Mechanism of Repression of the aroP P2 Promoter by the TyrR Protein of Escherichia coli

Abstract

Previously, we have shown that expression of the Escherichia coli aroP P2 promoter is partially repressed by the TyrR protein alone and strongly repressed by the TyrR protein in the presence of the coeffector tyrosine or phenylalanine (P. Wang, J. Yang, and A. J. Pittard, J. Bacteriol. 179:4206–4212, 1997). Here we present in vitro results showing that the TyrR protein and RNA polymerase can bind simultaneously to the aroP P2 promoter. In the presence of tyrosine, the TyrR protein inhibits open complex formation at the P2 promoter, whereas in the absence of any coeffector or in the presence of phenylalanine, the TyrR protein inhibits a step(s) following the formation of open complexes. We also present mutational evidence which implicates the N-terminal domain of the TyrR protein in the repression of P2 expression. The TyrR binding site of aroP, which includes one weak and one strong TyrR box, is located 5 bp downstream of the transcription start site of P2. Results from a mutational analysis show that the strong box (which is located more closely to the P2 promoter), but not the weak box, plays a critical role in P2 repression.

In bacterial cells, transcription initiation is a complex process which, in general, can be divided into four sequential steps comprising closed complex formation, open complex formation, initial transcribing complex formation, and promoter clearance (10, 18, 21, 29). At the first step, RNA polymerase recognizes and binds to a promoter in the region upstream of the transcription start site. This is then followed by an extension of interaction to DNA downstream of the transcriptional start site and separation of double-stranded DNA at the −10 region. At the third step, RNA polymerase is temporarily stalled at the promoter and repeatedly produces short abortive transcripts. Finally, RNA polymerase overcomes this barrier and proceeds to form a productive elongation complex. Many bacterial repressors repress transcription initiation by blocking the access of RNA polymerase to the promoter, inhibiting the formation of a closed complex (2). More recently, it has been shown that some repressors and RNA polymerase can bind simultaneously to DNA and that inhibition of transcription initiation occurs at a step(s) subsequent to closed complex formation (3, 12, 24). For example, the GalR repressor can prevent formation of the initial transcribing complex at the gal P1 promoter (3), whereas the P4 protein from Bacillus subtilis phage φ29 inhibits promoter clearance at the A2c promoter (24). Recent studies by Green and Marshall have suggested that the regulatory protein FNR can repress expression of the yfiD and ndh promoters by inhibiting formation of the open complex (12).

The TyrR protein of Escherichia coli controls expression of a number of genes involved in the biosynthesis, catabolism, or transport of aromatic amino acids (1, 27). The TyrR monomer is a polypeptide of 513 amino acids (aa) (44) which contains an N-terminal domain (aa 1 to 200), a central domain (aa 201 to 467), and a C-terminal region (aa 468 to 513) (9). The N-terminal domain is absolutely required for transcription activation at the tyrP and mtr promoters but is not important for repression of transcription at the tyrP and aroF promoters (8, 44). Results from both in vivo and in vitro studies have identified amino acid residues in both the N-terminal domain of TyrR and the C-terminal domain of the α subunit of RNA polymerase that are necessary for TyrR-mediated activation of the tyrP and mtr promoters (41–43, 45). The central domain of the TyrR protein is believed to be involved in ATP binding and dimer and oligomer formation (7, 19, 44). The C-terminal region contains a Cro-like helix-turn-helix motif which is required for binding to DNA sequences known as TyrR boxes (13, 14).

In vitro studies by Wilson et al. (40) have shown that the TyrR protein exists as a dimer in a solution which contains no aromatic amino acids or only phenylalanine but that it self-associates to form a hexamer in the presence of ATP and tyrosine or ATP and high levels of phenylalanine. Tyrosine and phenylalanine are the two major coeffectors involved in TyrR-mediated regulation of various genes in the TyrR regulon (27, 39).

