Post-transcriptional regulation of the Bacillus subtilis pst operon encoding a phosphate-specific ABC transporter

Nicholas E. E. Allenby1, Nicola O'Connor2, Zoltán Prágai2,{dagger}, Noel M. Carter2, Marcus Miethke3, Susanne Engelmann3, Michael Hecker3, Anil Wipat4, Alan C. Ward1 and Colin R. Harwood2

1 School of Biology, University of Newcastle upon Tyne, Newcastle upon Tyne NE2 4HH, UK
2 School of Cell and Molecular Biosciences, University of Newcastle upon Tyne, Newcastle upon Tyne NE2 4HH, UK
3 Institut für Mikrobiologie und Molekularbiologie, E.-M.-Arndt-Universität, Greifswald, F.-L.-Jahnstraße 15, D-17487 Greifswald, Germany
4 School of Computing Science, University of Newcastle upon Tyne, Newcastle upon Tyne NE2 4HH, UK

Correspondence
Colin R. Harwood
colin.harwood{at}ncl.ac.uk


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
During phosphate starvation, Bacillus subtilis regulates genes in the PhoP regulon to reduce the cell's requirement for this essential substrate and to facilitate the recovery of inorganic phosphate from organic sources such as teichoic and nucleic acids. Among the proteins that are highly induced under these conditions is PstS, the phosphate-binding lipoprotein component of a high-affinity ABC-type phosphate transporter. PstS is encoded by the first gene in the pst operon, the other four members of which encode the integral membrane and cytoplasmic components of the transporter. The transcription of the pst operon was analysed using a combination of methods, including transcriptional reporter gene technology, Northern blotting and DNA arrays. It is shown that the primary transcript of the pst operon is processed differentially to maintain higher concentrations of PstS relative to other components of the transporter. The comparative studies have revealed limitations in the use of reporter gene technology for analysing the transcription of operons in which the messenger RNA transcript is differentially processed.


Abbreviations: APase, alkaline phosphatase; ONP, o-nitrophenol; PNP, p-nitrophenyl phosphate; Pi, inorganic phosphate

{dagger}Present address: DSM Nutritional Product Ltd, Department of Biotechnology, VFB, Bldg 203/112B, CH-4002 Basel, Switzerland.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
An important feature of bacterial genomes is that the majority of genes are organized into polygenic operons that facilitate their coordinated regulation. However, the individual products of these operons can be required in differing (i.e. non-stoichiometric) amounts. Bacteria employ a variety of methods to control the production of proteins from the same operon, including the use of internal promoters and transcription termination sites, post-transcriptional processing and differential stability of the primary transcript, and variations in the efficiency of translation initiation (Condon, 2003). Despite its potential importance, few studies have analysed the post-transcriptional control of Bacillus subtilis operons in detail. In the case of the cggR/gapA operon (Ludwig et al., 2001), the cggR portion of the primary transcript is degraded rapidly (half-life <0·3 min) whereas the gapA portion is more resistant to degradation (half-life 3·5 min). Similarly, the dnaK operon of B. subtilis has also been shown to exhibit differential mRNA stability (Homuth et al., 1999). Recently, whole-genome mRNA decay rates were determined for ~1500 B. subtilis genes during entry into stationary phase. The mRNAs of approximately 1200 of these genes had half-lives of less than 7 min, while the mRNAs of 30 of them had half-lives of more than 15 min (Hambraeus et al., 2003).

B. subtilis responds to phosphate starvation stress by regulating genes of the phosphate starvation (Pho) stimulon comprising, at least, the PhoP and SigB regulons (Hulett, 1996; Hecker & Völker, 1998; Antelmann et al., 2000; Prágai & Harwood, 2002). The PhoP regulon is controlled by the PhoP–PhoR (PhoPR) two-component signal transduction system (Hulett et al., 1994; Prágai et al., 2004); in response to low concentrations of inorganic phosphate (Pi<0·1 mM), the PhoP response regulator is activated by its cognate sensor-kinase, PhoR. Phosphorylated PhoP (PhoP~P) binds to PHO-box sequences, direct repeats of TT(A/T/C)ACA with a 5±2 bp spacer located in the promoter regions of PhoP-regulated operons, to activate or repress their expression (Eder et al., 1999). Currently, the PhoP regulon comprises 34 genes in eight polycistronic (phoPR, phoB–ydhF, pstSAC,BA,BB, phoD–tatAD, tuaABCDEFGH, resABCDE, tagAB and tagDEF) and five monocistronic (glpQ, phoA, tatCD, ykoL and yttP) operons (Hulett, 2002; Prágai et al., 2004).

Proteomic analyses of B. subtilis (Eymann et al., 1996; Antelmann et al., 2000) indicate that the most abundant protein induced in response to phosphate starvation is PstS, the lipoprotein substrate-binding component of a high-affinity phosphate transporter belonging to the ABC (ATP-binding cassette) superfamily of transporters (Higgins, 1992; Boos & Lucht, 1996). The genes downstream and in the same operon as pstS are pstC, pstA, pstBA and pstBB (Fig. 1). PstC and PstA are permease-type membrane proteins that facilitate the passage of Pi across the membrane (Webb et al., 1992); PstBA and PstBB are ATP-binding proteins that energize this transport (Chan & Torriani, 1996). These proteins share considerable identity with the products of the functionally related pstSCABU operon of Escherichia coli: BsuPstS (PstS of B. subtilis, 309 amino acids in length) exhibits 25 % identity with EcoPstS (PstS of E. coli); BsuPstC (300 amino acids) has 28 % identity with EcoPstC; BsuPstA (294 amino acids) has 28 % identity with EcoPstA; BsuPstBA (269 amino acids) and BsuPstBB (260 amino acids) both have 57 % identity with EcoPstB. The E. coli pst operon additionally includes phoU, encoding a regulatory component of the transporter that is not present in B. subtilis (Muda et al., 1992; Steed & Wanner, 1993).



