Research Article

Inhibition of the B. subtilis Regulatory Protein TRAP by the TRAP-Inhibitory Protein, AT

See allHide authors and affiliations

Science  14 Sep 2001:
Vol. 293, Issue 5537, pp. 2057-2059
DOI: 10.1126/science.1062187


An anti-TRAP (AT) protein, a factor of previously unknown function, conveys the metabolic signal that the cellular transfer RNA for tryptophan (tRNATrp) is predominantly uncharged. Expression of the operon encoding AT is induced by uncharged tRNATrp. AT associates with TRAP, the trp operon attenuation protein, and inhibits its binding to its target RNA sequences. This relieves TRAP-mediated transcription termination and translation inhibition, increasing the rate of tryptophan biosynthesis. AT binds to TRAP primarily when it is in the tryptophan-activated state. The 53-residue AT polypeptide is homologous to the zinc-binding domain of DnaJ. The mechanisms regulating tryptophan biosynthesis in Bacillus subtilis differ from those used by Escherichia coli.

In microorganisms with the capacity to synthesize amino acids needed for protein synthesis, each amino acid and/or its corresponding charged or uncharged tRNA is often recognized as a regulatory signal. Many bacterial species sense both tryptophan and tRNATrp in regulating expression of the operons responsible for tryptophan biosynthesis. E. coli, for example, forms a tryptophan-activated repressor that regulates transcription initiation (1). It also responds to the accumulation of uncharged tRNATrp by reducing transcription termination in the leader region of the trp operon (1). In B. subtilis, coordinate expression of seven trp genes is required for tryptophan biosynthesis. Six of these are clustered in thetrp operon, trpEDCFBA, a contiguous segment of a 12-gene aromatic supraoperon (2). The seventh gene,trpG, is located in the unlinked folate operon (2, 3). Transcription of the trpoperon of B. subtilis is regulated by attenuation, by the tryptophan-activated trp RNA-binding attenuation protein, TRAP (2–9). Active TRAP binds to a specific segment of the nascent trp operon leader transcript, promoting the formation of an RNA terminator structure that causes transcription termination (10). Activated TRAP also binds to the ribosome-binding site of trpG messenger RNA and inhibitstrpG translation (11, 12).

In studies with B. subtilis, uncharged tRNATrp has also been implicated in regulation of the genes of tryptophan biosynthesis (13–15). The accumulation of uncharged tRNATrp in a temperature-sensitive tryptophanyl-tRNA synthetase mutant (trpS1) leads to trp operon overexpression (13, 14). In an attempt to explain this observation, an operon was identified, yczA-ycbK, that appeared to be responsible for the trpS1 effect (15). Induction of yczA-ycbK expression by uncharged tRNATrp was shown to occur via the T-box transcription antitermination mechanism in which an uncharged tRNA specifically pairs with leader RNA (15,16). Induction of yczA-ycbK expression led to inactivation of TRAP, explaining the trpS1 effect (15), but how this occurred was not established (15). Successive deletions within the yczA-ycbKoperon suggested that yczA expression could be responsible for TRAP inactivation (15). In this research article, we examine this possibility and show that the product of yczA, AT, does in fact function to inhibit TRAP activity.

Overexpression of yczA in vivo increasestrp operon expression.

We examined yczA function in vivo by overexpressing this gene in B. subtilis. We also overexpressed a yczA derivative in which the yczAstart codon was replaced by a stop codon. Overexpression was achieved by using an isopropyl-β-d-thiogalactopyranoside (IPTG)-inducible plasmid construct (17);trp operon expression was measured with an integratedtrp promoter-leader-trpE-lacZ translational fusion reporter (15, 18, 19). Overexpression of AT completely abolished TRAP regulation oftrp operon expression (Table 1), in the presence or absence of added tryptophan. When we examined the strain with the construct in which theyczA start codon was replaced by a stop codon, IPTG addition had no effect on TRAP function. The high level of trp operon expression associated with AT overproduction was not increased in a strain bearing a deletion in the TRAP coding gene (ΔmtrB) (Table 1), implying that AT's effect is dependent on the presence of a functional TRAP protein. We also observed that AT overproduction increased trpG-lacZ expression (20). AT overexpression did not affect mtrB-lacZexpression, establishing that AT does not regulate TRAP synthesis (20). Thus, AT is presumed to act by inhibiting the ability of TRAP to interact with its target RNA molecules and to regulatetrp gene and operon expression.