The aroP gene, which codes for a membrane protein involved in the transport of all three aromatic amino acids, has three promoters (P1, P2, and P3) (5, 37). The TyrR protein represses expression of the major promoter, P1, by recruiting RNA polymerase to the divergent and nonproductive promoter P3 (36, 38). The P2 promoter, on the other hand, is partially repressed by the TyrR protein alone and shows enhanced repression in the presence of the coeffector tyrosine or phenylalanine (37). When cells are grown in minimal medium, the TyrR protein exerts fivefold repression on P2 transcription which, in the presence of either phenylalanine or tyrosine, is increased to 35- or 50-fold (37). The TyrR binding site, which includes one weak and one strong TyrR box, is located 5 bp downstream of the transcription start site of P2 (37). DNase I footprinting studies have shown that both boxes are protected by TyrR in the presence of ATP and tyrosine (26). In contrast, only the strong box is protected by TyrR in the presence of phenylalanine or in the absence of any coeffector (26). Since the presence of phenylalanine does not enhance the binding affinity of the TyrR protein to the strong box (19a), it seems unlikely that phenylalanine-mediated repression of P2 transcription involves direct competition between the TyrR protein and RNA polymerase for binding to the DNA.

In this study, we have carried out both in vivo and in vitro experiments to investigate the mechanism of P2 repression by the TyrR protein. Our results show that, in the presence of tyrosine, the TyrR protein inhibits open complex formation at P2 and that, in the presence of phenylalanine or in the absence of any coeffector, the TyrR protein inhibits the step(s) following formation of the open complex.

MATERIALS AND METHODS

Reagents, strains, plasmids, and media.

Restriction enzymes and chemicals were purchased commercially. All bacterial strains used in this study were derivatives of E. coli K-12. The strains and plasmids are listed in Table 1. The minimal medium was prepared from buffer 56/2 (23) supplemented with 0.2% glucose and appropriate growth factors. To study regulation, we added tyrosine or phenylalanine to the minimal medium at a final concentration of 1 mM. Trimethoprim and kanamycin were each used at a final concentration of 10 μg/ml.

TABLE 1.

E. coli K-12 strain, plasmids, and phages used in this work

Strain, plasmid, or phage	Relevant characteristics	Source or reference
Strain
JP8042	ΔlacU169 recA56 tyrR366	44
Plasmids
pMU1065	Kmr; 1.77-kb EcoRV-PvuII fragment containing tyrR gene in the HindII site of pACYC177	6
pMU2385	TprgalK′-lac′Z IncW; low-copy-number transcriptional fusion vector	28
pMU3309	Kmr; 2.2-kb EcoRI-BamHI fragment containing mutant tyrR allele (Δ5-42) in pSU39	44
pMU3310	Kmr; 2.2-kb EcoRI-BamHI fragment containing mutant tyrR allele (Δ5-102) in pSU39	44
pMU6198	Kmr; 2.2-kb EcoRI-BamHI fragment containing mutant tyrR allele (RQ10) in pSU39	42
pMU6242	Tpr; 0.3-kb HindIII-SalI fragment containing aroP regulatory region with P2wt and mutant P1 and P3 promoters in pMU2385	This work
pMU6243	Tpr; 0.3-kb HindIII-SalI fragment containing aroP regulatory region with P2up and mutant P1 and P3 promoters in pMU2385	This work
pMU6244	Tpr; pMU6242 derivative; G→T mutation in the weak TyrR box	This work
pMU6245	Tpr; pMU6242 derivative; G→T mutation in the strong TyrR box	This work
pMU6246	Apr; 0.3-kb HindIII-SalI fragment containing aroP regulatory region with P2wt and mutant P1 and P3 promoters in pUC19	This work
pMU6247	Apr; 0.3-kb HindIII-SalI fragment containing aroP regulatory region with P2up and mutant P1 and P3 promoters in pUC19	This work
pUC19	Apr; multicopy plasmid with a ColE1-type replicon	35
Phages
M13tg131	lacPOZ′	17
mpMU39	0.3-kb HindIII-SalI fragment containing aroP regulatory region with P2wt and mutant P1 and P3 promoters in M13tg131	This work
mpMU40	0.3-kb HindIII-SalI fragment containing aroP regulatory region with P2up and mutant P1 and P3 promoters in M13tg131	This work
mpMU41	mpMU39 derivative; mutation (G→T) in the weak TyrR box	This work
mpMU42	mpMU39 derivative; mutation (G→T) in the strong TyrR box	This work

Recombinant DNA techniques.