View larger version (25K):
[in this window]
[in a new window]
 
Fig. 1. (A) Genetic map of the pentacistronic pst operon. Filled thick arrows illustrate structural genes. White bars indicate the location of ribo-probes used for the Northern analyses (see Figs 3 and 4). Dotted lines indicate transcript length, inferred by Northern analysis under conditions of phosphate starvation. Stem–loop structures and putative rho-independent terminators are shown with stem–loop structures labelled SL and TTE respectively. (B) The mRNA predicted structure for the intergenic region between the E. coli pstS and pstC genes. The calculated {Delta}G value for this structure is –24·8 kcal mol–1 (–103·8 kJ mol–1). (C) The mRNA predicted structure of the intergenic region between the B. subtilis pstS and pstC genes as calculated by the Zuker algorithm (Zuker, 1989). The {Delta}G value for this structure is –24·1 kcal mol–1 (–100·8 kJ mol–1).

 
During phosphate-replete growth Pi is thought to be transported into the cell via the product of the pit gene which, on the basis of homology with genes in E. coli, is a low-affinity Pi transporter (Harris et al., 2001). Using a combination of reporter gene and DNA array technology, we have shown that pit expression is independent of Pi concentration and PhoPR (N. E. E. Allenby, C. R. Harwood & Z. Prágai, unpublished). In contrast, the pst operon is induced specifically in response to Pi starvation in a PhoP-dependent manner (Qi et al., 1997). The products of this operon form the main transporter of phosphate at low Pi concentrations (<0·1 mM) and pst mutants exhibit an approximately sevenfold reduction in Pi uptake under these conditions (Prágai et al., 2001).

The pst operon is transcribed from a sigma A-type promoter located immediately upstream of pstS (Qi et al., 1997). Recently, the operon has been shown to be induced during alkali stress (Atalla & Schumann, 2003). However, alkali stress induction is PhoR-dependent, uses the same promoter as phosphate starvation induction and is reversed in the presence of increased concentrations of extracellular phosphate.

In this study we have analysed the transcription of the pst operon using reporter gene technology, Northern hybridization and DNA arrays. We show that the pst operon is transcribed as a primary transcript of 4·4 kb that is processed rapidly into smaller transcripts. One of the products of this processing is a stable 0·9 kb monocistronic transcript encoding pstS. A consequence of this processing is the maintenance of higher concentrations of PstS compared to other components of the transporter. Finally, we have used the pst operon as a model for the comparative analysis of gene expression technologies. This analysis reveals limitations in reporter gene technology for the analysis of the transcription of operons in which mRNA is extensively processed.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains, plasmids, primers and media.
Bacterial strains and plasmids are listed in Table 1 and primers in Table 2. Strains were grown in Luria–Bertani (LB) medium, low-phosphate medium (LPM) or high-phosphate medium (HPM) (Prágai & Harwood, 2000a). The concentration of phosphate was 0·42 mM in LPM and 5·0 mM in HPM. Antibiotics were added as required: ampicillin (100 µg ml–1) for E. coli and erythromycin (0·3 µg ml–1), lincomycin (25 µg ml–1), tetracycline (12·5 µg ml–1) and chloramphenicol (5 µg ml–1) for B. subtilis.


View this table:
[in this window]
[in a new window]
 
Table 1. Bacterial strains and plasmids

 

View this table:
[in this window]
[in a new window]
 
Table 2. Primers

 
DNA manipulation and general methods.
Plasmid DNA extraction, restriction endonuclease digestion, ligation, agarose gel electrophoresis, and transformation of electrocompetent E. coli cells were carried out as described previously (Prágai & Harwood, 2000b). Enzymes, molecular size markers and deoxynucleotides were purchased from Roche Diagnostics or Amersham Pharmacia Biotech. Extraction of B. subtilis DNA and transformation of B. subtilis by the Groningen method were according to Bron (1990).

Construction of reporter gene strains.
Primers PSTAKO-FOR and PSTAKO-REV (Table 2) were used for PCR amplification of a 440 bp internal fragment of pstA, primers PSTCF-FOR and PSTCF-REV for the amplification of a 252 bp fragment containing the putative ribosome-binding site (RBS) and 5' end of pstC, and primers PSTBB-FOR and PSTBB-REV for the amplification of a 462 bp internal fragment of pstBB. The PCR reactions were carried out with Taq DNA polymerase using chromosomal DNA of B. subtilis 168 as template. After HindIII and BamHI digestion, the PCR fragments were ligated into HindIII- and BamHI-digested pMUTIN4 integrational vector (Vagner et al., 1998) and transformed into electrocompetent cells of E. coli XL-1 Blue (Stratagene). Transformants were selected on LB agar medium supplemented with ampicillin. The resulting plasmids pPSTA, pPSTBB and pPSTC were confirmed by restriction digestion and PCR using the insert-specific primers and the plasmid-specific primers MUT-FOR and MUT-REV (Tables 1 and 2; see Prágai et al., 2001).

