Overexpression of tnaC of Escherichia coli Inhibits Growth by Depleting tRNA₂^Pro Availability

Abstract

Transcription of the tryptophanase (tna) operon of Escherichia coli is regulated by catabolite repression and tryptophan-induced transcription antitermination. Induction results from ribosome stalling after translation of tnaC, the coding region for a 24-residue leader peptide. The last sense codon of tnaC, proline codon 24 (CCU), is translated by tRNA₂^Pro. We analyzed the consequences of overexpression of tnaC from a multicopy plasmid and observed that under inducing conditions more than 60% of the tRNA₂^Pro in the cell was sequestered in ribosomes as TnaC-tRNA₂^Pro. The half-life of this TnaC-tRNA₂^Pro was shown to be 10 to 15 min under these conditions. Plasmid-mediated overexpression of tnaC, under inducing conditions, reduced cell growth rate appreciably. Increasing the tRNA₂^Pro level relieved this growth inhibition, suggesting that depletion of this tRNA was primarily responsible for the growth rate reduction. Growth inhibition was not relieved by overexpression of tRNA₁^Pro, a tRNA^Pro that translates CCG, but not CCU. Replacing the Pro24CCU codon of tnaC by Pro24CCG, a Pro codon translated by tRNA₁^Pro, also led to growth rate reduction, and this reduction was relieved by overexpression of tRNA₁^Pro. These findings establish that the growth inhibition caused by tnaC overexpression during induction by tryptophan is primarily a consequence of tRNA^Pro depletion, resulting from TnaC-tRNA^Pro retention within stalled, translating ribosomes.

Transcription of the tryptophanase (tna) operon of Escherichia coli is regulated by catabolite repression and tryptophan-induced transcription antitermination. Induction requires attempted translation of a 24-residue coding region, tnaC, located in the 319-nucleotide transcribed leader region preceding tnaA, the structural gene for tryptophanase (4, 19). Both Trp12 and Pro24 of TnaC are essential for induction (6, 15, 10). The key feature of this antitermination mechanism has been shown to be retention of uncleaved TnaC-tRNA₂^Pro within the translating ribosome (9). The translating ribosome stalls at the tnaC stop codon (UGA) because TnaC-tRNA₂^Pro resists cleavage, and this stalling blocks Rho factor's access to the rut site in the tna transcript (11), thereby preventing transcription termination in the leader region of the operon.

During normal protein synthesis each ribosome translating an open reading frame continues translation until it reaches a termination codon. Then, a release factor (RF) promotes hydrolysis of the peptidyl-tRNA, releasing the polypeptide product and its previously associated tRNA (2, 13). In some instances, a peptidyl-tRNA dissociates from the ribosome before hydrolysis of the ester bond (3, 12, 16). This “drop-off” peptidyl-tRNA is believed to be hydrolyzed by the enzyme peptidyl-tRNA hydrolase (Pth), allowing the tRNA to be recycled for continued protein synthesis. Pth has been shown to be essential in E. coli (1, 18), presumably because efficient tRNA recycling is necessary for continued protein synthesis and cell growth. For example, it has been shown that minigene overexpression in E. coli can inhibit both translation and cell growth (14, 20). Inhibition occurs because cells fail to cleave the peptidyl-tRNA released from ribosomes efficiently, resulting in depletion of essential tRNA species (14, 20).

In an earlier study it was observed that overexpression of the tnaC operon leader region from a multicopy plasmid inhibited expression of the chromosomal tna operon (6). The cause of this inhibition was not established. Subsequent in vitro studies demonstrated that the normal mechanism of tryptophan induction of tna operon expression involves inhibition of RF2-mediated peptidyltransferase cleavage of TnaC-tRNA₂^Pro, resulting in retention of this peptidyl-tRNA within the translating ribosome (8). If this inhibition of TnaC-tRNA₂^Pro cleavage occurred in vivo it might deplete the cell of sufficient free tRNA₂^Pro to continue normal protein synthesis. This depletion could explain why overexpression of tnaC results in reduced expression of the chromosomal tna operon (6).

