Notice: Wiley Online Library will be unavailable on Saturday 27th February from 09:00-14:00 GMT / 04:00-09:00 EST / 17:00-22:00 SGT for essential maintenance. Apologies for the inconvenience.
George B. Spiegelman. E-mail email@example.com; Tel. (604) 822 2036; Fax (604) 822 6041.
Spo0A is a two-domain response regulator required for the initiation of sporulation in Bacillus subtilis. Spo0A is activated by phosphorylation of its regulatory domain by a multicomponent phosphorelay. To define the role of the regulatory domain in the activation of Spo0A, we have characterized four of the sof mutations in vitro. The sof mutations were identified previously as suppressors of the sporulation-negative phenotype resulting from a deletion of the gene for one of the phosphorelay components, spo0F. Like wild-type Spo0A, the transcription stimulation properties of all of the Sof proteins were dependent upon phosphorylation. Sof mutants from two classes were improved substrates for direct phosphorylation by the KinA sensor kinase, providing an explanation for their suppression properties. Two other Sof proteins showed a phosphorylation-dependent enhancement of the stability of the Sof∼P–RNA polymerase–DNA complex. One of these mutants, Sof114, increased the stability of the Sof114∼P–RNAP–DNA complex without increasing its own affinity for the spoIIG promoter. A comparison of the location of the sof mutations with mutations in CheY suggests that phosphorylation of Spo0A results in the exposure of a region in the regulatory domain that interacts with RNA polymerase, thereby contributing to the signal transduction mechanism.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
At present, the mechanism by which phosphorylation affects Spo0A such that it binds DNA and stimulates transcription is unknown. The predicted structure of the N-terminal domain of Spo0A (Spiegelman et al., 1990) is based on the actual structure of CheY (Stock et al., 1989; Volz, 1995) and Spo0F (Feher et al., 1997; Madhusudan et al., 1997) and on the N-terminus of other response regulators, such as NTRC (Volkman et al., 1995) and NarL (Baikalov et al., 1996). The basic structure of Spo0F and CheY is an (α/β)5 barrel. The five α-helices surround the β-sheets, with α-helices 1–3 on one side of the barrel and α-helices 4 and 5 on the other side. The amino acid residues of CheY that are involved in the phosphorylation reaction consist of D12, D13, D57, T87 and K107. D57 is the actual phosphorylation site, and D13 is involved in the binding of Mg2+, which may be a requirement for phosphorylation (Lukat et al., 1990). Several protein–protein interaction regions have been identified within CheY (reviewed in Stock et al., 1995; Schuster et al., 1998) and Spo0F (Tzeng and Hoch, 1997; Feher et al., 1998), leading to the possibility that the N-terminal domains of response regulators similar to CheY and Spo0F may play a more active role in the signalling pathway than has previously been thought.
Many regulatory domain mutations of spo0A have been identified (reviewed in Spiegelman et al., 1995). The sof mutations were isolated as suppressors of a spo− strain carrying a deletion of the spo0F gene (Spiegelman et al., 1990). The 22 sof mutations that have been mapped were all single amino acid changes within the regulatory domain of Spo0A. The mutations were grouped into six classes: class 1 (N12K); class 2 (E14V); class 3 (E14A); class 4 (P60S); class 5 (D92Y) and class 6 (Q121R), which exhibit differential suppression of deletions of various stage 0 genes. Therefore, they are potentially within regions that affect protein–protein interactions between Spo0A and its regulators. We characterized the effect of four classes of sof mutations on protein function in vitro and found that the Sof proteins all required phosphorylation to stimulate transcription. This led to the possibility that the mutations induced phosphorylation-dependent secondary effects, which would further define the role of the regulatory domain in the transcription stimulation process.
Phosphorylation of the Sof proteins is required for efficient stimulation of transcription from the spoIIG promoter
One representative from each of four classes of sof mutations was studied: sof103 (N12K) from class 1; sof115 (E14A) from class 3; sof118 (P60S) from class 4; and sof114 (D92Y) from class 5. As one of the possible results of Spo0A regulatory domain mutations was a constitutively active protein, the purified Sof proteins were tested for their ability to stimulate transcription. We used an in vitro transcription assay, which detected transcripts produced by purified B. subtilisσA RNA polymerase from the Spo0A-dependent spoIIG promoter (Bird et al., 1993, 1996). Phosphorylation of Spo0A induces a dramatic increase in its ability to stimulate transcription from this promoter (Satola et al., 1992; Bird et al., 1993; Baldus et al., 1994). Wild-type Spo0A or the Sof proteins were incubated with the phosphorelay components (KinA, Spo0F and Spo0B) and ATP and then added directly to the in vitro transcription assay (Fig. 1). Phosphorylation of Spo0A and the Sof proteins enhanced transcription three- to fivefold, with the exception of Sof115, which caused an eightfold increase in transcription. These results demonstrated that the sof mutations did not abrogate the need for phosphorylation to activate the transcription stimulation properties.
