INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Organisms utilize a wide range of regulatory mechanisms to control gene expression. Bacteria have developed several sophisticated regulatory mechanisms that allow the organisms to modulate gene expression after transcription has initiated. In addition, several subtle mechanisms allow the organisms to fine tune the final level of any particular gene product. Expression of the Bacillus subtilis trpEDCFBA operon is regulated in response to changes in the intracellular concentration of tryptophan by the trp RNA-binding attenuation protein (TRAP). TRAP regulates expression of the trp operon by transcription attenuation and translational control mechanisms (reviewed in references 4 and 12). TRAP exists as a complex consisting of 11 identical subunits arranged in a single ring (3). Tryptophan cooperatively activates TRAP by binding between adjacent TRAP subunits (3, 6). When TRAP is activated by tryptophan, 11 KKR motifs that outline the periphery of the TRAP complex bind to 11 (G/U)AG repeats present in the nascent trp leader transcript, thereby wrapping the RNA around the periphery of the TRAP complex (2, 7, 33). In the transcription attenuation mechanism, TRAP binding prevents formation of an antiterminator structure, since six of the (G/U)AG repeats are present within this RNA structure (5, 7). In this case, TRAP binding promotes the formation of an overlapping intrinsic terminator, resulting in transcription termination before RNA polymerase reaches the structural genes. In the absence of TRAP binding, formation of the antiterminator permits transcription of the entire operon (5). A third stem-loop structure that forms at the extreme 5' end of the trp leader transcript also plays a role in the transcription attenuation mechanism. TRAP-5' stem-loop interaction increases the affinity of TRAP for trp leader RNA and reduces the number of (G/U)AG repeats that are required for stable TRAP-trp leader RNA association. Thus, TRAP-5' stem-loop interaction may increase the likelihood that TRAP will bind to the (G/U)AG repeats in time to block antiterminator formation (10, 30).

In addition to regulating transcription of the trp operon, TRAP also regulates translation of trpE (11, 17, 20). RNA structural studies of the trp operon readthrough transcript indicated that the most thermodynamically stable conformation of the leader RNA contains a large secondary structure that includes the last six (G/U)AG repeats in the 5' half of the stem (Fig. 1). TRAP binding to these repeats prevents or disrupts formation of this large secondary structure, thereby promoting the formation of a structure that sequesters the trpE Shine-Dalgarno (SD) sequence (the trpE SD blocking hairpin) (11, 17, 20). Formation of the trpE SD blocking hairpin inhibits TrpE synthesis by blocking ribosome access to the trpE ribosome binding site (11). In vivo and in vitro studies have established that multiple nucleotide substitutions that destabilize the trpE SD blocking hairpin, without altering the SD sequence itself, reduce the ability of TRAP to regulate TrpE synthesis (11, 20). In this study, we examined the effect of tryptophan concentration on transcription attenuation and translational control and found that translational regulation requires a tryptophan concentration higher than that required for transcription attenuation.

View larger version (13K): [in this window] [in a new window]   FIG. 1.   Model of the trpE translational control mechanism. (A) Under tryptophan-limiting conditions, TRAP is not activated and is unable to bind to the leader of trp operon readthrough transcripts. In this case the trpE SD sequence is single stranded, allowing efficient translation. (B) Under excess-tryptophan conditions, tryptophan-activated TRAP binds to the (G/U)AG repeats present between nucleotides 36 and 91 in the leader of a trp operon readthrough transcript. In this case, the secondary structure of the downstream RNA is altered such that the trpE SD sequence is sequestered in a stable RNA hairpin (trpE SD blocking hairpin). This RNA secondary structure inhibits TrpE synthesis by preventing 30S ribosomal subunit interaction with the trpE message (11). The nucleotides between 171 and 184 that are responsible for sequestering the trpE SD sequence are shown in outline type in both diagrams. In the absence of bound TRAP, these nucleotides basepair with a segment of the TRAP binding site. The (G/U)AG repeats and the trpE AUG initiation codon are indicated in boldface. The nucleotide substitutions in the SD trpL that prevent formation of the trpE SD blocking hairpin are indicated with arrows. The nucleotides that make up the mutually exclusive antiterminator and terminator structures extend from positions 60 to 111 or 108 to 133, respectively. Periods indicate that sequence information has been omitted. Numbering is from the start of transcription.