To determine whether tryptophan induction does lead to TnaC-tRNA₂^Pro accumulation in vivo, tnaC of E. coli was overexpressed from its own promoter by using a multicopy plasmid-based system. Accumulation of TnaC-tRNA₂^Pro in vivo was observed, and this TnaC-tRNA₂^Pro was shown to be associated with ribosomes. Cell growth rate was reduced when tryptophan was present, but only when cells contained multiple copies of the wild-type tnaC, not a nonfunctional tnaC, W12R tnaC, in which Trp12 of TnaC was replaced by Arg. Approximately 60% of the tRNA₂^Pro in tryptophan-induced cells was found to be retained as TnaC-tRNA₂^Pro; this TnaC-tRNA₂^Pro was within the ribosomal fraction. Overexpression of tRNA₂^Pro relieved this growth rate reduction. Substituting Pro codon CCG, read by tRNA₁^Pro, for the wild-type tnaC Pro24 codon, CCU, read by tRNA₂^Pro, also resulted in growth rate reduction; this reduction was relieved by overexpressing tRNA₁^Pro. These findings establish that tryptophan-induced sequestration of tRNA^Pro as TnaC-tRNA^Pro within ribosomes stalled during tnaC translation is the cause of the observed growth rate reduction.

MATERIALS AND METHODS

Bacterial strains and plasmids.

All strains and plasmids used in the present study are listed in Table 1. Plasmids bearing proK and proL, the structural genes for tRNA₁^Pro and tRNA₂^Pro, respectively, were generously provided by Michael O'Connor (17).

TABLE 1.

Bacterial strains and plasmids used in this study

Strain or plasmid	Genotypea	Source or reference
Strains
W3110	Wild type	Lab collection
W3110 tnaA2	tnaA2	23
SVS1144	W3110 bglR551 Δ(lac-argF)U169 (λ tnap tnaA′-′lacZ)	19
Plasmids
pUC18	Ampr	Lab collection
pGF25-00	pUC18 tnap tnaC; Ampr	9
pGF25-14	pGF25-00 tnap W12R tnaC; Ampr	This study
pGF25-CCG	pGF25-00 tnap tnaC(Pro24CCU to CCG); Ampr	This study
pHSG575	pSC101; Cmr	17
pProK4	pHSG575 proL; Cmr	17
pProL5	pHSG575 proK; Cmr	17

^aCm^r, chloramphenicol resistance; Amp^r, ampicillin resistance.

Media and growth conditions.

Cultures were grown in Vogel-Bonner minimal medium (22) supplemented with 0.2% glycerol and 0.05% acid casein hydrolysate (ACH), with or without 100 μg of l-tryptophan/ml. When necessary, one or more of the following antibiotics was added: ampicillin (50 μg/ml) or chloramphenicol (30 μg/ml). In some experiments 1% glucose was added to growing cultures. Bacterial growth was monitored by measuring the optical density at 600 nm (OD₆₀₀).

RNA isolation and Northern blot analysis.

To isolate total RNA from cells transformed with plasmid pGF25-14 or pGF25-00, cultures were grown in supplemented minimal medium, and cells were harvested in mid-log phase (OD₆₀₀ = 0.8), collected by centrifugation, resuspended in a fresh ribosome isolation buffer (10 mM NH₄Cl, 175 mM K acetate, 10 mM MgCl₂, 35 mM Tris-Cl [pH 8.0], 1 mM dithiothreitol, 2 mM l-tryptophan), and disrupted by sonication. The resulting cell extracts were then centrifuged at 10,000 × g for 30 min to remove cell debris. Cleared extracts were extracted with acidic phenol (pH 4.2 to 5.1), the aqueous phase collected after centrifugation, and total RNA was isolated from the aqueous phase by ethanol precipitation. Where indicated, cell extracts were first treated with proteinase K (100 μg/ml, 37°C, 5 or 10 min) to release tRNAs from peptidyl-tRNAs before phenol extraction. Then, 20 μg of isolated RNA was electrophoresed on denaturing 6.5% polyacrylamide RNA gels. Separated RNA species were electroblotted onto Hybond-N+ membranes (Amersham Pharmacia). Northern blot hybridization was carried out by using ³²P-end-labeled oligonucleotides specific to tRNA₂^Pro (5′-CCTCCGACCCCCGACACCCCAT-3′) (5). Blots were visualized and quantified by using a PhosphorImager (Molecular Dynamics).

