4. Regulation of Gene Transcription
Transcription of the five major structural genes of the trp operon of E. coli and S. enterica is regulated by both repression and transcription attenuation. As previously mentioned, these regulatory mechanisms allow a response, respectively, to changes in the intracellular concentrations of tryptophan and tryptophan-charged tRNATrp, Trp-tRNATrp. The intracellular tryptophan concentration is determined by several events: tryptophan import from the cell’s environment, tryptophan produced internally by biosynthesis, and the rate of use of tryptophan during protein synthesis. The concentration of charged tRNATrp also depends on several factors: the intracellular concentrations of tryptophan, tRNATrp, and tryptophanyl-tRNA synthetase (see Aminoacyl tRNA Synthetases), and the overall rate of protein synthesis. It is apparent that the regulatory strategies that control transcription of the trp operon of E. coli and S. enterica were designed to adjust the rate of synthesis of this amino acid in response to all extracellular and intracellular events that alter the availability of tryptophan and Trp-tRNATrp for protein synthesis. Feedback inhibition of anthranilate synthase provides an additional important level of regulation by tryptophan by controlling entry of chorismate into the biosynthetic pathway.
Repression of the trp operon is sensitive to slight to moderate decreases in intracellular tryptophan concentration. In contrast, attenuation is designed to allow transcription of the structural genes of the operon only when there is nearly complete depletion of the pool of charged tRNATrp. This occurs under two conditions; when the supply of intracellular tryptophan is grossly insufficient and/or when the demand for Trp-tRNATrp greatly exceeds its availability. As we shall describe, the leader peptide coding region, trpL, of the trp operon of these two bacterial species has only two tryptophan codons, which are adjacent. This design allows relief of termination only when the intracellular concentration of Trp-tRNATrp is so low that rapid translation of both of these tryptophan codons cannot occur.
When E. coli or S. enterica is grown in the presence of excess tryptophan, repression reduces the rate of transcription initiation at the trp promoter to about 1/80th the rate observed in the absence of repression (2). The extent of repression reflects several factors: the intracellular concentrations of tryptophan and repressor, the rate of repressor synthesis, and the fraction of repressor molecules that is available for trp operator binding. Repressor synthesis is autoregulated; thus the repressor level drops to one third when cells are grown with excess tryptophan. From the finding that increased production of repressor specified by a multicopy plasmid decreases operon expression appreciably, it may be inferred that the trp repressor concentration is normally limiting for repression.
Feedback inhibition of anthranilate synthase can have a significant impact on the extent of repression. Thus, when E. coli is synthesizing the tryptophan it needs for growth and thus, repression is incomplete, feedback inhibition by the available tryptophan partially inhibits anthranilate synthase activity. This results in the production of slightly higher levels of the tryptophan biosynthetic enzymes, which is achieved by partial relief of repression. Feedback inhibition is a particularly effective regulatory mechanism because it reduces tryptophan synthesis instantaneously, whereas repression and attenuation have delayed effects. By employing complementary regulatory mechanisms, E. coli and S. enterica exploit the advantages that can be derived from different means of sensing and responding to tryptophan availability for protein synthesis.
Transcription termination at the trp attenuator is relieved completely only when cells are severely deficient in charged tRNATrp (3). Under these conditions, there is about a six- to eight fold increase in the level of trp operon structural gene mRNA above the level observed when there is an adequate level of Trp-tRNATrp. Starvation for other amino acids contained in the leader peptide does not have a comparable regulatory effect, except for arginine. Severe arginine starvation also eliminates transcription termination at the trp attenuator. The arginine starvation effect is due to the presence of a single arginine codon immediately following the two tryptophan codons in the leader peptide coding segment of the trp transcript. Upon arginine starvation, the ribosome that translates the leader peptide coding region presumably also stalls on the critical segment of the trp transcript, that is, the segment that contains the tryptophan codons.
