INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

When its nutrients are depleted, the gram-positive bacterium Bacillus subtilis undergoes a morphologically and genetically complex differentiation process that culminates in the production of an endospore (18, 51). Early during this differentiation, the cell divides asymmetrically, creating a larger compartment called the mother cell and a smaller forespore compartment. These two cell types follow different developmental paths, exhibiting different patterns of gene expression (18, 38, 51). These patterns of gene expression are determined by the successive appearance of secondary sigma factors, which direct RNA polymerase to different sets of promoters (18, 41). Additional DNA binding proteins further regulate transcription by RNA polymerase containing these sigma factors.

The product of this differentiation process, the endospore, is encased in a proteinaceous coat which confers resistance to a number of environmental insults (2). The genes encoding most of these spore coat proteins are transcribed in the mother cell compartment, and their gene products are deposited around the forespore (1, 16, 18, 38, 41). Gene expression in the mother cell compartment consists of four temporally expressed gene sets (55). Specific members of each temporal set regulate expression of genes in subsequent gene sets. ^E is the first sigma factor active in the mother cell, and it directs the expression of the first gene set, which includes the gene encoding the DNA binding protein SpoIIID (5, 16, 31, 43). SpoIIID, in conjunction with ^E, activates transcription of the next set of genes which includes sigK, the gene that encodes ^K, the second and final sigma factor active in the mother cell (19, 29, 33, 47). RNA polymerase containing ^K transcribes the third gene set, which includes several spore coat genes and the gene encoding the DNA binding protein GerE (10, 11, 46, 52, 55). The fourth set of genes includes additional spore coat genes, and their expression is dependent on both ^K RNA polymerase and GerE (15, 24, 44, 45, 52-55).

GerE is an 8.5-kDa DNA binding protein which can act both as an activator and as a repressor of transcription (3, 13, 20, 22, 24, 44-46, 53-55). The work in this study focuses on its positive regulatory effects. It is not known how GerE stimulates promoter activity, but one possibility is the direct interaction of GerE with one or more subunits of RNA polymerase. Transcriptional activators in bacteria can be divided into two groups, based on the location of their binding sites in the promoter regions of the genes they are activating (23, 26). Class I activators bind to sites that are upstream from the 35 region of promoters, and several have been shown to contact the C-terminal domain of the alpha subunit of RNA polymerase (8, 17, 25, 26, 40). Class II factors bind to sites that overlap the 35 region. There is evidence that class II factors may interact with the sigma subunit or with the N terminus of the alpha subunit (4, 6, 7, 30, 37, 39, 42, 48). The location of GerE binding sites differs among the promoters that are stimulated by GerE. In some instances the sites overlap the 35 region of the promoter; in others, the sites are further upstream from the transcriptional start site (24, 53, 54). Therefore, it is possible that GerE makes different contacts with RNA polymerase, depending on which promoter it is affecting.

Two GerE binding sites are present in the cotX promoter, centered at 36.5 and 60.5 with respect to the transcription start site (53). Because the region protected by GerE from DNase I digestion extends into the 35 region, we hypothesized that GerE contacts ^K at this promoter. If GerE stimulation of cotX promoter activity requires the interaction of GerE with ^K, then it may be possible to identify amino acid substitutions in ^K that interfere with its interaction with GerE but have little effect on the activity of ^K RNA polymerase on promoters at which interaction between ^K and GerE is not required. Here, we identify two amino acid residues of ^K that are required for the efficient utilization of the cotX promoter.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains and plasmids. EUKW9711 was constructed for the purification of RNA polymerase containing ^K. Strain AH45, a JH642 derivative with a spoIIIG in-frame deletion, was transformed with chromosomal DNA from a bofA::cat mutant to create EUKW9710. EUKW9711 was constructed by transforming EUKW9710 with pPolHis1, creating a His tag on the C terminus of the ' subunit. The remaining B. subtilis strains used in this study are listed in Table 1 or described in the text. Escherichia coli DH5 (Bethesda Research Laboratories) or One Shot cells (Invitrogen) were used for routine molecular cloning work, and BL21(DE3)pLysS (Stratagene) was used for overexpression of the GerE protein.

Construction of promoter fusions. A 237-bp PCR fragment of the cotX promoter region was amplified with COTX-H3 and COTX-BCLI (Table 2) by Pfu polymerase and inserted into pTKlac (28), replacing a HindIII-BamHI fragment and creating a promoter fusion to the lacZ gene. This plasmid was linearized by digestion with PstI and transformed into B. subtilis ZB307A, creating an SP specialized transducing phage carrying the cotX-lacZ promoter fusion. A phage lysate was prepared by heat induction.

Random PCR mutagenesis. Random mutations were introduced into the C-terminal two-thirds of the sigK gene by PCR (35). The template DNA used was pKH2, and the primers used for amplification (SIGK2 and SRSIGK3') were within region 2.2 of ^K or within the vector sequences (Table 2). The sigK gene, flanking chromosomal DNA, and the tetracycline resistance marker were amplified under standard PCR conditions, with the exception of the addition of MnCl₂ to a final concentration of 0.1 to 0.25 mM. The reaction mixture differed from that described in the protocol of Leung et al. (35) by the omission of dimethyl sulfoxide (DMSO) and -mercaptoethanol, and nucleotide ratios were not altered.