Standard recombinant DNA procedures, as described by Sambrook et al. (31), were used. DNA sequences were determined by the chain termination method described by Sanger et al. (32) with T7 DNA polymerase (Pharmacia).

Site-directed mutagenesis.

In vitro mutagenesis with synthetic oligonucleotides was performed on M13tg131 derivatives containing the aroP regulatory region by using commercially available kits (Amersham Corporation and U.S. Biochemical Corporation). Mutations were confirmed by DNA sequence analysis.

β-Galactosidase assay.

β-Galactosidase activity was assayed as described by Miller (22). Specific activity is expressed in units described therein. The data are the results of at least three independent assays. The values obtained were averaged, and the result of each individual assay deviated from the average by no more than 20%.

DNase I footprinting.

The bottom strand (template) or the top strand (nontemplate) of the 0.3-kb aroP fragment containing the wild-type P2 promoter and mutant P1 and P3 promoters was labelled with ³²P at the 3′ end. This was achieved as described below. The pUC19 derivative (pMU6246) which carries the EcoRI-HindIII aroP fragment was digested with either EcoRI (for labelling the bottom strand) or HindIII (for labelling the top strand). The DNA was then radioactively labelled by filling in the restriction ends with Klenow enzyme, [α-³²P]dATP, and deoxynucleoside triphosphates. Following a second digestion with either HindIII or EcoRI, the aroP fragments were each purified on a 5% polyacrylamide gel. About 50 cps (measured on a hand-held monitor) of the labelled fragment was used in each footprinting reaction. When required, phenylalanine and tyrosine were each added at a final concentration of 0.4 mM and ATP was added at a final concentration of 100 μM. RNA polymerase and the TyrR protein were added at final concentrations of 100 and 150 nM, respectively. Binding reactions of the RNA polymerase-DNA or RNA polymerase-TyrR-DNA complexes were carried out for 30 min at 37°C in a total volume of 30 μl of transcription buffer (50 mM Tris-HCl [pH 7.8], 50 mM NaCl, 3 mM magnesium acetate, 0.1 mM EDTA, 0.1 mM dithiothreitol, and 25 μg of bovine serum albumin per ml). The samples were then each treated with 0.02 units of DNase I (Boehringer Mannheim) to allow partial digestion of the DNA; after incubation for 30 s at room temperature, the reaction was terminated by phenol extraction. The resulting DNA fragments were analyzed on a 6% sequencing gel against an A+G ladder produced by the method of Maxam and Gilbert (20).

KMnO₄ footprinting.

The aroP fragment used in DNase I footprinting was also used in the KMnO₄ footprinting experiment. The protein-DNA binding reactions are the same as described for DNase I footprinting (see above). Following the formation of protein-DNA complexes, the DNA was treated with KMnO₄ (3 μl of an 80 mM concentration) for 2 min at room temperature. The reaction was then quenched with 2 μl of β-mercaptoethanol (14.7 M). The DNA fragments were precipitated twice with ethanol and cleaved with piperidine as described by Maxam and Gilbert (20) and analyzed on a 6% sequencing gel.

In vitro transcription.

Runoff transcription assays were performed by a method based on the standard single-round conditions described by Igarashi et al. (15). A reaction mixture containing linear DNA template (300 ng) and RNA polymerase (100 nM) was incubated at 37°C for 25 min in a total volume of 20 μl of transcription buffer. The mixture, when required, contained TyrR protein (150 nM), ATP (100 μM), and tyrosine (0.4 mM) or phenylalanine (0.4 mM). Following incubation, 10 μl of start solution (containing 1× transcription buffer with heparin [0.67 mg/ml]; ATP, CTP, and GTP [0.53 mM each]; UTP [0.053 mM]; and [α-³²P]UTP [3 μCi]) was added to initiate RNA synthesis. Transcription was allowed to proceed for 5 min before the reaction was terminated by phenol extraction. A portion of each sample (10 μl) was mixed with sequencing dye mix and analyzed on a 6% sequencing gel.