Detection of TnaC-tRNA^Pro in ribosomes in vivo.

Cells were cultured in supplemented Vogel-Bonner minimal medium with appropriate antibiotics, with or without 100 μg of tryptophan/ml. Cells were harvested in mid-log phase (OD₆₀₀ = 0.8), washed, and resuspended in fresh ribosome isolation buffer. All subsequent steps were performed at 4°C. Each cell suspension was sonicated, and the debris was removed by centrifugation for 30 min at 10,000 × g. The resulting extracts were either examined directly (total extracts) or centrifuged (for 2 h at 100,000 × g) to separate ribosome pellets from S-100 supernatants. Proteins and tRNAs in the various preparations were separated by electrophoresis on 10% Tricine-sodium dodecyl sulfate (SDS) protein gels, transferred to nylon membranes, and probed for TnaC-tRNA₂^Pro using the tRNA₂^Pro-specific oligonucleotide probes mentioned above. In detail, after cross-linking for 5 min with UV, membranes were prehybridized at 42°C for 1 h in a solution (10 ml per 10-by-5-cm membrane) consisting of 0.60 M NaCl, 0.12 M Tris-HCl, 8 mM Na₂-EDTA (pH 8.0), 250 pg of sheared and denatured salmon sperm DNA/ml, 0.1% SDS, and 10× Denhardt solution (1× Denhardt solution = 0.02% bovine serum albumin, 0.02% polyvinylpyrrolidone 40, and 0.02% Ficoll). Hybridization was at 42°C overnight in the same solution (10 ml) with the 5′³²P-labeled oligonucleotide probes. Membranes were washed three times for 20 min each at 42°C with washing buffer (0.45 M NaCl, 0.09 M Tris-HCl, 6 mM Na₂EDTA [pH 8.0], and 0.1% SDS) and then autoradiographed (21).

RESULTS

Detection of TnaC-tRNA₂^Pro accumulation in vivo.

Tryptophan codon 12 of tnaC is essential for tryptophan-induced accumulation of TnaC-tRNA^Pro in S-30 cell extracts (9). When tnaC is overexpressed from a multicopy plasmid (Fig. 1A), accumulation of TnaC-tRNA₂^Pro in vivo could conceivably lead to tRNA₂^Pro depletion, since this tRNA is solely responsible for translation of the proline codon that is the last codon of tnaC, CCU. To detect TnaC-tRNA₂^Pro accumulation in vivo, E. coli cells containing tnaC on a multicopy plasmid (Fig. 1A) were grown under inducing conditions, harvested by centrifugation, and disrupted by sonication, and Northern blotting was performed with ³²P-labeled oligonucleotides specific for tRNA₂^Pro (see Materials and Methods). Our initial objective was to determine whether tRNA₂^Pro probing would detect tRNA₂^Pro as a component of TnaC-tRNA₂^Pro. The data in Fig. 1B demonstrate that TnaC-tRNA₂^Pro does accumulate, and is detectable, when wild-type tnaC is overexpressed in the presence of inducing levels of tryptophan (Fig. 1B, bottom panel, +Trp, lanes 7 and 9). In the absence of added tryptophan, only free tRNA₂^Pro was evident (Fig. 1B, upper panel, −Trp, all lanes). The TnaC-tRNA₂^Pro detected was shown to be exclusively associated with ribosomes; when ribosomes were separated from the supernatant by centrifugation, no TnaC-tRNA₂^Pro was detected in the S-100 supernatant; it was only in the pellet (bottom panel, compare lanes 8 and 9). As expected, TnaC-tRNA₂^Pro was not detected when the plasmid with W12R tnaC (the plasmid in which the Trp12 codon of TnaC was replaced by an Arg codon) was overexpressed in the presence of tryptophan (Fig. 1B, lanes 4, 5, and 6). Nor was TnaC-tRNA₂^Pro detected in preparations from cells with the vector-plasmid control (Fig. 1B, lanes 1, 2, and 3). Quantitation of the relative band intensities in Fig. 1B (lanes 7, 8, and 9) provided the estimate that when wild-type plasmid-borne tnaC was overexpressed in the presence of tryptophan, ∼60% of the tRNA₂^Pro in the cells was present as TnaC-tRNA₂^Pro. Assuming there are ∼720 copies of tRNA₂^Pro and ∼5,000 ribosomes per cell (5), the number of tRNA₂^Pro molecules and fraction of ribosomes sequestered per cell would be ∼430 and ∼9.0%, respectively, under our experimental conditions.

