Previous Article | Next Article ![]()
Journal of Bacteriology, March 2003, p. 1672-1680, Vol. 185, No. 5
0021-9193/03/$08.00+0 DOI: 10.1128/JB.185.5.1672-1680.2003
Copyright © 2003, American Society for Microbiology. All Rights Reserved.
Kiyoshi Tachikawa,1,
Abraham L. Sonenshein,1* and Mark Strauch2
Department of Molecular Biology and Microbiology, Tufts University School of Medicine, Boston, Massachusetts 02111,1 Department of Oral and Craniofacial Biological Sciences, University of Maryland Dental School, Baltimore, Maryland 212012
Received 18 October 2002/ Accepted 10 December 2002
|
|
|---|
|
|
|---|
RapA is a Spo0F
P phosphatase; its activity therefore interrupts the phosphorelay, and PhrA relieves the blockage (33, 34). Spo0A
P acts as an essential positive regulator of sporulation genes (20). The genes that are turned on by Spo0A
P allow the cell to form the polar division septum characteristic of sporulating cells (26) and to activate a cascade of RNA polymerase sigma factors for transcription of all subsequently expressed sporulation genes (44).
A second kind of signaling depends on nutrient availability (43). It is unclear which specific nutrients in a complex mixture, such as nutrient broth, are sensed by cells, but exhaustion of these nutrients presumably leads to changes in intracellular pools of certain critical metabolites. Conditions that cause a drop in the GTP pool, for instance, are associated with the induction of stationary-phase genes and sporulation (12, 27, 28, 30). A major contributor to such regulation is CodY, a GTP-dependent pleiotropic repressor of stationary-phase genes (35, 39, 40, 42). When the GTP pool drops, CodY becomes inactivated, and its target genes can be expressed (35). Nutrient exhaustion, like peptide signaling, activates sporulation through the Spo0A phosphorelay (21).
One of the activities of Spo0A
P is to repress transcription of abrB (48), a gene whose product is a pleiotropic repressor of early-stationary-phase genes (46, 47). Since AbrB is rapidly turned over (15), inhibition of its synthesis leads to a large decrease in its intracellular concentration and induction of some stationary-phase genes.
The B. subtilis genes encoding the enzymes of the tricarboxylic acid branch of the Krebs citric acid cycle (citrate synthase, aconitase, and isocitrate dehydrogenase) are subject to at least two forms of regulation. During growth in minimal medium, these genes are repressed synergistically when the medium contains a rapidly utilizable carbon source, such as glucose, and a ready source of 2-ketoglutarate, such as glutamine or glutamate (37). When cells are grown in a poorer medium, e.g., with succinate or citrate as the sole carbon source, or when ammonium serves as the sole nitrogen source, the tricarboxylic acid branch genes are derepressed (37). This response fits well with the metabolic roles of the enzymes in ATP generation, production of reducing equivalents, and synthesis of critical biosynthetic intermediates derived from 2-ketoglutarate.
The B. subtilis tricarboxylic acid branch genes are also repressed during rapid exponential growth in broth medium but become induced as cells exhaust certain nutrients (7, 10, 37). All three enzymes of the tricarboxylic acid branch are required for efficient sporulation (5, 9, 13, 14, 22, 38, 53). The metabolic activities of the tricarboxylic acid branch enzymes appear to be as critical for sporulation as for growth in minimal medium, but aconitase may have an additional role in sporulation that is independent of its enzymatic activity (1).