Abortive transcription assays were performed as described for runoff transcription assays (see above) except that each reaction mixture contained the initiating dinucleotide CpU (500 μM) to ensure precise initiation. Following transcription reactions, the samples were analyzed on a 20% denaturing polyacrylamide gel.

RESULTS

Construction of plasmids carrying various aroP fragments.

To analyze TyrR-mediated repression of transcription from the aroP P2 promoter, a number of M13tg131 derivatives carrying aroP fragments containing various mutations in the regulatory region (P2wt, P2up, P2wt-wb⁻, P2wt-sb⁻) were constructed by site-directed mutagenesis (see Materials and Methods). The designations are as follows: P2wt, wild-type P2 promoter; P2up, mutant P2 with improved spacing between the −35 and −10 hexamers (Fig. 1); sb⁻, a knockout mutation in the strong TyrR box (Fig. 1); wb⁻, a knockout mutation in the weak TyrR box (Fig. 1). To avoid interference of P2 transcription by events occurring at P1 and P3, mutations which completely inactivate both P1 and P3 were also introduced into each of the aroP fragments (37) (Fig. 1).

FIG. 1.

Nucleotide sequence of the 0.3-kb DNA fragment containing the aroP regulatory region. The two TyrR boxes (boldface); the −35 and −10 regions of P1 (overlined), P2 (boxed), and putative P3 (37) (underlined); the transcription start points for the P2 promoter (asterisks); the mutations in the −35 region of P1 (from GTGCAT to GCACAT) or P3 (from AAGACT to AACCAT); the −10 region of the P2up promoter (boldface); and the protected regions observed in DNase I footprinting experiments (open bars for the bottom strand and solid bars for the top strand) are indicated.

Each of the DNA fragments was cloned into plasmid pMU2385 (a low-copy-number plasmid containing a promoterless lacZ gene) (28) to construct various P2-lacZ fusions for in vivo analysis of P2 repression.

The aroP fragments P2wt and P2up were also cloned separately into pUC19 to generate templates for in vitro analysis of P2 repression.

The N-terminal domain of TyrR is involved in P2 repression.

As mentioned in the introduction, repression of P2 by TyrR may involve a mechanism that requires simultaneous binding of the TyrR protein and RNA polymerase to the P2 promoter. If this assumption is correct, interactions between the two proteins may be necessary for P2 repression. As the N-terminal domain is known to interact with RNA polymerase in TyrR-mediated activation of the mtr and tyrP promoters, we decided to check the ability of three tyrR mutants with mutations affecting the N-terminal domain of TyrR protein to repress P2 expression.

Plasmids pMU3309 and pMU3310 carry tyrR genes with deletions such that they encode proteins lacking aa 5 to 42 and 5 to 102, respectively (44). These two mutant TyrR proteins are completely defective in transcription activation at the mtr and the tyrP promoters but are still capable of repressing tyrP and aroF transcription, indicating that although the activation ability of these mutants is impaired, the DNA binding and hexamerization functions of the mutant proteins are intact (44). Plasmid pMU6198 carries a mutant tyrR gene which codes for a protein with a single amino acid change (arginine to glutamine) at position 10 (RQ10) (42). The phenotype of this tyrR mutant is similar to that of the two tyrR deletion mutants, in that it is active in repression but inactive in activation (42). These three plasmids, as well as the control plasmid pMU1065 bearing the wild-type tyrR gene, were each transformed into strain JP8042 (tyrR366 Δlac) containing either the P2wt-lacZ fusion (pMU6242) or the P2up-lacZ fusion (pMU6243). The effects of the mutations on repression of transcription from the P2wt and P2up promoters were analyzed by β-galactosidase assays. The results are shown in Table 2.

