzyme.20 Recent data show that it is the N-terminal domain of the synthetase which is responsible for the competition with the ribosome. ThrS thus displays a truly modular structure reminiscent of transcriptional regulators (M. Springer, pers. comm.). Finally, based on sequence and structural comparisons this molecular mimicry mechanism appears to be well conserved in other Gram-negative organisms such as Salmonella, Yersinia, Vibrio and Haemophilus (J. Caillet, pers. comm.).

Figure 3. Figure 3.

Figure 3

Figure 3. The thrS leader mRNA and the composite structure of tRNAThr isoacceptors in the L-shaped representation. The thrS leader structure between nucleotides -121 and +3 is shown in A. The Shine-Dalgarno sequence and the initiation codon are underlined (more...)

Two aaRS with the Same Specificity (lysS and lysU)

In E. coli, there are two lysyl-tRNA synthetases encoded by the genes lysS and lysU. They are located at 65.3 and 93.8 min, respectively, on the E. coli chromosome (Fig. 1). This is the only case known, in this organism, of two synthetases being able to activate the same tRNA isoacceptor family. Individually, each of the two synthetases is dispensable for growth but disruption of lysS causes a cold-sensitive phenotype. The lysS gene encodes the housekeeping synthetase; it is constitutively expressed and is subject to growth rate dependent control. The dicistronic transcripts of lysS originate upstream of the prfB gene, encoding peptide chain release factor RF2, which is translationally autoregulated (see below). Due to the proximity (9 nucleotides) of the stop codon of prfB to the downstream initiator codon of lysS, it is likely that lysS expression is influenced by the translation of prfB.

The lysU gene, on the other hand, is normally silent at low temperatures which explains the cold-sensitive phenotype of a lysS mutant strain. Its expression is induced by a variety of stimuli such as high temperature, anaerobiosis, low external pH, the stationary phase, or the presence of specific metabolites including leucine or leucine-containing dipeptides in the growth medium.9 Expression of lysU involves several genes including lrp, rpoH, and hns. The leucine-response regulatory protein Lrp is a transcriptional regulator which acts as a repressor or activator for numerous genes.21 The lysU gene is part of the leucine regulon and normally repressed by Lrp22 which binds just upstream of the promoter.23 Repression is relieved upon addition of L-leucine. Independent of Lrp, heat shock induction of lysU occurs only in the presence of a functional rpoH gene, coding for the heat shock sigma factor σ32. However, lysU does not have a Σ32 dependent promoter and induction of RpoH synthesis, without temperature shift, does not cause lysU derepression, indicating that its effect is indirect.24 The abundant, histone-like protein H-NS, is potentially involved in lysU regulation by binding to a curved DNA region upstream of lysU. Disruption of hns increases lysU expression at low temperatures 2.5-fold. Since H-NS also interferes with the expression of lrp25 its role in lysU expression might be more complex.

Why does E. coli have two lysyl-tRNA synthetases in the first place ? The most noticeable difference between the two enzymes is an affinity of LysU for lysine nearly 8-fold higher than that of lysS. In agreement with this observation, LysU is relatively less sensitive to the presence of cadaverine, an inhibitor of lysine binding, than lysS. Remarkably, this polyamine is produced at the expense of lysine under several stress conditions that also induce lysU expression.26 Increasing the synthesis of LysU under these conditions could thus be useful for the cell and explain the conservation of two lysine specific synthetases.

Glutamyl-tRNA Synthetase (gltX)

The gene for glutamyl-tRNA synthetase, gltX, shares an intergenic regulatory region with the divergent valU tRNA operon.27 This is the only aminoacyl-tRNA synthetase gene that is adjacent to a tRNA gene in E. coli. The factor for inversion stimulation (FIS), which is one of the major positive regulators of rRNA operons, binds to three sites within the promoter region of gltX and valU. FIS stimulates valU transcription about two-fold during steady-state exponential growth, but not during growth acceleration. In contrast, gltX transcription is repressed two-fold by the presence of FIS, but only during growth acceleration. This leads to a concomitant decrease of the glutamyl-tRNA synthetase level. This lack of synchrony between the influence of FIS on gltX and valU transcription indicates that their expression is not coordinated.28 Most gltX transcripts originate from the gltX proximal promoter which does not have a directly overlapping FIS binding site. Interestingly, FIS mediated repression of this major gltX promoter depends on the presence of the valU promoter which lies 120 bp upstream. FIS can bend DNA by 40° to 90° and due to the topology of its three binding sites can bring both promoters closer to each other. The influence of FIS on gltX transcription during growth acceleration is thus of modulatory nature. On the other hand, since FIS does not influence expression during steady-state growth it has been concluded that it is not involved in growth rate control of gltX expression.28

Methionyl-tRNA Synthetase (metG)

In its native form methionyl-tRNA synthetase is a homodimeric enzyme encoded by the metG gene. Expression of the synthetase is regulated by the level of aminoacylation of tRNAMet. Two promoters mediate metG transcription.29 The upstream promoter lies within the coding region of the divergently transcribed apbC gene (initially called mrp) which is involved in thiamine synthesis by the alternative pyrimidine biosynthetic pathway.30 The downstream promoter lies in the apbC-metG intergenic region so that the -35 promoter regions of apbC and metG overlap. In addition, a transcription terminator is located between the two metG promoters. S1 nuclease mapping experiments show that transcription from the upstream promoter is attenuated at the terminator. The presence of a potential tRNA-like secondary structure with a CAU anticodon-like sequence upstream of the terminator led to the proposal of an autoregulatory model. In this model, an excess of methionyl-tRNA synthetase could bind to this structure and somehow affect transcription termination.29 Further experiments are needed to prove this theory.

Aminoacyl-tRNA Synthetases in Gram-Positive Bacteria

Progress made during the last 15 years such as the completion of genome sequences from several Gram-positive organisms now gives a broader view on how eubacteria regulate and coordinate the expression of the aminoacyl-tRNA synthetases (aaRS). Bacillus subtilis, due to its highly developed genetics, has become the best studied representative of the Gram-positive bacteria.

There are strong structural similarities between the aaRS from Bacillus and E. coli and many Bacillus synthetases are functional in E. coli as judged by the successful complementation of diverse E. coli synthetase mutants. However, there are significant differences between B. subtilis and E. coli in the organization and expression of aaRS genes.

  1. The chromosomal locations and arrangements of the B. subtilis genes are completely different from those of its Gram-negative counterpart.
  2. B. subtilis has no glutaminyl-tRNA synthetase ; instead, a single glutamyl-tRNA synthetase aminoacylates both tRNAGlu and tRNAGln with glutamate.31 The mischarged tRNAGln is subsequently converted to Gln-tRNAGln by an amidotransferase.32,33
  3. In B. subtilis there are two cases of aaRS gene redundancy : there are two threonyl- and two tyrosyl-tRNA synthetase genes (thrS/thrZ and tyrS/tyrZ, respectively).
  4. In contrast to the diversity of regulatory mechanisms thus far identified in E. coli, the expression of 14 out of the 21 aaRS genes in B. subtilis is regulated in a common manner: a transcriptional antitermination mechanism that does not exist in Gram-negative bacteria.34,35,36

Genetic Organization

B. subtilis has 21 genes (excluding genes for subunits) coding aaRS specific for 19 amino acids (see Fig. 1). Two of them, glycyl- and phenylalanyl-tRNA synthetase are composed of two subunits α and β, encoded by distinct, but co-transcribed genes (glyQ, glyS and pheS, pheT). As mentioned above, duplicate genes exist for the threonyl- and tyrosyl-tRNA synthetases. However, this rather appears to be an exception; other fully sequenced Gram-positive genomes such as those of Staphylococcus aureus, Bacillus halodurans, Lactococcus lactis or Deinococcus radiodurans do not display such a genotype. A gene initially thought to encode a second B. subtilis histidyl-tRNA synthetase (hisZ) actually encodes an aaRS paralog which lacks aminoacylation activity. It has been characterized as an essential component of the first enzyme in histidine biosynthesis.37 None of the B. subtilis aaRS genes is in a chromosomal context similar to that of E. coli (compare Figs. 1 and 4). In addition, most of them are grouped in three regions of the chromosome (Fig. 4). With the exception of trpS and tyrZ, transcription of all of the B. subtilis aaRS genes is in the same orientation as replication. This is typical of highly expressed genes and probably prevents transcription from interfering with genome replication.38 There is only one documented case of cotranscribed aaRS genes in B. subtilis, the gltX-cysE-cysS operon. It contains genes for both the glutamyl- and cysteinyl-tRNA synthetases separated by a gene encoding serine acetyl-transferase, the first enzyme of cysteine biosynthesis. Transcription of this operon originates 45 nucleotides upstream of gltX from a σA promoter.39 The genes encoding histidyl- and aspartyl-tRNA synthetases (hisS and aspS) are probably also cotranscribed. Only 13 nucleotides separate aspS from the upstream hisS gene, putting it potentially under control of the hisS regulatory sequences, although this has yet to be shown. In both cases, the genetic organization of the described genes is conserved in closely related organisms like Bacillus halodurans or Staphylococcus aureus but different in more distantly related Gram-positive bacteria.

Figure 4. B.

Figure 4

B. subtilis aminoacyl-tRNA synthetase genes. The aaRS genes are shown in boldface type together with adjacent genes. Positions are given in degrees and are calculated values from the complete genome sequence (Subtilist database). Arrows indicate the directions (more...)