FIG. 1.

(A) Schematic representation of plasmids pGF25-00 and pGF25-14. A 949-bp fragment containing the tna operon region from −213 to +306 from the tna promoter start site was joined to a fragment containing the rpoBC terminator, and this combined fragment was cloned into pUC18 (9). (B) The presence of inducing levels of tryptophan leads to TnaC-tRNA₂^Pro accumulation in vivo. E. coli cells bearing plasmid pUC18, pGF25-14 (W12R tnaC) or pGF25-00 (WT tnaC) were grown in a medium with or without 100 μg of l-tryptophan/ml (see Materials and Methods for details). To detect TnaC-tRNA₂^Pro in E. coli cultures, cells were harvested by centrifugation and disrupted by sonication, and total cellular extracts or ribosome pellets and S-100 supernatants were separated by centrifugation. Various fractions (50-μg samples of total extract protein and/or the corresponding fraction of supernatant and pellet) were analyzed for tRNA₂^Pro or peptidyl-tRNA₂^Pro after electrophoresis on a Tricine-SDS protein gel. Northern blotting was performed to detect tRNA₂^Pro using ³²P-labeled oligonucleotides specific for tRNA₂^Pro. T, total extract; S, S-100 supernatant; P, ribosome pellet. The percentage of total tRNA₂^Pro present as TnaC-tRNA₂^Pro is shown.

Since electrophoresis on a Tricine-SDS protein gel was used to separate tRNA₂^Pro from TnaC-tRNA₂^Pro, the high pH of the buffer could have led to hydrolysis of some of the TnaC-tRNA₂^Pro that had accumulated. This would have reduced our estimate of the fraction of tRNA₂^Pro sequestered in ribosomes. To validate our quantitation, we used an additional procedure that would maintain the TnaC-tRNA₂^Pro stable during its isolation and separation. Phenol treatment of extracts is known to precipitate peptidyl-tRNAs, while tRNAs remain soluble. Upon acidic phenol extraction, the ester linkage in peptidyl-tRNAs remains stable. Strains overexpressing wild-type tnaC and W12R tnaC were grown in the presence of inducing levels of tryptophan, the cells were collected and were disrupted by sonication. Extracts containing ribosomes were then treated with phenol under acidic conditions to separate free tRNA₂^Pro from sequestered tRNA₂^Pro, i.e., TnaC-tRNA₂^Pro (Materials and Methods). After acidic phenol extraction, TnaC-tRNA₂^Pro was exclusively in the phenol phase (Fig. 2A, lane 2), whereas the majority of the free tRNA₂^Pro was soluble and was located in the aqueous phase (Fig. 2A, lane 3). To be certain that proteinase K could attack the TnaC-tRNA₂^Pro within intact ribosomes under our experimental conditions, cell extracts were treated with proteinase K before phenol extraction. Isolated material in the phenol phase and RNA in the aqueous phase were then analyzed after electrophoresis on 10% Tricine gels. Northern blot analyses indicated that the tRNA₂^Pro present initially as TnaC-tRNA₂^Pro was released by proteinase K and had moved to the aqueous phase (Fig. 2A, lanes 4 and 5). On the basis of these findings, some cell extracts were treated with proteinase K to release tRNA₂^Pro from TnaC-tRNA₂^Pro, and the resulting soluble cellular RNA was collected in the aqueous phase. Precipitated RNA (20 μg) from the aqueous phase was then electrophoresed on an acidic denaturing 6.5% polyacrylamide gel, electroblotted onto a Hybond-N+ membrane, and probed for tRNA₂^Pro (see Materials and Methods for details). 5S tRNA was also probed, and the 5S band was used as a reference in quantitating the relative levels of tRNA₂^Pro present. These results are shown in Fig. 2B. When wild-type tnaC was overexpressed, and extracts were not treated with proteinase K, only ca. 36% of the tRNA₂^Pro was present in the aqueous phase after phenol extraction (Fig. 2B, lane 4). Presumably, the remaining 64% of the tRNA₂^Pro was present in the phenol phase as TnaC-tRNA₂^Pro. Proteinase K treatment of cell extracts released the tRNA₂^Pro initially present as TnaC-tRNA₂^Pro (lanes 5 and 6). Under our experimental conditions, 5 min of digestion with proteinase K only partially released the tRNA₂^Pro present in TnaC-tRNA₂^Pro (lane 5); 10 min of proteinase K digestion appeared to release all of it (lane 6). In contrast, when the W12R mutant tnaC was overexpressed under the same conditions, all, or almost all, of the tRNA₂^Pro was found in the aqueous phase (Fig. 2B, lane 1). Proteinase K treatment of the cell extracts with W12R tnaC did not release any tRNA₂^Pro; thus, there apparently was no TnaC(W12R)-tRNA₂^Pro (Fig. 2B, lanes 2 and 3). These results suggest that the fraction (>60%) of tRNA₂^Pro retained in the phenol phase after expression of wild-type tnaC is covalently attached to TnaC, i.e., as TnaC- tRNA₂^Pro (Fig. 2B, lane 4). It is unlikely that there was any other peptidyl-tRNA with tRNA₂^Pro, since when W12R tnaC was overexpressed almost all of the tRNA₂^Pro was present in the aqueous phase after phenol extraction (Fig. 2B, lanes 1 to 3). These data agree with our calculations in Fig. 1 and confirm that when wild-type tnaC is overexpressed in the presence of tryptophan, more than 60% of the tRNA₂^Pro in the cell is sequestered as TnaC-tRNA₂^Pro.

