Negative Transcriptional Regulation of the ilv-leu Operon for Biosynthesis of Branched-Chain Amino Acids through the Bacillus subtilis Global Regulator TnrA

Abstract

The Bacillus subtilis ilv-leu operon is involved in the synthesis of branched-chain amino acids (valine, isoleucine, and leucine). The two- to threefold repression of expression of the ilv-leu operon during logarithmic-phase growth under nitrogen-limited conditions, which was originally detected by a DNA microarray analysis to compare the transcriptomes from the wild-type and tnrA mutant strains, was confirmed by lacZ fusion and Northern experiments. A genome-wide TnrA box search revealed a candidate box approximately 200 bp upstream of the transcription initiation base of the ilv-leu operon, the TnrA binding to which was verified by gel retardation and DNase I footprinting analyses. Deletion and base substitution of the TnrA box sequence affected the ilv-leu promoter activity in vivo, implying that TnrA bound to the box might be able to inhibit the promoter activity, possibly through DNA bending. The negative control of the expression of the ilv-leu operon by TnrA, which is considered to represent rather fine-tuning (two- to threefold), is a novel regulatory link between nitrogen and amino acid metabolism.

Branched-chain amino acids (isoleucine, valine, and leucine) are the most abundant amino acids in proteins and form the hydrophobic core of proteins. In addition, these amino acids are the precursors for the biosynthesis of iso- and anteiso-branched fatty acids, which represent the major fatty acid species of the membrane lipids in Bacillus species (3). The initial step in isoleucine and valine synthesis is the condensation of pyruvate and threonine and two pyruvates, respectively, leading to the formation of branched-chain keto acids (5). Leucine is synthesized from one of the branched-chain keto acids, α-ketoisovalerate.

The ilv-leu operon of Bacillus subtilis comprises seven genes necessary for the biosynthesis of isoleucine, valine, and leucine (Fig. 1). Previous studies demonstrated that the ilv-leu operon is regulated in response to leucine availability by the T-box transcription antitermination system (8, 9, 15). The common T-box-dependent regulatory mechanism for the ilv-leu operon and the aminoacyl-tRNA synthetase genes (10, 18) could result in the coregulation of these genes. However, regulation of the ilv-leu operon solely by leucine availability would create a potential problem for the cell, because excess leucine could cause the cell to be starved of isoleucine and valine, which are also the end products of the isoleucine-leucine biosynthetic pathway encoded by ilv-leu.

FIG. 1.

Organization and transcription of the ilv-leu operon and deletion analysis of its promoter region. The gene organization of the ilv-leu region has been determined previously (9, 11). The ilv-leu operon, consisting of seven genes (ilvBHC and leuABCD), was transcribed from 482 bp upstream of the coding region for ilvB as a 8.5-kb mRNA; P_ilv-leu is the promoter of ilv-leu. The ilv-leu terminator, designated Ter, is located in the leader region upstream of ilvB, and its presence resulted in a 0.4-kb transcript. The locations of the TnrA box and the introduced base substitutions (A to C and C to T) are indicated. The ysnD gene divergent with respect to ilv-leu was transcribed from the ysnD promoter (P_ysnD); the transcription initiation base corresponds to base −222 (base 1 is the transcription initiation base for ilv-leu). For deletion analysis of the promoter region, promoter regions at bases −248 to 26, −187 to 26, and −667 to 26 were fused with lacZ and then integrated into the amyE locus of B. subtilis, resulting in strains FU676, FU709, and FU710, respectively. The position of the DNA probe used for Northern analysis (l-ilvB) is also indicated.

Recently, global studies on B. subtilis gene expression in response to amino acid availability and subsequent analysis of the expression of the genes for branched-chain amino acid biosynthesis (13, 14, 17) revealed that the ilv-leu operon is downregulated in the presence of the 16 amino acids included in casein, which is independent of the T-box transcription antitermination system but dependent on negative regulation through CodY. Very recently, Shivers and Sonenshein (21) demonstrated that CodY is activated by direct interaction with branched-chain amino acids to regulate its targets. Furthermore, Mäder et al. (13) demonstrated that the ilv-leu operon modulated the amount of the encoded protein through mRNA processing and differential segmental mRNA stability.

In addition to the feedback repression of the ilv-leu operon through CodY, we investigated the negative transcriptional regulation of the ilv-leu operon through another global regulator of nitrogen metabolism (TnrA) (22), initially inferred from DNA microarray data, obtained from analysis of the transcriptomes of the wild type and a tnrA disruptant (28).