A Specific But Conserved Control Mechanism

Nineteen genes in B. subtilis have been identified as members of the so-called T-box family, on the basis of conservation of sequence and structural elements in their leader regions. They include fourteen aaRS genes, three biosynthetic operons (ilvB-leu, proBA, proI) one potential aminoacid permease (yvbW) and one gene whose product intervenes in the regulation of the trp operon (yczA). Their untranslated leader sequences are about 300 nucleotides long and include a factor-independent transcription terminator, just upstream of the translation initiation site, which attenuates transcription of the structural gene.4042 Deletion of the leader terminator leads to constitutively high levels of expression. A particular case is the normally silent thrZ gene coding the second threonyl-tRNA synthetase.43 Its leader spans over 800 nucleotides and can be considered as containing the thrS leader-terminator motif repeated three times. The presence of three strong transcription terminators upstream of the structural gene probably explains why this gene is normally not expressed.40

The aaRS genes of this family are specifically derepressed in response to starvation for their cognate amino acid by increasing transcriptional readthrough at the terminator. Induction ratios vary from 2.5 to about 30. The primary leader sequences of the various genes are very different but can be folded into similar secondary structures.44 A scheme of the thrS leader based on an experimental determination of the RNA structure is shown in Figure 6.45 Despite great sequence diversity, several elements crucial for regulation are conserved in both sequence and position.34,35,44 The most prominent is the T-box element, located immediately upstream of the leader terminator (Fig. 5). A sequence complementary to the central part of the T-box in the 5' half of the terminator stem allows the formation of a mutually exclusive but unstable antiterminator structure. The model for specific modulation of terminator read-through is based on the presence of a triplet, known as the specifier codon, which specifies the cognate amino acid for the synthetase in question (Table 1). The specifier codon is always found in the same strategical position within the so-called specifier domain (Fig. 6). The effector molecule which signals limitation for the specific amino acid is the cognate uncharged tRNA which can interact directly with the leader by making at least two contacts.44 The first is a codon-anticodon interaction with the specifier codon. The second occurs by basepairing between the 5'-UGGN-3' of the T-box in the antiterminator side bulge and the 5'-N'CCA-3' acceptor end of uncharged tRNA (Fig. 7). The discriminator base (N') of the tRNA acceptor stem and the variable position within the T-box (N) co-vary. Thus, the first interaction is responsible for the specificity of induction while the second is thought to stabilize the antiterminator structure allowing the RNA-polymerase to escape termination. Both interactions have been well documented by genetic experiments. In several cases, altering the specifier codon affords a switch in the specificity of induction, generally with lower efficiency.44,46,47 However, adaptation of both points of interaction in the leader to optimally accommodate a different tRNA does not always permit a switch in induction specificity.46,48 At present, no specific tRNA determinants have been identified that can explain why an attempt to change the specificity of control was successful or not. Nevertheless, the tRNA-mRNA interaction is highly constrained and probably takes into account the overall tertiary structure of the tRNA.49 Additional contacts between tRNA and mRNA might also be important for regulation. An exceptionally high complementarity exists between the D- and T-arms of tRNATrp and the leader of the T-box controlled biosynthetic trp operon in Lactococcus lactis. In this system overexpression of wild-type tRNATrp induces expression even in the absence of the codon-anticodon interaction, albeit at a 20-fold lower level. This suggests that the extensive complementarity between the tRNA and the leader actually helps to recruit the tRNA for antitermination.50 This phenomena cannot be generalized to other T-box leaders, notably in B. subtilis where potential complementarities are much less convincing. However, they would be a good starting point to look for additional tRNA-leader interactions.

Figure 6. Scheme of the B.

Figure 6

Scheme of the B. subtilis thrS leader. Locations of conserved sequences (see text) are shown. The threonine specifier codon ACC is also marked.

Figure 5. T-box and downstream complementary sequences from 14 aaRS and 5 other genes (see text) in B.

Figure 5

T-box and downstream complementary sequences from 14 aaRS and 5 other genes (see text) in B. subtilis. Proposed factor-independent transcription terminators are depicted by inverted arrows. The distances to the initiation codon are indicated. Uppercase (more...)

Table 1. Aminoacyl-ARNt synthetase genes in Bacillus subtilis.

Table 1

Aminoacyl-ARNt synthetase genes in Bacillus subtilis.

Figure 7. Model of transcriptional antitermination and processing of the thrS leader.

Figure 7

Model of transcriptional antitermination and processing of the thrS leader. The endonucleolytic cleavage site is depicted by a scissors symbol.

Currently, only four of the 14 nucleotides of the T-box itself have yet been assigned a role and other short conserved leader sequences depicted in Figure 6 are also crucial for control (unpublished results).51 It has been suggested that some of the T-box nucleotides are conserved because they are involved in base pairing interactions with nucleotides which are also used by the terminator.51A Most substitutions in the AG-, GNUG and F-box completely abolish regulation. At present their function is totally unknown. Structural probing of the thrS leader in vitro and in vivo clearly shows that the specifier domain is thermodynamically more stable and better defined in vivo than in vitro.45 This difference might be explained by the presence of interacting proteins in vivo. Binding of these factors, if required for the regulation of all T-box genes, would presumably occur at conserved sequence or structural elements such as the AG- or GNUG boxes. The resulting stabilization of the specifier domain could be important for a productive interaction with the tRNA. The importance of proteins in the T-box system has recently been highlighted by the successful reconstitution of this antitermination system (thrS) in vitro, using the wild type regulatory tRNAThr isoacceptor and a partially purified protein fraction. As predicted by the model, aminoacylation of the tRNA with threonine completely abolishes its ability to act as an effector.51B A functional in vitro system should greatly increase the chances of identifying the missing protein factors likely to be common to all T-box regulated genes.

Uncharged tRNA is the principal effector of antitermination in this system. However, it is the ratio between charged and uncharged tRNA that is sensed by the system. An increase in expression of the biosynthetic ilv-leu operon caused by a less chargeable leucine tRNA mutant can essentially be suppressed in a strain diploid for mutant and wild-type alleles.52 This suggests that both charged and uncharged tRNA may compete for interaction with the ilv-leu leader. Moreover, a 10-fold increase in the absolute amount of uncharged tRNATrp does not increase expression of the L. lactis trp operon, if the amount of charged tRNA increases in parallel. On the other hand, a low level of uncharged tRNA can be sufficient for induction if the ratio to charged tRNA is increased.53 Based on this observation, overexpression of any T-box regulated aaRS should repress its own expression. This is true for the threonyl-tRNA synthetase ThrS; a 10-fold overproduction of the synthetase causes a 10-fold repression of a thrS-lacZ transcriptional fusion. Moreover, a progressive reduction of ThrS synthesis leads to induction of the normally silent thrZ gene in a dose-compensatory manner. Overproduction of either gene on multicopy plasmids causes various degrees of auto- and cross-repression, with ThrS being the much more potent repressor.54 Similar results are obtained with leuS55 cys39 and valS56 but curiously overexpression of the PheST synthetase has no effect on the expression of a pheS-lacZ fusion.46 The most likely scenario is that the negative role played by the synthetases is indirect, and occurs by altering the ratio of charged to uncharged tRNA in the cell.

Growth Rate Dependent Control

Production of tRNA increases with growth rate and, logically, so too should the synthesis of the enzymes responsible for charging them. In E. coli the expression of most aaRS is induced two- to three-fold for a five-fold increase in growth rate. Few data are available in B. subtilis. Proteome and transcriptome data on the global expression pattern of B. subtilis (including the aaRS) have recently become available for a limited set of conditions. Decreasing the doubling time from 45 min to 25 min through addition of amino acids to a minimal medium does not significantly alter either the protein or the mRNA concentrations of 18 aaRS (U. Mäder and M. Hecker, pers. communication). On the other hand, expression of a thrS-lacZ transcriptional fusion is increased 3.5-fold when bacteria are grown in complex rather than minimal medium. This regulation does not occur at the level of transcription initiation but does depend on leader terminator read-through. Various mutants in the specifier codon and other regions of the leader apart from the terminator are still regulated by growth rate suggesting that this type of control is independent of the tRNA-leader mRNA interaction and tRNA charging levels.46 One possible explanation is that growth in complex medium increases the transcription elongation rate which is known to favor readthrough of intrinsic terminators.57

Leader mRNA Processing

The leader regions of several aaRS genes in B. subtilis are cleaved between the T-box and the terminator.58 The efficiency of the cleavage of the thrS leader is specifically increased by threonine starvation, suggesting that it occurs in the antiterminator conformation in close proximity to the tRNA: :T-box interaction (Fig. 7). Cleavage of the thrS leader at this site creates a functional thrS mRNA starting with a <<terminator>> (i.e., a putative 5' stabilizer motif, see chapters 9 and 10 by Beran et al and Dreyfus and Joyce, respectively) and ending with a terminator. This configuration of the processed transcript makes it up to five times more stable than the full-length mRNA. During threonine starvation, almost 90% of thrS expression comes from the more stable processed transcript. This cleavage event can thus be considered as an amplifier of antitermination by increasing the half-life and hence the steady state concentration of the read-through mRNA transcripts.58

When transferred on a plasmid, processing of the B. subtilis thrS leader can occur at the same site in E. coli. This cleavage is catalyzed by E. coli RNase E, both in vivo and in vitro, suggesting that a functional homologue of RNase E is responsible for thrS processing in B. subtilis.59 The Bacillus counterpart to RNase E with respect to maturation of the 5S ribosomal RNA is RNase M5. In contrast to the single strand cleavages by RNase E, the latter cleaves twice in a double-helical region of a precursor of 5S rRNA to yield mature 5S rRNA in a single step.60 Moreover, RNase M5 activity depends on the presence of ribosomal protein L18 which probably locks the precursor rRNA into a conformation recognizable by the nuclease.61 It is therefore not surprising that inactivation of its gene (rnmV) does not affect thrS leader processing.62 The identity of the nuclease responsible for this cleavage is still unknown.

Another cleavage, which differs in both position and mechanism from the thrS processing event occurs in the gltX-cysE-cysS operon. Expression of cysE/cysS is partially uncoupled from that of gltX by the presence of a tRNACys dependent transcriptional attenuator of the T-box type between gltX and cysE.39 Transcripts escaping termination at this site are cleaved just 3' to the terminator structure in vivo and in vitro, which explains why one finds mainly monocistronic gltX mRNA in the cell. The fact that cleavage occurs in vitro in the absence of proteins other than RNA polymerase suggests that it is due either to the latter or to self-cleavage. This processing leaves a stable secondary structure (the terminator) at the 3' end of the gltX transcript and also allows the formation of a putative hairpin structure at the 5' end of the cysE/cysS transcript. Cleavage also removes any single stranded residue 3' of the gltX terminator which probably renders the upstream mRNA particularly resistant to 3' exonucleases. Accordingly, the steady-state mRNA level of gltX is much greater than that of cysE/cysS.36,63

Structure and Expression of E. coli Translation Factor Genes

Decoding the genetic message is an active process requiring ribosomes, translation factors and the amino acids which are brought to the ribosome by specific tRNAs. Protein synthesis can be divided into four steps: peptide chain initiation, elongation, termination and the last step, disassembly of the termination complex to provide a pool of ribosomal subunits ready for new initiation events. Each step requires specific proteins that interact with the tRNAs, mRNA, and/or ribosomes. Expression of these proteins has to be regulated and adapted in order to obtain optimal protein synthesis rates under various conditions. Table 2 gives an overview of the principal functions of E. coli translation factors.