Table 1

In vivo overexpression of yczA completely abolishes TRAP regulation of trp operon expression.yczA overexpression was achieved by cloning yczAinto the B. subtilis replicative plasmid pDG148, immediately downstream of the IPTG-inducible spac promoter. B. subtilis strains carrying a trppromoter-leader-trpE-lacZ fusion (15) were transformed with different plasmids and β-galactosidase assays were performed, with and without IPTG addition. ΔmtrB, strain with a deletion of mtrB, the TRAP structural gene (7); pDG148, parental plasmid with no insert (control); pDGyczA, plasmid with a wild-type yczA insert; pDGSTOPyczA, plasmid with a yczA insert in which a stop codon replaces the yczA start codon.

View this table:

Purification and characterization of AT.

The AT polypeptide contains 53 amino acid residues. AT was overexpressed, by using an E. coli expression system (21), and the protein was purified by heat treatment, ammonium sulfate fractionation, ion-exchange chromatography, and sizing column chromatography (22). Estimates of molecular mass for the intact protein, based on the use of a Sephadex G-75 gel-filtration column, suggested that AT is a 28-kD species. This would correspond to a molecule consisting of five copies of the 5.6-kD AT polypeptide. Matrix-assisted laser desorption/ionization–time of flight (MALDI-TOF) mass spectrometry analyses (20) are consistent with this estimate. Mass measurements showed one major peak, corresponding to the AT monomer, with a molecular mass of 5648 daltons, and a series of minor peaks that represent different oligomers ranging in composition from two to six copies of the monomer. The mass spectral results are supported by cross-linking experiments performed with glutaraldehyde: the same range of oligomers was observed. We conclude that AT is a multimeric protein composed of five or six identical 5.6-kD subunits.

A BLASTP search of the nonredundant protein database revealed that the AT amino acid sequence is similar to that of the cysteine-rich zinc-binding domain of the chaperone protein DnaJ from a variety of organisms (23, 24). In particular, except for one G, the characteristic C-X-X-C-X-G-X-G sequence motif is conserved (Fig. 1) (25). This motif is repeated four times in the DnaJ polypeptide and twice in the AT polypeptide. Studies performed withE. coli DnaJ indicate that its cysteine-rich domain may be involved in binding to denatured protein substrates (23).

Figure 1

Multiple sequence alignment of AT with the cysteine-rich domain of DnaJ from different microorganisms (Bs,Bacillus subtilis; Ec, Escherichia coli; Sce,Saccharomyces cerevisiae; Pfa, Plasmodium falciparum; Meth, Meth-anobacterium thermoautotrophicum). The Cys and Gly residues of the C-X-X-C-X-G-X-G motif are highlighted in green and orange, respectively (25). The alignment was created by using Clustal/Jalview (30).

AT inhibits TRAP-mediated transcription termination and RNA band-shifting, in vitro.

Purified AT was examined in an in vitro transcription termination and readthrough assay (26), and AT was observed to completely abolish TRAP-dependent transcription termination (Fig. 2). The sensitivity of this inhibition was not affected by the preincubation time of TRAP with AT, implying that AT is not acting catalytically. Rather, it most likely forms a complex that inhibits TRAP's ability to bind RNA. AT could act either by binding target RNA directly, blocking TRAP's access to its binding sites, or by interacting with TRAP and interfering with its ability to bind RNA. RNA gel-retardation assays (27) discriminated between these two possibilities (Fig. 3). AT did not bind totrp leader RNA, rather, it prevented TRAP from binding to this RNA. Higher AT/TRAP molar ratios were required to achieve an appreciable effect in the gel-retardation assay (Fig. 3), compared with the transcription termination and readthrough assay (Fig. 2). This is most likely due to the affinity of AT for TRAP, because a concentration of TRAP was used in band-shift assays (23 nM) that was 1/15th that used in in vitro transcription analyses (340 nM). Moreover, the different requirements could reflect the nature of the two procedures: gel-retardation analysis is performed under equilibrium conditions, whereas in vitro transcription analysis is a kinetic assay.