TnrA is known to both activate and repress nitrogen-regulated genes during nitrogen-limited growth (24). When nitrogen sources are in excess, the concentrations of intracellular glutamine and other metabolites are thought to become high enough to cause feedback inhibition of glutamine synthetase (GlnA). The feedback-inhibited GlnA captures TnrA to form a protein-protein complex and thereby abolishes the DNA-binding ability of TnrA. By contrast, during nitrogen-limited growth, TnrA is released from the GlnA-TnrA complex and binds to its specific sites on the DNA for the regulation of transcription. Thus, TnrA exerts its regulatory function only in cells grown under nitrogen-limited conditions (24).

In contrast to CodY and TnrA, which are global regulators of nitrogen metabolism, CcpA plays a central role in the control of carbon metabolism, as recently reviewed (4). The ilv-leu pathway is thought to occupy the central position in both protein and fatty acid metabolism through synthesis of branched-chain amino acids. The fact that insufficient expression of the ilv-leu operon limits growth of a ccpA mutant suggested that the control of expression of this operon by CcpA provides a regulatory link between carbon metabolism and amino acid metabolism (12). CcpA forms a complex with P-Ser-HPr, and the complex regulates various operons involved in carbon metabolism (4). The amount of P-Ser-HPr increases as the intracellular concentration of fructose-1,6-bisphosphate increases, which is regarded as a gauge of carbon supply.

In this study, we verified that the expression of the ilv-leu operon is repressed by TnrA, as is the expression of the glnRA and gltAB operons involved in the central pathway of ammonium assimilation (2, 22). The TnrA box responsible for this regulation was located approximately 200 bp upstream of the ilv-leu transcription start site, implying that there is a regulation mechanism that evokes a three-dimensional conformation change in DNA that inhibits the ilv-leu promoter activity. The control of the expression of the ilv-leu operon by TnrA is a novel regulatory link between nitrogen metabolism and amino acid metabolism.

MATERIALS AND METHODS

Bacterial strains and their construction.

The B. subtilis strains used in this work are listed in Table 1. Strain FU659 carrying tnrA::Tn917 was obtained by transformation of strain 168 with chromosomal DNA from strain SF416T to erythromycin resistance (0.3 μg/ml). Strains FU698 and FU657 carrying tnrA::cat and having pMUTIN inserted into ilvB and ysnD, respectively, were obtained by transformation of strains FU382 and BFS2411 with DNA of strain FU459 to erythromycin resistance.

TABLE 1.

Strains used in this work

Strain	Genotype	Reference or source
168	trpC2	1
FU382	trpC2 ilvB::pMUTIN2	17
SF416T	trpC2 tnrA62::Tn917 ΔamyE::(nrgA-lacZ)416(neo)	22
FU459	trpC2 tnrA::cat	JAFANa
FU698	trpC2 ilvB::pMUTIN2 tnrA::cat	This study
BFS2411	trpC2 ysnD::pMUTIN	Micadob
FU657	typC2 ysnD::pMUTIN tnrA::cat	This study
FU659	trpC2 tnrA62::Tn917	This study
FU676	trpC2 amyE::[cat P(ilv-leu)(−248/26)-lacZ]	This study
FU709	trpC2 amyE::[cat P(ilv-leu)(−187/26)-lacZ]	This study
FU710	trpC2 amyE::[cat P(ilv-leu)(−667/26)-lacZ]	This study
FU679	trpC2 tnrA62::Tn917 amyE::[cat P(ilv-leu)(−248/26)-lacZ]	This study
FU712	trpC2 tnrA62::Tn917 amyE::[cat P(ilv-leu)(−667/26)-lacZ]	This study
FU713	trpC2 amyE::[cat P(ilv-leu)(−248/26)(A−207C C−196T)-lacZ]	This study

^aJAFAN, Japan Functional Analysis Network for B. subtilis (http://bacillus.genome.ad.jp).

^bMicado, Microbial Advanced Database Organization (http://locus.jouy.inra.fr/micado).

To construct strains carrying fusions of various regions upstream of ilvB, including bases −248 to 26, −187 to 26, and −667 to 26 (Fig. 1) (base 1 is the transcription initiation base of the ilv-leu operon [9]), with lacZ in the amyE locus, the regions were amplified by using primers D-248F and D+26R, primers D-187F and D+26R, and primers D-667F and D+26R (Table 2), respectively, and DNA of strain 168 as the template. The PCR products were trimmed with XbaI and BamHI and then ligated with the XbaI-BamHI arm of plasmid pCRE-test2 (16). The ligated DNAs were used for transformation of Escherichia coli strain DH5α to resistance to ampicillin (50 μg/ml) on Luria-Bertani medium plates (19). The correct construction of the fusions in the resulting plasmids was confirmed by sequencing. The plasmids were linearized with PstI and then used for double-crossover transformation of strain 168 to resistance to chloramphenicol (5 μg/ml) on tryptose blood agar base (Difco) plates containing 10 mM glucose (TBABG), which produced strains carrying the lacZ fusions FU676, FU709, and FU710, respectively. Strains carrying both the fusions and tnrA::Tn917, FU679 and FU712, were obtained by transformation of strains FU676 and FU710 with DNA of strain FU659 to erythromycin resistance, respectively.

