Bacillus subtilis Phosphorylated PhoP: Direct Activation of the Eσ^A- and Repression of the Eσ^E-Responsive phoB-P_S+V Promoters during Pho Response

Bacterial strains and growth conditions.All strains, plasmids, and primers used in this study are listed in Table 1. Bacterial stocks maintained at −80°C were used to inoculate LB agar plates for Escherichia coli strains or tryptose blood agar base containing 0.5% glucose for B. subtilis strains. The media were supplemented with selective antibiotics at the following concentrations: tetracycline, 10 mg/liter; erythromycin, 0.25 mg/liter; chloramphenicol, 5 mg/liter; and ampicillin, 200 mg/liter. To obtain isogenic wild-type and mutant B. subtilis strains harboring the required phoB promoter-lacZ reporter fusion, three pDH32 derivatives, pRC695, pRC696, and pPS2, containing the phoB promoter regions (Fig. 1) (relative to the translation start codon at position 1) at positions −178 to 34 (P_S+V), −118 to 34 (P_V), and −432 to −87 (P_S), respectively, were constructed (9) and used for transformation of wild-type (JH642), phoPR (MH5913), and sigE (EU8701) strains. Before transformation, plasmids pRC695 and pRC696 were linearized by ScaI digestion, and plasmid pPS2 was linearized by NruI digestion. Transformants were selected for chloramphenicol resistance and amyE phenotypes. The resultant nine strains were grown in low-phosphate defined medium (LPDM) for P_i starvation conditions or in modified Schaffer sporulation medium (43 mM P_i) amended with 0.1% glucose (SSG medium) for sporulation induction in P_i-replete conditions.

• Open in new tab
• Download powerpoint

FIG. 1.

Chromosomal organization of the phoB-ydhF operon and its promoter deletions. Genes are represented by arrows that indicate the direction of transcription. The original DNA clone containing the ydhG-phoB intergenic region was fused with a promoterless lacZ gene in pDH32, resulting in plasmid pCB619. The physical locations of the phoB-P_V and phoB-P_S transcription start sites (arrows), the putative σ^A and σ^E consensus −10 and −35 recognition sequences (solid boxes), and the PhoP core binding region (open boxes) are indicated. The line separation was used to reduce the physical distance between bases −178 and −361 (the putative location of the ydhG promoter) relative to the phoB translational start site at position 1. For promoter deletion analysis various DNA fragments containing the P_S and/or P_V transcription start sites were fused with a promoterless lacZ gene in pDH32, resulting in plasmids pRC695, pRC696, and pPS2, as indicated.

To construct a plasmid for overexpression of the mature active form lacking the transcription-inhibiting prosequence (25) of the σ^E protein (*σ^E), the sigE gene lacking the region encoding the prosequence was PCR amplified using JH642 chromosomal DNA with primers FMH490 and FMH491 containing cleavage sites for NdeI and BamHI, respectively. The PCR-amplified fragments were cloned into plasmid pCR2.1, resulting in plasmid pCH201. The mutant sigE gene was then released from pCH201 by digestion with NdeI and BamHI and cloned into plasmid pET16b at the same sites to generate plasmid pCH202. The accuracy of all plasmid constructs was confirmed by DNA sequencing. Plasmid pCH202 was used for transformation of E. coli BL21(DE3)/pLysS, and representative transformants were selected for the chloramphenicol resistance phenotype.

Enzyme assays.For determination of total alkaline phosphatase specific activity, 250 μl of an LPDM culture was added directly to 1 ml of the substrate, 0.1 M p-nitrophenyl phosphate in 1 M CHES (N-cyclohexyl-2-aminoethanesulfonic acid; pH 9.5), and the reaction rates were measured at an optical density at 420 nm (OD₄₂₀). One unit of APase activity was equivalent to 1 μmol of p-nitrophenol released per min at 37°C. The APase specific activity was expressed in units per unit of OD₅₄₀.

The β-galactosidase (β-Gal) specific activity was measured by the method of Ferrari et al. (17) using o-nitrophenyl-β-d-galactopyranoside as the substrate. One unit of β-Gal activity was equivalent to 0.33 nmol of o-nitrophenol released per min at 37°C. The β-Gal specific activity was expressed in units per milligram of total cellular protein. The amount of B. subtilis JH642 total cellular protein was calculated as previously described by Eder et al. (14).

Overexpression and purification of proteins.For overexpression of _His10PhoP, _His10*σ^E, _GST*PhoR, and _His6σ^A proteins, E. coli BL21(DE3)/pLysS was used as a host for plasmids pCH128, pCH201, pLS21, and pLC2, respectively (Table 1). Overexpression and purification of His-tagged and glutathione S-transferase (GST)-tagged proteins were performed as described previously (29). *PhoR is the soluble cytoplasmic portion of the PhoR protein. The His tags were cleaved by factor Xa treatment, whereas the GST tag was removed by thrombin cleavage (29). For preparation of the B. subtilis His₆-tagged RNA polymerase (_His6RNAP) core enzyme (E), B. subtilis strain MH5636 (38) was grown in 12 1-liter portions of LB medium. Cells were collected at the end of the exponential phase of growth. Cell lysis and RNAP core enzyme purification were performed as previously described (38).

Gel mobility shift assay.For preparation of probes, the phoB-P_S+V promoter fragment was amplified by PCR using B. subtilis JH642 genomic DNA as the template and primers FMH745 and FMH746, whereas the phoB-P_S promoter fragment was PCR amplified using plasmid pPS2 as the template and primers FMH124 and FMH128 (FMH124 and FMH128 are homologous to DNA in pDH32 adjacent to the EcoRI and BamHI cloning sites, respectively.). ³²P labeling was performed during PCR by incorporating [α-³²P]dATP and/or [α-³²P]dGTP (0.2 μCi/μl) into the newly synthesized DNA strand in the presence of limited dATP and/or dGTP concentrations (20 μM, usually one-tenth the concentration of dCTP and dTTP). The ³²P-labeled probes were ethanol precipitated and extracted from a 1.2% agarose gel using a gel extraction kit (QIAGEN), and the final concentration was adjusted to 0.1 μM in nuclease-free transcription buffer (NFTB) (see below). The binding reactions (8 μl) were performed in NFTB containing, in order of addition, 12 nM ³²P-labeled DNA (∼15,000 cpm), 180 μM ATP, 0.18 μM *PhoR, and PhoP at concentrations ranging from 0.35 to 2.8 μM. The reaction mixtures were incubated at 37°C for 30 min and were subsequently loaded onto 5% nondenaturing polyacrylamide gels (acrylamide/bisacrylamide, 29:1), which had been prerun in Tris-borate-EDTA buffer at 100 V/10°C for 30 min, and electrophoresed for 2 h at 100 V/10°C. The free probe and shifted PhoP∼P-DNA complex on dried gels were visualized using a PhosphorImager (Molecular Dynamics).