The sequence of the trpEDCFBA operon revealed that the coding sequences of all but two of the genes (trpC and trpF) overlap by several nucleotides (Fig. 2). This gene organization suggested that the trp genes might be regulated by translational coupling (14), a process in which translation of a gene is partially dependent on translation of the gene immediately upstream (19, 24). The first two genes of the trp operon, trpE and trpD, have the most extensive overlap in the operon (29 nucleotides). We found that these two genes are translationally coupled and that formation of the trpE SD blocking hairpin regulates TrpD synthesis via translational coupling. Since inhibition of translation can result in reduced mRNA stability (18) and/or transcriptional polarity (16), we performed experiments to determine if trpE translational control influences trp operon expression by reducing the stability of the message or by allowing Rho access to the nascent trp transcript. Our results establish that formation of the trpE SD blocking hairpin promotes Rho-mediated transcriptional polarity but does not influence the stability of the downstream mRNA.

View larger version (5K): [in this window] [in a new window]   FIG. 2.   Schematic representation of the B. subtilis trpEDCFBA operon highlighting the overlapping reading frames. With the exception of the 4-nucleotide intercistronic region between trpC and trpF, all of the open reading frames overlap by several residues. The length of each overlap (in nucleotides) is shown in parentheses. The positions of the trp promoter (p>) and leader (trpL) are shown (not drawn to scale).

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Plasmids and bacterial strains. The genotypes of the B. subtilis strains used in this study are listed in Table 1. Plasmid pPB20 containing the wild-type (WT) B. subtilis trp promoter, the leader, and the first 40 trpE codons was previously described (5). pINT-SDtrpL contains several trp leader point mutations (SDtrpL) that destabilize the trpE SD blocking hairpin without disrupting the trpE SD sequence itself (20). Plasmid pHD15 contains the trp promoter, the leader, the entire trpE coding sequence, and the N-terminal coding sequence of trpD, while pHD22 contains a trpE'-'lacZ translational fusion that is controlled by the trp promoter and leader (11). The integration plasmids ptrpBGI (WT trpL trpE'-'lacZ) (29) and pHD24 (SD trpL trpE'-'lacZ) were previously described (11). The integration vector, ptrpBGI-PLK, used for generation of trpE'-'lacZ translational fusions, was described previously (20). The plasmid constructions used to generate trpED'-'lacZ translational fusions were made in Escherichia coli. Integrative plasmids containing various translational fusions were linearized with PstI and separately integrated into the amyE locus of the B. subtilis chromosome by homologous recombination. For example, to construct strains containing trpE-'lacZ translational fusions in a tryptophan auxotrophic strain, we integrated ptrpBGI or pHD24 into strain 168 (trpC2) to yield strain PLBS176 or PLBS201, respectively. Transformation was by natural competence (1). Selection was for chloramphenicol resistance (5 µg/ml). Integration into amyE was confirmed by screening for the absence of amylase activity on starch plates by using iodine staining (28). In each case, the correct plasmid construction was confirmed by automated DNA sequencing.

-Galactosidase assay. B. subtilis cultures were grown in minimal-acid casein hydrolysate medium in the presence of 5 µg of chloramphenicol/ml. Tryptophan prototrophs were grown in the presence of 0 or 200 µM tryptophan. Tryptophan auxotrophs were grown in the presence of 20, 50, 100, 200, or 400 µM tryptophan. The cells were harvested during late exponential growth (110 Klett units; filter no. 54; Klett Manufacturing Co., Inc.). Aliquots were then assayed for -galactosidase activity as previously described (11).