TABLE 2.

Oligonucleotide primers used in this work

Oligonucleotide	Sequencea
D-248F	GCGCGCGCTCTAGATGATCTGTCAGACTCAATC
D-187F	GCGCGCGCTCTAGATATAAAATTAAATAATTCTG
D-667F	GCGCGCGCTCTAGAATATCTAAAGCCTCCGTGTA
D+26R	GCATGCGCGGATCCGTGAAGCTTGCATTTATCTT
MctF	CGTGTTCTAAAATCTTCTATGTTTAT
MagR	ATAAACATAGAAGATTTTAGTTCACG
NlbF	CTGGCCGAAAAAGATGCTAC
NlbR	GCCTCTTCCAGCTTTTCCTT
PE-F	GCTTTGGACGCAGTGTCAACAGCA
PE-R	GGGACAGCAGGATAGTACTGCTG
GP.F	GTGCCTATCCGAAATTGAAATG
GP.R	TTGGTACAAAAGGAATTGCAAA
GD.F	CGAACCATATCATGTTTATATA
GD.R	TATATAAACATGATATGGTTCG

^aThe underlined sequences are the sites for restriction enzymes used for cloning of the resulting PCR products.

To construct strain FU713 with a lacZ fusion carrying two base substitutions (A to C at base −207 and C to T at base −196) in the TnrA box located at bases −211 to −195, the PCR product of the region (bases −248 to 26 [Fig. 1]) carrying the substitutions was obtained as follows. First, two PCR products carrying the two base substitutions at the 5′ and 3′ ends were prepared by using oligonucleotides D-248F and MagR and oligonucleotides MctF and D+26R, respectively, and DNA of strain 168 as the template. Then a second PCR product was prepared by using oligonucleotides D-248F and D+26R and the first two PCR products as templates. The resulting PCR product was fused with lacZ and integrated into the amyE locus, as described above.

Cell growth and β-Gal assay.

B. subtilis cells carrying lacZ fusions were grown at 30°C on TBABG containing erythromycin (0.3 μg/ml) and/or chloramphenicol (5 μg/ml) overnight. The cells were inoculated into 50 ml of MM medium (26) containing 13.6 mM sodium glutamate instead of glutamine as the sole nitrogen source and 50 μg of tryptophan per ml and then incubated at 37°C. During growth, 1-ml aliquots of the culture were withdrawn, and the β-galactosidase (β-Gal) activity in the cells was assayed spectrophotometrically, as described previously (26).

Northern and primer extension analyses.

Total RNA was extracted and purified from B. subtilis cells as described previously (27). For Northern analysis, RNA was electrophoresed in a glyoxal gel and then transferred to a Hybond-N membrane (Amersham) (19). To prepare a probe for detection of the transcript carrying the leader region of ilv-leu, the product amplified by PCR with primers NlbF and NlbR (Table 2) and chromosomal DNA of strain 168 as the template was labeled with a BcaBEST labeling kit (Takara Shuzo Co., Ltd., Kyoto, Japan) and [α-³²P]dCΤP (Amersham). Hybridization and transcript detection were carried out as described previously (19).

Primer extension analysis was performed as described previously (25). Reverse transcription was initiated from the PE-R primer corresponding to bases 181 to 205 (base 1 is the transcription initiation base for ysnD, which had been labeled at its 5′ end by use of a Megalabel kit [Takara Shuzo] and [γ-³²P]ATP [Amersham]). A template for the dideoxy sequencing reactions for ladder preparation starting from the same end-labeled primer was prepared by PCR by using primers PE-F and PE-R (Table 2) and DNA from strain 168 as the template.

Gel retardation and DNase I footprinting experiments.

Gel retardation and DNase I footprinting experiments were performed as described previously (29). The TnrA protein extract used for the experiments was prepared from E. coli strain JM109 cells carrying plasmid pTNRA; the amount of TnrA was 0.8% of the total protein (28). The expression of tnrA cloned in plasmid pUC18 led to only 0.8% of the total protein even if isopropyl-β-d-thiogalactopyranoside (IPTG) was added to the medium (28), so it was not easy to purify active TnrA. Hence, we did not eliminate the possibility that binding of TnrA to the ilv-leu TnrA box required a second factor that should be present in the extract. For gel retardation analysis, the probe DNA was a PCR product amplified from DNA of strain 168 with primers GP.F and GP.R. Probe DNA with the ilv-leu TnrA box deleted was prepared as follows. First, two PCR products carrying a TnrA box deletion at the 5′ and 3′ ends were prepared by using oligonucleotides GP.F and GD.R and oligonucleotides GD.F and GP.R, respectively, and DNA of strain 168 as the template. Then a second PCR product was prepared by using oligonucleotides GP.F and GP.R and the two first PCR products as templates. The probes were labeled with the BcaBEST labeling kit and [α-³²P]dCΤP (Amersham). For DNase I footprinting, probe DNA was prepared by PCR amplification with specific primers GP.F and GP.R, one of which had been labeled at the 5′ terminus with a Megalabel kit and [γ-³²P]ATP (Amersham).