TABLE 2.

Repression of aroP (P2wt)-lacZ and aroP (P2up)-lacZ fusions by various tyrR alleles

TyrR mutant	β-Galactosidase sp act (U)a
	aroP (P2wt)-lacZ			aroP (P2up)-lacZ
	MM	Phe	Tyr	MM	Phe	Tyr
tyrR null mutant	110	110	122	400	397	428
tyrR+ control	20 (5.5)	3 (37)	3 (41)	65 (6.2)	8 (50)	5 (86)
Δ5-42	85 (1.3)	75 (1.5)	44 (2.7)	343 (1.2)	333 (1.2)	201 (2.1)
Δ5-102	73 (1.5)	80 (1.4)	27 (4.5)	230 (1.7)	250 (1.6)	129 (3.3)
RQ10	13 (8.5)	4 (28)	6 (20)	47 (8.5)	9 (44)	11 (39)

^aUnits are those defined by Miller (22). MM, minimal medium; Phe, minimal medium containing 1 mM phenylalanine; Tyr, minimal medium containing 1 mM tyrosine. Values in parentheses are fold repression: specific activity of β-galactosidase of the tyrR null mutant strain divided by that of the tyrR⁺ or tyrR mutant strains. 

As expected, expression of the P2wt promoter was repressed 5.5-fold by the wild-type TyrR protein in minimal medium. The repression was increased to about 40-fold when cells were grown in the presence of phenylalanine or tyrosine. In contrast, in the strains which expressed the proteins from the deletion mutants (Δ5-42 and Δ5-102), the minimal medium- and phenylalanine-mediated P2wt repression was almost completely destroyed and tyrosine-mediated repression was significantly reduced from about 40-fold to 2.7- and 4.5-fold. On the other hand, strong repression was observed with the mutant TyrR protein RQ10, which is unable to activate expression of mtr and tyrP.

In comparison with P2wt, there was a fourfold increase in the transcription activity of the P2up promoter in the strain with a tyrR null mutation on the chromosome (JP8042) (Table 2). However, similar regulatory effects by the wild-type and mutant TyrR proteins were seen with P2up and P2wt, suggesting that, although improving the spacing between the −10 and −35 regions of P2 increases the rate of transcription initiation, it does not alter the way in which the TyrR protein represses expression of the P2 promoter.

Effects of a mutation in either the strong or weak TyrR box on P2wt repression.

In the case of the genes tyrP and aroF, whose expression is repressed directly by TyrR protein with tyrosine as a coeffector, the weak and strong TyrR box combination is so arranged that the weak box is closest to and overlaps the RNA polymerase binding site. In these cases, it has been proposed that the hexamer is required to bind to both the strong and weak boxes and that the binding of the TyrR protein to the weak box is critical for repression (26). In the aroP P2 promoter, however, the strong TyrR box is closest to the RNA polymerase binding site (Fig. 1). In addition, this promoter can be repressed under growth conditions in which the TyrR hexamer is not formed (e.g., in minimal medium or minimal medium plus phenylalanine or by using the hexamerization-defective mutant TyrR-EQ274 [19, 47]). The role of each of the TyrR boxes in repression of the P2 promoter was therefore investigated.

Plasmids which contain P2-lacZ fusions with a mutation changing the invariant G into a T in the left arm of either the strong or weak TyrR box, pMU6245 (P2wt-sb⁻-lacZ) and pMU6244 (P2wt-wb⁻-lacZ), were each transformed into strains JP8042 (tyrR366 Δlac) and JP8042 containing pMU1065 (multicopy tyrR⁺). The resulting strains were assayed for β-galactosidase expression under different growth conditions. As shown in Table 3, the mutation in either of the TyrR boxes did not cause any disruption in transcription in the tyrR366 strain. In the strain with pMU1065 (multicopy tyrR⁺), the weak box mutation had no effect on P2 repression in minimal medium and both phenylalanine- and tyrosine-mediated repression remained strong (18- and 13-fold), albeit with a two- and threefold reduction in repression compared to that with the P2wt-lacZ control. In contrast, when the strong TyrR box was mutated, the TyrR-mediated repression of P2 was severely impaired under all growth conditions. These results show that whereas the strong TyrR box is critical for both TyrR-dimer- and TyrR-hexamer-mediated repression, the weak box plays a much less significant role in P2 repression, quite unlike the situation with the aroF and tyrP promoters.

