Analysis of Tryptophanase Operon Expression in Vitro

Expression of the tryptophanase (tna) operon in Escherichia coli is regulated by catabolite repression and tryptophan-induced transcription antitermination. The key feature of this antitermination mechanism has been shown to be the retention of uncleaved TnaC-peptidyl-tRNA in the translating ribosome. This ribosome remains stalled at the tna stop codon and blocks the access of Rho factor to the tnatranscript, thereby preventing transcription termination. In normal S-30 preparations, synthesis of a TnaC peptide containing arginine instead of tryptophan at position 12 (Arg12-TnaC) was shown to be insensitive to added tryptophan, i.e.Arg12-TnaC-peptidyl-tRNA was cleaved, and there was normal Rho-dependent transcription termination. When the S-30 extract used was depleted of release factor 2, Arg12-TnaC-tRNAPro was accumulated in the absence or presence of added tryptophan. Under these conditions the accumulation of Arg12-TnaC-tRNAPro prevented Rho-dependent transcription termination, mimicking normal induction. Using a minimal in vitro transcription system consisting of a tna template, RNA polymerase, and Rho, it was shown that RNA sequences immediately adjacent to thetnaC stop codon, the presumed boxA andrut sites, contributed most significantly to Rho-dependent termination. The tna boxA-like sequence appeared to serve as a segment of the Rho “entry” site, despite its likeness to the boxA element.

contains a single tryptophan residue. Synthesis of TnaC is essential for induction (3,8,9). Replacing the tnaC start codon by a stop codon (9,10) or replacing Trp codon 12 by a codon for some other amino acid (10) prevents induction.
Evidence supporting the essential role of Rho factor in mediating transcription termination in the tna operon leader region was provided by analyses of Rho mutants (8), examination of Rho-inhibiting drugs (11), and deletion of a leader region sequence-rich in C residues (10). Mutations in rho that reduce Rho factor activity as well as addition of bicyclomycin, an inhibitor of Rho action, increase basal expression of the tna operon significantly (7,8,11). Similarly, deletion of a 23-nucleotide C-rich tna RNA sequence immediately following the tnaC stop codon (Fig. 1A) reduces transcription termination (10). Comparable sequences, called Rho utilization (rut) sites, have been identified in Rho transcription termination studies with other systems (12)(13)(14). Little is known about the features of the rut site required for Rho binding other than it generally has relatively little secondary structure (15) and is rich in cytosine residues (12,16). Interaction between Rho and an RNA rut sequence is essential for Rho-dependent termination (17,18).
An additional sequence element shown to be essential for maximum termination is located near the distal end of tnaC (8,19). This sequence is homologous to the boxA elements found in studies with the phage N/nut antitermination system (20,21) and with bacterial rrn operons (22,23) (Fig. 1B). boxA of these operons is required for prevention of Rho-dependent termination, not for facilitation of Rho-dependent termination. The N protein of phage relieves both Rho-dependent termination and intrinsic transcription termination (24). Several host factors called Nus factors are involved in antitermination at sites of Rho-dependent termination (21,25). For example, rrn boxA is the loading site for the E. coli S10 (or NusE) and NusB proteins (26). boxA of rrn operons appears to be sufficient to prevent Rho-dependent termination in vivo (27), while in the phage N/nut antitermination system, boxA and boxB sequences are both required for specific N binding and subsequent formation of the RNA polymerase-N-NusA-NusB-NusG-S10-nut antitermination complex (21,25,28). In contrast, the boxA sequence in the tna operon does not behave like a typical boxA sequence, mutations in this presumed boxA sequence decrease rather than increase transcription termination in the tna operon leader region (8).