Plasmids pPSTA, pPSTBB and pPSTC were transformed into competent cells of B. subtilis strain 168 and transformants selected on LB agar plates containing erythromycin and lincomycin. The resulting strains, PSTA, PSTBB and PSTC, respectively, were analysed by PCR to confirm the integration of a single copy of the plasmids into the target genes on the chromosome, using the strategy described previously (Harwood et al., 2002).

Northern blotting and mRNA half-life determinations.
Total RNA was extracted from B. subtilis strains 168, 168-PR and ML6 (Table 1) as described previously (Eymann et al., 2002). Northern blot analysis was performed according to the manufacturer's instructions (DIG Northern Starter kit, Roche Diagnostics), using 5 µg total RNA per lane. PCR products were generated for sites internal to the pstS, pstC, pstA and pstBB genes using the primers described in Table 2. The REV primers included a promoter sequence recognized by T7 polymerase. The resulting amplified fragments were used as a substrates for the in vitro T7 RNA polymerase-directed synthesis of the digoxigenin-labelled pstS-, pstC-, pstA- and pstBB-specific RNA probes.

To determine the half-lives of pst-specific mRNA transcripts, wild-type B. subtilis was grown in LPM and rifampicin added to the culture medium 2 h after transition to phosphate-starvation-induced transition phase (T2). Samples were harvested immediately before the addition of rifampicin (0 min) and at 2, 4, 7, 10, 13, 25 and 45 min after the addition of the antibiotic. The samples were hybridized with the same probes as were used for the Northern blot analysis (see Fig. 4).



View larger version (77K):
[in this window]
[in a new window]
 
Fig. 4. Transcriptional stability of the pst operon. RNA was isolated from wild-type B. subtilis (168). Bacteria were grown in LPM and samples were taken 2 h after entry into the stationary growth phase provoked by phosphate starvation. RNA was isolated from samples harvested before and 2, 4, 7, 10, 20, 45 min after the addition of rifampicin. Five micrograms of RNA was applied per lane then, after capillary blotting the filters were hybridized to gene-specific ribo-probes pstS (A), pstC (B), pstA (C) and pstBB (D). Transcript size was determined by comparison with DIG-labelled RNA size markers (Roche Diagnostics).

 
Individual bands from Northern blots were quantified using the computer program QuantityOne (Bio-Rad), and the data used to calculate the half-lives of individual transcripts, averaging the data from three separate experiments.

Enzyme and Pi assays.
Alkaline phosphatase (APase) and {beta}-galactosidase activities were determined as described previously (Prágai & Harwood, 2002).

Transcriptome analysis by DNA macroarray hybridization.
Cell harvesting, preparation of RNA, synthesis of radioactively labelled cDNA and hybridization of B. subtilis macroarrays (Sigma-Genosys) were performed as described by Eymann et al. (2002). Each analysis was carried out twice, using two independently isolated RNA preparations and two different batches of DNA array membranes. Exposed PhosphorImager screens were scanned with a Storm 860 PhosphorImager (Molecular Dynamics) at a resolution of 50 µm and a colour depth of 16 bit. For quantification of the hybridization signals and background subtraction, the ArrayVision software (version 5.1, Imaging Research) was used. Normalized intensity values of the individual spots were calculated using the overall-spot-normalization function of ArrayVision. To avoid extreme expression ratios for genes close to or below the detection limit, genes with signal intensity values corresponding to less that twice the background were not counted in the analysis.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Proteomic analyses indicate that the proteins encoded by the pst operon are present in the cell in very different amounts (Eymann et al., 1996; Antelmann et al., 2000). While PstS (the substrate-binding component) represents the most prominent protein spot in both the cytoplasmic and extracellular proteome fractions, significantly lower amounts of PstBA and PstBB (membrane-associated ATP-binding components of the transporter) were detected exclusively in the cytoplasmic fraction. The relative amounts of these proteins are consistent with their biological role, since the substrate-binding components of ABC transporters are required in molar excess of the components of the translocator itself. We were therefore interested in establishing how an operon transcribed from a single promoter was able to mediate the differential expression of its individual genes.

One possibility we considered was that the observed differential production of Pst proteins was due to the efficiency of translation initiation, resulting from differences in either start codons or the binding energies between their ribosome-binding sites (RBS) and sequences at the 3'OH end of 16S rRNA. However, all of the genes use ATG as the start codon and calculation of free binding energy ({Delta}G) values (Tinoco et al., 1973) indicated that the {Delta}G of the pstS RBS ({Delta}G –10·2 kcal mol–1; –42·7 kJ mol–1) was lower than that of pstC ({Delta}G –14·8 kcal mol–1; –61·9 kJ mol–1), pstA ({Delta}G –16·6 kcal mol–1; –69·5 kJ mol–1), pstBA ({Delta}G –11·0 kcal mol–1; –46·0 kJ mol–1) and pstBB ({Delta}G –16·6 kcal mol–1; –69·5 kJ mol–1). The absence of an obvious mechanism for differential translation led us to analyse the transcription pattern of the pst operon, using a combination of lacZ reporter gene, Northern blotting and DNA array technology.

Reporter gene analysis of the pst operon
The transcriptional activity of the pst operon was monitored by creating a series of lacZ transcriptional fusions, using the pMutin integration vector (Vagner et al., 1998). The resulting strains, PSTSK (Prágai et al., 2001), PSTC, PSTA, PSTBA (Atalla & Schumann, 2003) and PSTBB, each had a copy of the E. coli lacZ gene transcriptionally fused within the pstS, pstC, pstA, pstBA and pstBB genes, respectively (Table 1). These strains were grown in LPM and samples taken to monitor growth and to determine APase and {beta}-galactosidase activities. Three independent growth experiments gave comparable results and a representative LPM dataset is shown in Fig. 2.