FIG. 2.

Quantitation of the relative level of tRNA₂^Pro sequestered as TnaC-tRNA₂^Pro during tnaC overexpression in vivo. (A) E. coli cells bearing pGF25-14 (W12R tnaC) or pGF25-00 (wild-type [WT] tnaC) were grown in a medium with 100 μg of l-tryptophan/ml. Cells were harvested, disrupted by sonication, and centrifuged to remove debris, and extracts were either mock treated or treated with proteinase K at 37°C for 5 min or 10 min. Total protein-containing fractions and RNA from cells harboring pGF25-00 (WT tnaC) were isolated by using acidic phenol treatment, and components were analyzed by electrophoresis on 10% Tricine gels. Northern blotting was performed by using anti-tRNA₂^Pro probes. (B) The components in 20 μg of RNA from each sample were separated by electrophoresis on denaturing 6.5% polyacrylamide RNA gels, and Northern blotting hybridization analyses were performed with 5′ ³²P-end-labeled oligonucleotides specific for tRNA₂^Pro or tRNA₂^Pro-containing compounds. 5S tRNA was probed by using an appropriate labeled oligonucleotide and was used as a loading control. The ratios of counts in the tRNA₂^Pro band to counts in the 5S tRNA band are shown. Relative levels of tRNA₂^Pro were normalized to the tRNA₂^Pro from cells harboring pGF25-14 (W12R tnaC) without treatment with PK, which was set at 1.0.

Estimation of the in vivo stability of overproduced TnaC-tRNA₂^Pro.

There are various mechanisms that the cell might use to cleave accumulated TnaC-tRNA₂^Pro and release the stalled ribosomes and provide free tRNA₂^Pro for general protein synthesis. Experiments were therefore performed to examine the in vivo stability of accumulated TnaC-tRNA₂^Pro. Using our plasmid-based tnaC overexpression system, TnaC-tRNA₂^Pro was overproduced in a growing culture, and then glucose was added to activate catabolite repression and shut off tna operon transcription. Glucose addition to cultures containing glycerol and tryptophan is known to reduce tna operon expression by >100-fold (19). Samples were removed at various times, and the stability of the previously synthesized TnaC-tRNA₂^Pro was determined (Fig. 3). It can be seen that the accumulated TnaC-tRNA₂^Pro was somewhat unstable; it had a half-life of 10 to 15 min under the conditions examined. These findings demonstrate that this peptidyl-tRNA is sufficiently stable in vivo that its synthesis and survival could deplete the pool of free tRNA₂^Pro needed for additional protein synthesis.

