|Home | About | Journals | Submit | Contact Us | Français|
Tryptophan, phenylalanine, tyrosine, and several other metabolites are all synthesized from a common precursor, chorismic acid. Since tryptophan is a product of an energetically expensive biosynthetic pathway, bacteria have developed sensing mechanisms to downregulate synthesis of the enzymes of tryptophan formation when synthesis of the amino acid is not needed. In Bacillus subtilis and some other Gram-positive bacteria, trp operon expression is regulated by two proteins, TRAP (the tryptophan-activated RNA binding protein) and AT (the anti-TRAP protein). TRAP is activated by bound tryptophan, and AT synthesis is increased upon accumulation of uncharged tRNATrp. Tryptophan-activated TRAP binds to trp operon leader RNA, generating a terminator structure that promotes transcription termination. AT binds to tryptophan-activated TRAP, inhibiting its RNA binding ability. In B. subtilis, AT synthesis is upregulated both transcriptionally and translationally in response to the accumulation of uncharged tRNATrp. In this paper, we focus on explaining the differences in organization and regulatory functions of the at operon's leader peptide-coding region, rtpLP, of B. subtilis and Bacillus licheniformis. Our objective was to correlate the greater growth sensitivity of B. licheniformis to tryptophan starvation with the spacing of the three Trp codons in its at operon leader peptide-coding region. Our findings suggest that the Trp codon location in rtpLP of B. licheniformis is designed to allow a mild charged-tRNATrp deficiency to expose the Shine-Dalgarno sequence and start codon for the AT protein, leading to increased AT synthesis.
Bacteria regulate the expression of their tryptophan (trp) biosynthetic genes and operons by using various strategies that sense the levels of free tryptophan (Trp) and/or uncharged tRNATrp (29). Trp is synthesized from chorismic acid, which is also the precursor of phenylalanine, tyrosine, p-aminobenzoic acid (PABA), and several other metabolites (50). Trp biosynthesis involves catalysis by the protein products of seven genes or genetic segments. Many bacilli have all seven trp genes in one operon, and this operon is regulated transcriptionally by an uncharged-tRNATrp-sensing T-box sequence or by tandem T-box sequences (23, 29). In some bacilli, including Bacillus subtilis and Bacillus licheniformis, the trp operon is organized differently. A cluster of six of the seven trp genes, trpEDCFBA, is located as a trp suboperon within a larger aromatic (aro) supraoperon (19). Initiation of transcription of the trp suboperon occurs at two promoters, one at the beginning of the aro supraoperon and the second preceding trpE of the trp suboperon (50). The seventh trp gene, trpG-pabA, specifying a bifunctional protein involved in both Trp and PABA synthesis, is located in the unlinked folate operon (19, 21, 29). Expression of trpG-pabA is also regulated in response to the availability of Trp (17, 46).
Prior studies (10, 20, 21, 24) identified many genes, operons, and regulatory molecules and events that are involved in Trp biosynthesis and its regulation in B. subtilis (Table (Table1).1). These same genes and other cell components appear to be present in B. licheniformis, a closely related bacterium (29). However, the aro supraoperon, containing a trp suboperon, is a relatively uncommon organizational strategy in the bacilli (29) or in other bacteria, and the presence of a regulatory at operon, responding to uncharged tRNATrp as a regulatory signal, is even rarer (29).
Transcription of the trp suboperons of B. subtilis and of B. licheniformis, initiated at either the aro or trp suboperon promoter, is regulated by transcription attenuation (termination) in the region immediately preceding trpE. The principal regulator is the Trp-activated RNA-binding attenuation protein, TRAP (15, 19-21, 46). When free Trp is plentiful and available, it binds to—and activates—TRAP. The TRAP protein contains 11 identical protein subunits, 11 Trp binding sites, and 11 Lys-Lys-Arg motifs on the periphery of the protein complex (2, 3, 13, 49). When activated by Trp, each Lys-Lys-Arg motif is capable of binding to a (G/U)AG repeat in a target transcript, resulting in the RNA being wrapped around TRAP's perimeter (8, 9, 30). This prevents the formation of an RNA antiterminator structure, thereby promoting the formation of an RNA terminator structure that causes transcription termination (8, 9, 30). Activated TRAP also binds to (G/U)AG repeats in mRNA segments of other genes and inhibits initiation of their translation. This occurs in transcripts of the following genes: trpE (16), trpG-pabA (17, 46); trpP-yhaG, encoding a putative Trp import protein (32); and ycbK, encoding a putative Trp efflux protein (32). The ycbK gene is located within the rtpA-ycbK (at) operon (45, 47). The rtpA gene encodes the anti-TRAP protein, AT, which is capable of binding to Trp-activated TRAP at its RNA-binding surface (35, 36, 43) and preventing TRAP from binding to its target RNAs. Thus, by binding to TRAP, AT can regulate transcription or translation of all the operons regulated by TRAP (39, 40). AT synthesis is highly regulated itself, both transcriptionally and translationally, in response to the accumulation of uncharged tRNATrp (11, 12).