For the B. subtilis aconitase (citB) gene, only one mechanism of regulation has been uncovered to date. Repression of citB during growth in minimal medium containing glucose and glutamine is primarily due to the activity of CcpC, a LysR family transcription factor (23). CcpC binds to the citB promoter region in two locations. The primary binding site is a dyad symmetry element centered at position -66 with respect to the transcription start site (10, 11, 23). A secondary binding site, whose utilization is dependent on the -66 dyad, is located near position -30 (23; S.-I. Kim, C. Jourlin-Castelli, S. R. Wellington, and A. L. Sonenshein, submitted for publication). When CcpC is bound to both sites, the DNA between them bends, occluding access by RNA polymerase (23; Kim et al., submitted). Mutations in the -66 and -30 regions reduce binding of CcpC and derepress citB transcription (10, 11; Kim et al., submitted). In the presence of the inducer, citrate, binding to the -30 region is inhibited, bending is partially relieved, and RNA polymerase enters the promoter and initiates transcription (23; Kim et al., submitted). Glucose affects citB expression by indirectly activating CcpA, a repressor of the citrate synthase (citZ) gene (24), and by inhibiting citrate uptake (49, 50, 52). Glutamine serves as a source of 2-ketoglutarate, which in turn inhibits the activity of the small amount of citrate synthase enzyme that is made in the presence of glucose (H.-J. Kim, unpublished results). Thus, when cells are growing in glucose-glutamine medium, there is insufficient accumulation of citrate to inactivate CcpC, and citB is repressed.
In cells growing in broth medium, however, neither interruption of the ccpC gene nor mutation of the -66 dyad element derepresses citB expression (10). Thus, there must be at least one additional citB regulatory protein that is active in broth-grown cells. Since citB expression is induced in late-exponential-phase cells in broth culture (7, 10, 37) and by treatment of rapidly growing cells with decoyinine, an inhibitor of guanine nucleotide synthesis (7, 30), we suspected that either AbrB or CodY might be the second regulator. We show here that CodY and AbrB as well as CcpC bind to the citB promoter region and that mutations that inactivate CcpC and CodY have small derepressing effects individually on citB expression in broth-grown cells but give nearly total derepression when combined. While AbrB was also capable of binding to the citB promoter region in vitro, an abrB mutation did not contribute to derepression either alone or in combination with codY and ccpC mutations. In fact, disruption of abrB led to a small decrease in citB expression.
|
|
|---|
|
View this table: [in a new window] |
TABLE 1. B. subtilis strains used
|
![]() View larger version (25K): [in a new window] |
FIG. 1. Growth phase-dependent expression of a citB-lacZ fusion in wild-type and codY, ccpC, and abrB mutant strains. ß-Galactosidase activity was measured in wild-type and mutant strains sampled at various times during growth in DS medium. Arrows indicate the end of the exponential growth phase. (A) Strains: AF21 (wild type [wt]), HKB95 (codY), CJB9 (ccpC), HKB98 (codY ccpC), and HKB99 (codY ccpC abrB). (B) Strains: AF21 (wild-type), HKB94 (abrB), HKB95 (codY), HKB96 (abrB codY), and HKB99 (codY ccpC abrB). (C) Strains: AF21 (wild-type), CJB9 (ccpC), HKB94 (abrB), and HKB97 (ccpC abrB).
|
![]() View larger version (26K): [in a new window] |
FIG. 4. Effects on citB expression of citrate synthase, aconitase, and spo0A mutations. Cultures were grown in DS medium, and ß-galactosidase activity was measured in samples harvested at various time points during growth. Arrows indicate the end of the exponential growth phase. (A) Strains: AF21 (wild type [wt]), HKB76 (citA citZ), HKB165 (citB), and HKB166 (citA citZ citB). (B) Strains: AF21 (wild type), HKB168 (spo0A), HKB165 (citB), HKB169 (citB spo0A), and HKB170 (citA citZ citB spo0A).
|
A C-terminal, 6x histidine-tagged version of CodY (CodY-His6) was expressed in Escherichia coli from plasmid pKT1 under the control of the araBAD promoter. To construct pKT1, the codY gene was amplified from the B. subtilis chromosome by PCR with a forward primer that incorporated a SacI site, the codY ribosome-binding site, and the codY start codon and a reverse primer that included the last six codons of codY (including a C-terminal histidine), an additional five histidine codons, a stop codon, and an SphI site. After digestion with SacI and SphI, the PCR product was ligated to SacI- and SphI-treated pBAD30, a vector useful for placing genes under the control of the araBAD promoter (19). An ampicillin-resistant transformant of E. coli strain KS272 (45) had the appropriate insert, as verified by DNA sequencing of the codY open reading frame.