TABLE 3.

Effect of a mutation in either the strong or weak TyrR box on P2 repression

lacZ fusion	β-Galactosidase sp act (U)a
	tyrR366	Multicopy tyrR+
		MM	Phe	Tyr
P2wt-lacZ control	110	20 (5.5)	3 (37)	3 (37)
P2wt-wb−-lacZ (weak box mutation)	124	21 (6)	7 (18)	10 (13)
P2wt-sb−-lacZ (strong box mutation)	114	78 (1.5)	69 (1.7)	66 (1.7)

^aDesignations and parenthetical values are as described in Table 2, footnote a. 

Analysis of closed and open complex formation by DNase I footprinting.

DNase I footprinting experiments were next carried out to analyze the effects of the TyrR protein and its coeffectors on the binding of RNA polymerase to the P2 promoter. The DNA fragment containing the P2 regulatory region (P2wt) was labelled with ³²P in either the bottom (template) or top (nontemplate) strand (see Materials and Methods). The DNA was incubated with E. coli RNA polymerase with or without purified TyrR protein and the coeffectors tyrosine and phenylalanine. The protein-bound fragment was then treated with DNase I, and the resulting DNA was analyzed on a sequencing gel.

Data from the bottom strand (Fig. 2A and Fig. 1) show that, in the absence of the TyrR protein, a region between positions −44 and +24 (relative to the start site of P2 transcription) is protected by RNA polymerase and there is a hypersensitive site at −23. This footprint is typical of open complexes (33). In the presence of TyrR alone or TyrR plus phenylalanine, the upstream boundary of the protected region remained the same, at −44, but the downstream boundary was extended (to +29) to cover the entire strong TyrR box. In the presence of TyrR and tyrosine, however, a quite different protection pattern, which includes two distinct regions, can be seen. One region covers the sequence between −56 and −5, which corresponds to the footprint of a closed complex (33), and the other spans the sequence between +8 and +53, which includes the double TyrR boxes.

FIG. 2.

DNase I footprinting of the aroP P2 promoter. Results shown are from experiments using the DNA fragment labelled in the bottom (A) and top (B) strands. When required, phenylalanine and tyrosine were each added at a final concentration of 0.4 mM. RNA polymerase and the TyrR protein were added at a final concentrations of 100 and 150 nM, respectively. The protected regions observed under different conditions are indicated.

The results from the top strand (Fig. 2B and Fig. 1) are in agreement with those from the bottom strand and confirm that the TyrR protein and RNA polymerase can bind simultaneously to the P2 promoter region. In the absence of any aromatic amino acid coeffector or in the presence of phenylalanine, the binding of TyrR protein does not affect the formation of an open complex, whereas in the presence of tyrosine, the RNA polymerase-promoter complex appears to be held in the “closed configuration” by the TyrR protein. The disappearance of the hypersensitive bands at −23 of both the bottom and top strands in the presence of tyrosine (Fig. 2A, lane 7, and Fig. 2B, lane 5) confirms that this result is not an artifact arising from a mixture of two populations of protected molecules, one with RNA polymerase alone and the other with the TyrR protein and tyrosine.

Analysis of open complex formation by KMnO₄ footprinting.

