The trp RNA-Binding Attenuation Protein of Bacillus subtilis Regulates Translation of the Tryptophan Transport Gene trpP (yhaG) by Blocking Ribosome Binding

Bacterial strains and plasmids.The plasmids pPB77 containing the B. subtilis trp operon leader (3), pPB31 containing pabA (7), pYH14 containing a rho::neo allele (40), and pYH28 containing a trpE′-′gfp translational fusion in which tryptophan codon 57 in gfp was changed to a phenylalanine codon (29) have been described previously. The cloning vectors pTZ19R and pTZ18U each contain a T7 RNA polymerase promoter upstream from a polylinker (United States Biochemical Corp.). The plasmid pHZB6 was constructed by cloning a chromosomally derived PCR fragment containing +1 to +137 relative to the start of trpP transcription into pTZ18U. Plasmid pYH34 was constructed by replacing the trp operon sequences in pYH28 with a chromosomally derived PCR fragment containing +1 to +112 relative to the start of trpP transcription. The resulting trpP′-′gfp translational fusion contained the 10th trpP codon fused in frame with the first gfp codon.

The B. subtilis strains used in this study are described in Table 1. B. subtilis strain PLBS338 was constructed by transforming strain 168 (trpC2) with W168 (tryptophan prototroph) chromosomal DNA, selecting for tryptophan prototrophy, and screening for rifampin sensitivity (0.25 μg/ml). Strains PLBS339 and PLBS340 were constructed by transforming PLBS338 with chromosomal DNA from CYBS400::pJS648 (amyE::P_trpPtrpP′-′lacZ Cm^r) and PGBS11 (amyE::P_pabpabB-pabA′-′lacZ Cm^r), respectively. Selection was for chloramphenicol resistance (5 μg/ml). Integration into amyE was confirmed by screening for the absence of amylase activity (30). Transforming strains PLBS339 and PLBS340 with chromosomal DNA from strain BG4233 (ΔmtrB) resulted in strains PLBS341 and PLBS342, respectively. Selection was for 5-fluorotryptophan resistance (200 μg/ml). rho::neo from pYH14 was used to replace the wild-type rho allele in strains PLBS339, PLBS340, PLBS341, and PLBS342 to yield strains PLBS343, PLBS344, PLBS345, and PLBS346, respectively. Selection was for kanamycin resistance (10 μg/ml).

β-Galactosidase assays. B. subtilis cultures were grown in minimal-acid casein hydrolysate medium containing 5 μg of chloramphenicol/ml in the absence or presence of 200 μM tryptophan. Growth medium for rho mutant strains also contained 10 μg of kanamycin/ml. The cells were harvested during late exponential phase. Aliquots were then assayed for β-galactosidase activity as previously described (14).

Gel mobility shift assay.TRAP was purified as previously described (39). Quantitative gel mobility shift assays used to examine TRAP-RNA interactions followed a previously published procedure (39). RNA was synthesized in vitro using the Ambion MEGAscript kit. Linearized plasmid pPB31 was used as the template to generate pabA RNA containing nt −72 to +109 relative to the AUG start codon. Linearized plasmid pHZB6 was used to synthesize trpP RNA containing nt −82 to +55 relative to the AUG start codon (+1 to +137 relative to the start of transcription). Gel-purified transcripts were dephosphorylated and subsequently 5′-end labeled with T4 polynucleotide kinase and [γ-³²P]ATP. Labeled transcripts were gel purified, ethanol precipitated, and suspended in Tris-EDTA (TE). Transcripts were renatured by heating to 80°C for 1 min followed by slow cooling prior to use in binding reactions.

Binding reaction mixtures (8 μl) contained 50 mM Tris-acetate (pH 8.0), 4 mM magnesium acetate, 5 mM dithiothreitol (DTT), 10% glycerol, 0.2 mg of Escherichia coli tRNA/ml, 400 U of RNasin (Promega)/ml, 0.1 nM 5′-end-labeled pabA RNA or 0.5 nM 5′-end-labeled trpP RNA, 1.2 mM l-tryptophan, purified TRAP (various concentrations), and 0.1 mg of xylene cyanol/ml. Competition assays also contained unlabeled RNA competitor (see Results for details). TRAP-RNA complexes were allowed to equilibrate at 37°C for 20 min. Samples were then fractionated on native 6% (pabA) or 8% (trpP) polyacrylamide gels in 375 mM Tris-HCl (pH 8.8), 5% glycerol, and 1 mM EDTA. Radioactive bands were visualized using a PhosphorImager (Molecular Dynamics). Free and bound RNA species were quantified using ImageQuant (Molecular Dynamics), and the apparent equilibrium binding constants (K_d) of TRAP-RNA complexes were calculated by fitting to the simple binding equation as previously described (39).