In vitro transcription analysis.All components used for in vitro transcription reactions were adjusted to the required concentrations by dilution in NFTB (Tris-HCl [pH 7.8], 10 mM; KCl, 50 mM; MgCl₂, 5 mM; CaCl₂, 1 mM; EDTA-Na₂, 0.1 mM) containing 5% glycerol, unless indicated otherwise. To obtain the template DNA for in vitro transcription analysis, the full-length phoB-P_S+V promoter region (positions −222 to 106 relative to the translation start codon at position 1 [see Fig. 4A]) was PCR amplified using JH642 chromosomal DNA and primers FMH745 and FMH746 (Table 1), and the ykoL promoter region (positions −106 to 151 relative to the transcription start site at position 1 [see Fig. 6A]) was PCR amplified using primers FMH769 and FMH768. The phoB-P_S promoter region was obtained by PCR using plasmid pPS2 as the template and primers FMH124 and FMH128. Promoter DNA was extracted from a 1.2% agarose gel using a gel extraction kit (QIAGEN), ethanol precipitated, and resuspended in NFTB to a final concentration of 100 nM. To prepare RNAP holoenzyme, core RNAP was incubated with each σ factor on ice for 30 min. For each transcription reaction (final volume, 15 μl), template DNA (33 nM) was mixed with ATP (180 μM), *PhoR protein (0.18 μM), and PhoP protein at different concentrations (0.02 to 10 μM in twofold increments) and incubated at 37°C for 20 min. *PhoR and PhoP dilutions were prepared in phosphorylation buffer (HEPES [pH 8.0], 50 mM; KCl, 50 mM; MgCl₂, 50 mM; 7% glycerol) prior to addition to the reaction mixtures. To produce various concentrations of PhoP∼P in the in vitro transcription reaction mixtures, PhoP phosphorylation was performed during the initial DNA binding step using *PhoR at a fixed concentration, 0.18 μM. After the binding reaction, the reconstituted RNAP holoenzyme (core enzyme concentration, 0.09 μM; σ factor concentration, 0.7 μM) was added to the reaction mixture; this was followed by addition of a nucleoside triphosphate mixture (180 μM each of ATP, GTP, and CTP, 18 μM UTP, 0.1 μM [α-³²P]UTP [3,000 Ci/mmol; 10 μCi/μl], and 1 U/μl RNasin RNase inhibitor in NFTB) and incubation at 37°C for another 30 min. In vitro transcription reactions were stopped by addition of 7.5 μl stop buffer (7 M urea, 100 mM EDTA-Na₂, 0.05% xylene cyanol, 0.05% bromophenol blue, 5% glycerol), and the mixtures were heated at 75°C for 5 min. Ten microliters of each reaction mixture was loaded onto a sequencing gel (6% polyacrylamide [19:1], 8 M urea) and electrophoresed at 350 V for 1 h. The RNA transcripts on dried gels were visualized with a PhosphorImager (Molecular Dynamics) and subjected to quantitative analysis using the volume report function of the ImageQuant software, version 5.1, with corrections for background by subtraction of intensities below and above each specified band. The radiolabeled RNA marker was prepared according to the manufacturer's instructions (Novagen).

Estimation of cellular PhoP concentrations by Western immunoblotting.Protein concentrations were determined using the Bio-Rad protein assay (Bio-Rad Laboratories) with bovine serum albumin as the standard. For preparation of a PhoP standard curve, dilutions of purified standard PhoP were prepared in phoPR mutant cell lysates and subjected to immunoblot detection using PhoP_CTD-specific antibody (a polyclonal antibody developed against the PhoP C-terminal DNA binding domain) as previously described (8). To determine the cellular PhoP concentrations, wild-type strain MH6143 was grown in LPDM, and sampling began when the OD₅₄₀ reached 0.5 ± 0.1. Culture samples (15 ml) were collected hourly from T₋₂ to T₅ during phosphate starvation (where T₀ represents sporulation stage 0). Microscopic cell counting of culture samples was performed using a Petroff-Hausser counter (Hausser Scientific Partnership) and a phase-contrast microscope at a magnification of ×400. Whole-cell extracts were prepared by resuspension of cell pellets in appropriate volumes of lysis buffer (50 mM Tris-HCl [pH 7.0], 10 mM EDTA, 15 mg/ml lysozyme, 10 μg/ml DNase I, 100 μg/ml RNase A, 1 mM phenylmethylsulfonyl fluoride) to bring the final cell concentration to 5 × 10⁷ cells/μl and were incubated at 37°C for 15 min; this was followed by addition of sodium dodecyl sulfate (SDS) sample buffer (0.5 M Tris-HCl [pH 7.0], 12% SDS, 0.1% bromophenol blue, 20% 2-mercaptoethanol, 30% glycerol). Samples were incubated in a boiling water bath for 10 min prior to loading. Efficient cell lysis was monitored microscopically. Samples were diluted in phoPR cell lysates to obtain a final concentration equivalent to 2 × 10⁷ cells/μl. From calibrated samples, 4 μl (equivalent to lysate from 8 × 10⁷ cells) or 10 μl (equivalent to lysate from 2 × 10⁸ cells) was electrophoresed on 12% SDS-polyacrylamide gel electrophoresis gels and subjected to immunoblot detection as previously described by Chen et al. (7). Developed blot membranes were scanned using a computer scanner (MICROTEK ScanMaker 5), and bands were quantified using the volume report function of the ImageQuant software. Band intensities were normalized by using standard PhoP protein applied to separate lanes in the same gel. For estimation of cellular PhoP concentrations, we assumed that the cell volume was 10 μm³ (44).

Determination of inorganic phosphate concentration.The extracellular inorganic phosphate concentrations were determined for LPDM cultures using the method of Ames and Dubin (1). For analysis, 300 μl of properly diluted culture supernatant (usually 10% in deionized H₂O for an LPDM culture) was mixed with 700 μl of ascorbic acid reagent (10% ascorbic acid in H₂O and 0.4 g of ammonium molybdate liter in 1 N H₂SO₄, mixed at a ratio of 1:6 [vol/vol]), and the reaction mixtures were incubated at 45°C for 20 min. The inorganic phosphate concentration (μM) was equivalent to the OD₈₂₀ of the reaction mixture divided by the slope obtained from a P_i standard curve. Under our experimental conditions the slope of the P_i standard curve was ∼0.008 OD₈₂₀ unit/μM.