Figure 2

Purified AT inhibits TRAP-dependent transcription termination. A DNA template containing the trppromoter-leader region was transcribed in vitro by the B. subtilis vegetative RNA polymerase, producing a labeled transcript of 320 nucleotides (nts) (readthrough, RT). In the presence of 0.5 mM tryptophan and 340 nM TRAP, the principal product is a terminated transcript of ∼140 nt (T). The AT/TRAP molar ratios shown (1× to 15×) were calculated by assuming a AT molecular mass of 28 kD. In the last lane of the gel, AT was tested alone (no TRAP) at the same concentration as in the previous lane.

Figure 3

AT addition prevents the TRAP:RNA band-shift. The 140-nt trp leader RNA containing the TRAP binding site, labeled by in vitro synthesis, was incubated with TRAP and/or AT (in the presence of 0.5 mM tryptophan) and loaded onto a native polyacrylamide gel. The AT concentration was calculated by assuming a molecular mass of 28 kD.

AT-TRAP cross-linking.

Proof that AT interacts with TRAP directly was provided by performing cross-linking experiments with AT, TRAP, and glutaraldehyde (28) (Fig. 4). Cross-linking of AT to TRAP at the concentrations used was only observed when tryptophan was added to the reaction mixture. Although our results are insufficient to allow determination of the stoichiometry of the complex, several conclusions can be drawn. First, formation of the AT-TRAP complex does not require the presence of target RNA, a possibility that our previous experiments had not ruled out. Second, because the residues that react with glutaraldehyde are mainly lysine, and most of the lysine residues of TRAP are on its surface and some are crucial for RNA-binding (9), AT may act by masking TRAP's RNA-binding surface. The additional finding that the AT-TRAP complex forms only in the presence of tryptophan argues that TRAP's AT binding surface may only form when TRAP is activated by tryptophan. Equilibrium dialysis experiments indicate that AT alone does not bind tryptophan (20). Therefore, this TRAP inhibitory protein appears to be capable of distinguishing between the active and inactive conformations of TRAP.

Figure 4

AT can be cross-linked to TRAP, in the presence of tryptophan. SDS-polyacrylamide gel (4 to 20% gradient) electrophoresis of TRAP (0.75 μg) and/or AT (1.2 μg, 5× molar ratio over TRAP), in the presence or absence of 0.5 mM tryptophan, after cross-linking with 0.2% glutaraldehyde. M, molecular size standards; units at right are kilodaltons.


Our findings explain how B. subtilis recognizes uncharged tRNATrp as a regulatory signal. This signal is integrated as a regulatory command that indirectly modulates the activity of the TRAP regulatory protein by inducing the synthesis of the AT protein. Thus the mechanisms of regulation of expression of the genes of tryptophan biosynthesis in B. subtilis differ considerably from those used by E. coli, although each organism recognizes both tryptophan and tRNATrp as regulatory signals (Fig. 5). In E. coli, tryptophan activates the trp aporepressor, and the active repressor binds at the trp operon operator and regulates transcription initiation (1). In B. subtilis, tryptophan activates the TRAP protein and active TRAP binds to trpleader RNA (3). This promotes the formation of a transcription terminator structure, causing transcription termination. In E. coli uncharged tRNATrp accumulation leads to ribosome stalling during translation of a 14-residue leader peptide, and this stalling induces the formation of an antiterminator structure that prevents transcription termination (1); thus,trp operon expression is increased by uncharged tRNATrp. In B. subtilis, as described in this article, uncharged tRNATrp accumulation leads to AT production, and AT inactivates TRAP, leading to antitermination and increased trp operon expression. Each of these regulatory mechanisms appears to be effective in regulating trp operon expression. The regulatory differences observed presumably reflect evolutionary adjustments of ancestral species in their attempts to optimize gene expression in relation to operon organization and overall metabolism (1–3, 16, 29).

Figure 5

The different mechanisms used by E. coli and B. subtilis to regulate trp operon transcription (see text for details).

  • * To whom correspondence should be addressed. E-mail: yanofsky{at}


View Abstract

Stay Connected to Science

Navigate This Article