Filter binding assay.The labeled transcripts used in filter binding reactions were identical to those described for the gel mobility shift assay. Filter binding assays were carried out using a 96-well dot blot apparatus by modifying a two-filter method reported previously (36). Filters were equilibrated in 40 mM Tris-HCl (pH 8.0) and 250 mM KCl prior to use. Binding reaction mixtures (45 μl) contained 40 mM Tris-HCl (pH 8.0), 250 mM KCl, 5 mM DTT, 0.2 mg of E. coli tRNA/ml, 400 U of RNasin/ml, 0.1 nM 5′-end-labeled RNA, 1.2 mM l-tryptophan, and purified TRAP (various concentrations). TRAP-RNA complexes were allowed to equilibrate at 37°C for 20 min. Samples (40 μl) were then filtered and subsequently rinsed twice with 100 μl of 40 mM Tris-HCl (pH 8.0) and 250 mM KCl. Radioactive spots were visualized and quantified as described for the gel mobility shift assay.

Footprint assay.5′-End-labeled trpP RNA used in this analysis was generated as described for the gel mobility shift assay. Titrations of RNase T₁ (Roche), RNase T₂ (Sigma), RNase A (Ambion), and RNase V₁ (Pierce) were performed to optimize the amount of each reagent to prevent multiple cleavages in any one transcript. RNA suspended in TE was renatured by heating to 80°C for 1 min followed by slow cooling. Binding reaction mixtures (10 μl) contained 40 mM Tris-HCl (pH 8.0), 30 mM KCl, 8 mM MgCl₂, 32.5 ng of total yeast RNA, 100 μg of bovine serum albumin/ml, 7.5% glycerol, 1 mM l-tryptophan, 2 nM trpP RNA, and various concentrations of TRAP. Reaction mixtures were incubated for 30 min at 37°C to allow TRAP-trpP RNA complex formation prior to the addition of RNase T₁ (8 × 10⁻³ U/μl), RNase T₂ (2 × 10⁻⁴ U/μl), RNase V₁ (3 × 10⁻⁵ U/μl), or RNase A (10⁻⁶ μg/μl). Incubation was then continued for 15 min at 37°C. Reactions were terminated by the addition of 5 μl of stop solution (95% formamide, 20 mM EDTA, 0.025% sodium dodecyl sulfate [SDS], 0.025% xylene cyanol, 0.025% bromophenol blue), and samples were fractionated through 6% sequencing gels. Radiolabeled bands were visualized by phosphorimagery.

Toeprint assay.Toeprint assays were performed by modifying published procedures (10, 14, 19). trpP RNA was synthesized using linearized pHZB6 as template. Gel-purified RNA (250 nM) in TE was hybridized to a ³²P-end-labeled DNA oligonucleotide (500 nM) complementary to the 3′ end of the transcript by heating to 80°C for 1 min and slow cooling. Toeprint assays were carried out with 3 μM TRAP and 1 mM l-tryptophan and/or 10 pmol of E. coli 30S ribosomal subunits and 50 pmol of E. coli tRNA^fMet (Sigma). Toeprint reaction mixtures (10 μl) contained 2 μl of the hybridization mixture, a 375 μM concentration of each deoxynucleoside triphosphate, 10 mM DTT, and 100 μg of bovine serum albumin/ml in toeprint buffer (50 mM Tris-HCl [pH 8.3], 75 mM KCl, 3 mM MgCl₂). TRAP toeprint reactions were incubated for 30 min at 37°C to allow TRAP-trpP RNA complex formation. 30S ribosomal subunit toeprint reactions were performed by incubating RNA with 30S ribosomal subunits and tRNA^fMet as described previously (19). After the addition of 10 U of Moloney murine leukemia virus reverse transcriptase (U.S. Biochemical), incubation was continued at 37°C for 15 min. Reactions were terminated by the addition of 6 μl of stop solution (see “Footprint assay”). Samples were fractionated through 6% sequencing gels. Sequencing reactions were performed using pHZB6 as the template and the same end-labeled DNA oligonucleotide as a primer. Radiolabeled bands were visualized by phosphorimagery.