E. coli KS272(pKT1) was grown at 37°C in 2 liters of L broth supplemented with 50 µg of ampicillin per ml. When the culture reached an optical density at 600 nm (OD600) of approximately 0.6, expression of CodY-His6 was induced by the addition of arabinose to a final concentration of 0.2%. After incubation for 6 to 8 h, cells were harvested by centrifugation at 3,800 x g at 4°C for 10 min. The cell pellets were washed once with 15 ml of solution A (50 mM Tris-HCl [pH 8.0], 2 mM disodium EDTA, 1 mM dithiothreitol) supplemented with 1 mM phenylmethylsulfonyl fluoride and resuspended in 15 ml of sonication buffer (200 mM KCl, 50 mM Tris-Cl [pH 7.5], 10% glycerol, 0.1% Igepal [Nonidet P-40], 1 mM phenylmethylsulfonyl fluoride, 2 µM pepstatin A, 0.5 mM dithiothreitol, 0.2 mM disodium EDTA). The cells were lysed by freezing and thawing and then subjected to sonication with a Branson Sonifier cell disrupter, model 200, for five cycles of 30 s each with 15-s rests. Cell debris was removed by centrifugation at 20,000 x g for 20 min at 4°C. The supernatant fluids were collected, and DNA and ribosomes were precipitated by the addition of streptomycin sulfate (0.01 g/ml) with stirring for 2 h at 4°C. The suspension was centrifuged at 20,000 x g at 4°C for 15 min, and the supernatant fluids were dialyzed against 1 liter of sonication buffer without dithiothreitol or disodium EDTA for 2 h at 4°C, with buffer replacement after 1 h.
The dialyzed lysates were mixed with 2 ml of Talon (Clontech) metal (Co+) affinity resin that had been equilibrated with buffer I (20 mM Tris-HCl [pH 8.0], 5 mM ß-mercaptoethanol, 10% glycerol, 1 mM phenylmethylsulfonyl fluoride, 2 µM pepstatin A, and 0.1% Igepal) supplemented with 150 mM KCl and incubated for 20 min at 4°C with tumbling. The resin and bound His-tagged protein were collected by centrifugation at 1,000 x g for 10 min at 4°C and washed sequentially at 4°C with 15 ml each of buffer I containing 500 mM KCl and 5 mM imidazole and buffer I containing 125 mM KCl and 5 mM imidazole. The resin slurry was then added to a 5-ml disposable column (Clontech) and allowed to settle to form a 2-ml column. CodY-His6 was eluted with 5 ml of buffer I containing 125 mM KCl and 50 mM imidazole but lacking phenylmethylsulfonyl fluoride and pepstatin A. Fractions (0.5 ml) were collected and mixed with 5 µl of 100 mM dithiothreitol-100 mM EDTA. All fractions were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The fractions containing CodY-His6 (estimated purity, >95%) were concentrated by ultrafiltration with an Ultrafree- 10,000 NMWL membrane (Millipore) and stored at -20°C (CodY-His6 is stable for more than 2 months at -20°C, as judged by SDS-PAGE and the ability to bind to the dpp promoter).
DNase I footprinting assays.