View larger version (28K):
[in this window]
[in a new window]
 
Fig. 2. Growth and reporter activity of B. subtilis pst : : lacZ fusion mutants grown in LPM. (A) Growth curves (OD600) of lacZ-fusion mutants PSTSK ({bullet}), PSTC ({blacklozenge}), PSTA ({blacksquare}), PSTBA ({blacktriangledown}) and PSTBB ({blacktriangleup}) are shown with filled symbols. APase activity of these are shown with the corresponding open symbols. (B) Transcriptional activities of the pst operon; specific {beta}-galactosidase activities of PSTSK ({bullet}), PSTC ({blacklozenge}), PSTA ({blacksquare}), PSTBA ({blacktriangledown}) and PSTBB ({blacktriangleup}).

 
APase, a native intrinsic reporter of the PhoP regulon, was induced in LPM during transition from exponential to stationary phase (T0). The pst operon was induced concomitantly, as determined by the {beta}-galactosidase activities of the various reporter gene constructs (Fig. 2B). Genes pstS, pstC and pstBB were induced to similar levels of activity (~700 nmol ONP min–1 per OD600 unit), while the pstA and pstBA genes exhibited about one-third of this activity (~200 nmol ONP min–1 per OD600 unit). The observation that the first and last genes (pstS and pstBB) show similar levels of activity implies that the operon is transcribed as a single full-length product from a promoter upstream of pstS. However, it is not clear why two internal genes, pstA and pstBA should exhibit significantly lower levels of activity. When the strains were grown in HPM, very low levels of {beta}-galactosidase and APase activities were detected during transition to stationary phase (data not shown).

Northern blot analysis of the pst operon
Northern blot analysis was used to monitor the pattern of transcripts produced by the pst operon in response to phosphate starvation. The analysis was carried on the wild-type strain and phoR and sigB mutants, to determine the influence of the PhoP and SigB regulators. Total RNA was extracted from the wild-type strain and phoR and sigB mutants before, during and after transition to phosphate-limited growth. The extracted RNA was analysed by Northern blotting, using antisense-RNA probes specific to sequences throughout the pst operon (Fig. 3). Irrespective of the strain or probe, little or no pst-specific RNA was detected in samples taken 2 h before the onset of phosphate-limited growth, confirming the low expression of this operon in phosphate-replete conditions (Pi>0·1 mM). Similarly, little or no pst-specific RNA was detected in RNA extracted from the phoR mutant, confirming that the expression of this operon was dependent on the PhoPR two-component signal transduction system. In the case of the wild-type and sigB mutant, a transcript of 4·4 kb was detected during transition to phosphate limitation and 5 h later. The size of this transcript is consistent with the predicted length of the entire pst operon. In addition to the full-length transcript, a number of smaller pst-specific transcripts were detected, the pattern of which was dependent on the location of the probe. Consistent with our previous observations (Prágai & Harwood, 2002), while the transcription pattern was similar in the wild-type and sigB mutant, the level of expression was higher in the latter.



View larger version (79K):
[in this window]
[in a new window]
 
Fig. 3. Northern blot analyses of the pst operon. RNA was isolated from wild-type B. subtilis (168), sigB-null and phoR-null mutants. Bacteria were grown in LPM and samples were taken 2 h before (T–2), 0 h (T0) and 5 h (T5) after entry into the stationary growth phase provoked by phosphate starvation. Five micrograms of RNA was applied per lane, then after capillary blotting the filters were hybridized to gene-specific ribo-probes pstS (A), pstC (B), pstA (C) and pstBB (D). Transcript size was determined by comparison with DIG-labelled RNA size markers (Roche Diagnostics). The arrows labelled 16S rRNA and 23S rRNA indicate the locations of these rRNA species, which are known to trap smaller RNA species (Homuth et al., 1997).

 
When a hybridization probe specific to the pstS gene was used, a prominent band of 0·9 kb was detected at T0 and T5 in addition to the full-length transcript (Fig. 3A). The size of the prominent mRNA band corresponded to that of the pstS gene. The intensity of 0·9 kb transcript, compared with that of the full-length transcript, is consistent with proteomic studies showing that PstS is the most prominent protein encoded by the operon (Eymann et al., 1996; Antelmann et al., 2000). The transcript patterns obtained with the pstC- and pstA-specific probes (Fig. 3B, C) showed, in addition to the full-length transcript, the presence of two additional prominent bands, while the pstBB-specific probe revealed the presence of several additional bands (Fig. 3D). The additional bands could either be processing products or mRNA entrapped in the rRNA bands.

In order to determine whether the prominence of the 0·9 kb transcript was due to differential mRNA stability, we determined the half-lives of various pst-specific transcripts following treatment with rifampicin, an inhibitor of transcription initiation (Fig. 4). The 4·4 kb full-length transcript (see Fig. 4B–D) had a calculated half-life of 0·9±0·15 min while the 0·9 kb transcript, encoding the pstS gene, had a calculated half-life of 11·0±1·0 min (Fig. 4A). The calculated mean half-lives of other putative degradation products varied from 1·2 to 1·7 min.