mRNA half-life and steady-state level determinations. B. subtilis cultures were grown in minimal-acid casein hydrolysate medium in the presence of 20 or 200 µM tryptophan (trpC2 auxotrophic strains) or 0 or 200 µM tryptophan (trp prototrophic strains). The growth medium was supplemented with appropriate antibiotics (5 µg of chloramphenicol/ml or 10 µg of kanamycin/ml). When the cultures reached late exponential phase (110 Klett units; filter no. 54), 100 µg of rifampin/ml was added to inhibit transcription initiation. Six-milliliter aliquots were removed 0, 1, 2, 4, 8, and 16 min after the addition of rifampin and added to an equal volume of frozen killing buffer (8.5 mM Tris-HCl, pH 7.2, 5 mM MgCl₂, 25 mM sodium azide, and 500 µg of chloramphenicol/ml). Total RNA was isolated using the RNeasy protocol (Qiagen), and genomic DNA was eliminated using RNase-free DNase I (Promega). RNA was extracted with an equal volume of phenol-chloroform, precipitated with ethanol, and suspended in Tris-EDTA. Quantification of mRNA was performed by slot blot hybridization. Ten micrograms of total RNA from each sample was mixed with an equal volume of denaturing solution (0.55 ml of formamide, 0.2 ml of 37% formaldehyde, 0.2 ml of morpholinepropanesulfonic acid, pH 7.0) and incubated for 15 min at 65°C. Samples were subsequently chilled on ice and spotted onto Hybond-N+ membranes (Amersham Pharmacia Biotech) using a Minifold I dot blot apparatus (Schleicher & Schuel). After filtration, RNA was covalently linked to the membrane using a UV Stratalinker (Strategene). Prehybridization was performed at 65°C for 1 h in a buffer containing 7% sodium dodecyl sulfate (SDS), 0.5 M sodium phosphate buffer, pH 7.2, and 10 mM EDTA (9). Hybridization was carried out at 65°C for 12 h by adding lacZ, trpE, or trpD DNA probes that were radiolabeled using the random-primer DNA-labeling kit (Boehringer Mannheim Biochemical). The filters were washed twice for 30 min each time with 1× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate) containing 0.1% SDS at 62°C followed by two 30-min washes with 0.1× SSC containing 0.1% SDS at the same temperature. mRNA levels were quantified with a PhosphorImager (Molecular Dynamics, Inc.) and the ImageQuant software package. Determination of the steady-state levels of trpE and trpD mRNAs was performed as described above except that only 0-min time points were analyzed and rifampin was omitted.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

The trpE translational control mechanism requires a concentration of tryptophan higher than that required for transcription attenuation. Expression of the trpEDCFBA operon is regulated by TRAP-mediated transcription attenuation and translational control mechanisms. In the transcription attenuation mechanism, only transient TRAP interaction with the (G/U)AG repeats would be necessary to block formation of the antiterminator. Once transcription has terminated, it would be advantageous for the TRAP-trp leader RNA complex to dissociate to prevent titration of TRAP. However, for translational control it appears that TRAP must remain associated with the triplet repeats to maintain formation of the trpE SD blocking hairpin because the trpE SD sequence is single stranded in the absence of bound TRAP (11). It was also shown that the affinity of TRAP for trp leader RNA increases with increasing tryptophan concentrations (25, 32) and that the dissociation rate of TRAP-trp leader RNA complexes decreases as the concentration of tryptophan increases (8). Because of these findings, we reasoned that the tryptophan concentration required for translational control would be higher than that required for transcription attenuation.

View larger version (16K): [in this window] [in a new window]   FIG. 3.   -Galactosidase activity from trpC2 auxotrophic strains as a function of tryptophan concentration in the growth media. trpE'-'lacZ translational fusions were under control of the WT trpL (solid circles) or SD trpL (open circles). Both the transcription attenuation and trpE translational control mechanisms function in the WT trpL strain (PLBS176), whereas only the transcription attenuation mechanism functions in the SD trpL strain (PLBS201). -Galactosidase activity is shown in Miller units (21). This figure is derived from the data presented in Table 2.