Table 2. Function and nomenclature of translation factors.

Table 2

Function and nomenclature of translation factors.

Initiation Factors

Translation initiation comprises a series of events ensuring the correct recognition of the initiation signals on the mRNA. It is an important control point and probably the rate-limiting step in protein biosynthesis. A key intermediate of initiation is the 30S initiation complex consisting of the 30S ribosomal subunit, mRNA, the initiator tRNA (fMet-tRNAfMet) and the three initiation factors IF1, IF2-GTP, and IF3.64 IF3 assisted by IF1 promotes the dissociation of vacant 70S ribosomes thereby providing the pool of free ribosomal subunits ready for initiation. IF2 specifically recruits fMet-tRNAfMet and positions it correctly in the ribosomal P site. IF3 ensures the exclusive usage of AUG, GUG or UUG as start codons through specific recognition of the anticodon stem of the initiator tRNA. IF1, by binding to the ribosomal A site, confers specificity on the formation of the 30S initiation complex by occluding premature access of an elongator tRNA to the A site. Binding of the 50S subunit with the concomitant release of IF1 and IF3 gives rise to the 70S initiation complex with fMet-tRNAfMet in the ribosomal P site. The GTP carried by IF2 is hydrolyzed, IF2-GDP is released and the ribosomal A site is now ready to accept the first elongator aminoacyl-tRNA carried by elongation factor Tu (for details, see chapter 19 by de Smit and van Duin).

The three E. coli initiation factors are present in approximately equimolar amounts in the cell at about 0.2 to 0.3 molecule of each factor per ribosome.4 This is enough to saturate free 30S particles.64 The steady-state levels of the initiation factors are coordinately regulated relative to one another and relative to ribosome levels in exponentially growing bacteria.65 The genes for the three initiation factors (infA/IF1, infB/IF2, infC/IF3) are dispersed on the E. coli chromosome. With the exception of infA, they are associated with other components of the translational apparatus (Fig. 8).

Figure 8. E.

Figure 8

E. coli translation factor genes. The translation factor genes are shown in boldface type together with adjacent genes. Positions are given in min and are calculated values from the complete genome sequence (Colibri database). Arrows indicate the directions (more...)

Initiation Factor 1 (infA)

IF1 is the smallest initiation factor (8.1 kD) and is an essential protein.66 When IF1 levels in the cell are reduced, polysomes become smaller and cell growth decreases. The infA gene is located at 20 min on the E. coli chromosome. It is transcribed counterclockwise from two σ70 promoters, P1 and P2, yielding monocistronic mRNAs of 525 and 330 nucleotides, respectively. The smaller transcript is about two-fold more abundant than the larger one, but both end at the same Rho-independent terminator located immediately downstream of the infA coding region.67 Measurements of transcriptional fusions to lacZ show that P2 is indeed twice as active as P1. Growth rate-dependent control occurs exclusively at the P2 promoter. P2 directed β-galactosidase activity increases 1.7-fold with a doubling of the growth rate. Therefore, the entire increase in IF1 levels under these conditions can be attributed to an increased rate of IF1 synthesis. The infA gene is not autogenously regulated; neither transcription nor translation of infA is affected by high cellular levels of IF1.67

Initiation Factor 2 (infB)

IF2, the largest initiation factor is an essential GTP binding protein.68,69 In E. coli three natural forms of IF2 exist in the cell, IF2α (97,2 kD), IF2β1 (79,7 kD) and IF2β2 (7 amino acids shorter than IF2β1). The shorter IF2β species result from translation at in-frame start codons in the E. coli infB gene.70,71 The N-terminal part of the protein accounts for the affinity of IF2 to the 30S subunit72 but each of the three forms of IF2 can assure cell survival. However, expression of only one form slows growth at 37°C and results in a cold-sensitive phenotype.71 Even truncation of the entire N-terminal domain of IF2 still permits cell survival provided the protein is overexpressed.69 The infB gene is located at 69 min. on the E. coli chromosome and is part of the complex, multi-functional metY-nusA-infB operon (Fig. 9). The first gene of this operon, metY, encodes a minor form of the initiator tRNAf2Met. The rest of the operon consists of protein-coding genes including yhbC (p15A, a protein of unknown function), nusA (NusA, involved in the modulation of transcription termination), infB, rbfA (RbfA, a ribosome binding factor) and truB (a tRNA-modifying pseudouridine synthase). The nusA-infB core of the operon is conserved in evolutionarily very distant organisms such as B. subtilis, B. stearothermophilus, M. xanthus and T. thermophilus.

Figure 9. Gene organisation and transcripts of the metY-infB and the adjacent rpsO-pnp operons.

Figure 9

Gene organisation and transcripts of the metY-infB and the adjacent rpsO-pnp operons. The gene names are given above the boxes. Promoters (P) and terminators (t) are indicated, the transcripts are shown by dashed arrows. Genes known to be induced in the (more...)

Three promoters (P-1, P0 and P2) direct transcription of the operon8 (Fig. 9). Transcripts can extend into the downstream rpsO-pnp operon but the latter is essentially expressed from its own strong promoters. P-1 and P0, located upstream of metY, are the principal promoters of the operon.73,74 An upstream activating sequence (UAS) is recognized by the protein FIS, albeit with a ten-fold lower efficiency than that of the tufB operon75 (see below); The effect of FIS binding on the expression of the nusA-infB operon in vivo is not known. Most of the transcripts initiating at P-1 and P0 are terminated at the Rho-independent terminators t1 and t2 located between metY and yhbC producing short initiator tRNA precursors. Readthrough of t1 and t2 allows for the expression of nusA and infB. Most transcripts further extend past the weak Rho-independent terminator t3 downstream of infB allowing expression of rbfA and truB (Fig. 9). In addition, the various readthrough transcripts can be cleaved by RNase III immediately downstream of metY. This has two consequences: it separates the initiator tRNA from the coding polycistronic mRNA and, by removing the 5' terminal structural motif (the tRNA) from the mRNA causes a more rapid decay of its 5' end. However, this has no effect on the expression of NusA and IF2 since they are not overproduced in a RNase III mutant strain.76 Dissociation of the tRNA from the mRNA is also achieved directly by terminating metY transcripts at t1 or t2 and using the minor P2 promoter downstream of metY.

The NusA protein negatively controls the expression of its operon at the transcriptional level. This was shown by overexpressing NusA from a multicopy plasmid. Under these conditions the expression of lacZ-fusions containing various regions of the operon is repressed two-fold. Protein and gene fusions behave the same way indicating that NusA acts at the transcriptional level.77 Similarly, in a NusA mutant strain expressing only 30% of the wild-type NusA level, IF2 synthesis is increased five-fold.78 Overexpression of IF2 does not affect the regulation of the operon.

NusA, IF2, RbfA and the products of the downstream rpsO-pnp operon, the ribosomal protein S15 (rpsO) and polynucleotide phosphorylase (pnp) are all cold shock-induced proteins.79,80 At 37°C, the rpsO-pnp operon is essentially expressed from its own transcriptional signals. However, a cold shock results in a dramatic change in the spectrum and length of transcripts originating upstream of metY. Several major cold shock proteins, namely CspA, CspC and CspE, which are induced upon temperature decrease, increase readthrough of the t1 and t2 terminators and allow efficient cotranscription of both the nusA-infB and rpsO-pnp operons.81

Initiation Factor 3 (infC)

Like the other initiation factors, IF3 is an essential protein in E. coli.82 The infC gene, located at 37 min on the E. coli chromosome, is part of the thrS-infC-rpmI-rplT gene cluster encoding threonyl-tRNA synthetase, IF3, and the ribosomal proteins L35 and L20, respectively.83 Only three base pairs separate the thrS stop codon from the atypical AUU initiation codon of infC.83,84 Transcription of infC is initiated from three promoters, PthrS, P0 and P0185 (Fig. 10). PthrS, located upstream of thrS, is the source of bicistronic thrS-infC and tetracistronic thrS-infC-rpmI-rplT transcripts. P0 and P0' lie within thrSand the latter is the major promoter for the expression of infC, rpmI and rplT. The t1 terminator located immediately downstream of infC is not very efficient and roughly 50% of the transcripts generated from the upstream promoters are continued through rplT. Finally, transcription of rpmI and rplT can also initiate at the weak P1 promoter within the infC reading frame (Fig. 10). Thus, initiation and termination signals are arranged to yield a set of overlapping transcripts from this operon.

Figure 10. Gene organisation and transcripts of the thrS-infC operon.

Figure 10

Gene organisation and transcripts of the thrS-infC operon. The gene names are given above the boxes. Promoters (P) and terminators (t) are indicated, and the corresponding transcripts are shown by dashed arrows whose thickness is roughly proportional (more...)