Analysis of the expression of the pst operon with DNA arrays
DNA arrays were used to analyse the whole-genome response of B. subtilis to phosphate starvation (unpublished data). We have extracted the data for the transcription of the pst operon from these arrays to provide a comparative analysis with the reporter gene and Northern blot analyses. The DNA array data confirmed the induction of the pst operon in response to phosphate starvation (Fig. 5A, B). The activity of the operon peaked between T0 and T1 and declined thereafter. The amount of the pstS transcript was more than fivefold higher than that of the other genes in the operon, an observation that was more consistent with data from the Northern blots rather than the lacZ reporter gene experiments.



View larger version (22K):
[in this window]
[in a new window]
 
Fig. 5. DNA array analyses of the pst operon. RNA was isolated from wild-type B. subtilis (168) (A) and sigB-null mutant (B). Strains PSTSK ({bullet}), PSTC ({blacklozenge}), PSTA ({blacksquare}), PSTBA ({blacktriangledown}) and PSTBB ({blacktriangleup}) were grown in LPM and samples were taken before, during and after entry into the stationary growth phase provoked by phosphate starvation. Total RNA was isolated and used as the template for reverse transcriptase incorporating radiolabelled [33P]dATP which was hybridized to whole genome macroarray (Sigma-Genosys). Normalization and quantification were as described in Methods.

 
We also used the DNA arrays to analyse the transcription profile of the phoR and sigB mutants. As expected, there was little or no induction of the pst operon in the phoR mutant (data not shown). In the case of the sigB mutant, while the pattern of expression of the pst operon was similar to that of the wild-type, the expression levels of individual genes were significantly higher. These data provide additional support for previous studies showing interactions between the SigB and PhoP regulons (Prágai & Harwood, 2002).


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The B. subtilis pst operon is induced in response to phosphate starvation from a single sigma A-type promoter upstream of pstS in a PhoPR-dependent manner (Qi & Hulett, 1998). This is confirmed by reporter gene experiments in which the first (pstS) and last (pstBB) genes in the operon showed virtually identical activity profiles (Fig. 2B). However, in Northern blot experiments, the full-length transcript of the pst operon (4·4 kb) was a minor component of the mRNA species detected with gene-specific probes. By far the most prominent band was a 0·9 kb product hybridizing exclusively to the pstS-specific probe. Moreover, DNA macroarray analysis (Fig. 5A, B) indicated that pstS-specific mRNA was significantly more abundant than that of other members of the operon.

The question arises as to whether the 0·9 kb transcript results from differential processing of the primary transcript, a transcription termination event occurring immediately downstream of pstS or a combination of both processes. We therefore determined the half-lives of the full-length transcript and smaller transcripts within the 5' and 3' ends of the operon. The full-length transcript was rapidly degraded, with a half-life of less than 1 min. Processed transcripts within the 3' end of the operon had similarly short half-lives. In contrast, the 5' located 0·9 kb transcript had a half-life of 11 min, at least 12 times longer that of the primary transcript. Taken together with the reporter gene experiments showing similar levels of expression of pstS, pstC and pstBB, our data suggest that the pst operon is transcribed as a single transcript of 4·4 kb that is subject to rapid mRNA processing. As a consequence of this processing, the pstS-specific mRNA exhibits a significantly longer half-life than the other genes encoded by the operon. The data also suggest that transcription termination between pstS and pstC plays, at most, a minor role in the differential production of PstS.

Next we addressed the question of the increased stability of the 0·9 kb transcript. Analysis of the intergenic region between pstS and pstC, which is likely to include the 3' end of the 0·9 kb transcript, revealed the presence of an inverted repeat capable of forming a stem–loop structure with an asymmetric unpaired bubble (Fig. 1A). While the predicted secondary structure lacks features associated with rho-independent transcription terminators, we cannot rule out its possible role as an intrinsic transcription terminator. However, this stem–loop shows similarities to repetitive extragenic palindromic (REP) sequences in E. coli that increase mRNA stability by conferring resistance to 3'-5' exoribonucleases. In B. subtilis it has been shown that, while the 5' ends of mRNA play an important role in controlling the initiation of mRNA degradation (DiMari & Bechhofer, 1993; Jürgen et al., 1998; Sharp & Bechhofer, 2003), the processive degradation itself is exclusively the result of 3'-exonuclease activity. Consequently, the 0·9 kb transcript could result from a combination of endoribonucleolytic cleavage of the primary transcript in the intergenic region between pstS and pstBB and the subsequent resistance of the resulting 5' product to 3'-5' exoribonucleolytic degradation (Belasco & Higgins, 1988; Higgins et al., 1992).

The pattern of expression of the B. subtilis pst operon shows similarities to the pst operon of E. coli (Aguena et al., 2002). The E. coli pst operon is transcribed as a full-length transcript of 4·7 kb that is also subject to rapid post-transcriptional processing. A putative inverted repeat sequence, located immediately downstream of pstS, shows 73 % identity to REP sequences (Higgins et al., 1988; Bachellier et al., 1996) that impede 3'-5' exoribonuclease activity and therefore stabilize upstream mRNA (Newbury et al., 1987a, b; Belasco & Higgins, 1988). A comparison of the inverted repeat sequences downstream of the pstS genes in E. coli and B. subtilis is shown in Fig. 1.

Endoriboncleolytic cleavage followed by progressive exoribonucleolytic degradation is thought to be a common mechanism of mRNA decay in bacteria (Arraiano et al., 1988). In E. coli RNaseE is the major endoribonuclease involved in initiating decay. While no homologues of RNaseE have been found in B. subtilis, there is biochemical evidence for the presence of a similar enzymic activity in this bacterium (Condon et al., 1997; Vázquez-Cruz & Olmedo-Alvarez, 1997). In the absence of 5'-3' exoribonuclease activity in bacteria, mRNA degradation must start either from the 3' ends of primary transcripts or from 3' ends generated by endribonucleolytic cleavage within the transcript.