Sof103 and Sof115 are phosphorylated directly by KinA
As the sof mutations bypass the phosphorelay, one of their potential effects could be an alteration in the affinity of the mutant Spo0A proteins for one or more of the sensor kinases (Spiegelman et al., 1990; LeDeaux and Grossman, 1995). To test whether the Sof proteins were phosphorylated directly by the sensor kinase KinA, Spo0A or the Sof proteins were incubated with either the phosphorelay proteins or KinA in the presence of [γ-32P]-ATP. The quantification of an SDS–PAGE analysis of the radioactively labelled proteins is shown in Table 1. All of the Sof proteins were phosphorylated by the phosphorelay to approximately the same level as a similar protein input of Spo0A. However, Sof103 and Sof115 were phosphorylated directly by KinA at levels 2.5- to sevenfold higher than Spo0A. The phosphorylation of Sof118 and Sof114 by KinA was unaltered compared with Spo0A.
Table 1. . Phosphorylation of Spo0A and the Sof proteins. a. The percentage phosphorylation of the proteins by KinA was calculated relative to the phosphorylation of Spo0A by the phosphorelay.b. PR is an abbreviation for phosphorelay.c. The values are from two separate experiments, each normalized to their own Spo0A control.
To test whether the Sof103 and Sof115 proteins were improved substrates for the KinA sensor kinase, the rate of phosphorylation of these Sof proteins by KinA was compared with that of Spo0A and the KinA substrate, Spo0F. Results showed that KinA phosphorylated both Sof103 and Sof115 at a faster rate than Spo0A, but that the rate did not approach that of Spo0F (Fig. 2). After a 30 min incubation, phosphorylation of the Sof proteins was fivefold higher than the phosphorylation of Spo0A. This indicated that Sof103 and Sof115 were substantially better substrates for KinA than Spo0A.
The phosphorylated Sof proteins are not resistant to the Spo0E phosphatase in vitro
The in vivo studies of Spiegelman et al. (1990) demonstrated that the sof mutations suppressed the spo− phenotype of a strain carrying the spo0E11 mutation, which is a hyperactive variant of the Spo0A∼P phosphatase. This finding suggested that the phosphorylated Sof proteins could be resistant to dephosphorylation by Spo0E. We tested this possibility using an in vitro phosphatase assay with purified, wild-type Spo0E. Spo0A and the Sof proteins were phosphorylated by incubation with the phosphorelay and [γ-32P]-ATP. Spo0E was then added for a 15 min incubation, the labelled proteins separated by SDS–PAGE and the levels of phosphorylation quantified (Fig. 3). As the findings were similar for all three of the Sof proteins, the results for Sof115∼P are shown as a representative. The Sof proteins tested, Sof115, Sof103 and Sof114, were not more resistant to dephosphorylation by Spo0E compared with Spo0A∼P. Therefore, we concluded that resistance to phosphatase activity does not contribute to the suppression phenotype of the sof mutations tested.
The sof114 mutation stabilizes the interaction with RNA polymerase
The phosphorylated Sof proteins and wild-type Spo0A∼P were tested for their ability to form complexes that were resistant to increasing salt concentrations (Table 2). The in vitro transcription assays were performed routinely using a potassium acetate (KAc) concentration of 80 mM (Bird et al., 1993). Above 80 mM KAc, transcription stimulation is affected by the activator used. Transcription activation by both Spo0A and the isolated C-terminal, DNA binding domain of Spo0A, Spo0A-C, is highly sensitive to concentrations of KAc above 80 mM, compared with Spo0A∼P (Rowe-Magnus and Spiegelman, 1998). To monitor the sensitivity of the complexes of wild-type or the Sof forms of Spo0A and RNA polymerase to KAc, complexes were formed at low salt concentrations and then challenged by increasing the KAc concentration. Like Spo0A∼P, transcript production in the presence of the Sof proteins did not vary when challenged with KAc concentrations up to 80 mM. Furthermore, for equivalent concentrations of Sof∼P or Spo0A∼P, transcript levels were comparable when challenged with 20 mM KAc in the case of Sof114∼P, or with 80 mM KAc in the case of Sof118∼P. However, at 120 mM KAc, Sof114∼P and Sof118∼P stimulated transcription levels 10-fold higher than comparable concentrations of Spo0A∼P. The potential explanation that the sof mutations simply exposed the C-terminal domain was unlikely, as stimulation by Spo0A-C was sensitive to the salt challenge. Thus, the salt resistance phenomenon was a characteristic of the regulatory domains of the sof114 and sof118 mutant forms of Spo0A.