Both P_V and P_S are expressed during phosphate starvation, and a phoPR mutation or a sigE mutation affects both promoters, although differently.During growth in LPDM, the parental wild-type strains containing each of the three phoB promoter fusions induced expression from both the P_V and P_S promoters (Fig. 2A). Expression from the P_V or P_S+V promoter fusions was induced at T₀, when the P_i concentration was limiting (≤100 μM), and peaked at T₅, when maximum expression levels of ∼6,000 β-Gal units were detected. Expression from the P_S promoter fusion was induced at T₄, and the levels were >50-fold lower than the levels from the P_V promoter; however, they failed to reach the maximum levels in the first 12 h of growth. No expression from the P_V promoter fusion was observed in the phoPR mutant grown in LPDM (Fig. 2B), indicating that the PSI expression from the strong P_V promoter is phoPR dependent. In addition, phoPR deletion affected the timing of induction from the weak P_S promoter. Expression from the P_S promoter in phoPR mutants, from either the P_S or P_S+V promoter fusions, exhibited a further 2- to 3-h lag compared to the wild-type strain. During growth of the three sigE mutant strains in LPDM (Fig. 2C), no expression was observed from the P_S promoter fusion, indicating that the PSI expression from the P_S promoter is sigE dependent. Like expression in the wild-type strains, expression from the P_V promoter in the sigE mutants, from either the P_V or P_S+V promoter fusions, exhibited the strong induction pattern that was induced at T₀ and peaked at T₅ during phosphate starvation. This expression was ∼40% higher than that in the parental strains.

Sporulation-inducible phoB expression is solely from the sigE-dependent P_S promoter.During growth of wild-type, phoPR, or sigE mutant strains in SSG medium (P_i-replete conditions), no expression was observed from the strong P_V promoter fusion (Fig. 2D through F), indicating that the sporulation-inducible expression of phoB is from the weak P_S promoter only. Both the wild-type and phoPR mutant strains (Fig. 2D and E, respectively) induced expression from the P_S promoter, either from the P_S+V or the P_S promoter fusions, at T₄ and peaked at T₆, when the maximum expression levels (<80 β-Gal units) were detected. No expression from the P_S promoter fusion was observed during growth of the sigE mutant strains in either SSG medium (Fig. 2F) or LPDM (Fig. 2C).

phoPR mutation affects the level and timing of expression from the P_S promoter during phosphate limitation.To determine the extent to which a mutation in phoPR affected the timing and inducing strength of the P_S promoter under phosphate deficiency stress conditions, we compared the parental and phoPR mutant strains for growth and expression from the P_S promoter fusion for 24 h in LPDM (Fig. 3). In the parental strain, expression from the P_S promoter fusion was induced at T₄ and peaked at T₈ with a maximum expression of ∼110 β-Gal units. Expression from the P_S promoter fusion was delayed 2 to 3 h in the phoPR mutant strain, and the maximum expression level was reduced about 25%. In addition, the maximum expression levels detected from the P_S promoter in LPDM (Fig. 3) were approximately twofold greater than the level detected in SSG medium (Fig. 2D), suggesting that medium composition also influences the strength of expression from the sigE-dependent P_S promoter.

• Open in new tab
• Download powerpoint

FIG. 3.

Expression of phoB-P_S lacZ in wild-type and phoPR mutant B. subtilis strains during 24 h of growth in LPDM. Expression of P_S-lacZ was determined in wild-type strain MH6143 (squares) and phoPR MH6146 (triangles) grown under phosphate-limiting conditions in LPDM. Solid symbols, growth; open symbols, β-galactosidase specific activity. Time zero was the transition from the exponential to the stationary phase of growth.

phoB-P_S promoter is an Eσ^E-responsive promoter whose transcription in vitro is inhibited by high PhoP∼P concentrations.To determine if the role of PhoP-PhoR and the mother cell specific sigma factor (σ^E) in transcription of the phoB-ydhF operon was direct or indirect, an in vitro transcription analysis was performed using a 325-bp linear DNA fragment containing the full-length phoB promoter (102 bp upstream of P_S [position 1] and 142 bp downstream of P_V [position 1]) (Fig. 4A). The expected transcript initiating from the P_S transcription start site was 223 nucleotides long. As shown in Fig. 4B, when PhoP was not present in the in vitro transcription reaction mixtures (lanes 1 through 8), σ^E-containing RNA polymerase (Eσ^E) was sufficient for transcription initiation from the phoB-P_S promoter. No detectable P_S transcripts were observed when either core RNA polymerase or σ^E was not present in the in vitro transcription reaction mixtures (data not shown). Eight reaction mixtures (lanes 1 to 8) containing decreasing concentrations of core RNAP (0.18 to 0.01 μM) and increasing concentrations of σ^E (0.09 to 0.18 μM) yielded similar concentrations of transcript (lanes 1 to 4). Eσ^E (molar ratio, 0.09:0.7) was used in the following studies. When PhoP was present in the in vitro transcription reaction mixtures at concentrations ranging from 0.02 μM to 2.8 μM (Fig. 4B, lanes 9 through 16), no significant effect on transcription from the σ^E-dependent P_S promoter was observed at PhoP concentrations of ≤0.35 μM (lanes 9 through 13), but concentrations of >0.7 μM resulted in a reduction in transcription. In the presence of *PhoR, PhoP concentrations from 0.02 μM to 2.8 μM (lanes 17 through 24) resulted in greatly reduced transcription from the P_S promoter. Thus, PhoP was not required for transcription initiation from the P_S promoter, but the presence of low concentrations of PhoP∼P or high concentrations of PhoP repressed transcription from the σ^E-dependent P_S promoter.

• Open in new tab
• Download powerpoint

FIG. 4.