Our data are consistent with a model in which the rapid degradation of the 3'-region of the pst primary transcript is initiated by endoribonucleolytic cleavage events that generate entry site(s) for 3'-5' exoribonucleases. A similar mechanism has been proposed to account for the stability of cry1Aa mRNA in B. subtilis (Vázquez-Cruz & Olmedo-Alvarez, 1997). The detection of several putative degradation intermediates in the Northern blots with the pstBB-specific probe (Fig. 3D) indicates that there are likely to be several endoribonucleolytic processing sites. Exoribonucleases are responsible for the rapid degradation of the resulting secondary transcripts except for the region encoding pstS, presumably because they are stalled at the stem–loop structure downstream of pstS.

From a functional point of view, the processing of the pst mRNA permits the co-ordinated expression of the operon while allowing differential production of the phosphate-binding and pore-forming components of the transporter. The resulting molar excess of binding protein facilitates efficient scavenging of Pi from the growth medium, as many loaded copies of this protein are able to use a single transporter. Similar results have been reported for operons encoding periplasmic-binding-protein-dependent transport systems of Gram-negative bacteria (Newbury et al., 1987a; Horazdovsky & Hogg, 1987; Boos & Lucht, 1996; Hardham et al., 1997). This indicates that differential mRNA decay provides a mechanism for controlling the relative amounts of the protein components of ABC transporters in both Gram-negative and Gram-positive bacteria.

During these studies we used three different technologies to study the expression of the pentagenic pst operon. While there was a good correlation between the data obtained from Northern blot and DNA array experiments, with both able to detect marked differences in the quantities of the mRNA at the 5' and 3' ends of the operon, the lacZ reporter gene experiments showed a different pattern of expression. These results illustrate that, while lacZ reports faithfully the activity of the promoter, it less effectively reflects subsequent mRNA processing. This is particularly relevant, as in the case of the pst operon, when specific transcripts are differentially processed.

The lacZ transcriptional fusions to pstS, pstC and pstBB exhibited almost identical patterns of {beta}-galactosidase production. This implies that the lacZ-encoding regions of their mRNA transcripts, located at their 3' ends, are processed independently of other regions of the transcript. If, as our data suggest, the pst operon has internal endoribonuclease sites, subsequent 3'-5' exoribonuclease activity would not influence the amounts of lacZ mRNA. This is particularly clear in the case of the pstS : : lacZ and pstC : : lacZ fusions in which the absence of the putative stabilizing stem–loop in the former, and its presence in the latter, has no influence on the {beta}-galactosidase production. However, although {beta}-galactosidase production does not accurately reflect the amounts of pst mRNA, it does provide clear evidence that the intragenic stem–loop does not function as an intrinsic transcription terminator. An interesting but as yet unexplained observation is the lower amounts of {beta}-galactosidase produced by the pstA : : lacZ and pstBA : : lacZ strains.

In conclusion, our data support a model in which the preferential production of the binding component of the high-affinity phosphate transporter is the result of differential processing and stability of the full-length pst mRNA transcript. This was clearly shown with a combination of Northern blotting and DNA array experiments. While lacZ transcriptional fusion experiments were useful in confirming the absence of significant intrinsic transcription termination downstream of the pstS gene, they were not effective in monitoring the differential stability of the native pst mRNA.


   ACKNOWLEDGEMENTS
 
We thank W Schumann (University of Bayreuth, Germany) for the gift of strain PSTBA. This work was funded by the European Commission (QLG2-CT-1999-01455) and the UK Biotechnology and Biological Sciences Research Council (BBSRC; 13/PRES/12179). N. E. E. A. was in receipt of a studentship from the BBSRC.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Aguena, M., Yagil, E. & Spira, B. (2002). Transcriptional analysis of the pst operon of Escherichia coli. Mol Genet Genomics 268, 518–524.[CrossRef][Medline]

Anagnostopoulos, C. & Spizizen, J. (1961). Requirements for transformation in Bacillus subtilis. J Bacteriol 81, 741–746.

Antelmann, H., Scharf, C. & Hecker, M. (2000). Phosphate starvation-inducible proteins of Bacillus subtilis: proteomics and transcriptional analysis. J Bacteriol 182, 4478–4490.[Abstract/Free Full Text]

Arraiano, C. M., Yancey, S. D. & Kushner, S. R. (1988). Stabilization of discrete mRNA breakdown products in ams pnp rnb multiple mutants of Escherichia coli K-12. J Bacteriol 170, 4625–4633.[Medline]

Atalla, A. & Schumann, W. (2003). The pst operon of Bacillus subtilis is specifically induced by alkali stress. J Bacteriol 185, 5019–5022.[Abstract/Free Full Text]

Bachellier, S., Gilson, E., Hofnung, M. & Hill, C. W. (1996). Repeated sequences. In Escherichia coli and Salmonella: Cellular and Molecular Biology, pp. 2012–2040. Edited by F. C. Neidhardt and others. Washington, DC: American Society for Microbiology.

Belasco, J. G. & Higgins, C. F. (1988). Mechanisms of mRNA decay in bacteria: a perspective. Gene 72, 15–23.[CrossRef][Medline]

Boos, W. & Lucht, J. M. (1996). Periplasmic binding protein-dependent ABC transporters. In Escherichia coli and Salmonella: Cellular and Molecular Biology, pp. 1175–1235. Edited by F. C. Neidhardt and others. Washington, DC: American Society for Microbiology.