In a number of bacilli, AT-dependent mechanisms for regulating trp operon expression appear to exist. They include B. subtilis, B. licheniformis, Bacillus amyliquefaciens, Bacillus mojavensis, and Bacillus spizizenii (14). Most studies of AT and TRAP synthesis and function have been performed with B. subtilis (15, 19, 39-41, 44, 49). In that organism, AT synthesis is regulated both transcriptionally and translationally by sensing the accumulation of uncharged tRNATrp. Transcription regulation of at mRNA synthesis is achieved in one segment of the operon's leader region by an uncharged-tRNATrp-sensing T-box transcription antitermination mechanism (23, 32). In B. subtilis, AT synthesis is also regulated translationally at a 10-residue leader peptide-coding region, rtpLP, located immediately upstream of rtpA, the structural gene for the AT protein. The rtpLP-coding region of B. subtilis contains three adjacent Trp codons, and its stop codon is located 6 nucleotides preceding the rtpA Shine-Dalgarno (SD) sequence (Fig. (Fig.1)1) (12). Completion of the translation of rtpLP mRNA of B. subtilis inhibits initiation of AT synthesis, presumably by ribosome blockage of the rtpA SD sequence. However, when there is a cellular charged-tRNATrp deficiency, the ribosome translating rtpLP mRNA presumably stalls at any one of its three Trp codons, exposing the SD region of rtpLP mRNA, allowing rtpA mRNA translation and AT synthesis (11). In B. licheniformis, AT synthesis is also regulated transcriptionally by the T-box mechanism and translationally by rtpLP, the at operon's leader peptide-coding region. However, in B. licheniformis, rtpLP is a 22-residue-coding region rather than a 10-residue-coding region, as it is in B. subtilis. In addition, in B. licheniformis, the rtpLP coding region includes the SD sequence, as well as the start codon of rtpA, the structural gene for the AT protein (Fig. (Fig.1).1). Most importantly, the three Trp codons of the rtpLP mRNA of B. licheniformis are dispersed throughout this coding region, rather than adjacent to one another as they are in rtpLP of B. subtilis. In the studies described in this article, we analyzed the significance of the different Trp codon locations within the rtpLP leader peptide-coding region of B. licheniformis. We focus on explaining the differences in organization and function of this rtpLP coding region relative to rtpLP of B. subtilis. Our findings suggest that the Trp codon location and other features of the rtpLP leader mRNA of B. licheniformis are designed to allow the organism to respond predominantly to conditions leading to the accumulation of low levels of uncharged tRNATrp.
The strains used in this study are listed in Table Table1.1. CYBS400 is a B. subtilis prototroph, Bl54A is a B. licheniformis prototroph, and CYBS318 [CYBS400 Δ(rtpA-ycbK)::Spr] is a B. subtilis strain lacking its AT-coding region and the region encoding the 5′ end of the ycbK open reading frame (ORF); it does not produce either the AT or the YcbK protein. The strain was constructed by replacing a 614-bp chromosomal segment of the rtpA-ycbK region with a gene conferring spectinomycin resistance (32). The plasmid Pat-LR-ΔT-rtpLP-rtpA-ycbK′-lacZ was used to construct CYBL derivative strains (Table (Table2).2). It contained a 736-bp leader region encompassing the rtpA-ycbK promoter, the leader region, the rtpA ORF, the intergenic region (33), and 9 nucleotides of the ycbK coding region, followed by the lacZ gene. Part of the terminator region was deleted to disrupt terminator function, as described by Sarsero et al. (33). Strain CYBL has the B. subtilis rtpA-ycbK promoter with the terminator region deleted (33), followed by B. licheniformis rtpLR replacing B. subtilis rtpLP, followed by B. subtilis rtpA-ycbK′-lacZ. In other constructs, each of the Trp codons of B. licheniformis rtpLP in the hybrid plasmid was replaced with an arginine codon (Trp1→Arg or Trp2→Arg) or with a cysteine codon (Trp3→Cys), thereby preserving the predicted rtpLP RNA secondary structure and the normal location of the rtpA ATG start codon (strains CYBL1, CYBL2, and CYBL3). A 15-nucleotide spacer (5 GAT repeats) introducing repeat UGAs in the coding region was also inserted at nucleotide position 15 of B. licheniformis rtpLP (strain CYBL1 spacer) (see Fig. Fig.44).