RNA-directed cell-free translation.The trpP′-′gfp translational fusion transcript used for this analysis was synthesized using linearized pYH34 as template. A TRAP-deficient B. subtilis S-30 extract was prepared from strain CYBS306 by following a published procedure (12). Cell-free translation reactions were carried out by modifying published procedures (14, 16, 29). The S-30 extract was preincubated with RNase-free DNase I for 15 min at 37°C to remove endogenous mRNA and DNA. Reaction mixtures (24 μl) contained 60 mM Tris-HEPES (pH 7.5), 60 mM NH₄Cl, 15 mM MgCl₂, 12 mM KCl, 0.5 mM EGTA, 5 mM DTT, 2 mM ATP, 0.6 mM GTP, 0.08 mM calcium folinate, 4 μg of aprotinin/ml, 4 μg of leupeptin/ml, 4 μg of pepstatin A/ml, 4 μl of S-30 extract (12 μg of total protein), 800 U of DNase I/ml, 500 U of RNasin/ml, 10 mM phosphoenolpyruvate, 35 U of pyruvate kinase/ml, 0.4 mg of E. coli tRNA/ml, 100 nM trpP′-′gfp mRNA, 10 μCi of [³⁵S]methionine, 5 mM potassium glutamate, 5 mM glutamine, and a 0.1 mM concentration of each of the other amino acids except tryptophan. Tryptophan was added at a concentration of 1 mM when used. Reaction mixtures were incubated for 25 min at 37°C and terminated by adding 6 μl of SDS-stop buffer (125 mM Tris-HCl [pH 6.8], 5% SDS, 25% glycerol, 2% 2-mercaptoethanol, and 12.5 mg of bromophenol blue/ml). Aliquots (10 μl) were heated at 95°C for 5 min, and proteins were fractionated on SDS-14% polyacrylamide gels. Radiolabeled bands were visualized by phosphorimagery and quantified using ImageQuant.

TRAP-mediated regulation of trpP and pabA expressionPrevious results demonstrated that TRAP regulates translation of trpP (27) and pabA (16, 41). We compared TRAP-mediated regulation of these two genes by measuring β-galactosidase activities in strains containing pabA′-′lacZ or trpP′-′lacZ translational fusions that were otherwise isogenic. The effect of exogenous tryptophan was assessed from the ratio of expression when cells were grown in the absence or presence of tryptophan (−Trp/+Trp ratio). TRAP-dependent regulation was observed for both fusions with inhibition ratios (−Trp/+Trp) of 5 for pabA and 150 for trpP (Table 2). As expected, when these experiments were repeated in a TRAP-deficient genetic background (ΔmtrB), regulation in response to tryptophan was abolished (Table 2). The extent of TRAP-mediated regulation in vivo was determined by comparing expression in ΔmtrB strains with that in wild-type strains grown in the presence of tryptophan. The TRAP-dependent inhibition ratios for pabA and trpP were 16 (545/34) and 900 (1,200/1.3), respectively (Table 2). These results indicate that TRAP-mediated regulation of trpP expression is more tightly controlled than that of pabA.

TRAP binding to trp operon readthrough transcripts regulates TrpE synthesis by promoting formation of an RNA secondary structure that sequesters the trpE S-D sequence (14). It was also shown that Rho causes transcriptional polarity of the trp operon under conditions that promote translation control (tryptophan excess); expression was elevated in rho mutant strains (40). Since both trpP and pabA are regulated by TRAP at the level of translation, we carried out experiments to determine whether expression of these two genes was altered in a rho mutant background. Expression from the trpP′-′lacZ fusion was unchanged in the Rho-deficient strain, whereas expression from the pabA′-′lacZ fusion was reduced twofold in the rho mutant strain (Table 2). When expression from each fusion was examined in mtrB rho double mutant strains, β-galactosidase activity was similar to that observed in the mtrB single mutant strains. These results indicate that transcriptional polarity, if present, is not sufficient to provide regulation.