In vitro transcription analysis of the phoB promoter. (A) Nucleotide sequence of the phoB promoter coding strand. Regulatory elements are indicated as follows: arrows, transcription start sites for P_S and P_V; subscripts, −10 sequences for both promoters and −35 sequence for the P_S promoter; superscripts, consensus PhoP tandem binding sequences; SD, Shine-Dalgarno sequence; Met, methionine (translational start codon). The lowercase letters indicate mismatches with the consensus sequence. The thick line and the dashed line indicate the DNase I-protected regions for PhoP(∼P) on the coding and noncoding strands, respectively, and the arrowheads indicate the previously identified hypersensitive sites in the footprints. Primers FMH746 and FMH745, used to amplify the phoB promoter, are indicated by arrows (5′ to 3′ direction) for the coding and noncoding strands, respectively. The numbers below the sequences indicate positions relative to the phoB translational start site (position 1). The numbers above the sequences indicate positions relative to the transcription start site of the P_S or P_V promoter. (B) In vitro transcription analysis of phoB promoter using various concentrations of the core (E) and sigma factor (σ) (lanes 1 through 8) in the reconstituted RNAP. (Upper panel) Eσ^E; (lower panel) Eσ^A. Linearly increasing concentrations of unphosphorylated PhoP (in twofold increments) were used in both the Eσ^E- and Eσ^A-driven transcription reactions in the absence of *PhoR (lanes 9 through 16) or in the presence of *PhoR at a final concentration of 0.18 μM (lanes 17 through 24). The protein concentrations (μM) are indicated above the lanes. Lane M contained a radiolabeled RNA marker. nts, nucleotides.

PhoP∼P and Eσ^A requirement for transcription from the phoB-P_V promoter in vitro.For in vitro transcription analysis of the P_V promoter, we used the reaction conditions described above, except that we initiated transcription with σ^A-saturated RNAP (Eσ^A). The expected transcript initiating from the P_V transcription start site (position 1) was 142 nucleotides long. As shown in Fig. 4B (lower panel), in the absence of PhoP, σ^A-containing RNA polymerase (Eσ^A) failed to stimulate transcription from the phoB promoter (lanes 1 through 8) at any concentration used. Addition of PhoP at concentrations ranging from 0.02 to 2.8 μM (lanes 9 through 16) stimulated a dose-dependent increase in transcription from the P_V promoter only when the unphosphorylated PhoP concentration was ≥0.18 μM (lanes 12 through 16). When *PhoR was present in the in vitro transcription reaction mixtures (lanes 17 through 24), PhoP concentrations ranging from 0.02 to 0.18 μM (lanes 17 through 20) stimulated a dose-dependent increase in transcription from the P_V promoter, and the maximal yield was obtained at a PhoP concentration of 0.18 μM (lane 21). Further increases in PhoP concentrations (lanes 21 through 24) resulted in reduced transcription. The use of PhoP at concentrations of ≥1.4 μΜ significantly inhibited transcription activation from the P_V promoter (lanes 23 and 24).

PhoP inhibition of P_S promoter transcription is specific and involves PhoP binding to the P_S promoter in addition to the P_V promoter.A phoB promoter fragment containing only the P_S promoter (from pPS2 [Fig. 1]) was used as a template for in vitro transcription to determine if reduction of the P_S transcript from the phoB-P_S+V promoter in the presence of phosphorylated PhoP was exclusively the result of PhoP binding to the P_V promoter core binding region, which blocked RNAP passage (Fig. 5A, top panel). Lane 1 contained the transcript from the Eσ^E-dependent P_S promoter without PhoP or *PhoR. Lanes 2 through 6 contained increasing concentrations of PhoP (0.09 to 1.4 μM) plus *PhoR, which were shown to increasingly inhibit P_S transcription from the phoB-P_S+V promoter in the presence of *PhoR (0.18 μM) and ATP (Fig. 4B, upper panel). Under these conditions higher concentrations of PhoP increasingly inhibited the P_S promoter, although not as strongly as identical concentrations did when the phoB-P_S+V promoter was used as the template. A dominant role for PhoP∼P in P_S inhibition was suggested by the strong transcript in the reaction with 2.8 μM PhoP in the absence of *PhoR (lane7). Quantitation of the P_S transcript with increasing concentrations of PhoP plus *PhoR from the P_S template alone is compared in Fig. 5A (bottom panel) to quantitation of the transcripts from the phoB-P_S+V promoter fragment (Fig. 4B) with the same concentrations of PhoP or PhoP and *PhoR or with Eσ^E alone. Note the difference in P_S transcription in the reaction mixture containing 1.4 μM PhoP with the full-length P_S template in the presence of *PhoR.

• Open in new tab
• Download powerpoint

FIG. 5.

Effect of deleting the downstream direct repeats on PhoP∼P-mediated repression of transcription from the σ^E-dependent phoB-P_S promoter in vitro. (A) The upper panel shows in vitro transcription of the phoB-P_S promoter lacking the downstream direct repeats, using Eσ^E alone (lane 1) and various concentrations of PhoP∼P (lanes 2 through 6) or unphosphorylated PhoP (2.8 μM) (lane 7). The histogram in the lower panel shows the relative amounts of phoB-P_S transcripts in vitro (in PhosphorImager output units) obtained using the P_S+V promoter with unphosphorylated PhoP (open bars) or PhoP∼P (solid bars) and using the P_S promoter with PhoP∼P (shaded bars). For comparison, the maximum transcript output obtained from the P_S+V or P_S promoter with Eσ^E alone was normalized to 100%. (B) Specificity of the PhoP∼P-mediated repression of transcription from the phoB-P_S promoter. The phoB-P_S promoter and the spoIIID promoter (a PhoP∼P-insensitive, σ^E-dependent promoter) were used as templates for in vitro transcription reactions using Eσ^E alone (lane 1), Eσ^E with unphosphorylated PhoP (lane 2), and Eσ^E with PhoP in the presence of a low *PhoR concentration (0.18 μM) (lane 3) or a high *PhoR concentration (0.35 μM) (lane 4). (C) PhoP∼P binds to the phoB-P_S promoter independent of the downstream direct repeats. Gel mobility shift assays were performed using a 325-bp ³²P-labeled phoB-P_S+V or 500-bp ³²P-labeled phoB-P_S promoter fragment in the absence of PhoP∼P (lane 1) or in presence of various PhoP∼P concentrations (lanes 2 through 8). The positions of the free DNA probe (Fp) and shifted PhoP∼P-DNA complex (C_PhoP∼P) are indicated on the right. The protein concentrations (μM) are indicated above the lanes. nts, nucleotides.

To determine if inhibition of P_S transcription by high concentrations of PhoP and *PhoR was specific, we compared P_S transcription to transcription with another Eσ^E-dependent promoter, spoIIID, under conditions described above (Fig. 5A, bottom panel) that inhibited the phoB-P_S+V promoter. Lane 1 in Fig. 5B shows the level of transcript from the phoB-P_S (top panel) or spoIIID (lower panel) template with Eσ^E alone. Eσ^E with PhoP (2.8 μM) resulted in some decrease in the P_S promoter transcript, but in the presence of *PhoR (0.18 μM) and PhoP the transcript was completely inhibited, similar to the results obtained for the phoB-P_S+V promoter (Fig. 4B, top panel, lane 24). Under the same conditions and when the *PhoR concentration was increased from 0.18 to 0.35 μM, the level of the spoIIID transcript (lanes 3 and 4, respectively) was similar to the level observed with Eσ^E without PhoP and *PhoR (lane 1), indicating that the role of PhoP and *PhoR in inhibition of the phoB promoter is specific.