Attempts to detect a gene encoding a presumed trans-acting factor that is responsible for tryptophan-induced expression of the tna operon of E. coli have been unsuccessful (29). The findings obtained in in vivo and in vitro studies of tna operon regulation suggest that all the genetic information necessary for tryptophan-induced expression of the tna operon is located in its leader region. The in vivo regulatory features of tna operon expression have been confirmed in vitro using an S-30 system (7). Most importantly, it was shown that translation of tnaC in the presence of added tryptophan leads to the accumulation of TnaC-peptidyl-tRNA Pro in the translating ribosome (7). The uncleaved TnaC-peptidyl-tRNA Pro stalls the translating ribosome at the tnaC stop codon (30), and this stalling is believed to block Rho factor binding and action (30).
In this paper we show that depleting RF-2 1 from an S-30 extract results in the accumulation of uncleaved Arg 12 -TnaC-tRNA Pro in the translating ribosome in the absence of added tryptophan. This accumulation of uncleaved Arg 12 -TnaC-tRNA Pro prevents Rho factor action establishing that ribosome stalling at the tnaC stop codon is sufficient to eliminate Rhomediated transcription termination. We provide additional in vitro evidence demonstrating that the RNA sequence masked by the ribosome stalled at the tnaC stop codon serves as the Rho "entry" site. Our findings suggest that the tna boxA sequence, despite its sequence similarity to boxA elements of bacteriophage and rrn operons, contributes to Rho-dependent termination by serving as a segment of the Rho entry site. Cell-free Transcription-Translation-S-30 extracts were prepared as described by Zubay (31) using cells of E. coli strain A19 RNaseI Ϫ containing trpR ⌬lacZ ⌬trpEA2 tnaAbgl::Tn10. Reaction conditions and procedures for preparation of the self-ligated DNA templates used to direct in vitro S-30 reactions have been detailed (7). Single round analyses of transcription coupled to translation, using cell-free extracts, have been described (7).

Enzymes and Reagents-Restriction
Single Round in Vitro Transcription Analyses Using a Purified System-DNA templates bearing mutations were prepared by megaprimer PCR methods (32) and were gel-purified. Buffering conditions used in the coupled transcription/translation system (7) were adopted for use with the purified transcription system. The standard transcription reaction mixture (in a total volume of 25 l) contained 2 units of RNAP (Epicentre), 35 mM Tris acetate, pH 8.0, 10 mM magnesium acetate, 200 mM potassium glutamate, 30 mM ammonium acetate, 2 mM dithiothreitol, 2 mM cAMP, 1 mM ATP, and 35 g/ml CRP protein, 10 g/ml bovine serum albumin, and 10 nM linear DNA template. CTP and GTP were present at 150 M. Single round transcription reactions were initiated by adding 20 M UTP, 20 Ci of [␣-33 P]UTP, and 50 g/ml rifampicin after 10 min of preincubation at 37°C (33). Reactions were terminated by phenol extraction after another 10-min incubation and analyzed on a 6% polyacrylamide, 7 M urea gel. When specified, a high level of Rho factor (1 M hexamer) was included in the preincubation reaction mixture. Gels were dried and exposed to a phosphorimaging screen or x-ray film. Levels of 33 P-labeled readthrough (RT) bands were quantified by using a phosphorimaging device (molecular Imager System G3 363, Bio-Rad).
Immunoprecipitation of RF-2-To remove RF-2 from the S-30 extract directly, a Protein A prebinding step was performed to enrich for antibodies. 150 l of Protein A-Sepharose 4B slurry (ϳ100 l of packed beads) was prewashed with TMNK S-30 buffer (35 mM Tris acetate, pH 7.8, 10 mM magnesium acetate, 30 mM ammonium acetate, 60 mM potassium glutamate) three times. After each wash, the beads were recovered by centrifugation at 2500 ϫ g for 1 min. 100 l of anti-RF-2 antiserum was added, and the mixture was placed on a rotating wheel for 1 h at room temperature. The beads were recovered, washed with TMNK buffer containing 5 g/ml leupeptin three times, and distributed evenly into three siliconized tubes. 100 l of S-30 was added to one tube and mixed with the beads for 2 h at 4°C. After centrifugation at 10,000 ϫ g for 2 min, the supernatant was recovered. This immunoprecipitation step was repeated twice with the same S-30 extract. The S-30 extract depleted of RF-2 was stored at Ϫ80°C. An S-30 extract treated identically with a pre-bleed serum was used as a control.