4.1. The Operator Region of the trp Operon and the Mechanism of Repression
The promoter region of the trp operon of E. coli and S. enterica contains multiple repressor binding sites, or operators (Fig. 8). The tryptophan-activated trp repressor can bind at these sites and inhibit transcription initiation at the major trp promoter. Three-dimensional structures have been determined by X-ray crystallography and by NMR of the tryptophan-free trp aporepressor, the tryptophan-activated trp repressor, and the repressor-operator complex. Both the aporepressor and repressor exist as dimers composed of interlocking identical polypeptide chains. Each trp aporepressor dimer has two tryptophan binding sites. Bound tryptophan displaces helix E of the two helix-turn-helix domains of the dimer and positions these domains so that the dimer can effectively bind the symmetrical operator. Binding of tryptophan at the two sites of the aporepressor is cooperative. The structure of the repressor-operator complex shows that the repressor makes multiple contacts with target operator sequences and that some important specific contacts are water-mediated. This mechanism of recognition has been termed indirect readout (28). Although crystallographic analyses suggest that the preferred operator binding sequence is 5 ‘ACTAGT3′ (28), other studies indicate that the consensus operator sequence is 5′ GNACT3′ and that there are three operators in the trp promoter/operator region (Fig. 8) (29). The presence of multiple bound repressor dimers in the operator region could facilitate inhibition of transcription initiation of the operon by reducing the liklihood that the promoter/operator will be repressor-free. Repressor dimers, it has been observed, associate in solution, but the significance of this association is not yet known. The trp repressor also regulates initiation of the transcription of several other operons concerned with tryptophan metabolism.
Figure 8. The trp operon promoter/operator region. The -35, -10, and +1 sequences of the promoter are marked. The sequences in the two DNA strands that are assumed to be recognized specifically by each trp repressor dimer are indicated in outline form. Note that the diagram implies that three repressor dimers can be bound per operator region.
4.2. The Mechanism of Transcription Attenuation
The principal trp promoter, trpP1, is a relatively strong promoter; in tryptophan-deficient cells, RNA polymerase can initiate transcription at this promoter every 4-5 seconds (3). Inasmuch as a transcribing polymerase molecule could reach the attenuator region within a few seconds, it is essential that the decision be made quickly whether or not to terminate transcription at the attenuator. Because this decision is based on the transcript position of the ribosome that is attempting synthesis of the leader peptide, it is essential to synchronize translation of the leader peptide coding region with transcription of the leader region. Synchronization is achieved by features of the nucleotide sequence of the leader transcript; other features of the transcript determine the appropriate transcription termination response, which depends on the cellular level of charged tRNATrp (3).
4.3. Transcription Pausing, the Initial Important Event in Transcription Attenuation
After initiation of transcription at the trp promoter, RNA polymerase synthesizes an RNA segment that forms a hairpin structure that causes the transcribing polymerase to pause in the leader region of the operon (Fig. 9). The paused polymerase probably remains at the pause site for less than 1 second (3). The sequence of the pause RNA hairpin, particularly its upper base-paired half, it is believed, is the predominant pause signal (3). However, the DNA region immediately downstream from the pause RNA-encoding region also contributes to pause stability, as does the RNA polymerase accessory factor, NusA. In the paused complex, the 5′ end of the nascent transcript is exposed and is available for ribosome loading and initiation of leader peptide synthesis. In fact it is the act of synthesis of the initial portion of the leader peptide that disrupts the transcription pause complex and reactivates transcript elongation. Thus, transcription pausing is the event that is essential to regulation by this form of transcription attenuation.
Figure 9. Organization and features of the trp leader region and the trp leader transcript. The operon map at the top expands the trp leader region and shows the locations of trpL, the leader peptide coding region, the attenuator, and the five structural genes. Below the operon map are transcripts that have the three alternative hairpin structures, the pause, antiterminator, and terminator. The terminated and the read-through transcripts are marked. Part of the leader peptide coding region, trpL, overlaps the pause structure, as indicated. The start and stop codons of trpL and the positions of its adjacent Trp codons are shown. Nucleotide positions that define the various transcript segments are indicated.