Because transcription from the P_S promoter was reduced in the presence of PhoP∼P, we asked if PhoP directly bound to the P_S promoter fragment in a gel mobility shift assay and how the binding compared to PhoP binding to the full-length phoB promoter (Fig. 5C). Lane 1 in Fig. 5C contained the full-length phoB-P_S+V free probe in the top panel and the P_S free probe in the bottom panel. Lanes 2 to 8 in each panel contained increasing concentrations of PhoP in the presence of ATP and 0.18 μM *PhoR. Both probes revealed a partial shift at the lowest concentration of PhoP (lane 2), although the percentage of the phoB-P_S+V probe shifted was significantly greater than the percentage of the P_S probe shifted. Roughly twofold-higher concentrations of PhoP were required for a complete shift of the phoB-P_S probe (1.4 μM PhoP) (Fig. 5C, bottom panel, lane 6) than for a complete shift of the phoB-P_S+V probe (0.7 μM PhoP) (Fig. 5C, top panel, lane 4). These data indicate that PhoP binds directly to the P_S promoter independent of the core binding region in the P_V promoter previously reported. In light of the current data, we reexamined the PhoP footprinting protection on the phoB-P_S+V promoter (29). While nearly complete protection of the large core binding region of the P_V promoter was observed even at the lowest PhoP∼P concentration (55 nM), there was also a previously unnoticed, smaller protected region upstream of the P_s transcription start site on the noncoding strand. Partial protection was observed starting at 110 nM PhoP∼P, and there was increasing protection at PhoP∼P concentrations of 220 and 440 nM. The protected region in the promoter sequence is indicated in Fig. 4A. Interestingly, the protected region on the noncoding strand contains two putative consensus repeats for a PhoP dimer-binding sequence on either side of the −10 element of the Eσ^E-responsive P_S promoter (⁻²TAAATAATGGTTATTCTTT⁻¹⁹), with a conservation match of 5/6 and 4/6, respectively. PhoP protection at this P_S promoter site required PhoR and ATP for PhoP phosphorylation, just as retardation of the P_S promoter probe in the studies described above did (data not shown).

PhoP activates in vitro transcription from the phoB-P_V promoter in a phosphorylation- and/or concentration-dependent manner.The concentration range of PhoP used was expanded up to ∼10 μM to compare the maximum transcription of P_V with PhoP compared to that with PhoP∼P and to determine if high concentrations of PhoP inhibited transcription from the P_V promoter (Fig. 6A). Because these experiments were designed to investigate the effects of various unphosphorylated PhoP and PhoP∼P concentrations on transcription from the P_V promoter, the only experimental variable was the PhoP concentration, and all other reaction components were unchanged. This strategy resulted in three changes in the ratio of PhoP to PhoR. In the first case, the concentration of PhoP (0.02 to 0.09 μM) was less than that of PhoR (0.18 μM). Consistent with the data shown in Fig. 4B, this concentration range of PhoP stimulated transcription from the P_V promoter only when PhoR was present (Fig. 6A). In the second case, PhoP and PhoR were present at equimolar concentrations or nearly equimolar concentrations. The concentration of PhoP∼P (0.18 μM) was the optimum concentration for maximum transcriptional yield from the P_V promoter. In the last case, PhoP concentrations between 0.35 and 10 μM were higher than the PhoR concentration. In contrast to the other cases, PhoP at concentrations of >0.18 μM in the presence of PhoR resulted in linearly decreasing transcriptional yields, and complete repression was detected at concentrations of ≥1.4 μM. Interestingly, the same range of PhoP concentrations was able to stimulate transcription from the P_V promoter in the absence of PhoR. The optimum unphosphorylated PhoP concentration that stimulated the maximum transcriptional yield from the P_V promoter was 2.8 μM or near 2.8 μM, a repressing concentration in the case of PhoP∼P. Decreased transcription from the P_V promoter was also observed at higher concentrations of unphosphorylated PhoP (concentrations above 5.25 μM). Thus, both unphosphorylated PhoP and PhoP∼P stimulated transcription from the P_V promoter in a concentration-dependent manner for a maximum transcriptional yield, and addition of higher concentrations of either form decreased transcription from the P_V promoter. The maximum transcriptional yield with the unphosphorylated PhoP was ∼85% that obtained with PhoP∼P and required ∼16-fold-higher PhoP concentrations.

• Open in new tab
• Download powerpoint

FIG. 6.

Effects of various concentrations of unphosphorylated PhoP, PhoP∼P, and σ^A on transcription activation from the phoB-P_V promoter in vitro. (A) The upper panels show the results of an in vitro transcription analysis of the phoB promoter using reconstituted Eσ^A in the presence of various concentrations of unphosphorylated PhoP (Un) or phosphorylated PhoP (∼P). The graph in the lower panel shows the amounts of the phoB-P_V transcripts (in PhosphorImager output units) plotted as a function of the unphosphorylated PhoP (open circles) or PhoP∼P (solid circles) concentration. The PhoP-to-PhoR molar ratios (plus signs) are plotted as a function of the various PhoP concentrations used in the in vitro transcription reactions. (B) Dependence of unphosphorylated PhoP-driven transcription reactions on σ^A concentration: in vitro transcription of the phoB promoter using various concentrations of σ^A in reconstituted RNAP in the presence of the optimum concentration of either unphosphorylated PhoP (Un*) (2.8 μM) or PhoP∼P (∼P*) (0.18 μM). The protein concentrations (μM) are indicated above the lanes.

To determine if phosphorylation of PhoP might influence its interaction with the Eσ^A form of RNAP, we varied the concentration of σ^A in the reconstituted RNAP (Fig. 6B) while PhoP and PhoP∼P were used at their optimal concentrations, determined as described above. At the optimal PhoP∼P concentration (0.18 μM), limiting concentrations of σ^A (0.02 μM) stimulated transcription, although the transcriptional yield was reduced compared to that obtained using higher σ^A concentrations. Increasing the σ^A concentration so that it was up to eightfold greater than the PhoP∼P concentration had no significant effect on the transcriptional yield. In contrast, the σ^A concentration significantly influenced the ability of the unphosphorylated PhoP to stimulate transcription from the P_V promoter in vitro. Concentrations of σ^A of <0.35 μM were not sufficient to allow the optimal unphosphorylated PhoP concentration to stimulate transcription. σ^A concentrations of ≥0.35 μM allowed maximum transcription activation at the optimal unphosphorylated PhoP concentration. Interestingly, σ^A was required at a ≥17-fold-higher concentration to allow transcription activation at the optimum PhoP concentration compared to the concentration required at the optimum PhoP∼P concentration.