RESULTS
Preparation of an RF-2-depleted S-30 -In previous studies using an in vitro S-30 system it was shown that tryptophan induction of tna operon expression is a consequence of tryptophan inhibition of cleavage of newly synthesized TnaC-pepti-FIG. 1. A, the tna leader transcript and leader region deletions. The locations of the deletions used to examine the relative contributions of boxA and rut site sequence elements in Rho-dependent termination are indicated. The 3Ј nucleotides (nt) at which tna transcripts are believed to be terminated by Rho action are in bold and are numbered (6). The tnaC nucleotide sequence and the sequence of its encoded leader peptide also are shown. In specific experiments, the crucial Trp codon, UGG, at position 12 of the tnaC coding sequence was replaced by an Arg codon CGG. SD, Shine-Dalgarno sequence. B, comparison of boxA sequences from bacteriophage tR1 region, rrn operons, and the tna operon. Nucleotides that are conserved between at least two of the sequences are in bold. C, predicted RNA secondary structures present at the 5Ј end of the tna leader transcript. The RNA secondary structures were predicted using the MFOLD program of Zuker (35). Calculated free energies of formation are indicated below each structure. tnaC start codon, stop codon, and Trp codon 12 are in boldface. Arrows mark the first two Rho-dependent termination sites. dyl-tRNA Pro (7). This peptidyl-tRNA remained in a complex with the stalled translating ribosome and its associated tna transcript. It was demonstrated that both the presence of tryptophan and synthesis of TnaC-tRNA Pro were necessary for stalling of the translating ribosome at the tnaC stop codon. The stalled ribosome appears to prevent Rho factor from binding to the tna transcript, thereby preventing transcription termination. In the absence of added tryptophan the synthesized TnaC-tRNA Pro was cleaved, the translating ribosome dissociated from the tna transcript, Rho bound to the tna transcript, and transcription was terminated (30). On the basis of these findings we presumed that if the S-30 extract were depleted of RF-2 the translating ribosome would stall at the tnaC stop codon in the absence of tryptophan and that Rho action would be blocked. To examine this possibility, an anti-RF-2 antiserum was used to deplete the S-30 of RF-2, as detailed under "Materials and Methods." Although we did not experimentally determine that all of the RF-2 in the extract had been removed, the in vitro translation results presented below demonstrate that there is insufficient RF-2 in the depleted S-30 to carry out normal translation termination.

Cleavage of TnaC-tRNA Pro in an RF-2-depleted S-30 Preparation Is Dependent on Added Functional RF-2; Tryptophan
Addition Inhibits the Action of Added RF-2-We showed previously, and confirmed in this study, that TnaC-tRNA Pro accumulates in the presence of inducing levels of tryptophan in the control S-30 system. In this system, in the absence of added tryptophan, only free TnaC was detected ( Fig. 2A, compare lanes 1 and 2) (7). When an RF-2-depleted S-30 was used in this transcription-translation coupled system, TnaC-tRNA Pro was also detected in the absence of added tryptophan (lane 3). Addition of tryptophan increased the intensity of the TnaC-tRNA Pro band slightly (compare lanes 3 and 4). When purified RF-2 was added to the RF-2-depleted S-30, in the absence of added tryptophan, no TnaC-tRNA Pro was detected, only free TnaC was observed (lane 5). As expected, addition of tryptophan restored the accumulation of TnaC-tRNA Pro in the presence of purified RF-2 (lane 6). These results show that RF-2 must be added to the RF-2-depleted S-30 to achieve TnaC-tRNA Pro cleavage.