RESULTS

TnrA-dependent negative regulation of expression of the ilv-leu operon.

Recently, candidate TnrA-regulated genes were screened by a DNA microarray analysis to compare the transcriptomes from the cells of a tnrA mutant and the parental wild-type strain (28) in the middle of the logarithmic growth phase under nitrogen-limited conditions, with glutamate as the sole nitrogen source. When the expression ratios with respect to the ilv and leu genes were examined with DNA microarray data (28), which are available on a website (http://www.genome.ad.jp/kegg/expression/), the ilvB, ilvH, ilvC, leuA, leuB, and leuC genes exhibited significantly altered expression ratios (2.4, 2.6, 2.2, 2.7, 1.9, and 2.9, respectively), whereas the ratios for the ilvA and ilvD genes were nearly 0.9; the signal intensities for leuD were too low to obtain a significant ratio. The results implied that expression of the ilv-leu operon comprised of ilvB, ilvH, ilvC, leuA, leuB, leuC, and leuD (Fig. 1) might be under TnrA-dependent negative regulation. To confirm the DNA microarray results, we constructed isogenic wild-type and tnrA::cat strains carrying ilvB::pMUTIN2 to monitor β-Gal activity as an indicator of expression of the ilv-leu operon during growth of cells in MM medium containing sodium glutamate as the sole nitrogen source. As shown in Fig. 2, β-Gal synthesis was derepressed in tnrA::cat cells during the logarithmic phase compared to β-Gal synthesis in the wild-type cells, indicating that the expression of the ilv-leu operon is under TnrA-dependent negative control. (However, we did not further characterize the stationary-phase increase in β-Gal accumulation, especially in a TnrA⁺ background.) To further confirm this negative regulation of the ilv-leu operon, we also performed a Northern analysis using the wild-type and tnrA::cat strains with a probe (l-ilvB) derived from the leader region of this operon (Fig. 3). We clearly observed the band of the 0.4-kb transcript covering only the leader region of the ilv-leu transcript, the density of which was clearly lower for TnrA⁺ (Fig. 3, lane 1) than for TnrA⁻ (lane 2). However, we could not detect the full-length 8.5-kb transcript covering the entire ilv-leu operon, which was rapidly degraded, producing a smear, as shown in lane 2. This observation is consistent with the fact that the half-life of the full-length transcript is 1.2 min (13). Thus, we concluded that the expression of the ilv-leu operon was under TnrA-dependent negative control during logarithmic growth under nitrogen-limited conditions.

FIG. 2.

TnrA-dependent negative regulation of β-Gal synthesis directed by the ilv-leu promoter in B. subtilis carrying pMUTIN integrated into ilvB. Cells of strains FU382 (ilvB::pMUTIN2) (circles) and FU698 (ilvB::pMUTIN2 tnrA::cat) (squares) were grown in MM medium containing 13.6 mM sodium glutamate instead of glutamine as the sole nitrogen source. β-Gal was assayed as described in the text. The open and solid symbols indicate optical density at 600 nm (OD₆₀₀) and β-Gal activity, respectively. OD, optical density unit.

FIG. 3.

Northern analysis of the ilv-leu transcript. RNA samples prepared from cells of strains 168 (wild type) (lane 1), FU459 (tnrA::cat) (lane 2), BFS2411 (ysnD::pMUTIN) (lane 3), and FU657 (tnrA::cat ysnD::pMUTIN) (lane 4), which had been grown in MM medium containing 13.6 mM sodium glutamate instead of glutamine to an optical density at 600 nm of 0.6, were subjected to Northern analysis with an ilv-leu leader-specific ³²P-labeled DNA probe (l-ilvB) (Fig. 1). Each lane contained 10 μg of total RNA. The positions of size markers are indicated on the left. The arrow indicates the position of the 0.4-kb transcript terminating before ilvB.

Identification of a TnrA box located upstream of the promoter of the ilv-leu promoter.

Since the expression of the ilv-leu operon was found to be under TnrA-dependent negative control, we searched for a TnrA box in the ilv-leu promoter region by using a web-based application, GRASP-DNA (20), and the 17 sequences of TnrA boxes listed recently (28). This search revealed a TnrA box-like sequence (TGTTATAAAATCTTCCA) that exhibited a relatively good GRASP score (17.9) at bases −211 to −195 upstream of the transcription initiation base for ilv-leu (base 1), which was mapped previously (9) (Fig. 1). This location is closer to the ysnD gene oriented divergently to ilv-leu than it is to ilvB.