Expression of infC is controlled independently from that of thrS. There is five times more IF3 protein and mRNA in the cell than threonyl-tRNA synthetase and thrS mRNA. Moreover, the steady state level of infC mRNA does not vary with cell growth indicating that growth rate dependent synthesis of IF3 is not under transcriptional control nor due to differential mRNA stability.86

IF3 uses its ability to differentiate between ‘normal’ and ‘abnormal’ initiation codons to autoregulate its translation, which begins with an AUU initiation codon. Such an initiation codon is unique in E. coli but well conserved in other IF3 genes even in distantly related organisms. An exception is Myxococcus xanthus which initiates infC translation at an equally unusual AUC codon. The AUU start codon is not only essential,87 but also sufficient,88 for autoregulation. Changing the AUG initiation codons of the thrS or rpsO genes to AUU increases their expression in an infC mutant background. However, under conditions of IF3 excess, repression of these mutant genes is weaker than that observed for infC, suggesting that the infC message has specific features that render its expression particularly sensitive to regulation.88 One of them could be a very rare second potential ribosomal binding site between the probable Shine-Dalgarno region and the AUU start codon. Finally, parts of the infC mRNA can be folded into a long range secondary structure occluding the ribosomal binding site of the downstream rpmI gene. This structure is specifically recognized and stabilized by the L20 protein (rplT) and allows for translational repression of the downstream rpmI and rplT genes.89

Elongation Factors

After formation of the 70S initiation complex, three major protein factors, EF-Tu, EF-Ts and EF-G, catalyze the next step in translation, peptide chain elongation. EF-Tu transports the aminoacylated tRNAs to the A-site of the ribosome in the form of the ternary complex EF-Tu·GTP·aa-tRNA. Subsequent GTP hydrolysis causes its release in the form of a binary complex, EF-Tu·GDP. EF-Ts catalyses the GDP/GTP nucleotide exchange on EF-Tu in order to reactivate the factor. An analog of EF-Tu, SelB, brings the only elongator tRNA not recognized by EF-Tu, selenocysteyl-tRNASec, to the ribosome. After peptide bond formation, EF-G complexed with GTP promotes the translocation of the mRNA bound peptidyl-tRNA from the A to the P site of the ribosome, a process requiring hydrolysis of the GTP molecule (for details, see chapter 20 by P. Nissen et al). A less well-characterized translation factor, EF-P, probably allows peptide bond formation to occur more efficiently with some aminoacyl-tRNAs that are poor acceptors for the ribosomal peptidyltransferase.

Elongation Factor Tu (tufA, tufB)

Gene Organization and Structure

Two unlinked genes tufA and tufB encode the elongation factor EF-Tu, a monomeric protein of 43.1 kD. The two genes are located in different operons and differ at only 13 nucleotides. EF-TuA and EF-TuB are identical except for the C-terminal residue.90,91 The tufA gene located at 74.6 min on the E. coli chromosome is part of the streptomycin (str) operon rpsL-rpsG-fusA-tufA encoding ribosomal proteins S12 and S7, EF-G and EF-TuA, respectively (Fig. 11). This operon is one of the most conserved in prokaryotic evolution. The tufB gene maps at 90 min and is preceded by the four tRNA genes thrU, tyrU, glyT and thrT coding for tRNA4Thr, tRNA2Tyr, tRNA2Gly and tRNA3Thr, respectively (Fig. 11).

Figure 11. Gene organisation and transcripts of the two operons carrying a gene for EF-Tu.

Figure 11

Gene organisation and transcripts of the two operons carrying a gene for EF-Tu. (A) str operon comprising rpsL, rpsG, fusA and tufA encoding ribosomal proteins S12, S7 and elongation factors EF-G and EF-TuA, respectively. (B) thrU-tufB operon showing (more...)

Regulation of Expression

EF-Tu is one of the most abundant proteins in the cell, representing up to 10% of the total protein content. There is about one EF-Tu molecule for every tRNA and this ratio is constant under different growth conditions.92 Normally, 75–90% of the tRNAs are charged and bound to EF-TuσGTP, forming a reactive ternary complex.93 Two gene copies are probably maintained to achieve efficient expression of EF-Tu. Either of the tuf genes can be deleted without loss of viability. However, deletion of tufA increases the generation time by one third.94 At least one half of the cellular EF-Tu is derived from tufA95 and this proportion remains constant under different growth conditions.

Three to four times more EF-TuA is synthesized than S12, S7 and EF-G whose genes are cotranscribed with tufA from a common upstream promoter (P1, Fig. 11A). Two observations account for this differential expression. First, there are two additional promoters (P2 and P3) located within the upstream fusA gene (EF-G) which are functional in vivo and together contribute up to 25% to tufA transcription.96 Second, tufA expression is not affected by S7 which acts as a translational repressor for the first three genes of the operon (see below, EF-G). Thus, increased transcription and uncoupled translation allows for a selectively enhanced expression of EF-TuA. Depletion or overexpression of EF-Tu has no effect on tufA expression. However, some results suggest that overproduction of EF-Tu negatively autoregulates EF-TuB expression.97

The tufB gene is transcribed with four upstream tRNA genes yielding a multicistronic mRNA of about 1.8 kb. This transcript can be processed in the thrT-tufB intergenic region separating the structural genes from the tufB mRNA (Fig. 11B).98 A partial deletion of the four tRNA genes has no significant effect on tufB expression. The major control element of this operon is a cis-acting AT-rich upstream activator sequence (UAS), 5' to the major P1 promoter.75 Deletion of this sequence results in a 10- to 15-fold drop in transcription. The UAS constitutes a binding site for the pleiotropic regulator, FIS,99 which also recognizes similar sequence elements upstream of rRNA and tRNA operons (see chapter 23 by M.M. Barker and R.L. Gourse). FIS binds to two sites in the UAS of the tufB operon and activates transcription by facilitating the binding of RNA polymerase to the promoter. Activation of the tufB operon upon a nutritional upshift is drastically reduced in fis mutant strains.100 In addition, expression of this operon is subject to stringent control. In vitro, selective inhibition of tufB transcription by ppGpp depends on the presence of a G/C rich discriminator sequence between the -10 region of the promoter and the first base of thrU. Similarly located G/C elements are crucial for many stringently controlled promoters. Thus, the tufB operon is very well adapted to respond quickly and efficiently to nutritional changes.

Elongation Factor SelB (selB)

The elongation factor SelB is the key molecule for the specific incorporation of the amino acid selenocysteine into polypeptides. It specifically recognizes the selenocysteine charged tRNASec, which has a UCA anticodon, in an EF-Tu like manner. This allows insertion of selenocysteine at in-frame UGA stop codons. In E. coli SelB binds GTP, selenocysteyl-tRNASec and a stem-loop structure immediately downstream of the UGA codon (the SECIS sequence).101 The absence of active SelB prevents the participation of selenocysteyl-tRNASec in translation.102

The selB gene is cotranscribed with selA encoding the selenocysteine synthase, which converts the serine attached to tRNASec to selenocysteine. The selAB operon maps at 81 min on the E. coli chromosome and is expressed from a rather weak σ70 promoter affording transcription initiation 48 bases upstream of selA.103 The selA termination codon overlaps the selB initiation codon suggesting tight translational coupling and hence coordinate synthesis of the two proteins. Consistent with this idea, both SelA and SelB are present in the cell at 1200 to 2000 copies. Expression of selAB and other sel genes is constitutive in both aerobically and anaerobically grown E. coli cells. This reflects the need to synthesize selenoproteins under both conditions even though some of them are differentially expressed. Weak transcription and low stability of the transcripts account for the low level of full-length selAB mRNAs in the cell. These features probably obviate the requirement for regulation, for example, in response to the intracellular selenium concentration.103

Elongation Factor G (fusA)

EF-G in complex with GTP stimulates the translocation of the mRNA/tRNA complex from the A to the P site of the ribosome. It is an essential monomeric protein of 77.4 kD and is present in the cell at about one molecule per ribosome.4

The fusA gene encoding EF-G is located upstream of the tufA gene within the str operon (see Fig. 11). All four genes of this operon are cotranscribed from a common promoter. Expression of the first three, rpsL (S12), rpsG (S7) and fusA is translationally coupled. The ribosomal protein S7 acts as a translational repressor by binding to a complex RNA stem-loop structure in the intergenic region between rpsL and rpsG, probably inducing a conformational change in this structure.104,105 A crucial element required for repression is an anti-SD sequence that sequesters the ribosomal binding site of rpsG in a double stranded structure thereby inhibiting ribosome binding. This translational feedback regulation by S7 inhibits translation of its own gene and that of fusA whose start codon is only 28 bp downstream of rpsG. Interestingly, the stem-loop structure is also important for the coupled translation of S12 and S7. It has thus been proposed that ribosomes skip the intercistronic loop in order to reinitiate at the rpsG initiation site which is brought right next to the rpsL stop codon by RNA folding. In addition, S7 represses only coupled translation of its gene allowing for a small basal level of unregulated expression of S7 and presumably also EF-G.106 All three proteins S12, S7 and EF-G are thus coordinately expressed. This is consistent with the fact that the cellular level of EF-G corresponds to that of ribosomes. EF-Tu expression from the distal tufA gene is not under control of the translational repressor S7 and is translationally uncoupled. However, synthesis of both EF-Tu and EF-G proteins is stringently controlled and subject to growth rate dependent regulation (see above, EF-Tu).

Elongation Factor Ts (tsf)

The EF-Tu nucleotide exchange factor EF-Ts is present in the cell at about 0.2 molecule per ribosome which is 30 times less than its << substrate >> EF-Tu.4 EF-Ts is expressed from a bicistronic operon, located at 4.1 min on the E. coli map, together with ribosomal protein S2 (rpsB).107 A single major promoter immediately upstream of rpsB directs transcription of the rpsB-tsf operon. Two potential Rho-independent transcription terminators are present in the operon, one in the intergenic region between the two genes and one distal to tsf. Like EF-Tu and EF-G, EF-Ts synthesis is under stringent control and coordinated with ribosomal protein synthesis in a growth rate dependent manner.108

Elongation Factor P (efp)

The role of the EF-P protein in the elongation process has been little considered thus far. Yet, it is an essential well-conserved factor required for protein synthesis and has a eukaryotic counterpart, eIF5A. EF-P stimulates the peptidyltransferase activity of fully assembled 70S ribosomes,109,110 preferentially promoting first peptide bond synthesis. EF-P activity requires the ribosomal protein L16, suggesting that the binding site of EF-P may overlap the peptidyltransferase center. In contrast to the other elongation factors it does not require GTP for its action. This 20.5 kD protein is encoded by the efp gene located at 94.3 min on the E . coli genome. There are 800–900 EF-P molecules in the cell (0.1–0.2 copy per ribosome) and this ratio is independent of the stage of cell growth.111

Release and Ribosome Recycling Factors

The elongation cycle of protein biosynthesis on the ribosome is brought to an end when a stop codon appears in the A (decoding) site of the ribosome. In contrast to translation initiation and elongation, stop codon recognition is not achieved by a tRNA, but by two specialized proteins called release factors (RF). RF1 recognizes UAG and UAA, and RF2 UGA and UAA stop codons. Binding of either one of the two factors triggers hydrolysis of peptidyl-tRNA. A third release factor, the GTP binding protein RF3, catalyses the release of RF1/RF2 and thereby accelerates the transition from termination to ribosome recycling. The latter step requires the ribosome recycling factor (RRF) which together with EF-G or RF3 dissociates the post-termination complex, a process involving GTP hydrolysis. Finally, the deacylated tRNA still present on the 30S particle is displaced by IF3 allowing recycling of the 30S subunit (for details, see chapter 21 by R.H. Buckingham and M. Ehrenberg).