The anti-RF-2 antiserum employed cross-reacts very weakly with RF-1 2 ; thus, depletion of RF-1 as well as RF-2 could be contributing to the results obtained above. To rule out this possibility, purified RF-1 was added to the RF-2-depleted S-30 to determine whether it could mediate TnaC-tRNA Pro cleavage (Fig. 2B). TnaC-tRNA Pro was accumulated in the RF-2-depleted S-30 preparation in the absence of added tryptophan (lane 1). The accumulation of TnaC-tRNA Pro was unchanged in the presence of added purified RF-1 (lane 2). Consistent with the results shown in Fig. 2A, addition of purified RF-2 led to TnaC-tRNA Pro cleavage, and this cleavage was inhibited by added tryptophan (lanes 3 and 4). These findings indicate that strict stop codon recognition is maintained in the S-30 system. Recall that it was previously observed that when centrifugation-purified TnaC-tRNA Pro -ribosome complexes were subjected to RF-1 treatment, RF-1 addition did lead to cleavage of the peptidyl-tRNA despite the fact that the tnaC stop codon is UGA (30).
It was shown previously that replacing Trp 12 of TnaC by Arg eliminates tryptophan induction in vivo (10). As expected, when an Arg 12 template was used in control S-30 reactions, no accumulation of Arg 12 -TnaC-tRNA Pro was detected either in the absence or presence of added tryptophan; only free Arg 12 -TnaC peptide was observed (Fig. 2C, lanes 1 and 2). By con-trast, Arg 12 -TnaC-tRNA Pro was accumulated in the RF-2-depleted S-30 preparation in the absence of tryptophan (lane 3). Arg 12 -TnaC-tRNA Pro was mostly cleaved and barely detected when purified RF-2 was added to the reaction mixture (lane 4). As expected, there was no effect of added tryptophan on Arg 12 -TnaC-tRNA Pro cleavage (lane 5).
Mimicry of Tryptophan Induction: Constitutive Expression of the tna Operon in the RF-2-depleted S-30 -When the S-30 extract is depleted of RF-2, the ribosome translating tnaC presumably stalls at the tnaC UGA stop codon. As occurs upon induction, the stalled ribosome should prevent Rho factor from binding to the transcript, thereby allowing transcription of the tna operon to continue. To test this possibility, single round transcription analyses (coupled with translation) were monitored in S-30 extracts directed by either the wild type Trp 12 template (Fig. 3) or the mutant Arg 12 template (Fig. 4). Fig. 3A shows the 33 P-labeled mRNA bands observed in a control S-30 reaction with or without added tryptophan. In the absence of added tryptophan, due to the action of Rho factor followed by RNA degradation (7), multiple pause transcript species are visible in early time samples, and these bands subsequently 2 K. Ito, personal communication. disappear. In the presence of added tryptophan, intermediate length bands appear, continue to elongate, and consequently the RT band becomes more prominent. When a similar experiment was performed using an RF-2-depleted S-30, transcription termination did not occur, even in the absence of added tryptophan (Fig. 3B, ϪTrp ϪRF2 lanes). Intermediate length bands appear, they eventually disappear, and ultimately reach the RT position. The short, doublet transcript species (D) observed are believed to result from degradation of ribosomebearing RT transcripts (7) (Fig. 3B). Most importantly, addition of purified RF-2 to the RF-2-depleted S-30 system restores Rho-dependent transcription termination; labeled pause transcripts are visible in early time point samples, then they disappear. The doublet did not accumulate in the presence of purified RF-2 (Fig. 3B, ϪTrp ϩRF2 lanes). When a translation inhibitor, chloramphenicol (20 g/ml), was added to the RF-2depleted system, a RNA expression pattern similar to that obtained in Fig. 3B (ϪTrp ϩRF2) was observed (Fig. 3B, ϪTrp ϩCm lanes); pause RNA appeared and was degraded, and there was little readthrough RNA. Also, no doublet RNA appeared.