FIG. 3.

Stability in vivo of overproduced TnaC-tRNA₂^Pro. (A) Detection of TnaC-tRNA₂^Pro after glucose addition to a culture in vivo. Cells bearing plasmid pGF25-00 were grown in minimal medium supplemented with 0.05% ACH and 0.2% glycerol with 100 μg of l-tryptophan/ml. Cell growth was monitored by measuring OD₆₀₀. A total of 1.0% glucose was added at an OD₆₀₀ of 0.6. Samples were taken at the indicated times (in minutes). After sonication and centrifugation, samples containing 50 μg of protein were analyzed for tRNA₂^Pro or peptidyl-tRNA₂^Pro after electrophoresis on a Tricine-SDS gel. Northern blotting was performed to detect tRNA₂^Pro using ³²P-labeled oligonucleotides specific for tRNA₂^Pro. The percentage of total tRNA₂^Pro present as TnaC-tRNA₂^Pro is shown. (B) Time course of TnaC-tRNA₂^Pro decay calculated from the data in panel A.

Growth rate inhibition by tnaC overexpression and its relief by increased tRNA^Pro production.

If the cellular level of free tRNA₂^Pro is appreciably depleted by tnaC overexpression, this might reduce the cell growth rate. To explore this possibility, the cell growth rates of relevant strains were determined in media with or without added tryptophan. The cell growth rate was severely reduced when wild-type W12 tnaC was overexpressed in the presence of inducing levels of tryptophan (Fig. 4A, compare pGF25-00 −Trp with pGF25-00 +Trp). When the W12R tnaC construct or the empty vector was introduced into E. coli (Fig. 4A, pUC and p25-14) the cultures grew normally, with or without added tryptophan. No growth rate reduction was observed with the E. coli wild-type strain W3110 or the mutant lacking tryptophanase, W3110 tnaA2, with or without added tryptophan (Fig. 4A).

FIG. 4.

Growth inhibition and tRNA₂^Pro overproduction in E. coli. (A) Cells bearing the different plasmids shown in Fig. 4A were grown in minimal medium supplemented with 0.05% ACH and 0.2% glycerol with or without 100 μg of l-tryptophan/ml. Cell growth was monitored by measuring OD₆₀₀. (B) Measurement of tRNA₂^Pro levels in cultures containing different plasmids. E. coli cells containing pProL5 (tRNA₂^Pro) or pHSG575 (vector) were grown in minimal medium supplemented with 0.05% ACH and 0.2% glycerol, without 100 μg of tryptophan/ml, in the presence of chloramphenicol (30 μg/ml). Supernatants and ribosomes were separated by centrifugation, samples were electrophoresed, and the components were probed using tRNA₂^Pro- and 5S tRNA-specific oligonucleotide probes. (C) E. coli cells bearing plasmids pGF25-00 (WT tnaC) and pRroL5 (tRNA₂^Pro) or pHSG575 (vector) were grown in medium with 100 μg of tryptophan/ml. S-100 supernatants and pellets were treated as described in Materials and Methods to measure the presence of TnaC-tRNA₂^Pro and/or tRNA₂^Pro in the supernatant or ribosome pellet. The percentage of total tRNA₂^Pro present as TnaC-tRNA₂^Pro is shown.