Release Factors: RF1 (prfA) RF2 (prfB) and RF3 (prfC)

Of the three release factors, only RF3 is non-essential for cell growth. However, it is necessary for optimal translational activity, particularly under environmental stress conditions.112,113 The genes for RF1 and RF2 can both be knocked out if a mutant form of RF2, capable of terminating translation at all three stop codons, is present.114 The number of RF1 molecules per cell increases from 1200 to 4900, and that of RF-2 from 5900 to 24900 as growth rates increases from 0.3 to 2.4 doublings per hour. The strict one to five ratio between RF1 and RF2 is maintained independently of growth rate and corresponds well to the one to four ratio of UAG and UGA stop codons in E. coli. The concentration of RF1/2 in the cell is thus similar to that of the initiation factors. This seems logical since they are required at the same rate at opposite ends of the translational process. On the other hand, RF3 is present at a roughly 60-fold lower level than RF1/2 and its concentration can vary significantly from one strain to the other.115 Thus different strains can accommodate low RF3 concentrations very well. Some species, such as Mycoplasma have simply dispensed with the RF3 gene altogether.116

RF1 (40.5 kD) is encoded by the prfA gene, located at 27.3 min on the E. coli genome. Transcription probably initiates at a σ70 promoter 65 base pairs upstream of the initiation codon and results in a monocistronic transcript ending at a terminator distal to prfA. The termination codon for RF1 translation is UGA, recognized by RF2.117

RF2 (41.2 kD) is encoded by the prfB gene located at 65.4 min. A strong σ70 promoter with a stringent discriminator, immediately upstream of prfB, directs transcription of a 2.8 kb bicistronic transcript, including lysS and which ends at the lysS distal terminator.118 RF2 translation is terminated with a UGA codon recognized by itself. A premature in-frame UGA codon at position 26 is crucial to RF2 expression. A +1 frameshift at this position is necessary for complete translation of RF2. Two elements are required to achieve the remarkably high frameshifting rate of 50% : a correctly placed Shine-Dalgarno-like sequence and a particular codon context (Fig. 12). A CUU (leucine) codon preceding the UGA provides for a uracil-rich stretch, favorable for frameshifting. Furthermore, tRNALeuGAG is a <<shifty >> tRNA with the ability to cause a 4-base translocation.119 Pairing of the Shine-Dalgarno-like sequence with the 16S rRNA leads to stalling or slowing down of the elongating ribosome which probably favors and orients the shift to the leucine codon (Fig. 12, see also chapter 22 by I.P. Ivanov et al). This provides a natural mechanism of autogenous control at the translational level. When RF2 levels in the cell are low, the U of the UGA codon is made more available for re-pairing by tRNALeuGAG and frameshifting is favored over termination and vice versa. Autoregulation has been demonstrated to occur in vitro120 and in vivo.121

Figure 12. Frameshift region in the prfB gene.

Figure 12

Frameshift region in the prfB gene. Shown are the Shine-Dalgarno interaction and the CUU UGA C sequence where the +1 frameshift occurs. Shifting of tRNALeu from CUU to UUU is indicated by an arrow.

Release factor 3 (59.6 kD) is encoded by the prfC gene located at 99.3 min on the E. coli map. Transcription probably originates about 60 base pairs upstream of the initiation codon at a σ70 promoter.112 The presence of a terminator downstream of prfC suggests monocistronic transcription. However, nothing is known about the regulation of prfC as yet.

Ribosome Fecycling Factor (frr)

Ribosome recycling factor (RRF) is an essential protein.122 It acts after RF1/RF2 mediated peptidyl-tRNA hydrolysis to participate in the disassembly of the post-termination complex. In addition, it maintains translational fidelity during chain elongation. The molar amount of RRF in the cell is about 50% of that of the ribosomes and approximately 30% of total RRF is bound to ribosomes.123

The frr gene, coding for RRF (20.5 kD), is separated by only one gene from tsf, encoding EF-Ts. Albeit transcribed in the same orientation (Fig. 8) they do not form an operon. A very atypical σ70 type promoter directs transcription initiation 58 bp upstream of the initiation codon. Spacing between the weak −35 and the −10 region is 20 nucleotides which is extremely rare and, in fact, only one other case is known in E. coli. Deletion of the -35 region still allows for RNA polymerase binding and reduced promoter activity.124 A potential regulatory role for this promoter configuration remains to be analyzed. Finally, overexpression of RRF only slightly affects growth, suggesting that it is not harmful for elongating ribosomes.

Conclusions and Perspectives

Expression of the genes described in this chapter is regulated by many different mechanisms and yet represents only a portion of the regulatory capabilities of the prokaryotic cell. It has become evident that the highly developed organisms that we study today dedicate appreciable resources to regulatory functions. By doing so, they ensure survival in various natural habitats and acquire the flexibility to respond efficiently to the many challenges they experience. Different organisms often use different strategies to regulate corresponding genes, as is the case for the aminoacyl-tRNA synthetases. In E. coli expression of all aaRS is regulated in a growth rate dependent manner, but specific mechanisms, such as those employed by the pheST and thrSgenes, allow induction in response to starvation for the cognate amino acid. In B. subtilis 14 aaRS genes (the T-box family) are specifically induced by tRNA-mediated transcription antitermination. To our current knowledge, this mechanism is essentially confined to Gram positive bacteria. As different and ingenious these controls may be, they share a common effector, the tRNA. In E. coli it provokes the crucial ribosome stalling in pheST attenuation, it titrates thrS away from binding to its operator on the mRNA and it directly favors formation of the antiterminator in B. subtilis. The conservation of many different regulatory mechanisms allows for both gene-specific regulation and more global coordinate control of genes with related functions. Some control mechanisms that evolved early, have possibly persisted because they performed well and there has been no selective pressure to improve or replace them.

The high induction ratios observed for some B. subtilis aaRS genes in the absence of one of their substrates appear counterintuitive. However, the cost of increased aaRS synthesis is probably outweighed by the increased efficiency acquired to scavenge for diminishing amino acid pools. This option might be advantageous to B. subtilis, which is often confronted with poor nutritional conditions, thereby avoiding premature commitment to the even more costly process of spore formation. The same control mechanism also allows a very limited adaptation of gene expression suggested by the moderate derepression of several other aaRS genes under similar conditions.

Many aspects of aaRS biosynthesis are still unresolved in both organisms. The mechanism underlying growth rate-dependent regulation of virtually all aaRS genes in E. coli is not understood; for pheST and thrSthis regulation can be explained at least partially by the specific control mechanisms. In B. subtilis a metabolic upshift can induce thrSexpression by increasing read-through of the leader terminator in the absence of a tRNA: :mRNA interaction. This might occur via a faster transcription elongation rate in rich medium which is known to reduce terminator efficiency. The tRNA alone is not sufficient to promote specific antitermination in vitro. The additional protein factors which are required to make this system work still remain to be identified.

The regulatory mechanisms described support the notion that, in many cases, the primary function of particular components of the translation apparatus has been exploited for the purposes of regulation of their synthesis. The mechanisms used to control translation factor expression clearly support this view. For example, IF3 autoregulates its own translation using its capacity to discriminate between “good” and “bad” initiation codons. It does this, not by binding to its mRNA, but to the ribosome. RF2 normally brings translation to an end when a final UGA stop codon is reached. However, when in excess it blocks its own translation at an in-frame stop codon on its mRNA. In order to make this mechanism work, an efficient alternate reading of the genetic code was designed: a + 1 frameshift allowing synthesis of the functional protein. These very specific mechanisms permit a perfect coordination of the synthesis of these translation factors to that of the ribosomes. The synthesis of several other factors involved in translation is controlled by mechanisms that are not fully understood yet. A great number of overlapping regulatory signals exists in different operons, i.e., multiple promoters, transcription terminators and RNA structures mediating translational coupling. It is likely that they all serve the ultimate purpose of coordinating the expression of the different components of the translational apparatus and rendering it as efficient as possible. A good example are the normally independently transcribed nusA-infB and rpsO-pnp operons. During a cold shock, when these proteins are all required at elevated levels, specific proteins promote transcription attenuation at the multiple terminators present and transform the two operons into a single coordinately regulated transcription unit. Studies in other organisms stimulated by newly available genome sequences will allow us to compare how different bacteria cope with similar challenges.


We thank Drs. C. Condon, J. Lapointe and M. Springer for critical reading of the manuscript. This work was supported by funds from the Centre National de la Recherche Scientifique (UPR9073 and UMR) and PRFMMIP from the Ministère de l'Education Nationale.