Identical experiments were performed using the Arg 12 tnaC template (Fig. 4). In control S-30 reactions with this template, Rho-dependent termination was observed both in the absence and presence of added tryptophan (Fig. 4A); multiple paused transcripts were visible in early time samples, and these bands subsequently disappeared. These results confirm that the Trp residue at TnaC position 12 is crucial for tryptophan-induced inhibition of Rho action (10, 30). When we examined the Arg 12 template in an RF-2-depleted S-30 extract, in the absence of added tryptophan, Rho-dependent termination was prevented (Fig. 4B, ϪRF2 lanes). Intermediate length bands appeared, they continued to elongate, and consequently the RT band became more prominent. Also evident was the RNA doublet band. The addition of purified RF-2 restored Rho-dependent termination; paused transcripts appeared in early time samples and subsequently were mostly degraded. A small amount of RT species was observed (Fig. 4B, ϩRF2 lanes). These findings establish that a translating ribosome stalled at the tnaC stop codon prevents Rho factor action.

Contributions of the boxA Sequence and the rut Site Adjacent to the tnaC Stop Codon to Rho-dependent Termination in
Vitro-A sequence resembling a boxA sequence, and a presumed rut site (Rho utilization), are located in the vicinity of the tnaC stop codon (Fig. 1). Their identification was based on sequence similarities and in vivo mutational studies (8,10). Deletion of the rut site results in semiconstitutive expression of the tna operon in vivo (10), as does deletion or genetic alteration of the boxA sequence. These findings suggest that both sequences play a role in mediating Rho-dependent transcription termination in the tna operon leader region (8,34).
It has been well established that boxA sequences of phage (20) and of rrn operons (22) are required for transcription antitermination, not termination. Thus the boxA-like sequence at the distal end of tnaC appears to behave differently. When templates with or without the boxA and rut sequences were tested in vitro in single round transcription experiments with purified RNA polymerase and Rho, both sites were observed to contribute significantly to Rho-dependent transcription termination (Fig. 5). Under the conditions used, Rho addition to the wild type template resulted in almost complete elimination of the RT species (Fig. 5A, WT ϩRho lane). The RT (%) levels with ⌬boxA-1 (⌬6 bp), ⌬boxA-2 (⌬13 bp), and ⌬rut (⌬23 bp) templates (Fig. 1A) were 13, 35, and 34%, respectively (Fig. 5, A and B). Most importantly, when the template lacked both the boxA and rut sites, ⌬boxA-rut (⌬36 bp), the RT level was 96%. (Fig. 5, A  and B). With a template bearing a 36-bp deletion of a sequence near the 5Ј end of the tna leader region (⌬ϩ7 to ϩ42, Fig. 1A) there was little reduction in the efficiency of Rho-dependent termination, compared with the WT template (Fig. 5B). This finding suggests that the differences found in the levels of RT transcripts when templates were used that bear specific tna leader deletions cannot simply be attributed to the changes in length of the corresponding transcripts.
Because the interaction of Rho with RNA is known to be very sensitive to RNA secondary structure (15), we used the MFOLD program of Zuker (35) to predict the secondary structures of the boxA and rut deletion transcripts. Computer modeling revealed that no new secondary structural elements were formed in the remaining sequences for any of the deletion constructs. In addition, the deletions introduced did not further stabilize any existing tnaC secondary structure; rather, most of the deletions (⌬boxA-1, ⌬boxA-2, and ⌬boxA-rut) destabilized a wild type structure (Fig. 1C). Taken together, our results suggest that the changes in Rho-dependent termination activity observed with our deletion templates must be due primarily to the loss of the specific deleted RNA segments and not to changes in length of the deletion transcript or to changes in an existing secondary structure. When templates bearing single or double C to A mutations in the boxA sequence were tested (AGCCCU and AGACCU), compared with a Ͻ1% RT level with the WT template (CGCCCU) in the presence of Rho protein, 3 and 7% RT levels were observed with the substitution templates (Fig. 5C). In addition, the principal Rho-dependent transcription termination sites with the WT template under the conditions used were at ϩ165 and ϩ176 (Fig. 1), whereas ϩ176 and ϩ234 were the main termination sites with templates AGCCCU and AGACCU, respectively. These point mutations were previously observed to have a more profound effect in vivo, for example the AGACCU construct led to a 6-fold increase in basal level expression compared with that of the wild-type control (34). The role of the presumed boxA site is difficult to evaluate fully, particularly because we do not yet understand the role, if any, of the RNA hairpin structure that includes the boxA sequence (Fig. 1C) (see "Discussion").