Table 2. . Transcription stimulation by Spo0A, spo0A-C or the Sof proteins in potassium acetate. a. Results are expressed as the percentage of transcription relative to the maximum transcription in 20 or 80 mM KAc.b. Protein inputs of 136 nM.c. Spo0A-C is the C-terminal, DNA binding domain of Spo0A.d. Protein inputs of 200 nM.
The apparently stable complexes formed by the Sof114 and Sof118 proteins, RNAP and DNA could result from differences in the dissociation and/or association rates. To test dissociation, we formed the complexes in 20 mM KAc, shifted them to 120 mM KAc and monitored decay of the complexes by testing their ability to produce transcripts from the spoIIG promoter (Fig. 4). For both wild-type and Sof proteins, there was an immediate drop in complexes and then a slow decay. The effect of the sof mutations was to increase the fraction of complexes that were stable to 120 mM KAc. If the KAc concentration was adjusted to 120 mM after the RNA polymerase was allowed to initiate by the addition of ATP and GTP, the increase in KAc had no effect on the levels of transcription stimulation by Spo0A and the Sof proteins (data not shown). This demonstrated that the stabilization of the complexes by the Sof mutants occurred before initiation. The ability to support higher levels of transcription in 120 mM KAc was unique to the Sof114 and Sof118 proteins, as the Sof115 transcription levels were similar to those of Spo0A (data not shown).
It was possible that the increased stability of complexes formed with Sof114 and Sof118 would be reflected in initial transcription rates and/or the binding affinity of the Sof proteins for the spoIIG promoter in the absence or presence of RNA polymerase. To test the transcription initiation rates, Sof114∼P, Sof118∼P or Spo0A∼P were incubated with the spoIIG promoter. The time course was started by the addition of RNA polymerase, and samples were removed at 10 s intervals and added to heparin plus UTP and CTP. After elongation, the level of transcription was determined. The initial rates of transcription in the presence of Sof114∼P, Sof118∼P and Spo0A∼P were identical (Fig. 5). Therefore, the increased levels of stable complexes seen with the Sof mutants were not caused by a change in the forward rates of complex formation.
Spo0A binds to two sites upstream of the spoIIG promoter. Each site contains two 0A boxes, and Spo0A appears to bind as dimers (Asayama et al., 1995). The upstream site (−82 to −97; site 1) is not required for normal regulation, whereas the downstream site (−37 to −56; site 2) is required (Baldus et al., 1994). The increased stability of Sof114∼P–RNA polymerase over Spo0A∼P–RNA polymerase complexes might be the result of enhanced binding of Sof114∼P to DNA. We analysed the binding of Sof114∼P and Spo0A∼P to the spoIIG promoter using a DNase I protection assay. Increasing concentrations of Spo0A∼P or Sof114∼P were incubated with the end-labelled spoIIG promoter, in the absence or presence of RNA polymerase, then treated with DNase I. The products of the DNase I digestion were separated by electrophoresis and detected by autoradiography (Fig. 6). Protection of the upstream region of the spoIIG promoter by Spo0A∼P showed a non-linear dependence on protein concentration, as reported in other studies (Strauch et al., 1990; Asayama et al., 1995; Greene and Spiegelman, 1996). Unexpectedly, the sof114 mutation decreased the binding of the mutant Spo0A on its own to the spoIIG promoter compared with wild-type Spo0A, as complete protection of the 0A boxes occurred with a concentration of 352 nM Spo0A∼P, whereas the same concentration of Sof114∼P only partially protected the 0A boxes.
As reported earlier, RNA polymerase alone bound to the spoIIG promoter providing protection upstream of the −10 site. Addition of both Spo0A∼P and Sof114∼P increased protection, both in the extent of protection and in the number of nucleotides protected. Both Spo0A∼P and Sof114∼P induced hypersensitive sites when bound with RNA polymerase (Fig. 6, arrows), as described by Bird et al. (1996). These hypersensitive sites could be detected at a concentration of 176 nM for either Spo0A∼P or Sof114∼P, indicating that the presence of RNA polymerase stimulated the binding of both Spo0A forms relative to their binding alone. In the presence of RNA polymerase, the binding of Sof114∼P and Spo0A∼P was similar. As Sof114∼P was an inherently weaker DNA binding protein than Spo0A∼P, the presence of RNA polymerase had a greater effect on Sof114 than it had on Spo0A, a finding compatible with increased interaction conferred by the sof114 mutation in transcription assays.