Three different radioactive DNAs were used for footprinting assays. For the experiment shown in Fig. 5, a PCR product corresponding to positions -320 to +202 relative to the citB transcription start site was synthesized with a 32P-labeled oligonucleotide that primed synthesis of the template (noncoding) strand, an unlabeled oligonucleotide for the complementary strand, and chromosomal DNA as the template. Primer labeling was achieved with T4 polynucleotide kinase (Gibco BRL, Life Technologies) and 150 µCi of [
-32P]ATP (6,000 Ci/mmol; New England Nuclear). For Fig. 6, plasmid pKSI64 (Kim et al., submitted), which contains the citB regulatory region from positions -166 to +123 cloned in pSK- (Stratagene), was the template. The PCR also contained the 32P-labeled T3 primer and the unlabeled T7 primer. As a result, the nontemplate (coding) strand was labeled. The PCR product was purified from a 1% agarose gel with the QIAquick gel extraction kit (Qiagen). For Fig. 7, the target DNA was an EcoRI-HindIII fragment from pAF13 (10) containing the citB promoter region from positions -84 to +36. The DNA fragment was labeled at the 3' end at either the EcoRI or HindIII site with [
-32P]dATP and the Klenow enzyme.
![]() View larger version (136K): [in a new window] |
FIG. 5. DNase I footprint assay of CodY interaction with the citB regulatory region. A 523-bp PCR product, corresponding to positions -320 to +202 with respect to the citB transcriptional start site and labeled at the 5' end of the template (noncoding) strand, was incubated with increasing amounts of purified CodY-His6 in the presence or absence of 2 mM GTP. After treatment with DNase I, the DNA was denatured and subjected to electrophoresis. A set of Sanger sequencing reactions (not shown) were primed with the same oligonucleotide used for synthesis of the template strand of the PCR product and used to establish the positions of the protected bands. Vertical lines on the left indicate regions protected by CodY.
|
![]() View larger version (143K): [in a new window] |
FIG. 6. Simultaneous binding of CcpC and CodY to the citB regulatory region. A PCR product encompassing positions -166 to +123 with respect to the citB transcription start site was labeled at the 5' end on the nontemplate (coding) strand. A set of Sanger sequencing reactions (A, C, G, and T) were primed with the same oligonucleotide used for synthesis of the nontemplate strand of the PCR product and used to establish the positions of the protected bands. The vertical bars indicate regions protected by CcpC or CodY, and the arrowhead points to a DNase I-hypersensitive site typically created by binding of CcpC to the citB promoter region (22). (A) The PCR product was incubated with increasing amounts of CcpC protein in the presence or absence of 900 nM CodY-His6. All reaction mixes also contained 2 mM GTP. After DNase I treatment, the DNA fragments were purified as described in Materials and Methods, denatured, and subjected to electrophoresis. CcpC concentrations (nanomolar) were as follows: lanes 1 and 11, 0; lanes 2 and 12, 12; lanes 3 and 13, 25; lanes 4 and 14, 50; lanes 5 and 15, 100; lanes 6 and 16, 200; lanes 7 and 17, 500; and lanes 8 and 18, 1,000. (B) The PCR product was incubated with 50 nM CcpC and increasing amounts of CodY with or without added GTP (2 mM). CodY concentrations (nanomolar) were as follows: lane 2, 0; lane 3, 300; lane 4, 600; lane 5, 1,000; lane 6, 1,800; and lane 7, 3,000. The sample in lane 1 contained neither CcpC nor CodY.
|
![]() View larger version (76K): [in a new window] |
FIG. 7. DNase I footprint of AbrB binding to the citB regulatory region. (A) The template strand was labeled at its 3' end. (B) The nontemplate strand was labeled at its 3' end. Lanes: 1, 10 µM AbrB; 2, 5 µM AbrB; 3, 1 µM AbrB; 4 and 5, no AbrB. Maxam-Gilbert purine (U) and pyrimidine (Y) sequencing ladders for each strand are shown for reference. The positions of the +1 and -35 nucleotides of the promoter are indicated.
|
-35S]dATP, and loaded on the same gel. After electrophoresis, the gel was dried and analyzed by autoradiography. DNase I footprinting of AbrB binding to the citB promoter region was assayed as described previously (47) except that the binding and digestion reactions were carried out at room temperature. Maxam-Gilbert sequencing reactions were performed by standard procedures (29).
|
|
|---|
|
View this table: [in a new window] |
TABLE 2. ß-Galactosidase activity of the citBp21-lacZ fusion in minimal mediuma
|
Cells lacking AbrB had a lower than normal level of induction during stationary phase (Fig. 1B). Addition of an abrB mutation to cells carrying a ccpC or codY mutation or both also caused a decrease in expression (Fig. 1B and C). An abrB codY double mutant had a particularly low level of citB-lacZ expression in stationary-phase cells (Fig. 1B and C).