In general, this in vitro transcription analysis illustrated two different conditions for PhoP-dependent transcription activation from the Eσ^A-responsive phoB-P_V promoter. The first condition is phosphorylation dependent, which allows a relatively low PhoP∼P concentration compared to the PhoP concentration to stimulate transcription even at limiting σ^A concentrations. The second condition is concentration dependent, requiring considerably higher concentrations of both PhoP and σ^A in order to stimulate transcription in vitro.

Unphosphorylated PhoP or PhoP∼P is essential for transcriptional activation of a second PhoP-regulated promoter, ykoL.Because the ability of unphosphorylated PhoP to activate transcription in vitro was observed for the first time during the course of this study, we explored the use of a second PhoP-regulated promoter. The ykoL promoter exhibits significant similarity to the phoB promoter (Fig. 7A). First, it is transcribed from a potential σ^A-dependent promoter which also lacks a consensus −35 sequence essential for Eσ^A recognition. Second, it contains the four TT(C/A/T)A(C/T)A repeats characteristic of PhoP-activated promoters, which are quite similar to the phoB promoter in terms of conservation, relative distances between tandem sequences, and position upstream of the transcription start site (41). Recently, DNase I footprint analysis of the ykoL promoter region with PhoP indicated that the levels of protection afforded by the phosphorylated and unphosphorylated forms of PhoP were similar (36), a situation that we observed for the phoB-P_V promoter. The in vitro transcription analysis of the ykoL promoter (Fig. 7B) showed that despite the ability of Eσ^A alone to direct transcription from a nonspecific start site (P_X), which produced an RNA transcript that was ∼100 nucleotides long (Fig. 7B, lanes 2 through 8), no P_V transcript was observed. Addition of PhoP in the absence of *PhoR (lanes 9 through 16) or in the presence of *PhoR (lanes 17 through 23) switched transcription initiation to a site that yielded a transcript that was ∼150 nucleotides long, consistent with the P_V transcription start site that was identified previously by primer extension analysis of the ykoL mRNA during growth of a wild-type strain in phosphate starvation medium (41) (Fig. 7A). Similar to the maximum transcriptional yield with the phoB promoter, the maximum transcriptional yield observed with the ykoL-P_V promoter in the presence of unphosphorylated PhoP was ∼75% that observed in the presence of PhoP∼P. Higher concentrations of unphosphorylated PhoP decreased transcription activation from the ykoL-P_V promoter start site, and there was a 10% reduction in the transcription yield from the ykoL-P_V start site at a PhoP concentration of 10 μM (lane 16). PhoP∼P (0.35 μM) showed maximal transcription activation. Figure 7C shows quantitation of the ykoL-P_V transcript with PhoP alone (lanes 9 to 16) and with PhoP plus *PhoR (lanes 17 to 23).

• Open in new tab
• Download powerpoint

FIG. 7.

In vitro transcription analysis of a second PhoP-regulated promoter, ykoL. (A) Nucleotide sequence of the ykoL promoter coding strand. Regulatory elements are indicated as follows: arrows, transcription start site for P_V; subscripts, −10 sequences for P_V promoter; superscripts, consensus PhoP tandem binding sequences; SD, Shine-Dalgarno sequence; Met, methionine (translational start codon). The lowercase letters indicate mismatches with the consensus sequence. Primers FMH768 and FMH769 used to prepare the ykoL promoter are indicated by arrows under the coding and noncoding strands, respectively. The numbers below the sequences indicate the positions relative to the ykoL translational start site (position 1). The numbers above the sequences indicate the positions relative the transcription start sites of P_V. (B) Runoff in vitro transcription analysis of ykoL promoter using various concentrations of purified B. subtilis sigma factor (σ^A) and core RNAP (E) (lanes 1 through 8). Linearly increasing concentrations of unphosphorylated PhoP (in twofold increments) were used in the Eσ^A-driven transcription reactions in the absence of *PhoR (lanes 9 through 16) or in the presence of *PhoR at a final concentration of 0.18 μM (lanes 17 to 23). The protein concentrations (μM) are indicated above the lanes. Lane M contained a radiolabeled RNA marker. (C) Amount of ykoL-P_V transcript (in PhosphorImager output units) plotted as a function of the unphosphorylated PhoP concentration (open circles) or the PhoP∼P concentration (solid circles). nts, nucleotides.

Cellular concentrations of PhoP increase threefold during phosphate-limited growth.To determine the fold increase in the cellular PhoP concentration induced upon P_i depletion, we carried out a quantitative Western blot analysis of PhoP synthesis during growth in LPDM. Figure 8A shows the detectable concentrations of PhoP in the linear range under the conditions used. PhoP exhibited steady-state levels of ∼1 μM (Fig. 8B) during the pre-Pho induction period (from T₋₂ to T₀). When the P_i concentration in the medium decreased below 100 μM, alkaline phosphatases were induced and PhoP concentrations began to increase. Total APase specific activity increased rapidly until T₂, while the PhoP concentrations continued to increase until T₄. At T₄, the PhoP concentration was approximately threefold higher than the concentration detected at T₀. These data are consistent with our understanding of the two APase promoters and the complex phoPR promoter. It is likely that production of PhoP continues into the stationary phase because the phoPR operon has equally strong, PhoP∼P-enhanced Eσ^A and Eσ^E promoters. In contrast, the APase genes each have a strong, PhoP-dependent Eσ^A promoter (phoA and phoB-P_V), but the Eσ^E promoter of phoB-P_S is very weak (roughly 2% the strength of the Eσ^A promoter) and it appears to be inhibited by PhoP∼P at high concentrations. Thus, the contribution of the weak Eσ^E phoB-P_S promoter to total APase activity should be negligible, like its contribution to phoB expression from the phoB-P_S+V promoters during phosphate starvation (Fig. 2A).

• Open in new tab
• Download powerpoint

FIG. 8.