4.4. Transcription Beyond the Pause Site
When the paused RNA polymerase resumes transcription, the position assumed by the ribosome engaged in synthesizing the leader peptide dictates whether the RNA antiterminator or terminator structure will form (Fig. 9). If a cell has adequate levels of Trp-tRNATrp (and other required charged tRNAs), the translating ribosome will reach the leader peptide stop codon while transcription of the leader region of the operon is proceeding. A ribosome at this position would block formation of the antiterminator structure, thereby allowing what would be the 3′ strand of the RNA antiterminator to pair with the immediately following RNA segment to form the terminator (Fig. 9). Formation of the terminator within the transcription complex would signal the transcribing polymerase to terminate transcription. If the Trp-tRNATrp concentration was very low, the translating ribosome would stall during attempted translation of the adjacent tryptophan codons. Such stalling would promote formation of the RNA antiterminator structure. The antiterminator would prevent the subsequent formation of the terminator because the 3 ‘ strand of the antiterminator must be free for it to serve as the 5′ strand of the terminator (Fig. 9). Thus the crucial location of the ribosome engaged in synthesizing the leader peptide dictates which RNA hairpin structure forms, and this, in turn, determines whether termination or read-through will occur.
4.5. Basal Level Control
When cells have adequate levels of Trp-tRNATrp to satisfy their needs for protein synthesis and synthesis of the trp leader peptide can be completed, only about 90% of trp transcripts are terminated in the leader region, there is about 10% read-through transcription. These read-through transcripts provide a "basal level" of all five trp polypeptides. Read-through is principally due to ribosome dissociation from the leader peptide stop codon when the synthesis of the leader peptide is completed. Ribosome release allows the leader segment of the transcript to form an antiterminator structure occasionally, which results in read-through. This read-through depends on the ability of cells to initiate and complete synthesis of the leader peptide. When initiation of synthesis of the trp leader peptide is defective, transcription termination at the trp attenuator increases from about 90% to 98%. This increased termination is due to increased formation of the terminator and decreased formation of the antiterminator. The mechanism responsible for this change is termed basal level control. Although basal level control is a natural consequence of the sequence and arrangement of the trp transcript, it is conceivable that this regulatory feature was introduced to reduce trp operon transcription further when a severe amino acid deficiency interfered with general protein synthesis.
4.6. Cessation of Leader Peptide Synthesis
Whether transcription stops at the attenuator or continues into the structural genes of the operon, subsequent translation of the leader peptide coding region would no longer be necessary. In fact, after the attenuation decision has been made, leader peptide synthesis is shut down; both the terminated and read-through transcripts contain leader nucleotide sequences that can pair with the ribosome binding site of the leader peptide coding region, and block further peptide synthesis (3).
5. Feedback Inhibition
Regulation of carbon flow in the tryptophan pathway is mediated by classical negative feedback inhibition of the various activities of the anthranilate synthase-APR transferase complex by L-tryptophan. Feedback inhibition results from the binding of one molecule of tryptophan to the feedback site of the TrpE subunit of the complex. Complete inhibition of anthranilate synthase activity can be attained with the intact TrpE-TrpG-TrpD complex, the TrpE-TrpG partial complex, and the uncomplexed TrpE subunit. This is also true for the individual ADIC synthase, ADIC lyase, and glutaminase activities of the complexes (Fig. 2). In contrast, inhibition of APR transferase activity is partial and never exceeds a maximum of about 60%. Tryptophan inhibition of APR transferase occurs only when TrpG-TrpD is complexed with TrpE, consistent with the location of the tryptophan binding site on the TrpE subunit. In all cases, inhibition is competitive with respect to chorismate and noncompetitive with respect to the other substrates, indicating allosteric communication between the inhibitor site and the active sites of the TrpE and TrpG-TrpD subunits.
Significant conformational effects accompany tryptophan binding to the complex. These are manifest in the cooperative kinetics of tryptophan binding, in the tryptophan-induced cooperative kinetics for chorismate utilization, and in the tryptophan-induced alterations in the chromatographic behavior of the enzyme. The global nature of the conformational effects of tryptophan binding are perhaps best demonstrated in the properties of a doubly mutant hybrid complex, assembled in vitro, that contains one catalytically active, feedback-insensitive TrpE subunit, and one catalytically inactive, feedback-sensitive TrpE subunit (30). The feedback sensitivity of the anthranilate synthase activity of the hybrid complex demonstrates that the binding of a single tryptophan molecule to one TrpE subunit elicits conformational and kinetic effects that are propagated to the active site of the unliganded companion TrpE subunit.