Addition of a high concentration of glucose to cells in DS medium caused a delay in citB-lacZ expression. That is, induction occurred at an OD600 of 1.0 rather than at an OD600 of 0.2 (Fig. 2). Repression by glucose was partly relieved in a ccpC mutant strain (Fig. 2), but a codY mutation had no detectable effect (data not shown). A codY ccpC double mutant was highly derepressed in DS-glucose medium (Fig. 2). Thus, CodY contributes to repression of citB in DS-glucose medium, but the effect of a codY deletion is masked by very strong repression by CcpC.
![]() View larger version (17K): [in a new window] |
FIG. 2. Effect of glucose on citB expression. Strains AF21 (wild type [wt]), CJB9 (ccpC), HKB98 (codY ccpC), and HKB99 (abrB codY ccpC) were grown in DS medium supplemented with 2% glucose, and ß-galactosidase activity was measured in samples harvested at various time points during growth. Arrows indicate the end of the exponential growth phase.
|
![]() View larger version (21K): [in a new window] |
FIG. 3. Effect on citB expression of deletion of the CcpC binding site in the citB promoter region. Strains AF23 (citBp23-lacZ), AF24 (citBp24-lacZ), HKB125 (citBp23-lacZ codY), and HKB126 (citBp24-lacZ codY) were grown in DS medium, and ß-galactosidase activity was measured in samples harvested at various time points during growth. Arrows indicate the end of the exponential growth phase. The citBp23-lacZ fusion contains the citB promoter region from positions -84 to +36 with respect to the transcriptional start site. The citBp24-lacZ fusion contains the citB promoter region from positions -67 to +36 and is thus missing the left arm of the dyad symmetry element (9).
|
In wild-type cells growing in DS medium (Fig. 4A), a double mutation in citA and citZ, the two genes that encode citrate synthases in B. subtilis (22), prevented expression of citB-lacZ. Whereas a citB mutation led to higher and more prolonged expression of citB-lacZ, combining the citA, citZ, and citB mutations reduced citB expression to a an undetectable level, showing that citrate accumulation is necessary for both the normal level of expression and hyperexpression in the citB single mutant (Fig. 4A).
Unusually high levels of citrate might cause prolonged and overabundant expression of citB by inactivation of the repressing activity of CcpC. However, a ccpC null mutant did not show prolonged or unusually high expression of citB (Fig. 1). Alternatively, the high intracellular and extracellular accumulations of citrate that occur in a citB mutant are known to prevent activation of the transcription factor Spo0A by chelating Mn2+ and Fe2+ (5). However, deletion of spo0A neither prevented the normal shut-off of citB expression in an otherwise wild-type cell nor suppressed the hyperexpression of citB seen in a citB mutant (Fig. 4B). Thus, neither repression by CcpC nor repression by Spo0A is responsible for limiting the extent of citB induction in wild-type stationary-phase cells.
Binding of CodY to the citB promoter region. A DNA fragment corresponding to positions -320 to + 202 with respect to the citB transcription start site was end labeled and incubated with purified CodY protein. Interaction of the protein and DNA resulted in protection of the DNA against subsequent treatment with DNase I (Fig. 5). Binding of CodY to the dpp promoter and repression of dpp transcription in vitro are enhanced in the presence of GTP (35). As shown in Fig. 5, GTP (2 mM) enhanced binding of CodY 3- to 10-fold. The concentration of CodY that protected 50% of the DNA molecules in the presence of 2 mM GTP was between 60 and 180 nM.