TRAP binds to trpP and pabA RNA with comparable affinities.The mechanism of TRAP-mediated inhibition of PabA synthesis was previously characterized; however, the affinity of TRAP for the pabA transcript was not examined. To characterize further the interaction of tryptophan-activated TRAP with pabA mRNA, we performed quantitative gel mobility shift assays with a pabA transcript containing nt −72 to +109 relative to its AUG start codon (Fig. 2A). Nonlinear least-squares analysis of these data yielded an estimated equilibrium binding constant (K_d) of 14 nM TRAP. Filter binding studies were also carried out as an alternative method to measure the affinity of tryptophan-activated TRAP for pabA RNA. This method yielded a K_d value of 33 nM TRAP. For comparison, the K_d value for the TRAP-trp leader RNA interaction was found to be approximately 7 nM by gel mobility shift analysis (data not shown) and 1 nM by filter binding (42). Similar gel mobility shift and filter binding assays were also carried out to examine TRAP interaction with a trpP transcript containing nt −82 to +55 relative to the AUG start codon (+1 to +137 relative to the start of transcription). In this case, estimated K_d values of 51 and 31 nM TRAP were obtained for gel mobility shift (Fig. 2C) and filter binding, respectively. It is important to note that while the affinities of TRAP for pabA and trpP RNA were similar, the extent of TRAP-mediated regulation of the two genes differed considerably (Table 2). Possible explanations for this apparent discrepancy will be addressed in the Discussion.

• Open in new tab
• Download powerpoint

FIG. 2.

Gel mobility shift analysis of TRAP complexed with pabA and trpP transcripts. 5′-End-labeled RNA was incubated with the concentration of TRAP shown at the bottom of each lane. Gel shift assays were performed in the absence or presence of various competitor RNAs. The concentration of each competitor RNA is indicated at the bottom of the corresponding lane. The positions of bound and free RNA are shown. (A) TRAP-pabA RNA complex formation. (B) Competition assay for TRAP-pabA RNA complex formation. (C) TRAP-trpP RNA complex formation. (D) Competition assay for TRAP-trpP RNA complex formation.

While the above binding studies determined the affinity of TRAP for the pabA and trpP transcripts, the specificities of the TRAP-pabA RNA and TRAP-trpP RNA interactions were investigated by performing competition experiments with specific and nonspecific unlabeled RNA competitors. The concentration of TRAP used in these experiments was chosen such that about 75% of the labeled RNA was shifted in the absence of competitor RNA (Fig. 2B, second lane, and D, second lane). As expected, pabA RNA was an effective competitor for TRAP-pabA RNA interaction, whereas a transcript derived from pTZ19R vector sequences was not (Fig. 2B). The specificity of TRAP-trpP RNA interaction was investigated by performing competition experiments with specific (trpP, trpL, and pabA) and nonspecific unlabeled competitors (Fig. 2D). The trp leader (trpL) transcript was the most effective competitor, while competition levels with the pabA and trpP transcripts were comparable to one another. As expected, the nonspecific competitor (pTZ19R) was unable to compete for TRAP-trpP RNA complex formation. These results establish that TRAP binds specifically to a trpP transcript containing its S-D sequence and translation initiation region.

TRAP binds to nine triplet repeats in the trpP transcript.Previous footprinting studies demonstrated that TRAP binds to 11 repeats in the trp operon leader transcript (7 GAG and 4 UAG) and to 9 repeats in the pabA message (7 GAG, 1 UAG, and 1 AAG). In general, the spacer residues separating adjacent repeats were not protected by bound TRAP (7, 16). Other published results have indicated that the affinity of TRAP is highest for GAG repeats (GAG > UAG > AAG > CAG) and that optimal spacing between repeats was 2 nt, with pyrimidines generally favored over purines (5, 8, 11, 37).