Comparison of the three-dimensional structures of the active and the inhibited enzyme forms suggests a likely molecular mechanism of feedback inhibition in the AS-APR complex (13). In the tryptophan-bound enzyme, a structural motif of the TrpE subunit, formed by residues 327363 and 387-403, is shifted about 7A away from the chorismate-binding pocket relative to its position in the chorismate-bound enzyme. Significantly, this movement displaces catalytic residues Thr329 and His398 as well as two metal-binding glutamate residues from the vicinity of the active site and results in the loss of catalytic activity. This rearrangement is the result of the ordering of a disordered loop made up of residues 42-49 in response to tryptophan binding (13, 14). In contrast, the binding of chorismate at the active site moves this same motif in the opposite direction toward the active site crevice and leads to activation of the enzyme. These reciprocal effects explain the competitive behavior of chorismate and tryptophan because binding of one of these molecules to TrpE necessarily precludes binding of the other. Also, the fact that the structural loops involved are situated at the heterotetramer interface explains how a single tryptophan molecule can inhibit both TrpE subunits; in view of the molecular twofold axis of the complex, a movement in one heterodimer must be accompanied by a corresponding movement in the other.
Binding of tryptophan by the TrpE subunit also induces relevant conformational changes in the TrpG subunit of the complex. Two b-strands of TrpG, made up of residues 103-109 and 130140, become disordered in the tryptophan-bound enzyme (13) in response to the tryptophan-induced movement of an interacting TrpE a-helix at the heterodimer interface. These changes disrupt the glutamine binding site of the TrpG subunit. Thus, a unitary mechanism for both tryptophan inhibition and chorismate activation of the glutaminase activity of the complexed TrpG subunit is manifest. The mechanism of the partial tryptophan inhibition of the APR activity of the complexed TrpG-APR subunit remains unknown.
Clearly, the unique allosteric properties of the anthranilate synthase-APR transferase complex, coupled with its critical location at the beginning of the pathway, provide a rapid and efficient mechanism of metabolic regulation that modulates and complements the elegant regulatory circuitry at work at the transcriptional and translational levels.
6. Other Features That Influence trp Operon Expression
6.1. Internal Promoter trpP2
A secondary, internal promoter, trpP2, exists near the 3′ end of the trpD gene of E. coli and S. enterica (Fig. 1) (31). trpP2 is a low-efficiency, constitutive promoter, that initiates transcription about 150 nucleotides upstream of the trpD stop codon (32). It is responsible for about 60% of the fully repressed expression of the trpC, trpD, and trpA genes. The trpP2 sequence lacks a conserved nucleotide in both the -35 and -10 regions of the canonical promoter sequence, accounting for its poor efficiency. The in vivo role of trpP2, it is thought, is to aid the cell in responding to an environment that has extreme variations in the supply of exogenous tryptophan. The elevated basal level of the three-terminal pathway enzymes resulting from trpP2 function might provide an advantage to the cell in adapting rapidly to conditions of tryptophan starvation. The conservation of this secondary transcriptional element in numerous enteric species indicates that it is physiologically significant, not merely a vestige of an earlier regulatory structure.
6.2. Translational Coupling
The multigene trp operon of E. coli and S. enterica is designed to permit regulated synthesis of a polycistronic mRNA that encodes the five polypeptides needed for tryptophan formation. The required concentration of each polypeptide would depend on its inherent catalytic activity, its affinity for its substrates, and its stability. Other potential variables that could influence the rate of tryptophan synthesis are the efficiency of translation of each mRNA coding segment and the mRNA segments’ functional half-lives. As mentioned, two pairs of trp polypeptides, TrpE and TrpD, and TrpB and TrpA, form enzyme complexes in which separate catalytic events are coordinated. To optimize catalysis, equimolar levels of each member of these complexes are achieved by translational coupling (2). Thus, the trpD and trpA ribosome binding sites are designed so they are inefficiently used unless translation of trpE and trpB, respectively, proceeds to the end of its coding region. Coupling ensures equimolar polypeptide synthesis.