Transformations were carried out by using natural competence (1). Gene fusions and cloned DNA fragments were integrated into the chromosomal amyE locus by homologous recombination after being introduced into the integration vector ptrpBG1-PLK (33). Mutant strains were isolated following transformation by selecting for chloramphenicol resistance, and disruption of amyE was confirmed by the absence of amylase production as shown by iodine staining (34).
All strains were grown in Vogel-Bonner minimal medium (42) supplemented with 0.5% glucose and trace elements at 37°C (10). Where indicated, the following supplements were included: various concentrations of indole acrylic acid (IA), 100 μg/ml phenylalanine, 100 μg/ml tryptophan, 100 μg/ml tyrosine, or 100 μg/ml PABA. Growth rates were determined by measuring cell density using a Klett-Summerson colorimeter equipped with a 660-nm filter.
Ten milliliters of cultures grown to 150-Klett-unit densities were harvested by centrifugation. RNA extraction was performed as described previously (15).
cDNA synthesis was carried out with a SuperScript III First Strand Synthesis System for RT-PCR from Invitrogen (catalog no. 180980-051) using 1 μg of RNA. One-tenth of the total cDNA reaction mixture was used for real-time PCR analyses. Gene-specific primers were designed to amplify 100-nucleotide fragments of target genes (Tables (Tables33 and and4).4). Each reaction was carried out in a 15-μl volume with 50% SYBR green mixture, according to the manufacturer's protocol (Bio-Rad IQ SYBR green Supermix; catalog no. 170-8882). Reactions were performed in a MyiQ Single-Color Real-Time PCR Detection System (catalog no. 170-9770) with the cycling conditions 72°C for 5 min and 95°C for 5 min, followed by 40 cycles of 95°C for 15 s, 52°C for 15 s, and 72°C for 20 s, plus 5 min at 72°C as a final step. The expression of each gene analyzed was normalized against rpoB gene expression, which served as the internal control (31). Relative mRNA levels were subsequently calculated using the 2−ΔΔCT threshold cycle method (26).
Cells were harvested by centrifugation and resuspended in 300 μl of 100 mM Tris-HCl, pH 8.0, and 50 mM NaCl. Samples were disrupted by sonic oscillation, and cell debris was removed by centrifugation. SDS Tris-tricine (300 μl) buffer was added, and protein extracts were boiled for 5 min. The Quick Start Bradford protein assay (Bio-Rad catalog no. 500-0201) was performed on the final samples, and equal amounts of protein were loaded in each lane. The samples were electrophoresed on SDS-15% polyacrylamide gels in Tris-tricine buffer and were electrophoretically transferred onto a Trans-Blot Transfer Medium nitrocellulose membrane (Bio-Rad catalog no. 162-0115). Immunoblotting was performed using rabbit polyclonal antibodies against B. subtilis AT prepared by the Covance Company and peroxidase-conjugated affinity-purified anti-rabbit antibodies from Rockland (catalog no. 611-1302). The bound antibodies were visualized using Super Signal WestPico chemiluminescent substrate from ThermoScientific (catalog no. 34077). Their levels were quantitated by the Adobe Photoshop program, version 7.0.
To compare trp operon sensitivity to Trp starvation in B. subtilis and B. licheniformis, we analyzed relevant mRNA levels in cultures grown in minimal medium with or without excess Trp and in minimal medium containing different levels of IA, a structural homolog of Trp that is an inhibitor of tryptophanyl-tRNA synthetase activity (27). The presence of IA reduces tRNATrp charging and therefore the availability of Trp-tRNATrp for new protein synthesis.