Chatterji D, Ojha AK. Revisiting the stringent response, ppGpp and starvation signaling. Curr Opin Microbiol. 2001;4(2):160–5. [PubMed: 11282471]
Kisselev LL, Justesen J, Wolfson AD. et al. Diadenosine oligophosphates (Ap(n)A), a novel class of signalling molecules? FEBS Lett. 1998;427(2):157–63. [PubMed: 9607303]
Martinis SA, Plateau P, Cavarelli J. et al. Aminoacyl-tRNA synthetases: a new image for a classical family. Biochimie. 1999;81(7):683–700. [PubMed: 10492015]
Bremer H, Dennis PP. Modulation of chemical composition and other parameters of the cell by growth rate In: Neidhardt FC, Curtiss R, Ingraham JL, Lin EC, Low KB, Magasanik B, Reznikoff WS, Riley M, Schaechter M, Umbarger HE, eds.Escherichia coli and Salmonella: cellular and molecular biology Washington, D.C.: American Society for Microbiology;19961553–1569.
Jakubowski H, Goldman E. Quantities of individual aminoacyl-tRNA families and their turnover in Escherichia coli. J Bacteriol. 1984;158(3):769–76. [PMC free article: PMC215508] [PubMed: 6373741]
VanBogelen RA, Abshire KZ, Moldover B. et al. Escherichia coli proteome analysis using the gene-protein database. Electrophoresis. 1997;18(8):1243–51. [PubMed: 9298644]
Morgan S, Söll D. Regulation of the biosynthesis of amino acid:tRNA ligases and of tRNA. Prog Nucleic Acid Res Mol Biol. 1978;21:181–207. [PubMed: 358278]
Grunberg-Manago M. Regulation of the expression of aminoacyl-tRNA synthetases and translation factors In: Neidhardt FC, Curtiss R, Ingraham JL, Lin EC, Low KB, Magasanik B, Reznikoff WS, Riley M, Schaechter M, Umbarger HE, eds.Escherichia coli and Salmonella: cellular and molecular biology Washington, D.C.: American Society for Microbiology;19961432–1457.
Putzer H, Grunberg-Manago M, Springer M. Bacterial aminoacyl-tRNA synthetases: Genes and regulation of expression In: Söll D, RajBhandary, eds.tRNA: Structure, Biosynthesis, and Function Washington, D.C.;1995293–333.
Putney SD, Schimmel P. An aminoacyl-tRNA synthetase binds to a specific DNA sequence and regulates its gene transcription. Nature. 1981;291:632–635. [PubMed: 6264314]
Springer M, Mayaux JF, Fayat G. et al. Attenuation control of the Escherichia coli phenylalanyl-tRNA synthetase operon. J Mol Biol. 1985;181:467–478. [PubMed: 3158742]
Fayat G, Mayaux JF, Sacerdot C. et al. Escherichia coli phenylalanyl-tRNA synthetase operon region.Evidence for an attenuation mechanism. Identification of the gene for the ribosomal protein L20. J Mol Biol. 1983;171:239–261. [PubMed: 6317865]
Springer M, Plumbridge JA, Butler JS. et al. Autogenous control of Escherichia coli threonyl-tRNA synthetase expression in vivo. J Mol Biol. 1985;185:93–104. [PubMed: 3930755]
Lestienne P, Plumbridge JA, Grunberg-Manago M. et al. Autogenous repression of E. coli threonyl-tRNA synthetase expression in vitro. J Biol Chem. 1984;259:5232–5237. [PubMed: 6325425]
Moine H, Romby P, Springer M. et al. Messenger RNA structure and gene regulation at the translational level in Escherichia coli: the case of threonine:tRNAThr ligase. Proc Natl Acad Sci USA. 1988;85:7892–7896. [PMC free article: PMC282304] [PubMed: 3054873]
Nogueira T, de Smit M, Graffe M. et al. The Relationship Between Translational Control and mRNA Degradation for the Escherichia coli Threonyl-tRNA synthetase Gene. J Mol Biol. 2001;310(4):709–22. [PubMed: 11453682]
Graffe M, Dondon J, Caillet J. et al. The specificity of translational control switched using tRNA identity rules. Science. 1992;225:994–996. [PubMed: 1372129]
Romby P, Brunel C, Caillet J. et al. Molecular mimicry in translational control of E.coli threonyl-tRNA synthetase gene. Competitive inhibition in tRNA aminoacylation and operator-repressor recognition switch using tRNA identity rules. Nucleic Acids Res. 1992;20:5633–5640. [PMC free article: PMC334396] [PubMed: 1280807]
Sacerdot C, Caillet J, Graffe M. et al. The Escherichia coli threonyl-tRNA synthetase gene contains a split ribosomal binding site interrupted by a hairpin structure that is essential for autoregulation. Mol Microbiol. 1998;29(4):1077–90. [PubMed: 9767575]
Romby P, Caillet J, Ebel C. et al. The expression of E.coli threonyl-tRNA synthetase is regulated at the translational level by symmetrical operator-repressor interactions. EMBO J. 1996;15:5976–5987. [PMC free article: PMC452390] [PubMed: 8918475]
Newman EB, Lin R. Leucine-responsive regulatory protein: a global regulator of gene expression in E. coli. Annu Rev Microbiol. 1995;49:747–75. [PubMed: 8561478]
Lin R, Ernting B, Hirshfield IN. et al. The lrp product regulates expression of lysU in E.coli K12. J Bacteriol. 1992;174:2779–2784. [PMC free article: PMC205928] [PubMed: 1569010]
Gazeau M, Delort F, Dessen P. et al. E.coli leucine-responsive regulatory protein (Lrp) controls lysyl-tRNA synthetase expression. FEBS letters. 1992;300:254–258. [PubMed: 1555652]
VanBogelen RA, Acton MA, Neidhardt FC. Induction of the heat shock regulon does not produce thermotolerance in E.coli. Genes Dev. 1987;1:525–531. [PubMed: 3315852]
Oshima T, Ito K, Kabayama H. et al. Regulation of lrp gene expression by H-NS and Lrp proteins in Escherichia coli: dominant negative mutations in lrp. Mol Gen Genet. 1995;247:521–528. [PubMed: 7603430]
Brevet A, Chen J, Leveque F. et al. Comparison of the enzymatic properties of the two Escherichia coli lysyl-tRNA synthetase species. J Biol Chem. 1995;270(24):14439–44. [PubMed: 7782306]
Brun YV, Sanfaçon H, Breton R. et al. Closely spaced and divergent promoters for an aminoacyl-tRNA synthetase gene and a tRNA operon in E.coliTranscriptional and post-transcriptional regulation of glutX, valU and alaW. J Mol Biol. 1990;214:845–864. [PubMed: 2201777]
Champagne N, Lapointe J. Influence of FIS on the transcription from closely spaced and non- overlapping divergent promoters for an aminoacyl-tRNA synthetase gene (gltX) and a tRNA operon (valU) in Escherichia coli. Mol Microbiol. 1998;27(6):1141–56. [PubMed: 9570400]
Dardel F, Panvert M, Fayat G. Transcription and regulation of expression of the E.coli methionyl-tRNA synthetase gene. Mol Gen Genet. 1990;223:121–133. [PubMed: 2259334]
Petersen L, Downs DM. Mutations in apbC (mrp) prevent function of the alternative pyrimidine biosynthetic pathway in Salmonella typhimurium. J Bacteriol. 1996;178(19):5676–82. [PMC free article: PMC178406] [PubMed: 8824612]
Lapointe J, Duplain L, Proulx M. A single glutamyl-tRNA synthetase aminoacylates tRNAGlu and tRNAGln in Bacillus subtilis and efficiently misacylates Escherichia coli tRNA1Gln in vitro. J Bacteriol. 1986;165(1):88–93. [PMC free article: PMC214374] [PubMed: 3079749]
Wilcox M. g-Phosphoryl ester of Glu-tRNAGln as an intermediate in Bacillus subtilis glutaminyl-tRNA synthesis. Cold Spring Harbor Symp Quant Biol. 1969;34:521–528. [PubMed: 4985889]
Curnow AW, Hong K, Yuan R. et al. Glu-tRNAGln amidotransferase: a novel heterotrimeric enzyme required for correct decoding of glutamine codons during translation. Proc Natl Acad Sci U S A. 1997;94(22):11819–26. [PMC free article: PMC23611] [PubMed: 9342321]
Henkin TM. tRNA-directed transcription antitermination. Mol Microbiol. 1994;13:381–387. [PubMed: 7527891]
Condon C, Grunberg-Manago M, Putzer H. Aminoacyl-tRNA synthetase gene regulation in Bacillus subtilis. Biochimie. 1996;78(6):381–9. [PubMed: 8915527]
Pelchat M, Lapointe J. Aminoacyl-tRNA synthetase genes of Bacillus subtilis: organization and regulation. Biochem Cell Biol. 1999;77(4):343–7. [PubMed: 10546897]
Sissler M, Delorme C, Bond J. et al. An aminoacyl-tRNA synthetase paralog with a catalytic role in histidine biosynthesis. Proc Natl Acad Sci U S A. 1999;96(16):8985–90. [PMC free article: PMC17719] [PubMed: 10430882]
French S. Consequences of replication fork movement through transcription units in vivo. Science. 1992;258:1362–1365. [PubMed: 1455232]
Gagnon Y, Breton R, Putzer H. et al. Clustering and co-transcription of the Bacillus subtilis genes encoding the aminoacyl-tRNA synthetases specific for glutamate and for cysteine and the first enzyme for cysteine biosynthesis. J Biol Chem. 1994;269:7473–7482. [PubMed: 7510287]
Putzer H, Gendron N, Grunberg-Manago M. Co-ordinate expression of the two threonyl-tRNA synthetase genes in Bacillus subtilis: control by transcriptional antitermination involving a conserved regulatory sequence. EMBO J. 1992;11:3117–3127. [PMC free article: PMC556796] [PubMed: 1379177]
Henkin TM, Glass BL, Grundy FJ. Analysis of the Bacillus subtilis tyrS gene: Conservation of a regulatory sequence in multiple tRNA synthetase genes. J Bacteriol. 1992;174:1299–1306. [PMC free article: PMC206425] [PubMed: 1735721]
Grandoni JA, Zahler SA, Calvo JM. Transcriptional regulation of the ilv-leu operon of Bacillus subtilis. J Bacteriol. 1992;174(10):3212–3219. [PMC free article: PMC205988] [PubMed: 1577690]