It was originally predicted that all of the sof increased the affinity of the proteins for one of the sensor kinases responsible for sporulation initiation to allow direct phosphorylation (Spiegelman et al., 1990). This hypothesis explained their suppression of the mutation in the phosphorelay (deletion of spo0F or spo0B ) and the fact that the activities of the proteins were still regulated during growth and development. It has since been demonstrated that the expression of constitutively active forms of Spo0A results in a phenotype different from that seen in the sof suppressor strains (Ireton et al., 1993). It was also possible that the Sof forms of Spo0A were more effective at transcription stimulation once phosphorylated, so that a lower cellular concentration would be needed to trigger sporulation. This low concentration might be achieved without the use of the phosphorelay. The discovery of sporulation-specific protein phosphatases (Ohlsen et al., 1994; Perego et al., 1994; 1996) provided another alternative for the effect of sof mutations. If Sof forms were highly resistant to dephosphorylation, phosphate input to the protein could result from the activity of a low-affinity kinase. In this study, we used our in vitro transcription assay to investigate these mechanisms for four of the six classes of sof mutants.
All four classes of Sof proteins required phosphorylation to stimulate transcription from the spoIIG promoter, supporting the prediction that the proteins were not constitutively active. The fact that the proteins showed wild-type sensitivity to the phosphatase Spo0E suggested that the sof mutations did not change the resistance of the proteins to dephosphorylation. For two of the sof mutant classes [class 1 (e.g. sof103, N12K) and class 3 (e.g. sof115, E14A)], we demonstrated conclusively that they were improved in vitro substrates for direct phosphorylation by KinA. These findings can probably be extended to class 2 mutations (e.g. sof104, E14K), which alter the same residue as in class 3. Thus, for some sof mutations, the primary effect was an alteration in the mechanism of phosphate input into the Spo0A protein. Based on the structures of Spo0F and CheY, these three mutations lie in the predicted α1 region. Furthermore, sof114 and sof118 did not show altered affinity for KinA, and the mutations in these proteins were not in α1. Altogether, these results support the extensive mutation analysis of Spo0F, which has also shown that the α1 region is the site of interaction with KinA (Tzeng and Hoch, 1997). These results strengthen the prediction that the overall structure of the N-terminus of Spo0A will be similar to that of Spo0F and CheY.
As the sof118 and sof114 mutations were able to suppress kinA and kinA,kinB mutants in vivo, yet required phosphorylation to stimulate transcription, it is clear that Sof118 and Sof114 can be phosphorylated by another source in vivo, possibly by KinC. The expected level of direct input of phosphorylation by KinC is low (Kobayashi et al., 1995; LeDeaux and Grossman, 1995; LeDeaux et al., 1995), suggesting to us that direct phosphorylation could not fully explain the suppression by these two mutations. We investigated the transcription activation properties of Sof114 and Sof118, and the results suggested that the mutations altered the signal transduction characteristics of the proteins, as complexes formed by RNA polymerase and Sof114∼P or Sof118∼P showed increased resistance to ionic strength. In the case of Sof114, we have shown that the increased resistance was not caused by the increased stability of binding of the Sof protein alone to the DNA.
Green et al. (1991) identified a mutation identical to sof118 in a screen for suppressors of the spo− mutation D10Q. As the D10 residue is predicted to form part of the acidic pocket necessary for phosphorylation, the D10Q mutation probably affects either the transfer of phosphate to Spo0A by Spo0B or the stability of Spo0A∼P. The effect of D10Q would be to lower the input of phosphate to Spo0A, so it is functionally similar to a deletion of spo0F.
Several deletion mutations of Spo0A that suppress an alteration of the amino acid residue predicted to be the site of phosphorylation reaction (D56) have been identified (Green et al., 1991; Ireton et al., 1993). These mutants, which are in α1, are unusual in that they are not constitutively active and therefore must be regulated in some manner. While the deletions have been predicted not to alter the structure of the N-terminus dramatically (Green et al., 1991), how this form of Spo0A is regulated in vivo is unknown. The sad mutations, which are deletions of varying sizes in the predicted α3 region, also suppress the D56Q mutation but are constitutively active (Ireton et al., 1993). This suggests that the large deletions of 5–20 amino acid residues in this region abrogate the inhibitory function of the N-terminus. The mechanism of action of sad54, a single amino acid deletion (D75), is less obvious, but this region of the N-terminal domain has been proposed to be critical for suppression of the activity of the C-terminal domain (Feher et al., 1998).
Perego and Hoch (1996) speculated that regulation of Spo0A∼P may include a component of autodephosphorylation that could be stimulated during interaction with RNA polymerase. If so, mutations such as sof114 and sof118 could decrease the autodephosphorylation, leading to an increase in the steady-state level of Spo0A∼P in vivo. However, if such a mechanism contributed to the salt stability in the in vitro transcription assays, one might also predict higher levels of transcription stimulation at low salt concentrations from Sof114∼P or Sof118∼P than from Spo0A∼P, which we did not observe.