6.3. Preferential Synthesis upon Tryptophan Starvation
TrpE and TrpA of E. coli and S. enterica lack tryptophan. A consequence of this deficiency is that upon severe tryptophan starvation—when synthesis of most tryptophan-containing proteins is reduced or prevented—synthesis of TrpE and TrpA continues, and the relative levels of these proteins increase (2). The disproportionate elevated level of TrpE would result in preferential sequestration of chorismate into the tryptophan pathway. This would occur despite the lack of a comparable increase in the formation of TrpG-D, which normally contributes the glutamine amidotransferase activity used for anthranilate formation, because anthranilate formation by anthranilate synthase does not absolutely depend on glutamine amidotransferase activity. The TrpE polypeptide acting alone can convert chorismate to anthranilate, using free ammonia as the amino source. Inasmuch as anthranilate synthesis proceeds by an irreversible reaction, the conversion of chorismate to anthranilate in vivo would lock chorismate into the tryptophan biosynthetic pathway.
6.4. trp mRNA Degradation
The half-life of trpmRNA is comparable to that of many E. coli mRNAs (see Messenger RNA). There is a modest gradient of decay along its length, trpEmRNA is about half as stable as trpAmRNA (2). As in other mRNAs, hairpin structures at the 5′ and 3′ ends of trpmRNA probably contribute to its stability.
6.5. trp Protein Turnover
During log phase growth in a minimal or rich medium, the five trp biosynthetic proteins of E. coli are relatively stable. Only upon prolonged starvation is there significant turnover of one of the trp proteins, the product of trpC. Apparently, E. coli adjusts the levels of the trp biosynthetic enzymes predominantly by regulating synthesis and by growth-dependent dilution.
7. Evolutionary Considerations
Although the same seven polypeptide domains appear to be responsible for tryptophan biosynthesis in all organisms that can synthesize this amino acid (33), in many organisms these domains are fused in different arrangements than in E. coli and S. enterica (15). In addition, chromosomal order and groupings of trp genes, as well as the mechanisms used to regulate trp gene/operon expression, vary in different species (15, 34). Providing explanations for these differences will require information that is difficult to obtain, such as understanding the functional demands that were imposed on each species during its evolution. There are species in which as many as five of the trp enzymatic functions reside in a single multifunctional polypeptide (35). In some species, novel metabolic reactions occur that use tryptophan pathway intermediates for other purposes; understanding these accessory reactions may provide explanations for the different gene arrangements and regulatory mechanisms that are observed (34).
The evolutionary origin of each of the trp polypeptide domains is also of considerable interest. The 3-D structures of six of the seven trp enzymatic domains are known now. As mentioned, three of these domains, TrpA, TrpF, and TrpC, form barrel structures that have similar characteristics; thus each could have evolved from one of the others or from other preexisting members of this family. In fact, the IGP synthase domain of E. coli has been mutationally altered to yield a polypeptide that can catalyze the PRA isomerase reaction (36). In addition, the HisA barrel protein of Thermotoga maritima, an enzyme that catalyzes an Amadori rearrangement of a phosphoribosylamine similar to the PRA isomerase reaction (Fig. 4), has been mutationally converted into an enzyme that catalyzes the PRA isomerase reaction (37). These findings illustrate how simple it would be for a protein to evolve with a new catalytic activity, starting with a copy of a gene encoding an enzyme that could catalyze a similar reaction. Alternatively, both enzymes may have evolved from an ancestral enzyme that had broad substrate specificity (37). The TrpB structure is homologous to proteins in a family of pyridoxal phosphate-dependent enzymes, many of which, like TrpB, can catalyze an amino acid dehydratase reaction. The TrpG domain is highly homologous to members of the family of Triad glutamine amidotransferases. A close homolog of TrpE participates in synthesizing an analog of anthranilate, p-aminobenzoate (15). Identifying the true origins of the trp genes and tryptophan proteins probably will be difficult because there are numerous ancestral possibilities, and the true evolutionary intermediates may no longer exist.
8. Dedication
We dedicate this article to the fond memory of Irving P. Crawford, who also loved the genes and enzymes of tryptophan biosynthesis.