Putzer H, Brackhage AA, Grunberg-Manago M. Independent genes for two threonyl-tRNA synthetases in Bacillus subtilis. J Bacteriol. 1990;172(8):4593–4602. [PMC free article: PMC213293] [PubMed: 2115870]
Grundy FJ, Henkin TM. Transfer RNA as a positive regulator of transcription antitermination in B.subtilis. Cell. 1993;74(3):475–482. [PubMed: 8348614]
Luo D, Condon C, Grunberg-Manago M. et al. In vitro and in vivo secondary structure probing of the thrS leader in Bacillus subtilis. Nucleic Acids Res. 1998;26(23):5379–87. [PMC free article: PMC148014] [PubMed: 9826762]
Putzer H, Laalami S, Brakhage AA. et al. Aminoacyl-tRNA synthetase gene regulation in Bacillus subtilis: induction, repression and growth-rate regulation. Mol Microbiol. 1995;16:709–718. [PubMed: 7476165]
Marta PT, Ladner RD, Grandoni JA. A CUC triplet confers leucine-dependent regulation of the Bacillus subtilis ilv-leu operon. J Bact. 1996;178:2150–2153. [PMC free article: PMC177919] [PubMed: 8606198]
Grundy FJ, Hodil SE, Rollins SM. et al. Specificity of tRNA-mRNA interactions in Bacillus subtilis tyrS antitermination. J Bacteriol. 1997;179(8):2587–94. [PMC free article: PMC179008] [PubMed: 9098057]
Grundy FJ, Collins JA, Rollins SM. et al. tRNA determinants for transcription antitermination of the Bacillus subtilis tyrS gene. RNA. 2000;6(8):1131–41. [PMC free article: PMC1369987] [PubMed: 10943892]
van De Guchts M, Ehrlich SD, Chopin A. Identity elements in tRNA-mediated transcription antitermination: implication of tRNA D- and T-arms in mRNA recognition. Microbiology. 2001;147(Pt 5):1223–33. [PubMed: 11320125]
Rollins SM, Grundy FJ, Henkin TM. Analysis of cis-acting sequence and structural elements required for antitermination of the Bacillus subtilis tyrS gene. Mol Microbiol. 1997;25(2):411–21. [PubMed: 9282752]
Grundy FJ, Moir TR, Haldeman MT. et al. Sequence requirements for terminators and antiterminators in the T box transcription antitermination system: disparity between conservation and functional requirements. Nucleic Acids Res. 2002;30(7):1646–55. [PMC free article: PMC101844] [PubMed: 11917026]
Putzer H, Condon C, Brechemier-Baey D. et al. Transfer RNA-mediated antitermination in vitro. Nucleic Acids Res. 2002;30(14):3016–3023. [PMC free article: PMC135736] [PubMed: 12136084]
Garrity DB, Zahler SA. Mutations in the gene for a tRNA that functions as a regulator of a transcriptional attenuator in Bacillus subtilis. Genetics. 1994;137:627–636. [PMC free article: PMC1206022] [PubMed: 8088508]
van de Guchte M, Ehrlich DS, Chopin A. tRNATrp as a key element of antitermination in the Lactococcus lactis trp operon. Mol Microbiol. 1998;29(1):61–74. [PubMed: 9701803]
Gendron N, Putzer H, Grunberg-Manago M. Expression of both Bacillus subtilis threonyl-tRNA synthetase genes is autogenously regulated. J Bacteriol. 1994;176:486–494. [PMC free article: PMC205072] [PubMed: 8288542]
Vanderhorn PB, Zahler SA. Cloning and nucleotide sequence of the leucyl-transfer RNA synthetase gene of Bacillus subtilis. J Bacteriol. 1992;174(12):3928–3935. [PMC free article: PMC206101] [PubMed: 1317842]
Luo D, Leautey J, Grunberg-Manago M. et al. Structure and regulation of expression of the Bacillus subtilis valyl-tRNA synthetase gene. J Bacteriol. 1997;179(8):2472–2478. [PMC free article: PMC178992] [PubMed: 9098041]
McDowell JC, Roberts JW, Jin DJ. et al. Determination of intrinsic transcription termination efficiency by RNA polymerase elongation rate. Science. 1994;266(5186):822–5. [PubMed: 7526463]
Condon C, Putzer H, Grunberg-Manago M. Processing of the leader mRNA plays a major role in the induction of thrSexpression following threonine starvation in B.subtilis. Proc Natl Acad Sci USA. 1996;93:6992–6997. [PMC free article: PMC38922] [PubMed: 8692931]
Condon C, Putzer H, Luo D. et al. Processing of the Bacillus subtilis thrS leader mRNA is RNase Edependent in Escherichia coli. J Mol Biol. 1997;268(2):235–42. [PubMed: 9159466]
Sogin ML, Pace B, Pace NR. Partial purification and properties of a ribosomal RNA maturation endonuclease from Bacillus subtilis. J Biol Chem. 1977;252(4):1350–7. [PubMed: 402365]
Pace B, Stahl DA, Pace NR. The catalytic element of a ribosomal RNA-processing complex. J Biol Chem. 1984;259(18):11454–8. [PubMed: 6432798]
Condon C, Brechemier-Baey D, Beltchev B. et al. Identification of the gene encoding the 5S ribosomal RNA maturase in Bacillus subtilis: Mature 5S rRNA is dispensable for ribosome function. RNA. 2001;7:242–253. [PMC free article: PMC1370082] [PubMed: 11233981]
Pelchat M, Lapointe J. In vivo and in vitro processing of the Bacillus subtilis transcript coding for glutamyl-tRNA synthetase, serine acetyltransferase, and cysteinyl-tRNA synthetase. RNA. 1999;5(2):281–9. [PMC free article: PMC1369759] [PubMed: 10024179]
Hershey JWB. Protein synthesis In: Ingraham JL, Low KB, Magasanik B, Schaechter M, Umbarger HE, eds.Escherichia coli and Salmonella typhimurium Washington, D.C.: American Society for Microbiology.;1987614–647.
Howe JG, Hershey JWB. Initiation factor levels are coordonately controlled in Escherichia coli growing at different rates. J Biol Chem. 1983;258:1954–1959. [PubMed: 6337147]
Cummings HS, Hershey JW. Translation initiation factor IF1 is essential for cell viability in Escherichia coli. J Bacteriol. 1994;176(1):198–205. [PMC free article: PMC205031] [PubMed: 8282696]
Cummings HS, Sands JF, Foreman PC. et al. Structure and expression of the infA operon encoding translational initiation factor IF1. Transcriptional control by growth rate. J Biol Chem. 1991;266(25):16491–8. [PubMed: 1909328]
Cole JR, Olsson CL, Hershey JW. et al. Feedback regulation of rRNA synthesis in Escherichia coli. Requirement for initiation factor IF2. J Mol Biol. 1987;198(3):383–92. [PubMed: 2448483]
Laalami S, Putzer H, Plumbridge JA. et al. A severely truncated form of translational initiation factor 2 supports growth of Escherichia coli. J Mol Biol. 1991;220(2):335–49. [PubMed: 1830345]
Plumbridge JA, Deville F, Sacerdot C. et al. Two translational initiation sites in the infB gene are used to express initiation factor IF2 alpha and IF2 beta in Escherichia coli. EMBO J. 1985;4(1):223–9. [PMC free article: PMC554173] [PubMed: 3894004]
Sacerdot C, Vachon G, Laalami S. et al. Both forms of translational initiation factor IF2 (alpha and beta) are required for maximal growth of Escherichia coli. Evidence for two translational initiation codons for IF2 beta. J Mol Biol. 1992;225(1):67–80. [PubMed: 1374802]
Moreno JM, Drskjotersen L, Kristensen JE. et al. Characterization of the domains of E.coli initiation factor IF2 responsible for recognition of the ribosome. FEBS Lett. 1999;455(12):130–4. [PubMed: 10428486]
Granston AE, Thompson DL, Friedman DI. Identification of a second promoter for the metY-nusA-infB operon of Escherichia coli. J Bacteriol. 1990;172(5):2336–42. [PMC free article: PMC208867] [PubMed: 1692017]
Ishii S, Kuroki K, Imamoto F. tRNAf2Met gene in the leader region of the nusA operon in E.coli. Proc Natl Acad Sci USA. 1984;81:409–413. [PMC free article: PMC344686] [PubMed: 6364142]
Verbeek H, Nilsson L, Baliko G. et al. Potential binding sites of the trans-activator FIS are present upstream of all rRNA operons and of many but not all tRNA operons. Biochim Biophys Acta. 1990;1050(13):302–6. [PubMed: 2207159]
Régnier P, Grunberg-Manago M. Cleavage by RNAse III in the transcripts of the metY-nusA-infB operon of E.coli releases the tRNA and initiates decay of the downstream mRNA. J Mol Biol. 1989;210:293–302. [PubMed: 2481042]
Plumbridge JA, Dondon J, Nakamura Y. et al. Effect of NusA protein on expression of the nusA, infB operon in E.coli. Nucleic Acids Res. 1985;13(9):3371–88. [PMC free article: PMC341241] [PubMed: 2987884]
Nakamura Y, Plumbridge J, Dondon J. et al. Evidence for autoregulation of the nusA-infB operon of Escherichia coli. Gene. 1985;36(12):189–93. [PubMed: 2998933]
Jones MD, Lowe DM, Borgford T. et al. Natural variation of tyrosyl-tRNA synthetase and comparison with engineered mutants. Biochemistry. 1986;25:1887–1891. [PubMed: 3011073]
Jones PG, Inouye M. The cold-shock response—a hot topic. Mol Microbiol. 1994;11(5):811–8. [PubMed: 8022259]
Bae W, Xia B, Inouye M. et al. Escherichia coli CspA-family RNA chaperones are transcription antiterminators. Proc Natl Acad Sci U S A. 2000;97(14):7784–9. [PMC free article: PMC16622] [PubMed: 10884409]
Olsson CL, Graffe M, Springer M. et al. Physiological effects of translation initiation factor IF3 and ribosomal protein L20 limitation in Escherichia coli. Mol Gen Genet. 1996;250(6):705–14. [PubMed: 8628231]