On the basis of our estimate of the extent of tRNA₂^Pro sequestration in TnaC-tRNA₂^Pro when wild-type tnaC is overexpressed from a multicopy plasmid, we reasoned that depletion of free tRNA₂^Pro is the likely cause of growth rate reduction. We therefore tested whether overexpression of tRNA₂^Pro could relieve the growth inhibition associated with tnaC overexpression. Initially, we examined the extent of plasmid-based tRNA₂^Pro overexpression in E. coli W3110 tnaA2. Figure 4B shows that the tRNA₂^Pro level in cells containing the tRNA₂^Pro overexpression plasmid, pProL5, was at least five times higher than in cells with pHSG575, the control, empty plasmid. As expected, additional tRNA₂^Pro was found in both the supernatant and the ribosome pellet (Fig. 4B and 4C). We then examined TnaC-tRNA₂^Pro accumulation in tryptophan-induced E. coli cultures overproducing both tRNA₂^Pro and tnaC. We observed that in the presence of elevated levels of tRNA₂^Pro, the extent of accumulation of TnaC-tRNA₂^Pro was dependent on the presence of inducing levels of tryptophan; ∼2-fold more TnaC-tRNA₂^Pro was observed per cell when tRNA₂^Pro was overexpressed (Fig. 4C, lane P with pProL5 versus lane P with vector). It was obvious that even when TnaC-tRNA₂^Pro was accumulated in the presence of pProL5, a large amount of free tRNA₂^Pro was present (Fig. 4C, lane S, pProL5 panel). When tRNA₂^Pro was overproduced, the percentage of tRNA₂^Pro as TnaC-tRNA₂^Pro decreased to 12.3% (Fig. 4C, left panel). In contrast, more than 63% of tRNA₂^Pro was associated with TnaC-tRNA₂^Pro when tRNA₂^Pro was not overproduced, and the parental plasmid was present (Fig. 4C, right P panel).

Having confirmed that tRNA₂^Pro is overexpressed in strains with pProL5, we next sought to determine whether this overexpression would relieve the observed growth inhibition. In Fig. 5A it is shown that overexpression of tRNA₂^Pro did reverse the growth inhibition attributed to wild-type tnaC overexpression (Fig. 5A). Neither overexpression of tRNA₁^Pro nor the presence of the tRNA vector plasmid overcame this growth inhibition (Fig. 5A). Growth inhibition by tnaC overexpression was also evident when the Pro24 codon of tnaC, CCU, was replaced by the Pro24 codon, CCG, a codon specifically translated by tRNA₁^Pro (Fig. 5B). It has been estimated that there are ∼900 molecules of tRNA₁^Pro per E. coli cell (5). As expected, TnaC-tRNA₁^Pro also accumulated when tnaC (CCG) was overexpressed in the presence of inducing levels of tryptophan (data not shown). Significantly, overexpression of tRNA₁^Pro, rather than tRNA₂^Pro, relieved this growth inhibition (Fig. 5B). These data demonstrate that sequestration of the specific tRNA^Pro that decodes the last sense codon of tnaC, as TnaC-tRNA^Pro, in the stalled ribosome complex, causes the growth inhibition associated with tnaC overexpression.

FIG. 5.

Overexpression of the tRNA^Pro that can translate the last sense codon of tnaC relieves growth inhibition caused by tnaC overexpression in the presence of tryptophan. E. coli cells bearing the different plasmids listed in the figure (also see Table 1) were grown in minimal medium supplemented with 0.05% ACH and 0.2% glycerol, with or without 100 μg of l-tryptophan/ml, and with shaking at 37°C. Cell growth was monitored at OD₆₀₀. (A) Cells with the plasmid overexpressing wild-type tnaC plus the parental plasmid, the plasmid encoding tRNA₂^Pro (pProL5), or the plasmid expressing tRNA₁^Pro (pProK4), were grown in the absence or presence of tryptophan. (B) Cells with the plasmid overexpressing tnaC with the Pro24 codon CCG, plus the parental plasmid, the plasmid encoding tRNA₁^Pro (pProK4), or the plasmid encoding tRNA₂^Pro (pProL5), were grown in the absence or presence of tryptophan.

DISCUSSION

Using an E. coli S-30 in vitro system it has been shown that tryptophan induction of tna operon expression leads to inhibition of RF2-mediated cleavage of TnaC-tRNA₂^Pro (8). The translating ribosome appears to stall at the tnaC stop codon, and this stalling blocks Rho factor's access to its rut binding site located on the tna mRNA in the vicinity of the tnaC stop codon. This prevents Rho-dependent transcription termination in the leader region of the tna operon (8). To validate this model, it was critical to show in vivo that inducing levels of tryptophan in the growth medium do result in TnaC-tRNA^Pro accumulation, and ribosome stalling. Previous attempts to detect TnaC-tRNA^Pro in vivo were unsuccessful (data not shown), leaving open the possibility that some mechanism of peptidyl-tRNA cleavage is operable within the growing cell.