DISCUSSION
In the present study, we provide in vitro evidence establishing that stalling of the ribosome translating tnaC at the tnaC stop codon is sufficient to prevent Rho-dependent transcription termination in the leader region of the tna operon. We show that expression of the tna operon appears "induced" in an RF-2-depleted S-30 extract, in the absence of tryptophan, the inducer. We also show that when a normally non-inducible mutant template with a Trp 3 Arg replacement at position 12 of TnaC is examined in this RF-2-depleted S-30 extract, Arg 12 -TnaC-tRNA Pro accumulates and Rho-dependent termination is prevented, presumably because the translating ribosome has stalled at tnaC stop codon. These findings are consistent with our model (Fig. 6) that the sites required for Rho binding are adjacent to the tnaC stop codon. In the absence of inducing levels of tryptophan, the ribosome translating the tnaC coding region completed synthesis of TnaC and then dissociated from the tnaC stop codon, rendering the transcript accessible to Rho. In the presence of the inducer features of the leader peptide lead to the formation of a tryptophan binding site in the ribosome, and, when tryptophan is bound, RF-2 cannot activate peptidyltransferase cleavage of the TnaC-peptidyl-tRNA Pro (30).
Our data show that depletion of RF-2 from the S-30 extract alters the requirements for prevention of Rho factor action. In a normal S-30 extract, inducing levels of tryptophan and the nascent wild-type TnaC-peptidyl-tRNA are required to prevent Rho-dependent termination (Fig. 3). In the RF-2 depleted S-30 extract, neither inducing levels of tryptophan nor synthesis of the nascent WT TnaC is required. Thus, when the Arg 12 -TnaC template was tested in the RF-2-depleted S-30 extract, accumulation of 35 S-labeled Arg 12 -TnaC-tRNA Pro was observed (Fig. 2), and Arg 12 -TnaC-tRNA Pro was shown to be associated with its translating ribosome. 3 This accumulation was correlated with prevention of Rho-dependent termination (Fig. 4). When translation of tnaC in the RF-2-depleted S-30 extract was inhibited by chloramphenicol addition, there was no inhibition of Rho-dependent termination (Fig. 3). Thus, translation of the Arg 12 -TnaC mutant template is required in the RF-2depleted S-30 extract for prevention of Rho action (Fig. 4).
We also show using the RF-2-depleted S-30 extract that elevated levels of tryptophan induce tna operon expression by inhibiting the action of RF-2, not RF-1. In previous studies with partially purified ribosome complexes we observed that RF-1 and RF-2 were equally effective in mediating TnaC-tRNA Pro cleavage at the tnaC UGA stop codon and that tryptophan inhibited the action of both release factors (30). Either some factor is removed during ribosome isolation that discriminates between the different release factors or stop codon recognition is no longer necessary for release factor participation in peptidyltransferase activation in our isolated ribosome complexes.