Inhibition of trpE translation mediated by the SD blocking hairpin does not alter stability of downstream mRNA. Inhibition of translation has been shown to increase the rate of mRNA decay in some cases (reviewed in reference 18). This presumably results from decreased ribosomal protection of the downstream message from ribonucleases. We tested the possibility that formation of the trpE SD blocking hairpin destabilized the downstream mRNA. We determined the lacZ mRNA half-life in strains containing trpE'-'lacZ fusions controlled by either the WT trpL (PLBS176) or the SD trpL (PLBS201). Since formation of the trpE SD blocking hairpin is dependent on tryptophan-activated TRAP, we expected that the mRNA half-life would be reduced in the WT trpL strain when it was grown in the presence of tryptophan. We were surprised to find that the lacZ mRNA half-lives were essentially identical in both strains when grown under limiting or excess tryptophan conditions (Fig. 4A). To ensure that the lacZ mRNA half-life data accurately reflected trp mRNA stability, we determined the half-life of trpE and trpD mRNAs in the presence of limiting or excess tryptophan. While the decay of trpE and trpD mRNAs was considerably more rapid than that of the lacZ message, formation of the trpE SD blocking hairpin did not influence the stability of the transcripts (Fig. 4B). Thus, trpE translational control does not lead to destabilization of trp mRNA.

View larger version (31K): [in this window] [in a new window]   FIG. 4.   Time course of lacZ, trpE, and trpD mRNA decay. (A) lacZ mRNA half-life experiments in strains containing trpE'-'lacZ translational fusions that were controlled by either the WT trpL (PLBS176) or the SD trpL (PLBS201) when each strain was grown in the presence of 20 or 200 µM tryptophan. Open circles, WT trpL with 20 µM tryptophan; solid circles, WT trpL with 200 µM tryptophan; open squares, SD trpL with 20 µM tryptophan; solid squares, SD trpL with 200 µM tryptophan. (B) trpE and trpD mRNA half-life experiments from the natural trp operon when strain 168 was grown in the presence of 20 or 200 µM tryptophan. Open circles, trpE with 20 µM tryptophan; solid circles, trpE with 200 µM tryptophan; open squares, trpD with 20 µM tryptophan; solid squares, trpD with 200 µM tryptophan. The relative levels of mRNA remaining at 0, 1, 2, 4, 8, and 16 min after the addition of rifampin were determined by dot blot analysis using the corresponding probe. Each value is an average from at least two independent experiments. The mRNA level corresponding to each 0-min time point was set to 1. The mRNA half-life ± the standard deviation for each strain and growth condition is shown next to the corresponding symbol.

trpE and trpD are translationally coupled. The coding regions of trpE and trpD overlap by 29 nucleotides (Fig. 2). To determine if this overlap allows translational coupling of these two genes, we constructed two trpED'-'lacZ translational fusions in which the central region of trpE was deleted (Fig. 5). In one of the fusions, we introduced a UAA termination codon in trpE. This stop codon would disrupt translational coupling unless it was translated by a tRNA suppressor. The other fusion contained the natural lysine codon (AAA) instead of the engineered stop codon. In both cases, the reading frame of the internal trpE deletion was maintained (Fig. 5). The translational fusions were integrated into the amyE locus of WT or sup-3 B. subtilis strains. It was previously shown that the sup-3 allele encodes a tRNA suppressor that can place a lysine residue at UAA stop codons approximately 15% of the time (22). Thus, placement of a lysine residue at this position by the tRNA suppressor would result in a trpE polypeptide that is identical in sequence to that encoded by the other fusion. We examined expression from the trpED'-'lacZ fusion and found that the levels of regulation in response to tryptophan for all four strains were comparable (compare the Trp/+Trp -galactosidase ratios for the first four strains in Table 3). Introduction of the engineered UAA stop codon reduced expression of the fusion sevenfold in the absence of tryptophan and fourfold in its presence (compare expression levels from PLBS142 and PLBS138 [Table 3]). Importantly, when expression was examined in the sup-3 strain, we found that expression was partially restored in the strain that contained the engineered stop codon; expression increased threefold in the absence of tryptophan and twofold in its presence (compare expression levels from PLBS138 and PLBS139 [Table 3]). The sup-3 allele had little effect on expression when the engineered stop codon was not present (compare expression levels from PLBS142 and PLBS143 [Table 3]). The finding that suppression of the UAA stop codon within the trpE coding sequence partially restored expression of the trpED'-'lacZ translational fusion demonstrates that trpE and trpD are translationally coupled.