The role of phosphorylation of multidomain response regulators is not completely understood. While it is generally accepted that the input domain is a negative regulator of the output domain and that phosphorylation blocks the inhibition, additional phosphorylation-dependent roles for the N-terminus in the signal transduction pathway are possible. For example, the phosphorylation of OmpR and NTRC stimulates multimerization and binding to DNA through the N-terminal interactions (Weiss et al., 1992; Porter et al., 1993; Chen and Reitzer, 1995). OmpR second-site mutations T83A or G94S restore the transcription regulation activity lost upon mutation of the phosphorylation site D55. This suggests that these mutations, which are within the β4–α4 region, mimic the effect of phosphorylation (Brissette et al., 1992). The assignment of the sof114 mutation (D92Y) to the γ-loop between β3 and α4 or to α4 must await determination of the structure of Spo0A. In either case, the mutation could alter the positioning of one or more of the neighbouring helices, affecting a protein–protein interaction surface.
Mutational analysis of CheY has demonstrated that binding to the downstream target protein FliM and transmission of the signal are separable events. Binding to FliM is mediated by the α5–β5 face of the CheY protein, and the structure of this face is regulated by the α4 region (Schuster et al., 1998). The similarities of CheY and Spo0F structure and the similarities of amino acid sequence between these proteins and the N-terminus of Spo0A suggest that the α5–β5 region of the Spo0A input domain is a potential site for protein interactions. We suggest that the sof114 mutation enhances a natural interaction between the phosphorylated N-terminus of Spo0A and the RNA polymerase, which is parallel to the increases in interactions seen between mutants of CheY and FliM. This increased interaction implies that the concentration of phosphorylated Sof protein needed to trigger sporulation would be lower and could potentially be achieved without the phosphorelay.
While the proposed N–terminal interaction is not the only one made with the RNA polymerase (Baldus et al., 1995; Schyns et al., 1997; Buckner et al., 1998; Hatt and Youngman, 1998) and cannot by itself explain transcription activation, it provides a rationale for differences between the intact phosphorylated Spo0A and the isolated output domain that we have observed previously (Grimsley et al., 1994; Rowe-Magnus and Spiegelman, 1998). We note that, while it has been shown that the N-terminal domains of response regulators are involved in signal transduction by controlling the formation of multimers, this is the first suggestion that the N-terminus of Spo0A∼P communicates directly with the RNA polymerase.
Bacteria, strains and media
The B. subtilis strains with the sof mutations that were used in this study have been described previously (Spiegelman et al., 1990): B. subtilis 12578::sof103, B. subtilis 12578::sof115, B. subtilis 12578::sof114. The B. subtilis strains were grown as described previously (Hoch, 1991). Escherichia coli strains DH5α and Epicurean Coli BL21(λDE3)pLysS were purchased as transformation-competent cells from Stratagene.
Proteins, plasmids and reagents
All restriction enzymes, T4 polynucleotide kinase and Taq T7 RNA polymerase were purchased from Gibco BRL. Vent polymerase was purchased from New England BioLabs. The pGEM-T polymerase chain reaction (PCR) cloning vector was purchased from Promega and the pET16b protein expression vector from Novagen. The NTPs and dNTPs used in the transcription assays and PCR reactions, respectively, were purchased as FPLC grade from Pharmacia. The (α-32P)-GTP (800 Ci mM−1) was from Mandel Scientific (NEN Life Science Products), and the [γ-32P]-ATP (7000 Ci mM−1) was from ICN Biochemicals. All chemicals were purchased from Sigma except IPTG and SDS–PAGE reagents, which were from Gibco BRL. Synthesis of PCR and sequencing primers and all sequencing reactions were performed by the Nucleotide and Protein Synthesis Unit, University of British Columbia, Vancouver, BC, Canada. Purified Spo0A, Spo0A-C, KinA, Spo0B and Spo0F were obtained from Dr James Hoch, Scripps Research Institute, La Jolla, CA, USA.
The spoIIG template DNA was isolated from the plasmid pUCIIGtrpA (Satola et al., 1991). This plasmid has 100 bp of B. subtilis DNA upstream of the spoIIG transcription start site, including the site 1 and site 2 0A boxes, 130 bp downstream of +1 and the trpA transcription terminator. For the transcription assays, pUCIIGtrpA was digested with PvuII, which liberated a 600 bp fragment containing the spoIIG promoter, 130 bp downstream of the +1 transcription start site, the trpA terminator and approximately 100 bp of vector sequences both upstream and downstream of spoIIG DNA.