Western immunoblot detection of cellular PhoP levels during the phosphate starvation response. (A) Dilutions of purified standard PhoP were prepared in phoPR cell lysates and used for Western immunoblot detection using PhoP-specific antibody (left panel). The standard curve in the graph on the right was prepared as described in Materials and Methods. (B) For Western immunoblot detection of intracellular PhoP concentrations, the wild-type strain (MH6143) was allowed to grow in LPDM, and samples were collected at the times indicated. Enzymatic cell lysis was performed as described in Materials and Methods. Microscopic cell counting was performed, and samples were calibrated in phoPR cell lysates to a final concentration equivalent to 2 × 10⁷ cells/μl. From the calibrated samples, 4 μl (equivalent to lysate from 8 × 10⁷ cells) and 10 μl (equivalent to lysate from 2 × 10⁸ cells) were electrophoresed on 12% SDS-polyacrylamide gel electrophoresis gels and subjected to immunoblot detection using PhoP_CTD-specific antibody. Growth (OD₅₄₀), microscopic cell counts, APase specific activities, and extracellular inorganic phosphate concentrations were determined at the times indicated. Time zero was the transition from the exponential to the stationary phase of growth.

PhoPR plays a dual role in regulation of the Eσ^E-responsive phoB-P_S promoter.In vitro transcription data reported here established that Eσ^E was necessary and sufficient for phoB-P_S promoter function. A direct role for Eσ^E in phoB-P_S promoter expression was corroborated by phoB-P_S promoter fusion data obtained with a sigE mutant strain during phosphate-limited growth in LPDM and during sporulation in SSG medium (Fig. 2). Previous phoB-lacZ fusion data (9) failed to identify P_S promoter expression during phosphate-limited growth because the promoter fusion contained the 50-fold-stronger P_V promoter in addition to the P_S promoter, which masked the contribution of the P_S promoter. P_S promoter expression during phosphate-limited growth was delayed ∼2 h in the phoPR mutant (Fig. 3), suggesting that either synthesis or maturation of σ^E is delayed under the P_i deficiency stress conditions in the absence of PhoPR. This implies that PhoPR or another Pho regulon gene is needed for efficient sporulation, which may be especially important in nature as inorganic phosphate is often the most limiting nutrient in the soil (34). P_S promoter expression from the P_S+V promoter fusion was not observed previously in phoP or phoR mutant strains grown in LPDM because experiments were terminated before delayed P_S promoter activation at T₅ to T₆ (9).

PhoP and PhoR have two roles in transcription from the phoB-P_S promoter. As stated above, the positive role for phoPR affects the timely expression from the Eσ^E-responsive promoter, P_S, during phosphate starvation. At the same time, our in vitro data suggest that PhoP∼P represses transcription from the P_S promoter by binding directly to the P_S promoter. These results suggest that PhoPR has dual roles in development affecting the timing of at least one Eσ^E promoter by positively affecting when Eσ^E is available during phosphate starvation and also preventing early expression via PhoP∼P before Pho regulation is turned off. The fact that the latter role was not detected during previous induction studies of P_S in phosphate-replete sporulation media when the P_V promoter was silent is understandable because there would have been no signal for PhoP phosphorylation and thus no repression of P_S (9).

Although the binding site in the phoB-P_S promoter, upstream of the P_V promoter, is positioned like other secondary binding sites for PhoP-activated promoters, it is clear that it has no role in P_V activation because expression from the phoB-P_S+V promoter fusion and expression from the P_V promoter are the same (Fig. 2A and C). A requirement for phosphorylation of PhoP for binding at P_S was observed for two other PhoP repressed promoters, resD and tagA (4, 28), that contain a single consensus PhoP dimer binding site.

Different mechanisms control genes and operons that have been classified as both Pho and Eσ^E regulon members.For genes classified as both Pho and Eσ^E regulon genes, the mechanism responsible for inclusion of the phoPR promoter has been explained by PhoP∼P amplification of σ^A- and σ^E-responsive phoPR operon promoters (35). Here we showed that the PhoP-regulated phoB promoter, P_V, is a σ^A promoter and that transcription from the second promoter, P_S, requires only Eσ^E, although the P_S transcription was delayed in a phoPR mutant (Fig. 2B and 3) and was negatively regulated by PhoP∼P. yhaX and yhbH are similar to the phoB-P_S promoter in three ways. Both these genes were expressed during P_i-limiting growth in a wild-type phoPR strain at a time later than the time that other Pho-regulated genes were expressed, neither gene was transcribed in a sigE mutant strain, and like the P_S promoter, neither yhaX nor yhbH has consensus tandem repeats for PhoP binding on the coding strand upstream of the putative Eσ^E promoter (37). Because the Pho regulon assignment for yhaX and yhbH was based solely on the absence of promoter expression in a phoR mutant strain, it seems that the possible delay of Eσ^E-dependent gene expression in that strain, not the lack of direct PhoPR promoter regulation, may have caused the negative results. Consistent with this idea, PhoP does not bind to the yhaX promoter (36). No further information is available on the mechanism of Pho regulation and σ^E-dependent regulation of yycP or glnQ. The putative yycOP and glnQ promoters are interesting in that both contain two putative consensus repeats for a PhoP dimer binding site on the coding strand on either side of the −35 element of the putative σ^E promoter. It should be interesting to determine if Eσ^E-responsive PhoP-activated genes exist (in addition to autoregulation) and if their requirement for PhoP binding is different than the requirement for all PhoP-activated σ^A promoters studied (13) or for the PhoP-enhanced autoregulation Eσ^E promoter (35).

PhoP or PhoP∼P plus Eσ^A is sufficient for phoB-P_V or ykoL promoter transcription in vitro, and the PhoP concentration is the key.The failure of Eσ^A to activate in vitro transcription of phoB-P_V (Fig. 4) or ykoL (Fig. 7) is likely due to the absence of a −35 σ^A consensus binding region in either promoter, as is true for other Eσ^A PhoP-activated promoters which absolutely require PhoP∼P for activation, including the phoA, pstS, and tuaA promoters (30, 40). Previous studies indicated that unphosphorylated PhoP was unable to activate transcription in vitro. The maximum concentrations of unphosphorylated PhoP and σ^A used in the previous studies were ∼0.25 μM and 0.02 μM, respectively. Consistent with the data presented here (Fig. 5B), these concentrations of unphosphorylated PhoP plus Eσ^A were not sufficient to activate transcription in vitro. Here, we showed that unphosphorylated PhoP stimulated transcription from the phoB-P_V promoter in vitro. However, the concentration of unphosphorylated PhoP required for maximum transcription was much higher than that required for PhoP∼P with either promoter (17-fold higher for phoB and 10-fold higher for ykoL), indicating that phosphorylation has an important role.