In the present study, we developed a novel method to monitor TnaC-tRNA₂^Pro accumulation in vivo. Cultures containing a multicopy plasmid bearing tnaC were grown under inducing and noninducing conditions, and cell extracts were prepared. Ribosome pellets were separated from supernatants, tRNA₂^Pro and TnaC-tRNA₂^Pro were separated by gel electrophoresis, and the separated material was transferred to a nylon membrane. Northern blotting was performed to detect tRNA₂^Pro, both as free tRNA₂^Pro and as TnaC-tRNA₂^Pro, by using ³²P-labeled oligonucleotides specific to tRNA₂^Pro (see Materials and Methods for details). Using this procedure, tryptophan-induced accumulation of TnaC-tRNA₂^Pro was observed in vivo (Fig. 1 and 4C). Moreover, ca. 60% of the tRNA₂^Pro in the cells was present as TnaC-tRNA₂^Pro. The TnaC-tRNA₂^Pro detected was exclusively associated with ribosome pellets (Fig. 1 and 4C). In addition, the total amount of tRNA₂^Pro in cells appeared to be unchanged even when more than ca. 60% of the tRNA₂^Pro was sequestered as TnaC-tRNA₂^Pro (Fig. 2B, lanes 1 and 6). Apparently, E. coli lacks a mechanism for increasing tRNA₂^Pro synthesis or release when free tRNA₂^Pro is depleted by tnaC overexpression.

When tnaC was overexpressed in the presence of tryptophan, both the growth rate and the rate of overall protein synthesis were reduced. Since tryptophan induces TnaC-tRNA₂^Pro accumulation in vivo (Fig. 1), tRNA₂^Pro sequestration was considered the likely cause of this inhibition. Overexpression of tRNA₂^Pro did in fact largely relieve this growth inhibition. Furthermore, when the Pro24 codon of tnaC, CCU, was replaced by CCG, and tnaC Pro24 CCG was overexpressed in the presence of inducing levels of tryptophan, overexpression of tRNA₁^Pro, rather than tRNA₂^Pro, relieved this tnaC Pro24 CCG inhibition (Fig. 5). In these experiments, TnaC-tRNA₁^Pro accumulation was also observed (data not shown). These data establish that depletion of the tRNA reading the last sense codon of tnaC is primarily responsible for the inhibition of cell growth associated with tryptophan induction.

Multicopy plasmids carrying tnaC in a strain with a chromosomal tnaA′-′lacZ fusion have been shown to inhibit tryptophan-induced TnaA-LacZ (β-galactosidase) production (6). Mutational studies established that this inhibition was not due to inhibition of transcription initiation, translation initiation, tryptophan transport, or enzyme activity (6). On the basis of our findings, we believe that this inhibition was a result of tRNA₂^Pro depletion. In fact, we have found that overproduction of tRNA₂^Pro does restore TnaA-LacZ production by the chromosomal operon (data not shown).

Inhibition of cell growth by tnaC overexpression mimics the inhibitory effect conferred by minigene expression (14, 20). However, the exact mechanisms causing tRNA depletion differ in these two systems. The deleterious effect of minigene expression is mediated by depletion of the corresponding pools of free tRNAs (14, 20). Inhibition occurs because cells fail to recycle efficiently the tRNAs from the peptidyl-tRNAs released from ribosomes, causing starvation for essential species of tRNAs. Thus, peptidyl-tRNA dropoff and inefficiency of tRNA recycling catalyzed by Pth are essential features of this minigene-conferred inhibition. Regarding growth inhibition by tnaC overexpression, our estimate of the stability of TnaC-tRNA₂^Pro within the cells of an induced culture suggests that it has a half-life of about 10 to 15 min (Fig. 3). It appears that TnaC-tRNA₂^Pro is retained within translating ribosomes for a period sufficiently long to create a free tRNA₂^Pro deficiency.