To verify that this sequence was the sequence of a TnrA box, we performed a gel retardation analysis using a crude protein extract of E. coli cells synthesizing B. subtilis TnrA and probes carrying this putative TnrA box and its deletion derivative (Fig. 4). As shown in Fig. 4A, the probe became increasingly retarded as the amount of TnrA-containing extract in the assay mixture was increased, forming a band corresponding to a protein-DNA complex. When the same amount of protein of a crude extract of cells carrying plasmid pUC18 or pTNRA (lanes 1 and 2) was added to the mixtures, retardation was observed only in lane 2, suggesting that it was due to the specific binding of TnrA to the probe. To determine if TnrA interacted with the putative TnrA box, we carried out another gel retardation analysis using probes carrying this putative box and its deletion derivative (Fig. 4B). We observed a clear retarded band with the probe carrying the putative TnrA box (lanes 2 and 3) but not with the probe carrying an internal deletion of this box (lanes 6 and 7). The results indicate that the TnrA box-like sequence actually interacted with TnrA.

FIG. 4.

Gel retardation analysis of TnrA binding to the TnrA box for negative regulation of the ilv-leu promoter. The ³²P-labeled DNA probes were PCR products amplified from the DNA of strain 168 and labeled, and protein extracts were prepared from E. coli JM109 cells carrying pTNRA or pUC18, as described in the text. The gel retardation analysis was performed as described in the text. (A) Lane 1 contained 30 μg of protein from cells carrying pUC18. Lanes 2, 3, 4, 5, 6, 7, and 8 contained 30, 15, 7.5, 3.6, 1.8, 0.9, and 0.45 μg, respectively. Lane 9 contained no protein. One microgram of total protein from the cells carrying pTNRA contained approximately 8 ng of TnrA. The concentrations of TnrA in the assay mixtures were calculated to be 382, 191, 96, 48, 24, 12, and 6 nM, respectively, as native TnrA is a dimer (23, 28). Each lane contained 0.02 pmol of the ³²P-labeled DNA probe. The bands for the probe (free) and the TnrA-probe complex (bound) are indicated by an arrow and an arrow with an asterisk, respectively. The equilibrium dissociation constant for the interaction of TnrA with the TnrA box under our assay conditions was estimated to be approximately 24 nM. (B) Lanes 1 to 4 and lanes 5 to 8 contained the full-length probe and the probe with an internal TnrA box deletion, respectively. Lanes 1 and 5 contained 15 μg of protein from the cells carrying pUC18, while lanes 2 and 6 and lanes 3 and 7 contained 15 and 7.5 μg of protein (191 and 96 nM TnrA) from the cells carrying pTNRA, respectively. Lanes 4 and 8 contained no protein. The bands for the probe and the TnrA-probe complex are indicated as described above.

The interaction between TnrA and the region containing this TnrA box was analyzed further by performing DNase I footprinting experiments with the same DNA fragment that was used for the gel retardation analysis (Fig. 5). A specific interaction on the noncoding and coding strands was observed, as shown in Fig. 5A; the regions protected against DNase I are indicated in Fig. 5A, and the sequences of these regions are indicated in Fig. 5B. As shown in Fig. 5B, the TnrA box is completely included in each of the protected regions of coding and noncoding strands. Notably, the interaction of TnrA with the TnrA box produced not only regions protected against DNase I but also sites hypersensitive to it.

FIG. 5.

DNase I footprinting analysis of the interaction of TnrA with the region containing the TnrA box. (A) The left and right halves of the gel are DNase I footprints of the 5′-end-labeled noncoding and coding strands, respectively, of the probe DNA prepared as described in the text. Lanes 1 to 5 contained 0.04 pmol of the ³²P-labeled probe DNA in the reaction mixture (50 μl). Lane 1 contained no protein extract. Lanes 2 and 3 contained 30 and 15 μg of protein (191 and 96 nM TnrA) from the cells carrying pTNRA, respectively, whereas lanes 4 and 5 contained 30 and 15 μg of protein from the cells carrying pUC18, respectively. Lanes G, A, T, and C contained the products of the corresponding sequencing reactions performed with the same primers that were used for probe preparation. The protected areas of noncoding and coding strands are enclosed in boxes. The arrowheads indicate hypersensitive sites formed through the interaction with TnrA. (B) Nucleotide sequences of the noncoding and coding strands of the region upstream of the ilv-leu promoter. The protected areas are indicated by bars, and the sequence of the TnrA box and the −35 and −10 regions of the ilv-leu promoter are also indicated (+1 indicates the transcription initiation base). The arrowheads indicate hypersensitive sites within the TnrA box.

TnrA-dependent positive regulation of ysnD expression is not involved in the TnrA-dependent negative regulation of ilv-leu.