A striking characteristic of Rho-dependent transcription ter-3 F. Gong and C. Yanofsky, unpublished data.   6. Model for tna operon regulation in E. coli. 1, in the absence of inducing levels of tryptophan the translating ribosome reaches the tnaC stop codon, RF-2-mediated TnaC-tRNA Pro cleavage occurs, and the translating ribosome subsequently releases from the template. Rho factor then has access to its entry site on the transcript in the vicinity of the tnaC UGA stop codon, it contacts a paused polymerase, and causes transcription termination. 2, in the presence of inducing levels of tryptophan, the combined action of ribosome-associated uncleaved TnaC-tRNA Pro and ribosome-bound tryptophan inhibits RF-2-mediated TnaC-tRNA Pro cleavage. As a result, the translating ribosome stalls at the tnaC stop codon. The stalled ribosome blocks the access of Rho to its entry site on the transcript. Transcription termination is thereby prevented, and RNA polymerase continues transcription into the tnaA-tnaB coding region. mination is the requirement for a specific entry site for Rho on the synthesized transcript. Binding to this site is a prerequisite for the activation of the ATPase activity of Rho. This is required for its subsequent helicase and transcription termination functions (14,36). Our data and previous findings suggest that the RNA sequence in the vicinity of tnaC stop codon constitutes the critical entry site for Rho. The previously identified boxA sequence, as well as the rut site, appear to be required for maximal Rho-dependent termination in our purified in vitro system, in the absence of Nus factors. Deletion of the 23-nucleotide rut site significantly reduced the action of Rho in vitro (⌬rut, 35% RT, Fig. 5). Most importantly, deletion of both the boxA and rut sites (⌬boxA-rut, 96% RT, Fig. 5) virtually eliminated transcription termination by Rho. The tna boxA sequence, like the rut site, is rich in C residues (Fig. 1).
The boxA sequence recognized in phage N/nut antitermination (37) and in rrn operon regulation (22) is required for prevention of Rho-dependent termination. It has been shown that nucleotides 2-6 of both rrn and boxAs are specifically involved in the assembly of the antitermination complex (26,28). For example, the boxA 4U 3 G mutation impairs the ability of rrn boxA to support transcription antitermination in vivo (26). The same mutation in boxA prevents the formation of the complete complex (RNA polymerase-N-NusA-NusB-NusG-S10-nut) (28). Note that tna boxA has a C instead of U at position 4 (Fig. 1B). Thus the resemblance of the tna boxA sequence to the classic boxA sequence may simply be coincidental. Indeed, previous in vivo studies with nusA1, nusA100, and nusE100 mutant strains (38,39) failed to detect an effect of these alterations on termination in the tna operon (34). The key role played by NusA in the phage N/nut (boxAϩboxB) antitermination system has been documented recently (24).
In view of the importance of ribosome stalling in tna operon induction, it is obvious that coupling of translation with transcription must play a significant role in tna operon regulation as it does in trp operon regulation (40). There are two presumed hairpin structures near the 5Ј end of tna mRNA (Fig. 1C), one containing the tnaC Shine-Dalgarno sequence and start codon and one formed from the very end of the tnaC coding region. It was shown previously that this latter secondary structure serves as a transcription pause structure in vitro (6). We assume there is some mechanism of disrupting the 5Ј structure, allowing ribosome binding and translation initiation, and that transcription pausing at the more 3Ј structure may provide the delay that achieves synchronization of transcription and translation. Thus progression of the translating ribosome may release the paused complex, allowing transcription to proceed into the downstream region. The transcribing polymerase would then pause at one or more of the multiple pause sites located between tnaC and tnaA. These pauses would provide the opportunity for Rho factor to bind to the boxA-rut site region of the tna transcript, contact a paused polymerase, and activate transcription termination. Access of Rho factor to the boxA-rut site region would depend on the rate of dissociation of the ribosome translating tnaC at the tnaC stop codon. Obviously when inducer is present the translating ribosome would not release, Rho factor could not bind, and the distal paused polymerase molecules would continue transcription into tnaA.
These kinetic features essential to tna operon regulation have not yet been addressed.
A second basic feature of tna operon regulation that remains to be elucidated is how, under inducing conditions, synthesis of the nascent TnaC peptidyl-tRNA participates in the creation of a tryptophan binding site. Also unknown is how the presence of ribosome-bound TnaC peptidyl-tRNA and bound tryptophan prevent RF-2 activation of ribosomal peptidyltransferase activity. Answers to these questions may provide an understanding of how features of a nascent peptide are recognized by a translating ribosome (41)(42)(43).