Sacerdot C, Fayat G, Dessen P. et al. Sequence of a 1.26-kb DNA fragment containing the structural gene for initiation factor IF3: presence of an AUU initiator codon. EMBO J. 1982;1:311–315. [PMC free article: PMC553041] [PubMed: 6325158]
Mayaux JF, Fayat G, Fromant M. et al. Structural and transcriptional evidence for related thrSand infC expression. Proc Natl Acad Sci U S A. 1983;80(20):6152–6. [PMC free article: PMC390161] [PubMed: 6353409]
Wertheimer SJ, Klotsky R-A, Schwartz I. Transcriptional patterns for the thrS-infC-rplT operon of Escherichia coli. Gene. 1988;63:309–320. [PubMed: 2838394]
Liveris D, Klotsky RA, Schwartz I. Growth rate regulation of translation initiation factor IF3 biosynthesis in Escherichia coli. J Bacteriol. 1991;173(12):3888–93. [PMC free article: PMC208021] [PubMed: 2050639]
Butler JS, Springer M, Grunberg-Manago M. AUU to AUG mutation in the initiator codon of the translation initiation factor IF3 abolishes translational autocontrol of its own gene (infC) in vivo. Proc Natl Acad Sci USA. 1987;84:4022–4025. [PMC free article: PMC305013] [PubMed: 2954162]
Sacerdot C, Chiaruttini C, Engst K. et al. The role of the AUU initiation codon in the negative feedback regulation of the gene for translation initiation factor IF3 in Escherichia coli. Mol Microbiol. 1996;21(2):331–46. [PubMed: 8858588]
Lesage P, Chiaruttini C, Graffe M. et al. Messenger RNA secondary structure and translational coupling in the Escherichia coli operon encoding translation initiation factor IF3 and the ribosomal proteins, L35 and L20. J Mol Biol. 1992;228(2):366–86. [PubMed: 1453449]
Arai K, Clark BF, Duffy L. et al. Primary structure of elongation factor Tu from Escherichia coli. Proc Natl Acad Sci U S A. 1980;77(3):1326–30. [PMC free article: PMC348487] [PubMed: 6990408]
Jones MD, Petersen TE, Nielsen KM. et al. The complete amino-acid sequence of elongation factor Tu from Escherichia coli. Eur J Biochem. 1980;108(2):507–26. [PubMed: 6997043]
Gouy M, Grantham R. Polypeptide elongation and tRNA cycling in Escherichia coli: a dynamic approach. FEBS Lett. 1980;115(2):151–5. [PubMed: 6995150]
Louie A, Ribeiro NS, Reid BR. et al. Relative affinities of all Escherichia coli aminoacyl-tRNAs for elongation factor Tu-GTP. J Biol Chem. 1984;259(8):5010–6. [PubMed: 6370998]
Zuurmond AM, Rundlof AK, Kraal B. Either of the chromosomal tuf genes of E. coli K-12 can be deleted without loss of cell viability. Mol Gen Genet. 1999;260(6):603–7. [PubMed: 9928940]
van der Meide PH, Vijgenboom E, Talens A. et al. The role of EF-Tu in the expression of tufA and tufB genes. Eur J Biochem. 1983;130(2):397–407. [PubMed: 6337847]
Zengel JM, Lindahl L. Mapping of two promoters for elongation factor Tu within the structural gene for elongation factor G. Biochim Biophys Acta. 1990;1050(13):317–22. [PubMed: 2207161]
Van Delft JH, Talens A, De Jong PJ. et al. Control of the tRNA-tufB operon in Escherichia coli. 2. Mechanisms of the feedback inhibition of tufB expression studied in vivo and in vitro. Eur J Biochem. 1988;175(2):363–74. [PubMed: 2456927]
Van Delft JH, Schmidt DS, Bosch L. The tRNA-tufB operon transcription termination and processing upstream from tufB. J Mol Biol. 1987;197(4):647–57. [PubMed: 2448475]
Travers A, Schneider R, Muskhelishvili G. DNA supercoiling and transcription in Escherichia coli: The FIS connection. Biochimie. 2001;83(2):213–7. [PubMed: 11278071]
Nilsson L, Vanet A, Vijgenboom E. et al. The role of FIS in trans activation of stable RNA operons of E. coli. EMBO J. 1990;9(3):727–34. [PMC free article: PMC551728] [PubMed: 1690124]
Bock A. Biosynthesis of selenoproteins-an overview. Biofactors. 2000;11(12):77–8. [PubMed: 10705967]
Leinfelder W, Stadtman TC, Bock A. Occurrence in vivo of selenocysteyl-tRNA(SERUCA) in Escherichia coli. Effect of sel mutations. J Biol Chem. 1989;264(17):9720–3. [PubMed: 2524495]
Sawers G, Heider J, Zehelein E. et al. Expression and operon structure of the sel genes of Escherichia coli and identification of a third selenium-containing formate dehydrogenase isoenzyme. J Bacteriol. 1991;173(16):4983–93. [PMC free article: PMC208187] [PubMed: 1650339]
Saito K, Nomura M. Post-transcriptional regulation of the str operon in Escherichia coli.Structural and mutational analysis of the target site for translational repressor S7. J Mol Biol. 1994;235(1):125–39. [PubMed: 8289236]
Spiridonova VA, Rozhdestvensky TS, Kopylov AM. A study of the thermophilic ribosomal protein S7 binding to the truncated S12-S7 intercistronic region provides more insight into the mechanism of regulation of the str operon of E. coli(1) FEBS Lett. 1999;460(2):353–6. [PubMed: 10544263]
Saito K, Mattheakis LC, Nomura M. Post-transcriptional regulation of the str operon in Escherichia coli.Ribosomal protein S7 inhibits coupled translation of S7 but not its independent translation. J Mol Biol. 1994;235(1):111–24. [PubMed: 7507167]
An G, Lee JS, Friesen JD. Evidence for an internal promoter preceding tufA in the str operon of Escherichia coli. J Bacteriol. 1982;149(2):548–53. [PMC free article: PMC216541] [PubMed: 6276361]
Miyajima A, Kaziro Y. Coordination of levels of elongation factors Tu, Ts, and G, and ribosomal protein SI in Escherichia coli. J Biochem (Tokyo) 1978;83(2):453–62. [PubMed: 344309]
Aoki H, Adams SL, Chung DG. et al. Cloning, sequencing and overexpression of the gene for prokaryotic factor EF-P involved in peptide bond synthesis. Nucleic Acids Res. 1991;19(22):6215–20. [PMC free article: PMC329127] [PubMed: 1956781]
Aoki H, Dekany K, Adams SL. et al. The gene encoding the elongation factor P protein is essential for viability and is required for protein synthesis. J Biol Chem. 1997;272(51):32254–9. [PubMed: 9405429]
An G, Glick BR, Friesen JD. et al. Identification and quantitation of elongation factor EF-P in Escherichia coli cell-free extracts. Can J Biochem. 1980;58(11):1312–4. [PubMed: 7011506]
Grentzmann G, Brechemier-Baey D, Heurgue V. et al. Localization and characterization of the gene encoding release factor RF3 in Escherichia coli. Proc Natl Acad Sci U S A. 1994;91(13):5848–52. [PMC free article: PMC44094] [PubMed: 8016077]
Mikuni O, Ito K, Moffat J. et al. Identification of the prfC gene, which encodes peptide-chain-release factor 3 of Escherichia coli. Proc Natl Acad Sci U S A. 1994;91(13):5798–802. [PMC free article: PMC44084] [PubMed: 8016068]
Ito K, Uno M, Nakamura Y. Single amino acid substitution in prokaryote polypeptide release factor 2 permits it to terminate translation at all three stop codons. Proc Natl Acad Sci USA. 1998;95(14):8165–9. [PMC free article: PMC20947] [PubMed: 9653158]
Holst-Hansen P, Kildsgaard J, MacDougall J. et al. Immunochemical determination of the cellular content of polypeptide chain release factor RF3 in Escherichia coli. Biochimie. 1997;79(12):725–9. [PubMed: 9523014]
Fraser CM, Gocayne JD, White O. et al. The minimal gene complement of Mycoplasma genitalium. Science. 1995;270(5235):397–403. [PubMed: 7569993]
Craigen WJ, Cook RG, Tate WP. et al. Bacterial peptide chain release factors : conserved primary structure and possible frameshift regulation of release factor 2. Proc Natl Acad Sci USA. 1985;82:3616–3620. [PMC free article: PMC397836] [PubMed: 3889910]
Kawakami K, Jönsson YH, Björk GR. et al. Chromosomal location and structure of the operon encoding peptide-chain-release factor 2 of E.coli.Proc Natl Acad Sci USA 1988855620–5624. [PMC free article: PMC281811] [PubMed: 2456575]
Weiss R, Gallant J. Mechanism of ribosome frameshifting during translation of the genetic code. Nature. 1983;302(5907):389–93. [PubMed: 6339944]
Craigen WJ, Caskey CT. Expression of peptide chain release factor 2 requires high-efficiency frameshift. Nature. 1986;322(6076):273–5. [PubMed: 3736654]
Kawakami K, Nakamura Y. Autogenous suppression of an opal mutation in the gene encoding peptide chain release factor 2. Proc Natl Acad Sci USA. 1990;87(21):8432–6. [PMC free article: PMC54970] [PubMed: 2236050]
Janosi L, Ricker R, Kaji A. Dual functions of ribosome recycling factor in protein biosynthesis: disassembling the termination complex and preventing translational errors. Biochimie. 1996;78(1112):959–69. [PubMed: 9150873]
Kaji A, Hirokawa G. Ribosome-recycling factor: an essential factor for protein synthesis In: Garrett RA, Douthwaite S, Liljas A, Matheson AT, Moore PB, Noller HF, eds.The Ribosome: Structure, function, antibiotics, and cellular interactions Washington, DC: ASM Press;2000527–539.
Shimizu I, Kaji A. Identification of the promoter region of the ribosome-releasing factor cistron (frr) J Bacteriol. 1991;173(16):5181–7. [PMC free article: PMC208211] [PubMed: 1860827]
International Union of Biochemistry and Molecular Biology(IUBMB)Prokaryotic and eukaryotic translation factors. Biochimie. 1996;78:1119–1122. [PubMed: 9150893]