Screening of potential mutants. Tet^r transformants were patched onto Difco sporulation medium (DSM) plates containing either X-Gal (5-bromo-4-chloro-3-indolyl--D-galactopyranoside; the substrate for the cotX-lacZ fusion) or X-Gluc (5-bromo-4-chloro-3-indolyl--D-glucuronic acid; the substrate for the gerE-gusA fusion). Colonies that were white on the X-Gal plates and blue on the X-Gluc plates were selected for further analysis. The frequency of cotransformation of the mutant phenotype with Tet^r marker was determined to be 88 to 90% by patching 100 Tet^r transformants each on DSM plates with X-Gal or X-Gluc.

Site-directed PCR mutagenesis. Site-specific changes were made in sigK by using the QuickChange kit from Stratagene (creating pE216K, pE216A, pF224L, pF224A, pH225Y, and pH225A). pKH2 was the DNA template, and primers used for the mutagenesis were complementary and divergent 30-bp oligonucleotides with the mutations located in the center of the sequences (Table 2). The DNA polymerase used in the reactions was the high-fidelity Pfu enzyme from Stratagene. The presence of the mutations was confirmed by sequencing the sigK gene of each plasmid.

Sporulation assay. The B. subtilis strains were grown for 24 h in DSM liquid. Samples from each culture (1 ml) were heated at 80°C for 10 min. The number of CFU in both heated and unheated samples was determined by plating cells diluted by 10², 10⁴, and 10⁶ onto LB medium containing tetracycline (7.5 µg/ml) and chloramphenicol (5 µg/ml).

Purification of RNA polymerases. Strains EUKW9711, EUKW9811, and EUKW9812 were grown for 7.5 h after the onset of sporulation in 1 liter of DSM containing chloramphenicol (5 µg/ml) and spectinomycin (50 µg/ml). The cells were harvested by centrifugation and resuspended in 10 ml of buffer 1 (10 mM Tris-HCl [pH 8], 0.1 M NaCl, 5% glycerol, 1 mM EDTA, 1 mM -mercaptoethanol, 1 mM phenylmethylsulfonyl fluoride [PMSF]) containing 2.5 mM imidazole. Cells were disrupted by two passages through a French pressure cell at 70,000 kPa and pelleted by centrifugation for 20 min at 4°C and 12,000 × g. Proteins from the supernatant were adsorbed at 4°C to 4 ml of a Ni²⁺-nitrilotriacetic acid (NTA) superflow matrix (Qiagen) previously equilibrated with buffer 1 containing 2.5 mM imidazole. After at least 1 h, the resin was packed into a disposable column and washed with 15 ml of buffer 1 containing 20 mM imidazole. The His-tagged RNA polymerase was eluted from the column with buffer 1 containing 300 mM imidazole. The fractions eluted from the column were tested for RNA polymerase activity in a nonspecific activity assay as previously described (49). The amount of active ^K RNA polymerase in each preparation was determined and normalized by its activity on the GerE-independent, ^K-dependent promoter gerE in run-off transcription assays.

Partial purification of GerE. BL21(DE3)pLysS cells containing pGerE-EX were grown to an optical density at 600 nm (OD₆₀₀) of approximately 0.7 in LB medium containing chloramphenicol (34 µg/ml) and kanamycin (30 µg/ml), at which time expression of GerE was induced by the addition of 1 mM isopropyl--D-thiogalactopyranoside (IPTG). The cells were grown for an additional 2.5 h at 37°C. The harvested cells were resuspended in lysis buffer containing 50 mM Tris-HCl (pH 8), 20% glycerol, 1 mM EDTA, 50 mM NaCl, 1 mM dithiothreitol (DTT), and 0.3 mg of PMSF. Cells were lysed at 4°C by a single passage through a French pressure cell at 70,000 kPa and centrifuged for 20 min at 4°C and 31,000 × g. Proteins from the supernatant were loaded onto a 1-ml Hi-Trap heparin column (Pharmacia) and eluted by a gradient from 100 to 500 mM NaCl by using a Pharmacia fast-protein liquid chromatography system. Fractions were electrophoresed on an 18% polyacrylamide gel to determine which fractions contained GerE. Fractions containing the majority of GerE protein were combined and dialyzed against 1 liter of storage buffer (50 mM Tris-HCl [pH 8], 20% glycerol, and 1 mM EDTA) for 4 h at 4°C. Aliquots were stored at 80°C.