The effect of the presence of TyrR protein on the formation of an open complex at the aroP P2 promoter was further investigated by KMnO₄ footprinting. The DNA fragment labelled at the 3′ end of either the top (nontemplate) or bottom (template) strand was incubated with RNA polymerase in the presence or absence of purified TyrR protein with or without a coeffector (tyrosine or phenylalanine). The protein-DNA complexes were treated with KMnO₄ to modify bases in the melted region in the open complex. Following cleavage at the modified positions with piperidine, the DNA was analyzed on a sequencing gel. The results from the experiment with the DNA fragment labelled at the 3′ end of the bottom (template) strand are shown in Fig. 3. In the presence of RNA polymerase alone, two strong hypersensitive bands corresponding to the thymine residues at positions −11 and −12 are seen (Fig. 1). Addition of the TyrR protein or TyrR plus phenylalanine had no significant effect on the formation of these hypersensitive bands. However, in the presence of TyrR and tyrosine, these hypersensitive bands were not seen. These results confirm the observation described in the previous section, that the TyrR protein in conjunction with the coeffector tyrosine inhibits formation of the open complex at P2 and that neither the TyrR protein alone nor TyrR plus phenylalanine is able to prevent formation of the open complex.

FIG. 3.

KMnO₄ footprinting of the aroP P2 promoter. The conditions used in this experiment are the same as those described for DNase I footprinting in the legend to Fig. 2. The positions sensitive to KMnO₄ treatment are indicated.

When the DNA fragment labelled in the top (nontemplate) strand was used in the experiment, no hypersensitive band was detected in the presence of RNA polymerase either with or without the TyrR protein and its coeffectors (results not shown).

In vitro transcription.

To examine the effect of the presence of the TyrR protein on production of either the full-length or abortive transcripts from P2, we carried out in vitro transcription experiments (see Materials and Methods). In these experiments, the P2up fragment (Fig. 1), which exhibits almost the same TyrR-mediated regulatory effect as P2wt (see above), was chosen as the template as it produces strong transcriptional signals in vitro, which allow better measurement of the inhibitory effect of the TyrR protein on P2 transcription (see below). The results of the full-length and abortive transcriptional analyses are shown in Fig. 4 and 5, respectively.

FIG. 4.

Runoff transcription. Single-round in vitro transcription was carried out by using the 0.3-kb fragment containing an improved P2 promoter (P2up) as a template. The concentrations of TyrR protein, RNA polymerase, and aromatic amino acids used in this experiment are the same as those described in the legend to Fig. 2.

FIG. 5.

Abortive transcription. The conditions used in this experiment are the same as those described for the runoff transcription experiment in the legend to Fig. 4. The initiating dinucleotide CpU (500 μM) was added to each reaction mixture to ensure precise initiation.

A strong mRNA band corresponding to the full-length runoff P2 transcript can be seen in the absence of the TyrR protein (Fig. 4). The production of this P2 transcript was significantly reduced by the presence of the TyrR protein alone (threefold reduction as measured by densitomitry) and was further reduced (to fourfold) when phenylalanine was present in the reaction. In the presence of TyrR and tyrosine, almost complete inhibition of P2 transcription was observed.

As shown in Fig. 5, production of the abortive transcripts from P2 was inhibited strongly in the presence of TyrR and tyrosine (fourfold) and also showed some reduction in the presence of the TyrR protein alone or TyrR plus phenylalanine. The TyrR-tyrosine-mediated effect at this step can be attributed to interference at the earlier step (open complex formation), as demonstrated in the previous sections. On the other hand, reduction of the abortive transcripts in the presence of TyrR alone or TyrR and phenylalanine suggests that the formation or activity of the initial transcribing complex can be affected under these conditions.

DISCUSSION

In this study, we have used both genetic and biochemical approaches to examine the mechanism of repression of the aroP P2 promoter by the TyrR protein and its coeffectors. From the data presented in this paper, the following conclusions can be drawn: (i) the binding of either the liganded or unliganded TyrR protein to the TyrR box(es) in the aroP regulatory region does not inhibit formation of a closed complex at the P2 promoter; (ii) in the presence of tyrosine, the TyrR protein represses P2 expression by preventing open complex formation; (iii) in the absence of any coeffector or in the presence of phenylalanine, the TyrR protein represses P2 expression by inhibiting a step(s) following the formation of open complexes; and (iv) an interaction between the N-terminal domain of TyrR and RNA polymerase is required for TyrR-mediated repression of P2 transcription.