It was previously pointed out that nine triplet repeats were present in the trpP transcript that overlapped and surrounded the trpP S-D sequence and translation initiation region (five GAG, three UAG, and one AAG) (Fig. 1) (27). However, in several cases the length of the spacers was suboptimal, ranging in size from 1 to 14 nt. Footprint experiments were carried out to determine which of the nine repeats constituted authentic TRAP recognition targets. We used partial RNase T₁ (cleaves following single-stranded G residues), RNase T₂ (preferentially cleaves following single-stranded A residues), and RNase A (cleaves following single-stranded pyrimidines) digestion to probe the TRAP-trpP RNA complex. The use of these three reagents would theoretically be capable of cleaving every nucleotide in the trpP transcript. The results of the footprint analysis are shown in Fig. 3 and are summarized in Fig. 4. RNase T₁ cleaved the majority of the G residues in the nine repeats in the absence of bound TRAP. Notable exceptions were the Gs in repeats 2 and 7. Tryptophan-activated (bound) TRAP protected G residues in repeats 1, 3, 4, 6, and 8 from RNase T₁ cleavage, whereas bound TRAP caused enhanced cleavage of the G residue in repeat 9. RNase T₂ cleaved the central A residue in repeats 1, 3, 5, 6, 8, and 9 in the absence of bound TRAP. Importantly, TRAP protected all of these A residues from RNase T₂ cleavage. RNase A cleaved the U residues in repeats 4, 7, and 9 in the absence of bound TRAP, while bound TRAP prevented cleavage at these three nucleotides. Thus, with the exception of repeat 2, the use of these three reagents allowed us to identify TRAP-dependent protection of each repeat. In addition to protecting residues within the triplet repeats, bound TRAP protected 7 out of 39 spacer nucleotides. In contrast, bound TRAP resulted in enhanced cleavage of seven spacer residues (Fig. 3 and 4). Thus, as was previously observed for the TRAP binding targets in the trp operon leader and in pabA, TRAP generally protected the residues in the triplet repeats but not those in the spacers separating the repeats.

• Open in new tab
• Download powerpoint

FIG. 3.

TRAP-trpP RNA footprint analysis. trpP RNA was treated with RNase T₁, RNase T₂, or RNase A in the absence or presence of TRAP. The concentrations of TRAP used were 0, 0.25, 0.5, 1, and 2 μM. Partial alkaline hydrolysis (OH) and RNase T₁ digestion (T1) ladders, as well as control (C) lanes in the absence of RNase treatment, are shown. The RNase T₁ ladder was generated under denaturing conditions so that every G residue in the transcript could be visualized. Residues in which RNase cleavage was reduced (−) or enhanced (+) in the presence of TRAP are marked. Apparent TRAP-dependent RNase T₁ cleavage of C90 is marked with an arrowhead. The relative positions of the triplet repeats (1 to 9), as well as the S-D sequence and translation start codon (AUG), are shown. Numbering at the right of each panel is relative to the start of trpP transcription.

• Open in new tab
• Download powerpoint

FIG. 4.

Summary of the trpP footprint and toeprint results. This figure is adapted from the data presented in Fig. 3 and 5. The composite RNase T₁, RNase T₂, and RNase A footprint shows the residues in which cleavage was reduced (−), enhanced (+), or unaffected by bound TRAP (^•). Residues that were not cleaved in the absence or presence of bound TRAP are indicated with an asterisk. Positions of TRAP and 30S ribosomal subunit toeprints are indicated by vertical arrows and inverted arrowheads, respectively. The positions of the trpP S-D sequence and AUG start codon are underlined. Triplet repeats 1 through 9 are shown in parentheses. An inverted repeat that is capable of forming an RNA secondary structure is indicated with horizontal arrows. Numbering is from the start of trpP transcription.

The RNA segment between nt 40 and 55 was refractory to RNase cleavage, suggesting that an RNA structure was present in the transcript. Computer predictions using MFOLD version 3.1 (23, 43) identified a large secondary structure extending from G18 through C54. Interestingly, the first triplet repeat is present in the loop of this hairpin and was extensively cleaved by RNases T₁ and T₂. Since the second trinucleotide repeat is present in the 3′ half of the stem within the secondary structure, we used RNase V₁ (specific for double-stranded RNA) as a probe to determine whether TRAP interacted with this triplet. RNase V₁ cleavage of G51 within the second repeat was observed in the absence of TRAP, whereas bound TRAP protected this residue from RNase V₁ cleavage (data not shown). Because RNA structure is known to inhibit TRAP binding to triplet repeats (8, 11, 29, 37), it is likely that breathing of this structure allows TRAP interaction with this repeat. Thus, the footprint results establish that TRAP interacts with all nine repeats that surround and overlap the trpP S-D sequence and translation initiation region. Because the TRAP-RNA cocrystal structure shows TRAP interacting with 11 repeats (1), we presume that TRAP can simultaneously interact with all 9 repeats in the trpP transcript.