View larger version (12K): [in this window] [in a new window]   FIG. 5.   Nucleotide and amino acid sequences of the trpED'-'lacZ translational fusion transcripts used in the translational coupling studies. Only the crucial features of the trpE and trpD coding regions are shown. The trpE and trpD amino acid sequences are shown above and below the nucleotide sequence, respectively. The trpE and trpD coding sequences overlap by 29 nucleotides. Two fusions were constructed that contained either an AAA lysine codon or an engineered UAA stop codon (***). The tRNA suppressor encoded by the sup-3 allele can place a lysine residue at the UAA stop codon with approximately 15% efficiency (22). The position of the natural trpE UGA termination codon is shown (*). The trpE codons 41 to 483 were replaced by a short linker. Thus, there are 31 codons between the engineered stop codon and the natural trpE stop codon. The trpE and trpD SD sequences are shown in boldface. The inverted arrows indicate the nucleotides involved in trpE SD blocking hairpin formation. Periods indicate that sequence information has been omitted.

The trpE SD blocking hairpin regulates translation of trpD via translational coupling. Since trpE and trpD are translationally coupled, it was possible that formation of the trpE SD blocking hairpin would also regulate translation of trpD. To test this possibility, we compared expression levels from trpED'-'lacZ fusions containing WT trpL or SD trpL trp leaders. We found that the Trp/+Trp ratios were similar in all four SD trpL strains and that the SD trpL reduced regulation of the trpED'-'lacZ fusion in response to tryptophan (compare the Trp/+Trp -galactosidase ratios for the last four strains in Table 3 with those of the first four strains). This reduction in regulation was primarily due to elevated expression of the SD trpL-containing strains when they were grown in the presence of tryptophan. As expected, we found that the translational coupling expression patterns were retained in the SD trpL strains. In this case, introduction of the engineered stop codon reduced expression eight- to ninefold (compare expression levels from PLBS146 and PLBS144 [Table 3]), while expression was restored two- to threefold when the sup-3 allele was present (compare expression levels from PLBS144 and PLBS145 [Table 3]). Again, the tRNA suppressor (sup-3) had little effect on expression when the UAA stop codon was not present (compare expression levels from PLBS146 and PLBS147 [Table 3]). Taken together, these results demonstrate that the trpE SD blocking hairpin regulates TrpD synthesis via translational coupling of trpE and trpD.

Rho termination factor influences expression of the trp operon. Polarity in operons can be caused by Rho-dependent termination within transcripts that are not being translated (reference 16 and references therein). Thus, inhibition of trpE translation caused by the SD blocking hairpin could result in Rho-dependent transcriptional polarity. Since we determined that trpE translational control only occurred at relatively high concentrations of tryptophan (see above), we expected Rho to have its greatest effect when cells were grown in the presence of excess tryptophan. To test this possibility, we engineered a Rho-deficient strain by disrupting rho with a kanamycin resistance gene. As was previously reported (16, 26), Rho is not essential for the viability of B. subtilis.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Expression of the tryptophan biosynthetic operon (trpEDCFBA) is regulated by TRAP-mediated transcription attenuation and translational control mechanisms. In the transcription attenuation mechanism, only transient TRAP interaction with the (G/U)AG repeats would be necessary to block formation of the antiterminator. However, for translational control it is probably necessary for TRAP to remain associated with the triplet repeats to maintain formation of the trpE SD blocking hairpin. Although the intracellular concentration of free tryptophan was not determined in this study, the results of our expression experiments indicate that the translational control mechanism requires a tryptophan concentration higher than that required for attenuation (Fig. 3 and Table 2). It was recently shown that yhaG encodes a putative transmembrane protein involved in tryptophan transport and that TRAP regulates translation of this gene (27). Since B. subtilis is a soil bacterium, it is reasonable to assume that the organism would encounter environments in which the amounts of available tryptophan differ. Thus, the ability of B. subtilis to utilize different regulatory mechanisms to modulate trp operon expression in response to changes in the environmental supply of tryptophan may provide a growth advantage.