In the presence of GTP, CodY, at a protein concentration of 180 nM, protected from positions -6 to +27 with respect to the transcription start site (Fig. 5). This localization is consistent with repression of a version of the citB promoter region deleted upstream of position -65 (Fig. 3). At higher concentrations of CodY, a second binding site, located between positions -43 and -83, was observed (Fig. 5). When GTP was omitted, protection of the weaker binding site was only seen at very high CodY concentrations (Fig. 5).
Simultaneous interaction of citB promoter region with CodY and CcpC. The secondary site of binding of CodY in the citB promoter region overlaps the dyad symmetry element that is critical for binding of CcpC. Therefore, we tested whether both proteins can bind simultaneously or compete for binding. When CodY was used at 900 nM in the presence of GTP, the addition of increasing amounts of CcpC led to binding of CcpC (as shown by protection of the -30 region and the appearance of a hypersensitive site at position -41) without displacement of CodY from the -6 to +27 site (Fig. 6A). Since virtually all of the DNA molecules were protected in both regions, CcpC and CodY must be binding simultaneously to the same DNA molecules. Formation of the hypersensitive site at position -41 was completely dependent on binding of CcpC to the -66 dyad symmetry element (Kim et al., submitted). Whether binding of CcpC displaces CodY from the -66 region could not be determined from this experiment.
When CcpC was held constant at 50 nM (a concentration that gives only partial protection of the -30 and -66 regions), CodY (in the presence of GTP) had different effects at different concentrations. When CodY was used at 300 nM, both proteins bound simultaneously, with CodY protecting the -6 to +27 region and CcpC protecting the -66 region (and creating its typical hypersensitive site at position -41) (Fig. 6B). At higher concentrations (500 to 1000 nM), however, CodY caused a reduction in hypersensitivity at position -41, which might be due to displacement of CcpC from the -66 region (Fig. 6B). In the absence of GTP, similar effects were seen at higher CodY concentrations. We estimated the intracellular concentration of CodY to be
2.5 µM (R. Shivers and A. L. Sonenshein, unpublished data).
Binding of AbrB to citB promoter region. An abrB mutation reduced the extent of induction of citB as cells approached stationary phase (Fig. 1). To see whether this effect was direct, we tested the ability of purified AbrB protein to interact with the citB promoter region in a sequence-dependent manner. As shown in Fig. 7, AbrB protected the citB promoter region against DNase I from positions -35 to +14. Surprisingly, this pattern of protection suggests that AbrB binds to a region that overlaps the binding sites for both CodY and CcpC.
|
|
|---|
In cells growing in a complex medium, however, at least one additional factor is implicated in citB regulation. Elimination of CcpC or its primary binding site yields only a slight derepression of citB expression in this case. In cells deleted for both ccpC and codY, however, citB expression is almost fully derepressed during rapid exponential growth, yet deletion of codY alone leads to only a slight derepression of citB. These results imply that either CcpC or CodY alone is sufficient to provide substantial repression of citB in cells growing in nutrient broth medium. In other words, induction of citB in such cells only occurs when the cells accumulate enough citrate to inactivate CcpC and become sufficiently depleted of GTP that CodY becomes inactivated. Since recent studies indicate that CodY responds to branched-chain amino acids as well as to GTP (18; R. Shivers and A. L. Sonenshein, unpublished data), the induction of citB presumably depends on the depletion of certain amino acid pools as well.
All genes known to be repressed by CodY are also regulated by other factors. For the dpp operon, the second regulator is AbrB (41); for the gabP gene, it is TnrA (8); for srfAA, it is ComA
P (31, 36); for the ure operon, it is PucR and GlnR (3, 51); for the bkd operon, it is BkdR (6); and for the roc operons, it is RocR (16). Thus, CodY seems to impose a general nutrient-responsive repression that is fine-tuned by regulators that sense other kinds of metabolic signals.
The different roles of CodY in cells grown in defined and complex media have been verified by microarray analysis. Several hundred genes are derepressed by a codY mutation in cells growing in DS medium or in defined medium containing glucose and a mixture of 17 amino acids, but very few genes are affected by a codY mutation during growth in minimal-glucose-glutamine medium (30a).