The proximity of the TnrA box to the divergently transcribed ysnD gene suggested one possible explanation for the observed negative effect that TnrA has on transcription from the ilv-leu promoter: that this effect was accomplished only indirectly via a positive effect of TnrA on ysnD expression. Although the experiments described here indicate that TnrA does indeed positively influence ysnD expression, the phenotype of the ysnD null allele allowed us to eliminate any indirect effects on ilvB expression, as explained below.

To examine if lacZ expression under the control of the ysnD promoter is regulated by TnrA during growth on glutamate as the sole nitrogen source, we constructed strain FU657 (ysnD::pMUTIN tnrA::cat) by introducing tnrA::cat into strain BFS2411 (ysnD::pMUTIN). Figure 6 shows that β-Gal synthesis began to increase in the middle of the logarithmic phase for both strains. However, the amount of β-Gal synthesized was greater in strain BFS2411 (wild type) than in strain FU657 (tnrA::cat), suggesting that ysnD expression might be controlled positively. To determine the 5′ end of the ysnD transcript, we carried out primer extension analysis using the same RNAs that were used for Northern analysis. As shown in Fig. 7, we detected a runoff cDNA band only with RNA from the wild-type strain, confirming that ysnD transcription is under TnrA-dependent positive regulation. Also, the size of the cDNA indicates that ysnD transcription likely started from C at position 115 upstream of the translation initiation nucleotide of ysnD, which corresponds to 222 bp upstream of the ilv-leu transcription initiation base. This location is very close to the TnrA box (bases −211 to −195), which corresponds to 10 to 26 bp upstream of the ysnD transcription initiation site. It was rather difficult to predict the candidate sequences of the −10 and −35 regions recognized by σ^A for the ysnD promoter. In spite of the fact that the TnrA box was located between these two regions, the interaction of TnrA with this box likely caused positive rather than negative regulation of ysnD expression.

FIG. 6.

TnrA-dependent positive regulation of β-Gal synthesis directed by the ysnD promoter in B. subtilis carrying pMUTIN integrated into ysnD. Cells of strains BFS2411 (ysnD::pMUTIN) (circles) and FU657 (ysnD::pMUTIN tnrA::cat) (squares) were grown in MM medium containing 13.6 mM sodium glutamate. β-Gal was assayed as described in the text. The open and solid symbols indicate optical density at 600 nm (OD₆₀₀) and β-Gal activity, respectively. OD, optical density unit.

FIG. 7.

Primer extension analysis of the ysnD transcript. RNA samples prepared from cells of strains FU459 (tnrA::cat) and 168, which had been grown in MM medium containing 13.6 mM sodium glutamate instead of glutamine to an optical density at 600 nm of 0.6, were subjected to primer extension analysis, as described in the text. Lanes 1 and 2 contained the cDNAs extended from the primer by using 50 μg of total RNA from strains FU459 and 168, respectively. Lanes G, A, T, and C contained the products of the corresponding dideoxy sequencing reactions, as described in the text. The part of the nucleotide sequence of the noncoding strand corresponding to the ladder is shown; the transcription initiation base determined in this analysis (+1) and the complement TnrA box (c-TnrA box) are indicated.

However, the positive regulation of ysnD expression through TnrA cannot be the cause of TnrA-dependent negative regulation of ilv-leu, because Northern analysis with the probe derived from the leader region of ilv-leu showed that ysnD disruption did not affect this negative regulation (Fig. 3, compare lanes 1 and 2 [YsnD⁺] and lanes 3 and 4 [YsnD⁻ ]).

Deletion and base substitution analysis of the region upstream of the ilv-leu promoter for the TnrA-dependent negative regulation of ilv-leu in vivo.

To localize the DNA stretch responsible for the TnrA-dependent negative regulation of ilv-leu within the region upstream of the ilv-leu promoter, we transcriptionally fused two DNA fragments containing the ilv-leu promoter with lacZ and integrated the resulting fusions into amyE of strains 168 (wild type) and FU659 (tnrA::Tn917). To monitor lacZ expression under the control of the ilv-leu promoter, samples taken at various times during growth in MM medium containing sodium glutamate as the sole nitrogen source were assayed for β-Gal activity. As shown in Fig. 8, when lacZ expression in strains FU676 (wild type) and FU679 (tnrA::Tn917) containing the lacZ fusion with the DNA fragment carrying the TnrA box and the ilv-leu promoter region (bases −248 to 26 [Fig. 1]) was monitored, TnrA-dependent negative regulation was observed after early logarithmic growth. Essentially the same results were obtained with fusion strains FU710 (wild type) and FU712 (tnrA::Tn917), which carried the ysnD gene, the TnrA box, and the ilv-leu promoter region (bases −667 to 26 [Fig. 1]) (Fig. 8). The results indicated that the region containing the TnrA box and the ilv-leu promoter region (bases −248 to 26 [Fig. 1]) was sufficient for the TnrA-dependent negative regulation of ilv-leu.