While it is possible that PhoP isolated from E. coli is phosphorylated, the activity of equimolar concentrations of PhoP compared to the activity of PhoP in the presence of PhoR plus ATP suggests that the percentages of PhoP phosphorylated in the two cases differ considerably. Furthermore, a PhoPN structural analysis (3) for crystals formed in 1 day at 4°C showed no phosphorylation. However, a number of E. coli-expressed unphosphorylated response regulators have been shown to activate transcription in vitro (PrrA [10], Spo0A [27], and DesR [11]), only to lose this ability in the variant with a mutation at the phosphorylatable aspartate residue and in strains containing this variant (PrrA [10] and DesR [11]). Spo0A (26) and DesR (11) were proven to be phosphorylated when they were isolated. UhpA or the UhpA D54N variant is constitutively active for expression of the sugar phosphate transporter UhpT in E. coli when it is overexpressed, but the variant is totally inactive as a single copy, suggesting that phosphorylation of UhpA is required for a single copy but not when UhpA is overexpressed (12, 33, 47). In the case of the PhoP/PhoQ two-component system of Salmonella, overexpression of the response regulator PhoP or the D52A variant activated regulon genes in vivo, albeit in a signal-independent manner, reportedly due to concentration-dependent activation of PhoP via dimerization. Further studies indicated that signal-dependent (low-Mg²⁺) PhoP DNA binding is phosphorylation dependent in vivo (46). In contrast to the examples described above, ResD D57A was shown to induce transcription of the ResDE regulon in response to oxygen limitation, suggesting that ResD senses oxygen-limiting conditions via an unknown mechanism in addition to being activated by phosphorylation by ResE (18).

Because of the high concentrations of unphosphorylated PhoP required for Pho regulon promoter activation in vitro and the modest threefold autoinduction of PhoP in vivo, we believe that a physiological role for unphosphorylated PhoP in a wild-type strain is unlikely. However, activation of Pho regulon genes has been reported in a phoR deletion strain. The ability to induce low-level expression from both tuaA and pstS promoters was retained in a phoR mutant strain during phosphate-deficient growth (30, 40). In the latter study (40), the authors proposed that PhoP was likely phosphorylated in a PhoR-independent manner (i.e., by unknown phosphodonors). In a second case, Ogura and colleagues (32) reported that overproduction of PhoP in vivo in a phoR deletion strain activated expression of phoB and a number of Pho regulon genes in a signal-independent manner; these genes are not normally activated in a phoR mutant during phosphate starvation. Although a number of possibilities may explain this observation, during PhoP overproduction under nonphysiological conditions concentration-dependent activation of PhoP is a potential factor.

Phosphorylation of PhoP minimizes the Eσ^A levels required for maximum transcription activation in vitro, consistent with in vivo conditions during Pho induction.Because both unphosphorylated PhoP and PhoP∼P are dimers in solution (29) and both forms of PhoP bind to phoB (29) or ykoL (36) promoters with similar affinity, we asked if phosphorylation of PhoP had an effect on its interaction with the transcriptional machinery (7). Consistent with this notion, when PhoP or PhoP∼P was used at the concentration determined to be optimal for phoB-P_V transcription (Fig. 5A), concentrations of σ^A (≤0.09 μM) required for PhoP∼P-stimulated transcription activity were significantly lower than the concentrations required by unphosphorylated PhoP (Fig. 5B). Analysis of the in vivo concentration of PhoP showed that autoinduction of PhoP synthesis and APase expression initiated in parallel, when PhoP levels were the lowest, at T₀. During the following transition period (T₁ and T₂), PhoP concentrations increased modestly, roughly onefold, while the abundance of Eσ^A is known to decrease due to sigma factor displacement (23). Taken together, these data suggest that the phosphorylated form of PhoP is probably responsible for in vivo activation of Pho regulon genes, including genes encoding APases.

Elevated concentrations of PhoP∼P inhibited transcription in vitro, providing a possible explanation for previous in vivo data.PhoP∼P is known to function not only as a transcription activator but also as a transcription repressor of at least three Eσ^A-responsive promoters, including tagA/D (30, 39) and resD (4). One difference observed in vitro between repressed and activated promoters is the customary requirement for PhoP phosphorylation for promoter binding at PhoP-repressed promoters (4, 30) compared to activated promoters, where either PhoP or PhoP∼P binds. Also, it appears that similar levels of PhoP∼P are required for activation and repression in vivo because repression of one set of genes and activation of another set by PhoP∼P occur simultaneously. For example, upon phosphate starvation the genes responsible for synthesis of teichoic acid (the tagA/D operons) are repressed (30), while the tuaA operon encoding proteins for the synthesis of teichuronic acid is activated (30), both in a phoPR-dependent manner. Thus, it seems unlikely that elevated levels of PhoP∼P have a physiologically relevant role in Pho gene repression in a wild-type strain. However, when overexpression studies are conducted, PhoP∼P may play a role. Data consistent with in vivo repression of Pho regulon genes by elevated PhoP∼P concentrations come from two in vivo studies, both involving nonphysiological overexpression of PhoP. It was observed that expression of PhoP from its own promoter on a multicopy plasmid in a phoPR wild-type background resulted in a reduced growth rate and repression of Pho regulon genes even under P_i-deficient conditions (W. Liu and F. M. Hulett, unpublished). In an independent study (32), overproduction of any one of three response regulators (PhoP, DegU, and ComA) in a wild-type background failed to result in expression of target genes, and in each case deletion of a cognate histidine kinase gene was required for expression. Our in vitro data are consistent with the hypothesis that repression by elevated concentrations of PhoP∼P is one possible explanation for the previously observed in vivo repression of Pho regulon genes.

In summary, data presented here show that phoB-P_S is an Eσ^E-responsive promoter that is expressed under phosphate-deficient conditions and that the timing of this expression is delayed in a phoPR mutant strain. These observations suggest that workers should be cautious when analyzing promoters believed to depend on both Eσ^E and phoPR. While a role for PhoP∼P in activation of an Eσ^E promoter has been reported previously (35), repression of the P_S promoter by PhoP∼P is the first example of a repressor role for PhoP∼P at Eσ^E-responsive promoters. Two Eσ^A promoters that require PhoP for expression in vivo and in vitro, phoB-P_V and ykoL, were shown to be activated at a high concentration of PhoP and to be repressed by high concentrations of PhoP∼P in vitro, thereby providing a possible explanation for previous in vivo observations (32, 39). Phosphorylation of PhoP resulted in decreased concentration requirements for both PhoP∼P and Eσ^A for stimulation of Pho regulon promoters in vitro.