Gene expression (mRNA) levels were related to expression levels for the rpoB gene, which served as the internal control (see Materials and Methods). The mRNAs analyzed (Table (Table5)5) were transcribed from the following genes: rtpA, encoding the AT protein (38, 39); mtrB, encoding the TRAP protein (20, 21); aroF, the first gene in the aromatic supraoperon (21, 50); trpE, the first gene of the trp suboperon (21); and trpS, the gene encoding tryptophanyl-tRNA synthetase, the enzyme responsible for charging of tRNATrp with Trp (37). Our RT-PCR analyses showed that in B. licheniformis, expression of rtpA and trpE is upregulated in response to IA addition, as it is in B. subtilis. However, most noticeably, B. licheniformis produces substantially higher relative trp operon (trpE) mRNA levels than B. subtilis when grown in the presence of IA (Table (Table5).5). One possible explanation for this response is that B. licheniformis may need to produce more of the trp operon enzymes under these conditions if it is to provide sufficient Trp for overall protein synthesis and near-normal growth rates (Table (Table55).
Our real-time PCR results suggested that Trp starvation caused by IA addition is more pronounced in B. licheniformis than in B. subtilis. Thus, the growth of B. licheniformis should be more sensitive to Trp starvation. To test this hypothesis, we grew cultures of both organisms under different mild Trp starvation conditions: low levels of IA. We observed that B. licheniformis was in fact more sensitive to growth inhibition by IA, an inhibitor of tryptophanyl-tRNA synthetase charging (Fig. (Fig.2).2). B. licheniformis' metabolism presumably is more sensitive to IA addition because it does not produce enough charged tRNATrp under these conditions.
To obtain additional understanding of the consequences of Trp starvation for B. licheniformis growth, we also analyzed the growth rates of IA-treated Trp-starved cells grown in the presence of different products of the aromatic biosynthetic pathway that have chorismic acid as a common precursor (50). Thus, we examined the effects of added phenylalanine, tyrosine, PABA, and Trp on IA-produced growth inhibition (Fig. (Fig.3).3). The addition of phenylalanine, tyrosine, and PABA did not reverse the IA-produced growth inhibition of B. licheniformis, whereas added Trp did, in the absence or presence of phenylalanine, tyrosine, and PABA (Fig. (Fig.3).3). Thus, IA addition appears to create a charged-tRNATrp deficiency that can be reversed by Trp addition.
We assume that synthesis of the AT protein is regulated both transcriptionally and translationally and that the locations of the Trp codons in rtpLP play a role in regulating AT synthesis. With these possibilities under consideration, experiments were performed (see Fig. Fig.5)5) to examine the effects on AT synthesis of replacing each of the Trp codons of the leader peptide-coding region with some other codon. As previously described in studies with B. subtilis, translation of the entire rtpLP coding region inhibits AT synthesis, whereas ribosome stalling at the rtpLP Trp codon cluster increases AT synthesis (12). Presumably, the ribosome reaching the rtpLP stop codon of B. subtilis masks the rtpA SD sequence, reducing initiation of AT synthesis, whereas a ribosome stalled at any one of the three rtpLP Trp codons should expose this SD sequence, allowing efficient initiation of AT synthesis (12). These experiments (see Fig. Fig.5)5) were performed to determine the regulatory significance of ribosome stalling at each of the 3 dispersed Trp codons in the B. licheniformis rtpLP sequence for the level of AT synthesized. Strains were constructed in which a hybrid plasmid bearing a modified at operon of B. subtilis containing rtpLP of B. licheniformis was integrated into the chromosome of a B. subtilis strain bearing a deletion of the resident at operon. In this plasmid, rtpLP of B. licheniformis replaced rtpLP of B. subtilis and the B. subtilis T-box terminator was deleted. Thus, at operon mRNA would be synthesized and rtpLP translational control of AT synthesis would determine the level of AT protein that was produced. In the 3 specific constructs, each of the three Trp codons of B. licheniformis rtpLP was replaced by a codon specifying another amino acid. Thus, the first Trp codon (UGG) was replaced by an Arg codon (AGG) in one construct, the second Trp codon (UGG) was replaced by an Arg codon (AGG) in a second construct, and the third Trp codon (UGG) was replaced by a Cys codon (UGU) in a third construct. These changes did not alter the predicted important RNA secondary structure of B. licheniformis rtpLP RNA (Fig. (Fig.4).4). Each plasmid construct was integrated into the amyE locus of a B. subtilis strain lacking the at operon, and thus, production of B. subtilis AT in these strains would be expected to be based on the organism's ability to translate the modified B. licheniformis rtpLP mRNA coding region (Fig. (Fig.1).1). The strains were grown both in minimal medium minus inducer and in minimal medium plus inducer (IA). The results obtained in Western blotting AT measurements performed with these strains are shown in Fig. Fig.5.5. Clearly, replacing the first Trp codon (Trp1) of