Repression involving a direct contact between a repressor and RNA polymerase has previously been reported in two other bacterial systems (the GalR repressor of E. coli [4] and the P4 protein of B. subtilis phage φ29 [24]). In both cases, the repressor binding sites (the operators) are located upstream of the promoters and the repressors interact with the α subunit of RNA polymerase (4, 25). Similarly, the binding sites for class I bacterial transcription activators are also located in the upstream regions of various promoters, and the activators stimulate transcription by interacting with the α subunit of RNA polymerase (11, 16). However, the TyrR-binding sites (the TyrR boxes) are situated close to and downstream of the transcription start site of the aroP P2 promoter, and preliminary results from in vitro transcription experiments indicate that the C-terminal domain of the α subunit of RNA polymerase is not involved in P2 repression (results not shown). It thus remains to be seen which subunit(s) of RNA polymerase is responsible for interaction with TyrR in P2 repression. The E. coli DnaA protein binds to a similar downstream position at the λ P_R promoter and activates transcription by interacting with the β subunit of RNA polymerase (34).

The inability of the mutant TyrR proteins (Δ5-42 and Δ5-102) to repress the aroP P2 promoter indicates that the TyrR protein requires an intact N-terminal domain, presumably to interact with RNA polymerase in repression of the P2 promoter as it does for activation of the mtr and tyrP promoters. However, both of the TyrR deletion mutants retain some low levels of repressibility in the presence of tyrosine (Table 2). This may result from an interaction of RNA polymerase with amino acid residues in some other part of the TyrR protein or, alternatively, may suggest that the TyrR hexamer can exert some repressive effect on P2 expression without an interaction with RNA polymerase. The arginine residue at position 10 has been shown to be a critical residue for the activation function of TyrR and has been proposed to interact directly with RNA polymerase. The observation that the mutant TyrR protein RQ10 is still able to exert phenylalanine- and tyrosine-mediated repression at the P2 promoter suggests that different amino acids are involved in interaction with RNA polymerase for activation of the mtr and tyrP promoters and for repression of the aroP P2 promoter.

The strong, but not the weak, TyrR box of aroP is essential for P2 repression (Table 3). This is in contrast to the situation with the aroF and tyrP promoters, where both the strong and weak boxes are critical for TyrR-mediated repression. The difference in the two systems is explained by the relative positions of the boxes, which in the latter case (aroF and tyrP) favors repression by tyrosine alone and in the former (P2) allows repression by either phenylalanine or tyrosine.

The results from the DNase I footprinting, KMnO₄ footprinting, and in vitro transcription experiments clearly show that the dimeric and hexameric forms of the TyrR protein can act at different steps to inhibit transcription initiation from the P2 promoter. The differential repression may reflect the differences between the TyrR dimer and hexamer in interaction with RNA polymerase and/or with DNA. The demonstration of the inhibitory effect on the synthesis of the abortive transcripts suggests that the TyrR protein can interfere with the formation or activity of the initial transcribing complex, in the presence or absence of phenylalanine. The failure to obtain a KMnO₄ footprint at the melted region with the DNA fragment labelled on the top (nontemplate) strand may suggest that the bases in the melted region of this strand are not available for modification by KMnO₄. This is in agreement with the contention by Roberts and Roberts (30) that the bases in the −10 promoter region of the nontemplate strand are the target for RNA polymerase and are persistently contacted by RNA polymerase during formation of the open complex bubble.

In the TyrR regulon, there is only one other gene, tyrB, whose TyrR boxes are also located downstream of the transcription start site (46). However, in this case, the weak box rather than the strong box is nearest the promoter, and both boxes are absolutely necessary for repression (46). Recently, we have found that repression at the tyrB promoter also requires an intact N-terminal domain of the TyrR protein (47). Although the detailed mechanism of TyrR-mediated repression at the tyrB promoter is yet to be established, it seems possible that this may also involve specific interactions between the TyrR protein and RNA polymerase.