Cloning and sequencing of the sof103, sof115, sof118 and sof114 genes
Isolation of chromosomal DNA from the B. subtilis::sof strains was carried out as described previously (Hoch, 1991), and the mutant spo0A genes were amplified by PCR. The PCR reactions (100 μl) contained approximately 1 μg of chromosomal DNA, 50 pM each primer, dATP, dCTP, dGTP and dTTP at a final concentration of 1 mM each, MgSO4 at a final concentration of 3 mM, 1× Vent polymerase react buffer and 2 units of Vent polymerase. PCR reactions were carried out for 30 cycles (one cycle comprised 95°C for 30 s; 57°C for 1 min; 72°C for 1 min), followed by 5 min at 72°C in an Ericomp TwinBlock system PCR machine. To clone sof103, sof115 and sof114, an 800 bp fragment was generated using the primers 0A-3 (5′-CGCCATGGGAGGAAGAAACGTG-3′) and 0A-4 (5′-CGGGATCCAAAGACGTTTGAT-3′). The upstream primer, 0A-3, created an NcoI site at the 5′ end of the fragment, and the downstream primer, 0A-4, created a BamHI site at the 3′ end. The 800 bp fragment included the ribosome binding site, located at −10 from the GTG start codon, the entire sof coding region, the spo0A stop codon and the transcription terminator. To amplify the sof118 gene by PCR, the upstream primer 0A-5 (5′-CGCCATGGAGAAAATTAAAGTTTGTGTTG-3′) was used in combination with the 0A-4 primer to generate an 800 bp fragment. The 0A-5 primer created an NcoI site at the 5′ end and changed the start codon from GTG to ATG. The 800 bp fragments were recovered by electroelution from a 0.7% agarose gel, ethanol precipitated and resuspended in 10 μl of TE (10 mM Tris-HCl, 1 mM EDTA, pH 8.0). To create a dATP overhang, the fragments were incubated at 72°C for 20 min with 1 mM dATP, 1 unit Taq polymerase in 1×Taq reaction buffer (final volume 15 μl). These fragments were then ligated directly into the pGEM-T cloning vector and sequenced in their entirety to confirm the presence of the particular mutation. The plasmids containing the cloned fragments were digested with NcoI and BamHI, the 800 bp fragments recovered by electroelution, ligated into the NcoI- and BamHI-digested pET16b expression vector and transformed into both E. coli DH5α and the protein expression host E. coli host BL21(λDE3)pLysS. The pET16b vectors containing the sof fragments were designated pMCsof103, pMCsof115, pMCsof118 and pMCsof114. BL21(λDE3)pLysS hosts containing the pMCsof plasmids were designated MCsof103, MCsof115, MCsof114 and MCsof118.
Overexpression and purification of the Sof proteins
Cultures [100 ml of Luria broth (LB) + 100 μg ml−1 ampicillin] of the MCsof103, MCsof115, MCsof118 and MCsof114 strains were grown overnight at 30°C with shaking at 180 r.p.m. These cultures were used to inoculate 2 l of LB broth + 100 μg ml−1 ampicillin, which were grown for approximately 3 h at 30°C with shaking at 180 r.p.m. until the culture reached an OD600 of 1.0. T7 RNA polymerase was induced by the addition of a final concentration of 1 mM IPTG. After a 4–5 h induction period, the cultures were harvested by centrifugation at 7000 × g for 15 min and the pellets frozen at −20°C. The protein purification up to and including the 50% ammonium sulphate precipitation step was performed as described by Grimsley et al. (1994). The pellet from the 20–50% ammonium sulphate cut was resuspended in 5–10 ml of buffer A [20 mM potassium phosphate buffer, pH 6.9, containing 150 mM NaCl, 1 mM dithiothreitol (DTT), 1 mM phenylmethylsulphonyl fluoride (PMSF)] and dialysed overnight at 4°C against 2 l of buffer A. The dialysate was filtered through a syringe filter (Millex-HA 0.45 mM; Millipore) to remove any insoluble material and then loaded directly onto a 10 ml heparin–agarose column (1 inch diameter, 20 cc syringe) equilibrated with buffer A. The column was washed with buffer A until the OD280 < 0.1. The column was washed with buffer A + 250 mM NaCl until the OD280 dropped below 0.1, and the Sof protein was then eluted with 60 ml of a 250–850 mM NaCl gradient in buffer A. Fractions (1 ml) were analysed by SDS–PAGE, and those containing the Sof proteins (≈11 fractions) were pooled and concentrated to 0.5–1.0 ml by covering a dialysis bag containing the fractions with PEG 20 000 for approximately 3 h at 4°C. The concentrated fractions were dialysed overnight at 4°C against 2 l of buffer C (20 mM NaP04 pH 8.0, 150 mM NaCl, 1 mM β-mercaptoethanol, 1 mM PMSF), and glycerol was added to a final concentration of 30%. The samples were aliquoted and stored at −20°C. The final product was approximately 85–90% pure and was used directly for the in vitro studies. Because the preparations were not completely pure, the concentration of the Sof proteins was determined by comparison with known amounts of Spo0A on SDS–PAGE gels.