B. licheniformis rtpLP had the greatest effect on increasing AT production. In this strain, the Trp1-to-Arg replacement would presumably allow the translating ribosome to reach the second Trp codon, and to stall there, when there is a deficiency of charged tRNATrp (Fig. (Fig.5).5). Ribosome stalling at this codon would presumably disrupt the RNA secondary structure that blocks the SD sequence needed for AT synthesis, and AT synthesis would be elevated. Apparently, even during growth in minimal medium (mild Trp starvation) there must be some increased stalling at Trp2 with this strain, since AT production in minimal medium was also higher in the Trp1 mutant than in the wild-type construct. When the wild-type rtpLP construct was grown with IA (severe Trp starvation), the AT level was also elevated, but it was not as high as with the Trp1 mutant. This result indicates that the presence of the Trp1 codon can prevent the large increase in AT production observed with the construct lacking Trp1 (Fig. (Fig.5).5). Therefore, in the wild-type construct, there must be little pausing at the Trp2 codon under the conditions tested. In the construct with the Trp2 codon replaced by an Arg codon, AT synthesis was comparable to that of the wild-type control culture in cultures grown in minimal medium and with IA (Fig. (Fig.5).5). Thus, there must be little or no ribosome stalling at the Trp2 codon in the wild type under these growth conditions. Clearly, the existence of Trp codon 2 is not solely responsible for the increased AT production associated with a charged-tRNATrp deficiency. Replacing the Trp3 codon with a Cys codon did not result in a significant change in AT production (Fig. (Fig.5).5). We suspect that a ribosome reaching this position or stalled at this codon would block initiation of AT synthesis. Additional experiments are required to explain the significance, if any, of the location of the Trp3 codon.
To confirm the significance of the close proximity of the second Trp codon of rtpLP to the presumed downstream rtpLP mRNA secondary structure that sequesters the SD sequence, we introduced a 15-nucleotide spacer downstream from the Trp2 codon in the context of the Trp1 codon mutant (Fig. (Fig.4).4). This spacer should prevent a ribosome stalled at the Trp2 codon from disrupting the RNA secondary structure that presumably limits AT production. Western blot analyses showed that, indeed, introduction of the spacer sequence reduced AT production to the same level observed with the wild-type rtpLP construct (Fig. (Fig.5).5). Complicating interpretation of these findings is the predicted formation of different RNA secondary structures in the various transcripts. However, these findings do suggest that in the rtpLP leader peptide-coding region of B. licheniformis the location of the Trp2 codon is critical in obtaining elevated AT protein production under certain Trp starvation conditions. The existence of the Trp1 codon at its location may reduce this elevation whenever a translating ribosome stalls at this codon and does not reach the Trp2 codon.
The purpose of this study was to examine trp operon regulation in two closely related species, B. licheniformis and B. subtilis, and to determine the significance of the organizational differences in Trp codon location in their respective at operon leader peptide-coding regions, rtpLP (Fig. (Fig.1).1). In both B. subtilis and B. licheniformis, the trp operon is located within an aro supraoperon, which is regulated by sensing the levels of Trp and uncharged tRNATrp and by the actions of the TRAP and AT proteins. These two signal molecules and two regulatory proteins influence both transcription and translation. In addition, expression of trpS, the gene encoding tryptophanyl-tRNA synthetase, is regulated in both organisms by the T-box mechanism in response to uncharged-tRNATrp accumulation (21). The tryptophanyl-tRNA synthetase level is a major determinant of how much charged tRNATrp will be produced and maintained per cell. However, the free-Trp level is also rate limiting for tRNATrp charging. In the studies described in this paper, we showed that the growth of B. licheniformis is more sensitive to the addition of IA, an inhibitor of tryptophanyl-tRNA synthetase activity, than is B. subtilis growth (Fig. (Fig.2).2). B. licheniformis responds to IA addition by producing higher trp operon mRNA levels (and presumably trp operon protein levels) than B. subtilis (Table (Table5).5). Furthermore, the lower aroF mRNA level (aroF is the first gene in the aro supraoperon) in B. licheniformis may indicate that the organism synthesizes less chorismate than B. subtilis. Therefore, it presumably must synthesize higher trp enzyme levels for the Trp pathway to compete effectively with the Phe and Tyr pathways for chorismate, their common precursor. Possibly, B. licheniformis must form higher levels of each of the Trp pathway enzymes in order to provide sufficient Trp to support growth. However, the relative contribution of Trp versus uncharged tRNATrp as a regulatory signal molecule and potential differences in Trp, Phe, and Tyr biosynthesis must be further analyzed in these two organisms before an explanation can be given for their differences in sensitivity to IA.