FIG. 8.

Analysis of the ilv-leu promoter region for TnrA-dependent negative regulation. Cells of strains FU676 [P(ilv-leu)(−248/26)-lacZ] (circles) and FU710 [P(ilv-leu)(−667/26)-lacZ] (diamonds) and cells of strains carrying tnrA::Tn91, as well as the three corresponding lacZ fusions, FU679 (squares), and FU712 (triangles), were grown in MM medium containing 13.6 mM sodium glutamate. β-Gal was assayed as described in the text. The open and solid symbols indicate optical density at 600 nm (OD₆₀₀) and β-Gal activity, respectively. OD, optical density unit.

To confirm that the TnrA box is responsible for the TnrA-dependent negative regulation of ilv-leu in vivo, we fused only the ilv-leu region with the TnrA box deleted (bases −187 to 26 [Fig. 1]) with lacZ and integrated the resulting fusion into amyE, producing strain FU709, and then we introduced two base substitutions into the TnrA box (A to C at base −207 and C to T at base −196), which resulted in strain FU713. Either deletion of the TnrA box or introduction of the base substitutions into the TnrA box abolished the TnrA-dependent negative regulation of ilv-leu (Fig. 9), indicating that the TnrA box located approximately 200 bp upstream of the ilv-leu promoter is actually involved in this negative regulation.

FIG. 9.

Deletion and base substitution analysis of the TnrA box for negative regulation of the ilv-leu promoter. Cells of strains FU676 [P(ilv-leu)(−248/26)-lacZ] (circles), FU709 [P(ilv-leu)(−187/26)-lacZ] (diamonds), and FU713 [P(ilv-leu)(−248/26)(A−207C C−196T)] (squares) were grown in MM medium containing 13.6 mM sodium glutamate. β-Gal was assayed as described in the text. The open and solid symbols indicate optical density at 600 nm (OD₆₀₀) and β-Gal activity, respectively. OD, optical density unit.

DISCUSSION

Recently, 17 TnrA targets were detected by a combination of DNA microarray hybridization, a genome search for TnrA boxes, and gel retardation assays (28). However, the ilv-leu operon was missed as a TnrA target in this genome-wide target screening, because it was arbitrarily assumed that there was a threefold threshold for the ratio of gene expression intensity in the tnrA mutant to gene expression intensity in the wild type, reflecting a significant alteration in gene expression; the ratios for the genes belonging to the operon were 1.9 to 2.9. We confirmed this TnrA-dependent negative regulation of expression of the ilv-leu operon by means of lacZ fusion and Northern blotting experiments (Fig. 2 and 3). The genome-wide search for TnrA boxes revealed a candidate TnrA box for negative regulation, overlapping the promoter of ysnD, a divergently oriented gene compared to the ilv-leu operon (Fig. 1 and 7). The gel retardation and footprinting analyses indicated that TnrA interacts with this box (Fig. 4 and 5). As shown in Fig. 5A, the interaction of TnrA with the TnrA box produced not only regions protected against DNase I but also sites hypersensitive to it. The DNase I-hypersensitive sites were previously observed at the same relative positions as those in TnrA footprints of the nrgAB (23) and ansZ (7) promoters.

TnrA also regulated ysnD expression positively (Fig. 6 and 7). However, this positive regulation of ysnD was not correlated with the TnrA-dependent negative regulation of ilv-leu, because disruption of ysnD did not affect the negative regulation (Fig. 3). Furthermore, when the DNA fragment carrying the TnrA box and the ilv-leu promoter region (bases −248 to 26 [Fig. 1]) was fused to lacZ and integrated into amyE, lacZ expression was under TnrA-dependent negative regulation (Fig. 8). Also, deletion of the TnrA box (bases −248 to −188) from this DNA fragment and base substitution abolished the TnrA-dependent negative regulation of ilv-leu (Fig. 9). We interpreted these results to indicate that the TnrA interaction with the TnrA box located 200 bp upstream of the ilv-leu transcription start site was able to inhibit the ilv-leu promoter activity. The molecular mechanism underlying the TnrA-dependent negative regulation of ilv-leu was inferred to be distinct from that mediating the TnrA-dependent positive regulation of other targets, such as nrgAB and ansZ, whose TnrA boxes are centered around 50 bp upstream of the transcription start site, as described previously (7, 23, 28).

The DNA fragment fused with lacZ contains only a short stretch of the region downstream of the transcription initiation site (bases 1 to 26), so the inhibition likely occurs before mRNA elongation. As implied by the formation of the hypersensitive site on the footprint (Fig. 5), DNA bending might bring TnrA interacting with the box located well upstream of the ilv-leu transcription start site to the promoter region, where RNA polymerase (and probably another DNA-binding regulatory protein[s] as well) bind. Such an interaction of TnrA with a DNA-binding protein(s), including RNA polymerase, in which DNA bending might occur, most likely causes the inhibition of the ilv-leu promoter activity. The candidate DNA-binding regulatory proteins which are known to regulate the expression of ilv-leu are TnrA (this study), CodY (21), and CcpA (12).