TRAP inhibits TrpP synthesis by blocking ribosome binding.The position of the TRAP binding target in the trpP transcript suggested a model in which bound TRAP would block ribosome access to the trpP ribosome binding site. We carried out TRAP and 30S ribosomal subunit toeprint experiments to test this prediction. The presence of bound TRAP or a 30S ribosomal subunit would block primer extension by reverse transcriptase, resulting in a toeprint band at a position near the 3′ boundary of the bound ribosome or TRAP. The toeprint results are presented in Fig. 5 and are summarized in Fig. 4. Prominent tryptophan-dependent TRAP toeprints were observed at positions G85, A96, and U102, corresponding to positions just downstream from repeat 7, within repeat 8, and just downstream from repeat 9, respectively. An additional TRAP-dependent toeprint band at U12 was also observed. Since stable RNA secondary structures are capable of blocking extension by reverse transcriptase (e.g., see reference 14), it appears that bound TRAP promotes formation of an RNA structure near the 5′ end of the transcript used in this analysis. In the case of 30S ribosomes, a cluster of three consecutive tRNA^fMet-dependent toeprint bands was observed that centered at U99 (Fig. 5). We also carried out toeprint experiments to determine whether TRAP could inhibit ribosome binding. When TRAP was bound to the trpP transcript prior to the addition of ribosomes and tRNA^fMet, the TRAP toeprints were observed while the ribosome toeprints were not. These results demonstrate that bound TRAP inhibits ribosome binding. We also observed a prominent toeprint in all lanes at C54, which is just downstream from the second triplet repeat. This result provides additional evidence for an RNA structure extending from G18 through C54.

• Open in new tab
• Download powerpoint

FIG. 5.

TRAP and 30S ribosomal subunit toeprint analysis of trpP RNA. The presence of TRAP, tryptophan (Trp), and/or 30S ribosomal subunits plus tRNA^fMet (30S Rib) is shown at the top of each lane. TRAP was added to the reaction mixture corresponding to the rightmost lane prior to the addition of 30S ribosomal subunits and tRNA^fMet. Arrows indicate bands corresponding to TRAP and 30S ribosomal subunit toeprints. C54 corresponds to an RNA structural toeprint in each lane, whereas U12 corresponds to a TRAP-dependent RNA structural toeprint. Positions of the trpP S-D sequence, the AUG initiation codon, and the triplet repeats (1 to 9) are shown at the left. Sequencing lanes to reveal U, G, C, or A residues are indicated.

Since the toeprint results indicated that bound TRAP competes with ribosomes for binding to the trpP transcript, RNA-directed cell-free translation experiments using a TRAP-deficient B. subtilis S-30 extract were performed to determine whether TRAP could inhibit synthesis of a TrpP-green fluorescent protein (GFP) fusion peptide from a preexisting mRNA template. This fusion was constructed such that it did not contain any tryptophan codons, so that in vitro translation could be performed in the absence or presence of tryptophan. A major protein species that was dependent on the addition of the trpP′-′gfp transcript was produced (Fig. 6). No translation product corresponding to the fusion peptide was observed without the addition of the trpP′-′gfp transcript. In the presence of tryptophan, addition of increasing concentrations of TRAP to the translation system resulted in a corresponding decrease in the level of TrpP-GFP synthesis. Inhibition of translation was not observed in the absence of added tryptophan and/or TRAP. In conjunction with the footprint and toeprint results described above, the cell-free translation experiments demonstrated that TRAP binding to the trpP message inhibits TrpP synthesis by blocking ribosome binding.

• Open in new tab
• Download powerpoint

FIG. 6.

Effect of tryptophan-activated TRAP on RNA-directed cell-free translation of trpP′-′gfp mRNA. A TRAP-deficient S-30 extract was prepared from B. subtilis strain CYBS306. Reactions were carried out with various concentrations of purified TRAP in the absence (−) or presence (+) of trpP′-′gfp transcript and/or 1 mM tryptophan (Trp). (A) TrpP-GFP translation products analyzed by SDS-polyacrylamide gel electrophoresis. The position of the full-length fusion protein is shown with an arrow. (B) Relative level of full-length TrpP-GFP polypeptide synthesis as a function of TRAP concentration. The level of polypeptide synthesis in the absence of TRAP was set to 1.0.