Some organisms that have their trp operons within an aro supraoperon contain an at operon, providing the anti-TRAP protein AT and allowing a regulatory response to a charged tRNATrp deficiency (29). They include B. subtilis, B. licheniformis, B. amyloliquefaciens, B. mojavensis, and B. spizizenii (14). In each of these organisms, the Trp-activated regulatory protein, TRAP, is primarily responsible for regulating trp operon transcription (5-7, 9, 15, 21, 46). Depending on how many TRAP molecules per cell are Trp activated and AT free, TRAP will be partially or fully active. Thus, TRAP can bind up to 11 molecules of Trp, to some extent cooperatively (4, 25, 36, 43), and each bound Trp molecule can contribute to TRAP's RNA-binding ability. However, even when TRAP is fully activated, AT can bind to TRAP and block TRAP's RNA-binding ability. AT synthesis is regulated both transcriptionally and translationally in response to the accumulation of uncharged tRNATrp. The existence of the at operon in B. subtilis and B. licheniformis allows posttranscriptional decisions to influence the regulation of the synthesis of the enzymes of the Trp biosynthetic pathway.
The leader peptide-coding regions of the at operons of B. licheniformis and B. subtilis are organized differently, and these differences are presumably designed to allow each organism to respond appropriately to a charged-tRNATrp deficiency. Both B. subtilis and B. licheniformis have three Trp codons in their rtpLP leader mRNA sequences. However, in B. subtilis, the three Trp codons are adjacent, and the SD sequence and start codon of AT are located downstream from the rtpLP stop codon (Fig. (Fig.4A).4A). Pausing at any one of these three codons would be expected to have a nearly equivalent severe effect, promoting maximal AT production. In B. licheniformis, its three Trp codons are spaced throughout the sequence, with the third Trp codon located just upstream of the stop codon, overlapping the SD sequence and start codon for AT (Fig. (Fig.4A).4A). The locations of the Trp codons in the rtpLP leader mRNA sequence are designed to allow the organism to be particularly sensitive to a slight reduction in the level of charged tRNATrp and to be less sensitive to a severe charged-tRNATrp deficiency (this study).
Ribosome stalling at the first Trp codon of rtpLP leader RNA (under severe Trp starvation conditions) of B. licheniformis should reduce leader peptide synthesis by allowing the rtpPL RNA secondary structure to form, reducing translation initiation at the rtpA (AT) SD sequence and start codon (Fig. (Fig.4B).4B). Stalling at the second rtpLP Trp codon, upon a mild charged-tRNATrp deficiency, should eliminate this secondary structure, allowing efficient ribosome binding and translation initiation at the rtpA SD sequence and start codon (Fig. (Fig.4C).4C). However, ribosome stalling at the Trp2 codon would allow a second potential RNA secondary structure to form, but presumably it is less effective in inhibiting ribosome binding at the rtpA SD sequence and start codon region. Thus, depending upon the severity of Trp starvation, the accumulation of uncharged tRNATrp, and the timing of charged-tRNATrp availability, the ribosome translating the rtpLP coding region could stall at either Trp codon, Trp1 or Trp2, and determine the efficiency of translation initiation at the rtpA start codon. Perhaps the objective of this design is to allow appreciable AT synthesis only when there is a slight charged-tRNATrp deficiency and to prevent additional AT synthesis when most of the tRNATrp in the cell is uncharged. This would be advantageous if under severe starvation conditions there was insufficient Trp to activate TRAP. Therefore, there would be no need to produce additional AT protein to inactivate TRAP. When a translating ribosome reached the third Trp codon of rtpLP mRNA, it would block AT synthesis until translation of rtpLP was completed. To explain the purpose of each of the three Trp codons in the rtpLP coding region of B. licheniformis, additional experiments are needed in which stalling at each of these Trp codons is related to the level of uncharged tRNATrp in the cell and the level of AT protein that is produced.
We are indebted to Paul Babitzke and Paul Gollnick for their excellent comments on the manuscript. We also thank Luis Cruz-Vera for help with the figures. This paper is the last publication based on research performed in my laboratory (C.Y.). I express my heartfelt appreciation to the many undergraduates, graduate students, postdoctoral students, visiting fellows, and other investigators who have contributed to our investigations. I have enjoyed every minute!
The studies described in this paper were performed with the support of the National Science Foundation (MCB-0615390).
Published ahead of print on 8 January 2010.