Curiously, a codY deletion mutant was unable to induce the citB gene in stationary phase to the same level as in wild-type cells. Two general hypotheses could explain this phenomenon. First, the absence of CodY may permit another regulator, such as CcpC, to repress expression more effectively by eliminating a competitor for binding to the citB regulatory region. The highly derepressed expression of citB in a codY ccpC double mutant is consistent with this hypothesis, as are in vitro studies that show that CodY, at high concentrations, interferes with the binding of CcpC. Alternatively, CodY may normally repress, during exponential growth phase, the expression of a gene whose product represses citB during stationary phase. This idea is consistent with the observation that citB expression at the end of the exponential growth phase is transient. These two hypotheses are neither exhaustive nor mutually exclusive.
Cells in DS medium sporulate efficiently after they enter stationary phase, and CodY appears to be the primary monitor of nutrient availability for sporulation. That is, a codY null mutant sporulates in a rich medium in which sporulation would normally be repressed (35). Since aconitase is required for sporulation, it is not surprising that the citB gene is under the control of CodY. The role of aconitase in sporulation is not completely understood. Part of the answer is that, without an active Krebs cycle, the cell is unable to make the ATP, reducing power, and biosynthetic intermediates needed for spore formation. The specific absence of aconitase, however, causes at least one additional defect. Failure to metabolize citrate leads to its accumulation, to a consequent drop in extracellular pH, and to chelation of divalent cations, such as Mn2+ and Fe2+, needed for critical enzymatic activities (5). In addition, B. subtilis aconitase protein appears to have a second, nonenzymatic function as an RNA binding protein (1). Whether this second activity is necessary for sporulation is unknown.
AbrB also appears to be implicated in citB regulation. AbrB binds to the citB promoter region, and an abrB null mutation causes a reduction in citB expression both in stationary-phase cells and in rapidly growing cells in defined medium. This result is surprising, since the binding site for AbrB overlaps the presumed RNA polymerase interaction site. Either AbrB, like MerR (2), stimulates transcription while binding to the promoter region, or the in vivo effect of AbrB is indirect. That is, AbrB may block binding of an as yet unidentified repressor of citB or may control the expression of a gene whose product participates in regulating citB. Since the effect of an abrB mutation was seen in cells carrying mutations in the ccpC and codY genes, the hypothetical target of AbrB cannot be ccpC or codY. Since Spo0A represses AbrB, we might have anticipated that a spo0A mutant strain, which overproduces AbrB (48), would have a higher level of citB expression than a wild-type strain. In fact, such an effect occurs in cells growing in minimal medium (M. Strauch, unpublished data), but we saw no such effect in DS medium.
The factor(s) responsible for reestablishing repression of citB in stationary-phase cells in DS medium remains unknown. Inactivation of this factor may be responsible for the hyperexpression of citB seen when citrate accumulates. Our results rule out CcpC, CodY, AbrB, and any protein whose synthesis depends on Spo0A. Aconitase itself can also be ruled out as a negative autoregulator, since citB hyperexpression also occurs in a citC (isocitrate dehydrogenase) mutant. (Aconitase protein is overproduced in a citC mutant but fails to block expression of a citB-lacZ fusion [data not shown]. Aconitase also cannot be a positive autoregulator, since hyperexpression was seen in a citB null mutant.) If the putative additional factor acts alone, it should be possible to isolate mutant strains in which citB expression is abnormally high during the late stationary phase in DS medium.
This work was supported by research grants from the U.S. Public Health Service to A. L. Sonenshein (GM36718 and GM42219) and to M. A. Strauch (GM46700).
Present address: Corporate-Sponsored Research and Licensing, Massachusetts General Hospital, Charlestown, MA 02129. ![]()
Present address: Torrey Mesa Research Institute, Syngenta, La Jolla, CA 92121. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»