Bron, S. (1990). Plasmids. In Molecular Biological Methods for Bacillus, pp. 75–174. Edited by C. R. Harwood & S. M. Cutting. Chichester, UK: Wiley.

Chan, F. Y. & Torriani, A. (1996). PstB protein of the phosphate-specific transport system of Escherichia coli is an ATPase. J Bacteriol 178, 3974–3977.[Abstract]

Condon, C. (2003). RNA processing and degradation in Bacillus subtilis. Microbiol Mol Biol Rev 67, 157–174.[Abstract/Free Full Text]

Condon, C., Putzer, H., Luo, D. & Grunberg-Manago, M. (1997). Processing of the Bacillus subtilis thrS leader mRNA is RNase E-dependent in Escherichia coli. J Mol Biol 268, 235–242.[CrossRef][Medline]

DiMari, J. F. & Bechhofer, D. H. (1993). Initiation of mRNA decay in Bacillus subtilis. Mol Microbiol 7, 705–717.[Medline]

Eder, S., Liu, W. & Hulett, F. M. (1999). Mutational analysis of the phoD promoter in Bacillus subtilis: implications for PhoP binding and promoter activation of Pho regulon promoters. J Bacteriol 181, 2017–2025.[Abstract/Free Full Text]

Eymann, C., Mach, H., Harwood, C. R. & Hecker, M. (1996). Phosphate-starvation-inducible proteins in Bacillus subtilis: a two-dimensional gel electrophoresis study. Microbiology 142, 3163–3170.[Medline]

Eymann, C., Homuth, G., Scharf, C. & Hecker, M. (2002). Bacillus subtilis functional genomics: global characterization of the stringent response by proteome and transcriptome analysis. J Bacteriol 184, 2500–2520.[Abstract/Free Full Text]

Hambraeus, G., von Wachenfeldt, C. & Hederstedt, L. (2003). Genome-wide survey of mRNA half-lives in Bacillus subtilis identifies extremely stable mRNAs. Mol Genet Genomics 269, 706–714.[CrossRef][Medline]

Hardham, J. M., Stamm, L. V., Porcella, S. F. & 7 other authors (1997). Identification and transcriptional analysis of a Treponema pallidum operon encoding a putative ABC transport system, an iron-activated repressor protein homolog, and a glycolytic pathway enzyme homolog. Gene 197, 47–64.[CrossRef][Medline]

Harris, R. M., Webb, D. C., Howitt, S. M. & Cox, G. B. (2001). Characterization of PitA and PitB from Escherichia coli. J Bacteriol 183, 5008–5014.[Abstract/Free Full Text]

Harwood, C. R., Wipat, A. & Prágai, Z. (2002). Functional analysis of the Bacillus subtilis genome. Methods Microbiol 33, 337–367.

Hecker, M. & Völker, U. (1998). Non-specific, general and multiple stress resistance of growth-restricted Bacillus subtilis cells by the expression of the sigmaB regulon. Mol Microbiol 29, 1129–1136.[CrossRef][Medline]

Higgins, C. F. (1992). ABC transporters: from microorganisms to man. Annu Rev Cell Biol 8, 67–113.[CrossRef][Medline]

Higgins, C. F., McLaren, R. S. & Newbury, S. F. (1988). Repetitive extragenic palindromic sequences, mRNA stability and gene expression: evolution by gene conversion? A review. Gene 72, 3–14.[CrossRef][Medline]

Higgins, C. F., Peltz, S. W. & Jacobson, A. (1992). Turnover of mRNA in prokaryotes and lower eukaryotes. Curr Opin Genet Dev 2, 739–747.[Medline]

Homuth, G., Masuda, S., Mogk, A., Kobayashi, Y. & Schumann, W. (1997). The dnaK operon of Bacillus subtilis is heptacistronic. J Bacteriol 179, 1153–1164.[Abstract]

Homuth, G., Mogk, A. & Schumann, W. (1999). Post-transcriptional regulation of the Bacillus subtilis dnaK operon. Mol Microbiol 32, 1183–1197.[CrossRef][Medline]

Horazdovsky, B. F. & Hogg, R. W. (1987). High-affinity L-arabinose transport operon. Gene product expression and mRNAs. J Mol Biol 197, 27–35.[Medline]

Hulett, F. M. (1996). The signal-transduction network for Pho regulation in Bacillus subtilis. Mol Microbiol 19, 933–939.[Medline]

Hulett, F. M. (2002). The Pho regulon. In Bacillus subtilis and its Closest Relatives: from Genes to Cells, pp. 193–201. Edited by A. L. Sonenshein, J. A. Hoch & R. Losick. Washington, DC: American Society for Microbiology.