TRAP-mediated regulation of the trpEDCFBA operon.TRAP binding to the nascent trp leader transcript plays a central role in regulating expression of the trpEDCFBA operon by transcription attenuation (3, 9, 22, 26, 38) and translation control mechanisms (14, 22, 24, 40). These dual TRAP-dependent regulatory mechanisms result in approximately 2,000-fold regulation of the trpEDCFBA operon. When cells are grown 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 termination in the leader region (transcription attenuation); however, since termination is never 100% efficient, in some instances RNA polymerase will escape termination despite TRAP binding. This would result in a TRAP-bound readthrough transcript that would sequester the trpE S-D sequence in the trpE S-D blocking hairpin (14). Inhibition of translation might also occur when cells are initially grown under limiting tryptophan conditions. In this case, a relatively high percentage of TRAP molecules would not be activated, resulting in transcription readthrough. Eventually, either by synthesis or transport, a sufficient level of tryptophan would build up in the cell and activate TRAP. Tryptophan-activated TRAP would then bind to the trp leader and promote formation of the trpE S-D blocking hairpin (14).

Another protein called anti-TRAP (AT) plays a role in regulating tryptophan biosynthesis and transport. AT is a zinc-containing protein that antagonizes TRAP activity by competing with TRAP's RNA-binding surface via direct protein-protein interaction (33-35). Transcription of the AT operon is activated by uncharged tRNA^Trp via a T-box antitermination mechanism (18, 28). Interestingly, translation of AT is also regulated by uncharged tRNA^Trp. A short 10-amino-acid leader peptide coding region containing three consecutive Trp codons just precedes the AT structural gene. It appears that low levels of charged tRNA^Trp cause the ribosome to stall at one or more of these Trp codons, which increases AT synthesis (13). Since expression of the gene encoding AT responds to the accumulation of uncharged tRNA^Trp, B. subtilis regulates tryptophan biosynthesis and transport by sensing the levels of both tryptophan and uncharged tRNA^Trp in the cell.

TRAP-mediated translation control of trpP and pabA.A putative TRAP binding site was identified in the trpP transcript that contained as many as nine triplet repeats (five GAG, three UAG, and one AAG) (Fig. 1) (27). While the central five repeats in the trpP transcript are separated by optimal 2-nt spacers, the first, second, seventh, and eighth spacers contain 10, 6, 14, and 1 nt, respectively. Our footprint results indicate that all nine of these repeats are involved in TRAP-trpP RNA interaction (Fig. 3 and 4). However, TRAP toeprints were observed within repeat 8 and following repeats 7 and 9 (Fig. 4 and 5), suggesting that TRAP interaction with the eighth and ninth repeats is relatively weak; presumably, reverse transcriptase was capable of disrupting TRAP interaction with these two repeats. Perhaps the suboptimal spacing between repeats 7 and 8 (14 nt) and between repeats 8 and 9 (1 nt) reduces the affinity of TRAP for the last two triplets. The position of the trpP TRAP binding site suggested a model in which bound TRAP would block ribosome binding. Our toeprint (Fig. 5) and in vitro translation (Fig. 6) results demonstrate that bound TRAP inhibits TrpP synthesis by blocking ribosome binding.

Since the trpP translation control mechanism is similar to what was previously identified for pabA (16, 41), we compared the extent of TRAP-mediated translation control of the two genes and found that TRAP exhibited much tighter control of TrpP synthesis (Table 2). One reasonable explanation for the less-extensive control of pabA expression is to allow some PabA synthesis in the presence of tryptophan-activated TRAP to maintain folic acid biosynthesis. Despite the considerable difference in TRAP-dependent translation control of these two genes, results from our gel mobility shift (Fig. 2) and filter-binding studies indicate that the affinity of TRAP for these two transcripts is similar. These findings imply that the two translation initiation regions are designed such that bound TRAP has a greater effect on translation initiation for trpP. What might be responsible for the difference in TRAP-dependent regulation of these two genes?