It is generally assumed that inhibition of translation can lead to a decrease in mRNA half-life by increasing the susceptibility of the message to nucleolytic attack. While this has been substantiated in several instances, there are examples where it has been shown that inhibition of translation did not alter the mRNA half-life or even had a stabilizing affect on the message (18). In this study, we found that inhibition of translation by the trpE SD blocking hairpin did not alter the stability of the downstream mRNA (Fig. 4).

The importance of translational coupling in maintaining stoichiometric production of biosynthetic pathway enzymes and ribosomal proteins is well established in E. coli (19, 24). In the majority of cases, the translation termination codon of the upstream cistron overlaps the initiation codon of the downstream cistron. In the case of the B. subtilis trpEDCFBA operon, the lengths of the overlaps are far more extensive (8 to 29 nucleotides) (Fig. 2). Our expression studies establish that trpE and trpD constitute a translationally coupled gene pair (Table 3). In addition, since we still observed substantial translation of trpD when we uncoupled translation by introducing the stop codon in trpE, our results indicate that TrpD synthesis is not dependent on translational coupling. Moreover, our coupling studies demonstrated that the trpE translational control mechanism, relying on TRAP-dependent formation of the trpE SD blocking hairpin, regulates TrpD synthesis via translational coupling (Table 3). However, since we observed the same general pattern of translational coupling in TRAP-deficient (mtrB) strains, our results indicate that the coupling mechanism itself is independent of TRAP (Table 3).

Translational coupling is generally thought to occur by one of two mechanisms (19, 24). In one case, the translating ribosome disrupts an RNA hairpin that can sequester the SD sequence of the downstream cistron. As a consequence, a second ribosome can gain access to the now single-stranded SD sequence and initiate translation. Since our computer predictions have not detected any evidence for a trpD SD blocking hairpin, we do not believe that this mechanism is involved in our case. In other instances, it is thought that the same ribosome that translates the upstream cistron can initiate translation of the downstream cistron because the respective stop and start codons overlap or are close to one another. While it is possible that coupling of trpE and trpD relies on this type of mechanism, it is difficult to envision how such an extensive overlap would be accommodated.

With the exception of trpC and trpF, all of the genes in the B. subtilis trp operon overlap by several nucleotides (Fig. 2). In the case of these two genes, the intercistronic region is only 4 nucleotides. Since it is known that genes with short intercistronic gaps can be coupled (19), it is possible that the entire B. subtilis trpEDCFBA operon is translationally coupled. Thus, it is conceivable that formation of the trpE SD blocking hairpin regulates expression of the entire operon via translational coupling. This possibility remains to be tested. Since overlapping genes are particularly common in B. subtilis, it is likely that translational coupling plays a significant role in the coordinate regulation of several genes in this organism (12).

Transcriptional polarity can occur when Rho is allowed access to a nascent transcript when translation is inhibited. Our results establish that formation of the trpE SD blocking hairpin leads to transcriptional polarity of the trp operon (Tables 4 and 5). Thus, it appears that Rho binds downstream from this RNA structure. It was previously shown that rho is a nonessential gene in B. subtilis (16, 26), in contrast to E. coli, Rhodobacter sphaeroides (13), and Micrococcus luteus (23), in which it is an essential gene. Interestingly, it was determined that B. subtilis Rho regulates expression of its own gene (16). To the best of our knowledge, this is the only other known example of Rho-dependent regulation in B. subtilis.

When cells are growing under conditions of tryptophan excess, TRAP would be activated and most likely bind to the message as it is being synthesized. In most cases this would promote Rho-independent (intrinsic) termination in the trp leader (transcription attenuation). However, in some instances, RNA polymerase will escape termination despite TRAP binding, since transcription termination is never 100% efficient. In other instances, TRAP might bind prior to transcription of the trpE SD sequence but not in time to promote termination. Both of these scenarios would result in a nascent TRAP-bound readthrough transcript that would contain the trpE SD sequence in the SD blocking hairpin. Rho could then bind to the nascent transcript and cause transcriptional polarity by promoting premature transcript release from RNA polymerase.