Hulett, F. M., Lee, J., Shi, L., Sun, G., Chesnut, R., Sharkova, E., Duggan, M. F. & Kapp, N. (1994). Sequential action of two-component genetic switches regulates the Pho regulon in Bacillus subtilis. J Bacteriol 176, 1348–1358.[Abstract]

Igo, M., Lampe, M., Ray, C., Schafer, W., Moran, C. P. & Losick, R. (1987). Genetic studies of a secondary RNA polymerase sigma factor in Bacillus subtilis. J Bacteriol 169, 3464–3469.[Medline]

Jürgen, B., Schweder, T. & Hecker, M. (1998). The stability of mRNA from the gsiB gene of Bacillus subtilis is dependent on the presence of a strong ribosome binding site. Mol Gen Genet 258, 538–545.[CrossRef][Medline]

Ludwig, H., Homuth, G., Schmalisch, M., Dyka, F. M., Hecker, M. & Stulke, J. (2001). Transcription of glycolytic genes and operons in Bacillus subtilis: evidence for the presence of multiple levels of control of the gapA operon. Mol Microbiol 41, 409–422.[CrossRef][Medline]

Moszer, I. (1998). The complete genome of Bacillus subtilis, from sequence annotation to data management and analysis. FEBS Lett 430, 28–36.[CrossRef][Medline]

Muda, M., Rao, N. N. & Torriani, A. (1992). Role of PhoU in phosphate transport and alkaline phosphatase regulation. J Bacteriol 174, 8057–8064.[Abstract]

Newbury, S. F., Smith, N. H. & Higgins, C. F. (1987a). Differential mRNA stability controls relative gene expression within a polycistronic operon. Cell 51, 1131–1143.[CrossRef][Medline]

Newbury, S. F., Smith, N. H., Robinson, E. C., Hiles, I. D. & Higgins, C. F. (1987b). Stabilization of translationally active mRNA by prokaryotic REP sequences. Cell 48, 297–310.[Medline]

Prágai, Z. & Harwood, C. R. (2000a). Screening for mutants affected in their response to phosphate. In Functional Analysis of Bacterial Genes: a Practical Manual, pp. 245–249. Edited by W. Schumann, S. D. Ehrlich & N. Ogasawara. Chichester: Wiley.

Prágai, Z. & Harwood, C. R. (2000b). YsxC, a putative GTP binding protein essential for the growth of Bacillus subtilis 168. J Bacteriol 182, 6819–6823.[Abstract/Free Full Text]

Prágai, Z. & Harwood, C. R. (2002). Regulatory interactions between the Pho and {sigma}B-dependent general stress regulons of Bacillus subtilis. Microbiology 148, 1593–1602.[Medline]

Prágai, Z., Eschevins, C., Bron, S. & Harwood, C. R. (2001). Bacillus subtilis NhaC, an Na+/H+ antiporter, influences expression of the phoPR operon and production of alkaline phosphatases. J Bacteriol 183, 2505–2515.[Abstract/Free Full Text]

Prágai, Z., Allenby, N. E., O'Connor, N., Dubrac, S., Rapoport, G., Msadek, T. & Harwood, C. R. (2004). Transcriptional regulation of the phoPR operon in Bacillus subtilis. J Bacteriol 186, 1182–1190.[Abstract/Free Full Text]

Qi, Y. & Hulett, F. M. (1998). PhoP-P and RNA polymerase sigmaA holoenzyme are sufficient for transcription of Pho regulon promoters in Bacillus subtilis: PhoP-P activator sites within the coding region stimulate transcription in vitro. Mol Microbiol 28, 1187–1197.[CrossRef][Medline]

Qi, Y., Kobayashi, Y. & Hulett, F. M. (1997). The pst operon of Bacillus subtilis has a phosphate-regulated promoter and is involved in phosphate transport but not in regulation of the Pho regulon. J Bacteriol 179, 2534–2539.[Abstract]

Sharp, J. S. & Bechhofer, D. H. (2003). Effect of translational signals on mRNA decay in Bacillus subtilis. J Bacteriol 185, 5372–5379.[Abstract/Free Full Text]

Steed, P. M. & Wanner, B. L. (1993). Use of the rep technique for allele replacement to construct mutants with deletions in the pstSCAB operon: evidence for the role of the PhoU protein in phosphate regulation. J Bacteriol 175, 6797–6809.[Abstract]

Tinoco, I., Jr, Borer, P. N., Dengler, B., Levin, M. D., Uhlenbeck, O. C., Crothers, D. M. & Bralla, J. (1973). Improved estimation of secondary structure in ribonucleic acids. Nat New Biol 246, 40–41.[Medline]

Vagner, V., Dervyn, E. & Ehrlich, S. D. (1998). A vector for systematic gene inactivation in Bacillus subtilis. Microbiology 144, 3097–3104.[Medline]

Vázquez-Cruz, C. & Olmedo-Alvarez, G. (1997). Mechanism of decay of the cry1Aa mRNA in Bacillus subtilis. J Bacteriol 179, 6341–6348.[Abstract]

Webb, D. C., Rosenberg, H. & Cox, G. B. (1992). Mutational analysis of the Escherichia coli phosphate-specific transport system, a member of the traffic ATPase (or ABC) family of membrane transporters. A role for proline residues in transmembrane helices. J Biol Chem 267, 24661–24668.[Abstract/Free Full Text]

Zuker, M. (1989). Computer prediction of RNA structure. Methods Enzymol 180, 262–288.[Medline]

Received 1 March 2004; revised 23 April 2004; accepted 26 April 2004.



This Article
Abstract
Full Text (PDF)
Alert me when this article is cited
Alert me if a correction is posted
Citation Map
Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Google Scholar
Articles by Allenby, N. E. E.
Articles by Harwood, C. R.
Articles citing this Article
PubMed
PubMed Citation
Articles by Allenby, N. E. E.
Articles by Harwood, C. R.
Agricola
Articles by Allenby, N. E. E.
Articles by Harwood, C. R.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
INT J SYST EVOL MICROBIOL MICROBIOLOGY J GEN VIROL
J MED MICROBIOL ALL SGM JOURNALS
Copyright © 2004 Society for General Microbiology.