The TRAP binding targets in pabA (seven GAG, one UAG, and one AAG) and trpP (five GAG, three UAG, and one AAG) each contain nine repeats. While the optimal spacing between repeats is 2 nt, several of the spacers in each transcript contain suboptimal spacing. It was previously pointed out that the trpP transcript contains five consecutive repeats with optimal spacing that overlap and surround its S-D sequence. In comparison, the pabA transcript contains two stretches of four repeats with optimal spacing, one of which overlaps and surrounds its S-D sequence (27). Of particular interest is the finding that the last two repeats in the trpP transcript are within the coding region, whereas all nine repeats in pabA are upstream from the start codon. Perhaps the relative arrangement of repeats in these two transcripts with respect to their cognate S-D sequences and translation initiation regions is at least partly responsible for TRAP having tighter control over trpP expression. Thus, extending the TRAP binding site into the coding region may be more effective at inhibiting ribosome binding.

Unlike the case for attenuation, the timing of TRAP binding does not appear to be critical for translation inhibition. Instead, translation control of pabA and trpP involves a competition between TRAP and ribosome binding. TRAP would only inhibit translation while it is bound to a transcript. In the absence of bound TRAP, a ribosome could bind and initiate translation. Once the ribosome clears the translation initiation region, either TRAP or another ribosome could bind to the transcript. Thus, a difference in the relative association or dissociation rates of TRAP for the two transcripts could contribute to the difference in translation inhibition that was observed for these two genes.

It is also possible that the gene arrangement of the two operons contributes to differences in TRAP-mediated control. pabA is the second gene in the folate operon, with pabB just upstream. Interestingly, the pabB translation stop codon lies within the pabA S-D sequence (Fig. 1). Since only two triplet repeats in the pabA TRAP binding site lie downstream from the pabB stop codon, it is possible that translation of pabB results in ribosome-mediated displacement of bound TRAP. This would result in a pabA S-D sequence that is transiently free of bound TRAP. Thus, ribosome-mediated displacement of bound TRAP might allow translation initiation of pabA. Since trpP is a single-gene operon, ribosome-mediated displacement of bound TRAP would not be a factor.

A putative TRAP binding site was also identified that overlaps the S-D sequence and translation initiation region of ycbK, a gene of unknown function (Fig. 1) (28). Thus, it is likely that TRAP regulates YcbK synthesis by a translation control mechanism similar to that of pabA and trpP. It is interesting that the stop codon for rtpA, the gene that encodes AT, overlaps the ycbK S-D sequence. However, in this case only one of the triplet repeats is upstream of the S-D sequence, while six are present in the ycbK coding region (Fig. 1). Thus, one might predict that ribosome-mediated displacement of bound TRAP would be less pronounced than for pabA, while the repeats in the coding sequence may increase the effectiveness of bound TRAP in blocking ribosome binding.

It is well established that RNA structure can inhibit TRAP binding (8, 11, 29, 37). In contrast, TRAP interaction with the 5′ stem-loop (5′SL) that forms at the extreme 5′ end of the trp leader transcript increases the affinity of TRAP for the trp leader by an unknown mechanism (15). Our footprint and toeprint results, combined with computer modeling, revealed an RNA secondary structure extending from nt 18 to 54 of the trpP leader transcript. The first triplet repeat is present in the loop of the hairpin, while the second repeat is in the 3′ half of the stem (Fig. 4). Interestingly, the trp leader 5′SL contains a GAG in the loop of the hairpin and an AAG in the 3′ half of the stem (15, 32); however, neither of these triplets is part of the 11-repeat TRAP-binding target identified by footprinting (7). A systematic mutational analysis of the 5′SL indicated that the overall structure was important for TRAP-dependent regulation. While changing the GAG in the loop of the 5′SL to GUG had only a small effect on TRAP-mediated regulation, mutating the AAG in the stem to AAC resulted in a considerable reduction in TRAP's ability to regulate trp operon expression (32). Thus, the 5′SL contains both structural and sequence elements that participate in TRAP binding. Perhaps in addition to contributing two triplet repeats, the trpP leader hairpin may provide a structural element that participates in TRAP interaction. Since the pabA transcript does not contain a similar structure, it is possible that the trpP RNA hairpin contributes to the difference in TRAP-mediated regulation of these two genes.