We expected a second TnrA box located closer to the promoter in addition to the TnrA box upstream of it, so a TnrA-TnrA interaction would cause the bending for the negative regulation of ilv-leu. However, we failed to find any candidate TnrA box near the ilv-leu promoter when we performed a GRASP search (20) and a careful sequence examination. Very recently, CodY was found to interact with four sites on the ilv-leu promoter region (21). The most upstream of these CodY sites lies immediately adjacent to the current TnrA box. Thus, we expected that a TnrA-CodY interaction might be involved in this negative regulation. However, the experiments performed with the fusion strains carrying the DNA fragment (bases −248 to 26) fused with lacZ in the CodY⁺ and CodY⁻ backgrounds indicated that CodY did not function under the nitrogen-limited growth conditions used, in which glutamate was the sole nitrogen source, where TnrA functions (Tojo and Fujita, unpublished observations), likely eliminating involvement of CodY in this TnrA-dependent negative regulation. Ludwig et al. (12) found by using Northern blotting that CcpA positively regulates ilv-leu expression, although they did not investigate further whether the ilv-leu operon is a direct target of CcpA for this positive regulation. Preliminary results obtained in our laboratory implied that the ilv-leu operon might be a direct target of CcpA for positive regulation, the binding site (cre) of which is likely located close to the ilv-leu promoter (Satomura, Tojo, and Fujita, unpublished data), so that the unproven interaction of TnrA with CcpA might be involved in the TnrA-dependent negative regulation of ilv-leu by blocking the positive action of CcpA.

The previous studies demonstrated that the ilv-leu operon is regulated in response to leucine availability by the T-box transcription antitermination system (8, 9, 15). Furthermore, Molle et al. (17) reported identification of new CodY-regulated genes based on a combination of chromatin immunoprecipitation, DNA microarray hybridization, and gel retardation assays. The genes for branched-chain amino acid biosynthesis encoded in the ilv-leu operon, including ilvD and ybgE, were among the newly identified direct CodY targets. More recently, it was found that CodY mediates regulation of the ilv-leu operon by GTP and branched-chain amino acids and binds to the ilv-leu promoter region (21). These amino acids increased the affinity of CodY for the ilv-leu promoter. This finding solves a potential theoretical problem for the cell. The previously described T box responds only to the charging of tRNA^Leu (8, 15). If the T box is the sole mechanism of regulation of this operon, cells might be starved of isoleucine and valine when there is excess leucine. The regulation provided by CodY in response to isoleucine and valine availability allows all three branched-chain amino acids to play a role in regulating the transcription of the operon.

In addition to the regulation of the ilv-leu operon by the end products of branched-chain amino acids through CodY described above, this operon was found in this work to be a negatively regulated target of TnrA, another global regulator of nitrogen metabolism. This regulation means that the expression of the ilv-leu operon is also regulated by the availability of nitrogen in the cell, because GlnA feedback inhibited by an intracellular metabolite such as glutamine causes the capture of TnrA to relieve TnrA-regulated genes from regulation by it (24), since glutamine is the major nitrogen donor for a large part of the nitrogenous molecules in the cell and thus serves as a gauge of nitrogen availability (6). Consequently, expression of the glnRA and gltAB operons is under TnrA-dependent negative regulation (2, 22), and both of these operons are involved in the central pathway of ammonium assimilation. When cells grow more slowly under nitrogen-limited conditions than under nitrogen-excess conditions, they need to synthesize relatively small amounts of branched-chain amino acids for protein synthesis, which presumably leads to a reduction in the pool size by repression of the expression of only the ilv-leu operon among the operons necessary for biosynthesis of amino acids other than glutamine and glutamate, possibly because branched-chain amino acids are thought to be in a central position in both protein metabolism and fatty acid metabolism. The intracellular concentration of branched-chain amino acids might serve also as a gauge for nutrient supply (21), which might be rigorously regulated and possibly linked with the cell growth rate, as implied by the link between carbon and amino acid metabolism, so that insufficient expression of ilv-leu limits growth of a ccpA mutant (12).

The cell downregulates ilv-leu expression through TnrA-dependent negative regulation under the growth conditions where TnrA functions, like glnRA and gltAB. Moreover, this downregulation of ilv-leu expression by TnrA is only two- to threefold, which reflects rather fine-tuning of expression of the operon according to nitrogen and nutrient availability. The negative control of expression of the ilv-leu operon by TnrA provides a novel regulatory link between nitrogen metabolism and amino acid metabolism.