In vitro phosphorelay reactions
Phosphorylation of Spo0A or the Sof proteins was carried out essentially as described previously (Burbulys et al., 1991; Bird et al., 1996). Reactions consisting of the complete phosphorelay contained 1 μM KinA, 0.2 μM Spo0B, 2.0 μM Spo0F and 1 mM ATP in 1× transcription reaction buffer (0.01 M HEPES, pH 8.0, 0.01 M MgAc, 1 mM DTT, 0.1 mg ml−1 BSA, 80 mM KAc) in a final volume of 20 μl. Unless otherwise stated, the phosphorelay reactions were incubated for 1.5 h at room temperature.
For the phosphorylation analyses, the indicated concentrations of Spo0A or the Sof proteins were incubated with either the complete phosphorelay or KinA alone as described, except that the ATP concentration was 25 μM and included 25 μCi of [γ-32P]-ATP. The reactions shown in Table 1 were incubated at room temperature for 1 h in the case of Spo0A, Sof103, Sof115 and Sof114, or 30 min in the case of Spo0A and Sof118. The reactions were then terminated by the addition of a final concentration of 1× SDS–PAGE sample buffer (2 × is 100 mM Tris-HCl, pH 6.8, 4% SDS, 20% glycerol, 0.2% bromophenol blue, 1.2% β-mercaptoethanol), and the labelled proteins were separated by electrophoresis through a 15% (29:1 acrylamide–N,N′-methylene-bis-acrylamide) SDS–PAGE gel. The gels were dried and exposed to a Molecular Dynamics phosphorimaging cassette for 1 h or X-ray film (Kodak XAR) overnight. Quantification of the percentage phosphorylation of the proteins was performed using the Molecular Dynamics SI PhosphorImager and IMAGEQUANT 1.0 software. Spo0A and the Sof proteins were phosphorylated by the phosphorelay to similar levels. The levels of phosphorylation by KinA were normalized with respect to the level of phosphorylation of Spo0A by the phosphorelay (100% value). Spo0A and Sof118 were added at concentrations of 3 μM; Spo0A and Sof115 were added at concentrations of 1.2 μM; Sof103 and Sof114 were added at concentrations of 500 nM.
To analyse the rates of phosphorylation (Fig. 2), Sof103, Sof115, Spo0A or Spo0F were incubated with KinA in 1× transcription buffer (final volume 10 μl) for the indicated time in the presence of 25 μM ATP and 25 μCi of [γ-32P]-ATP. Reactions were terminated by the addition of 1× SDS–PAGE sample buffer, and the labelled proteins were detected and quantified as described for the reactions in Table 1. The analyses of Sof103 and Sof115 were carried out separately with the accompanying Spo0A and Spo0F controls. The rates of phosphorylation of Spo0A and Spo0F were identical in both cases so, for convenience, the results are presented as a single graph.
Spo0E phosphatase assays
The spo0E clone and (His)6-tagged Spo0E purification procedure were obtained from Dr Marta Perego, Scripps Research Institute, La Jolla, CA, USA. For the phosphatase assays shown in Fig. 3, equal concentrations of Spo0A or Sof115 (4 μM) were phosphorylated by the phosphorelay as described for the reactions in Table 1. The indicated inputs of the Spo0E phosphatase were then added to the phosphorylation reactions, the incubation was continued for a further 15 min and the reactions were stopped by the addition of 1× SDS–PAGE sample buffer. The proteins were separated on a 15% SDS–PAGE gel, exposed to X-ray film, and the phosphorylation levels of proteins were quantified using IMAGEQUANT 1.0 software. All of the Sof proteins, except Sof118, were tested, but only the results for Sof115 are shown as an example, as the findings were all similar.
In vitro transcription assays
B. subtilisσA RNA polymerase was isolated as described previously (Dobinson and Spiegelman, 1987). Single-round in vitro transcription assays (Bird et al., 1993; 1996) were carried out in 1× transcription buffer in a final volume of 10 μl. RNA polymerase was incubated with unphosphorylated or phosphorylated Spo0A or Sof proteins and the initiation nucleotides ATP and GTP for 3 min at 37°C. This was followed by an elongation step brought about by the addition of UTP, CTP and heparin and incubation for 5 min at 37°C. The reactions were terminated by the addition of 5 μl of stop buffer (7 M urea in 2× Tris borate–EDTA, 1% of each of bromophenol blue and xylene cyanol; 1× TBE is 0.09 M Tris borate, 0.002 M EDTA; Sambrook et al., 1989). The transcripts were separated by electrophoresis through an 8% polyacrylamide gel (40:1.38. acrylamide–N,N′-methylene-bis-acrylamide) containing 7 M urea for 30 min at 600 V. Unless otherwise indicated, the transcripts were detected and quantified using the Molecular Dynamics PhosphorImager and IMAGEQUANT 1.0 software. Transcription assays were repeated more than twice, and representative experiments are reported.
The transcription assays contained the indicated concentrations of Spo0A or the Sof proteins, 2 nM spoIIG promoter template, 25 nM RNA polymerase, 0.4 mM each of ATP, CTP and UTP, 2 μCi [α-32P]-GTP and 10 μg ml−1 heparin. RNA polymerase was diluted in RNA polymerase dilution buffer (1× transcription reaction buffer, 10% glycerol); Spo0A∼P was diluted in 1× transcription buffer.
For the transcription assays presented in Table 2, transcription buffers contained 20 or 120 mM KAc (Spo0A∼P, Spo0A-C, Sof114) or 80 or 120 mM KAc (Spo0A∼P, Sof118∼P). Spo0A or the Sof proteins were incubated with the spoIIG promoter and RNA polymerase for 3 min at 37°C in buffer containing 20 or 80 mM KAc; then KAc was added to a final concentration of 120 mM. After a 2 min incubation, the initiation nucleotides ATP and GTP were added and the reactions incubated for 3 min, followed by the addition of the elongation nucleotides, UTP and CTP, and heparin. The transcripts were separated and analysed as described above. Transcription levels are presented as a percentage of the respective transcript levels under the 20 or 80 mM KAc conditions. Sof114 and Spo0A were added to the transcription reactions at a concentration of 136 nM. Spo0A-C was added at a concentration of 400 nM, which yielded transcription levels similar to those of 136 nM Spo0A∼P in 20 mM KAc. Spo0A∼P and Sof118∼P were added to the reactions at a concentration of 300 nM.
To determine the rate of dissociation of the Spo0A/Sof–RNA polymerase–DNA complexes (Fig. 4), phosphorylated Spo0A or the indicated Sof protein was incubated with spoIIG template DNA and RNA polymerase for 2 min at 37°C in 1× transcription buffer containing 20 mM KAc. Potassium acetate was then added to the reactions to a final concentration of 120 mM, and the incubation was continued at 37°C. At the indicated times, the initiation nucleotides ATP and GTP were added, the incubation continued for 3 min, and then the initiated complexes were allowed to elongate for 5 min at 37°C by the addition of heparin, UTP and CTP. Reactions were terminated by the addition of stop buffer, then electrophoresed and analysed as described. The percentage transcription was calculated relative to the respective transcription levels in 20 mM KAc. Experiments with Sof114 and Sof118 were performed separately with their own Spo0A∼P control. As the Spo0A curves were similar in both instances, the results were combined into a single graph.
To determine the rates of initiation of transcription, Sof114, Sof118 and Spo0A were phosphorylated by incubation with the phosphorelay components. A mixture containing Spo0A∼P, Sof114∼P or Sof118∼P, spoIIG template DNA and the initiation nucleotides ATP and GTP in 1× transcription buffer was incubated for 3 min at 37°C. RNA polymerase was then added and, at the indicated times, aliquots were removed and added to the prewarmed elongation mix containing UTP, CTP and heparin. After 5 min at 37°C for elongation, the reactions were stopped and the products, electrophoresed and analysed as described above.
DNase I footprint assays
DNase I footprint reactions were performed essentially as described previously (Bird et al., 1996). The DNA template was a 410 bp BamHI/PvuII fragment from pUCIIGtrpA end-labelled at the BamHI site by incubation with T4 polynucleotide kinase and [γ-32P]-ATP. DNA (2 nM) was incubated with increasing inputs of Spo0A∼P or Sof114∼P, in the absence or presence of RNA polymerase (25 nM) in 1× transcription buffer (final volume 20 μl) for 5 min at 37°C. The samples were then treated with DNase I (final concentration of 1.3 μg ml−1) for 10 s and stopped by the addition of three volumes of stop buffer (0.1% SDS, 4.0 mM EDTA, 270 mM NaCl and 40 μg ml−1 sonicated calf thymus DNA). Control reactions contained DNA and 2 μl of RNA polymerase dilution buffer or DNA incubated with RNA polymerase alone. The DNA was ethanol precipitated and resuspended in 5 μl of formamide loading buffer (95% formamide in 2× TBE, 1% of each of bromophenol blue and xylene cyanol). After boiling the samples for 3 min, an equal amount of radioactivity from each sample was loaded onto an 8% polyacrylamide sequencing gel containing 7 M urea. The DNA fragments were separated by electrophoresis at 45 W for approximately 4 h, the gels dried and then exposed to X-ray film overnight at −80°C. To determine the nucleotide positions relative to the + 1 transcription start site, the end-labelled DNA fragment was digested with HindIII (−100), AseI (−43) or AluI (−27) to produce size markers and electrophoresed in lanes adjacent to the footprint reactions.
We would like to thank Loverne Duncan and Janel Middelkamp for technical assistance, Dean Rowe-Magnus for helpful discussions, J. A. Hoch and members of his laboratory for the phosphorelay proteins, and M. Perego for the spo0E clone. This work was supported by grants to G.B.S. from the National Sciences and Engineering Research Council of Canada.