In vitro transcription. Plasmids pJZ6 (52) digested with BglII and pKW21 digested with EcoRI were used as templates for in vitro transcription reactions. Both templates, GerE, and RNA polymerase containing either wild-type or mutant ^K were preincubated for 10 min at 37°C in a 40-µl volume with a final concentration of 33 mM Tris-acetate (pH 7.9), 10 mM magnesium acetate, 0.5 mM DTT, and 0.15 mg of bovine serum albumin. Ribonucleotides (500 µM final concentration of ATP, CTP, and GTP [Boehringer Mannheim] and 2 µCi of [-³²P]UTP [400 Ci/mmol; Amersham]) were added for 1 min before reinitiation was prevented by the addition of 10 µg of heparin (Sigma). Five minutes later, unlabelled UTP (500 µM final concentration) (Boehringer Mannheim) was added for an additional 5-min incubation. Reactions were stopped by the addition of sodium acetate (0.3 M final concentration) and ethanol precipitation of nucleic acids. Before being loaded on a 6% polyacrylamide-7 M urea sequencing gel, nucleic acids were resuspended in 10 µl of a urea sequencing dye. Transcripts were quantitated by densitometry with the ImageQuant software coupled to a PhosphorImager 445 SI (Molecular Dynamics). Specific transcription from the cotX and gerE promoters produced nucleotide transcripts of 183 and 107 bp, respectively.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Mutant sigK alleles that reduce cotX expression but not gerE expression. We sought to identify mutations in sigK that specifically decreased cotX promoter activity while having no effect on the activity of the ^K-dependent, GerE-independent promoter, gerE. The gene encoding ^K consists of two gene fragments, spoIVCB and spoIIIC, that are separated by a 42-kb region of DNA called the skin element (50). The skin element sequences are deleted during sporulation from the mother cell chromosome by a site-specific recombination event to create an intact sigK gene (32). Cells engineered to have a skinless sigK were shown to follow normal timing for sporulation (32). For the purpose of easier manipulation of the gene and introduction of mutations into the chromosome, all strains and plasmids used in this study contain the skinless sigK.

View larger version (10K): [in this window] [in a new window]   FIG. 1.   Introduction of sigma mutations into the B. subtilis chromosome. B. subtilis EUKW9612 was transformed to tetracycline resistance with purified mutagenized PCR products (14). The chromosomal copy of sigK was replaced with the mutant copy by recombination within regions of homology. The asterisk denotes a mutation.

View larger version (9K): [in this window] [in a new window]   FIG. 2.   Amino acid substitutions in K that affect cotX activation. The amino acid sequence of the C terminus of K is shown with the position numbers of the first and last amino acids. Shaded boxes indicate residues that are conserved among sigma factors in the 70 family, and the bar above the sequence denotes the region of sigma factors responsible for recognition of the 35 region of promoters. The single-amino-acid substitutions that specifically reduced cotX expression are depicted by an arrow from the original amino acid residue to the substitution.

View larger version (25K): [in this window] [in a new window]   FIG. 3.   Effects of amino acid substitutions in K on transcription from the cotX and gerE promoters. B. subtilis MO1100 (wild type) (circles), VO558 (gerE36) (squares), and sigK mutant strains (triangles) containing promoter-lacZ fusions (cotX in panels A and B and gerE in panels C and D) were grown in DSM liquid medium. The mutant strains are EUKW9616 (K-E216K) (A), EUKW9628 (K-H225Y) (B), EUKW9617 (K-E216K) (C), and EUKW9629 (K-H225Y) (D). Samples were taken at the onset of stationary phase or sporulation (T0) and at 1-h intervals thereafter (T3 to T11) and assayed for -galactosidase accumulation. Two independent transductants for each of the above-mentioned strains were assayed for -galactosidase activity, and the averages are shown.

An alanine substitution at position 216 or 225 in ^K reduces cotX transcription but not gerE transcription. Since cotX promoter activity requires GerE, whereas gerE promoter activity does not, it seems likely that the E216K and H225Y substitutions in ^K interfere with its interaction with GerE. To test whether E216 and H225 in ^K are required for cotX promoter utilization, we introduced alanine substitutions at these positions, creating strains EUKW9619 and EUKW9702 (Fig. 2), respectively, and examined the effect of the mutations on promoter utilization in vivo. Alanine substitutions were chosen in order to replace the original amino acid residues without introducing large side chains with different charges. These strains were transduced with a specialized transducing phage SP that carried either the cotX-lacZ or the gerE-lacZ promoter fusion, and -galactosidase accumulation in the cultures was monitored throughout sporulation (Fig. 4). The E216A mutation reduced the peak of cotX activation to 32% of the wild-type peak activity (Fig. 4A), whereas the H225A mutation reduced peak expression to 55% of wild-type peak activity (Fig. 4B). Expression of gerE-lacZ was higher in both mutant strains than in the wild-type strain (Fig. 4C and D). These data correlate well with the results obtained with the original mutations and support the notion that the amino acid residues at positions 216 and 225 of ^K play a role in the utilization of the GerE-dependent promoter cotX.

View larger version (28K): [in this window] [in a new window]   FIG. 4.   Effect of alanine substitutions in K on transcription from the cotX and gerE promoters. B. subtilis MO1100 (wild type) (circles), VO558 (gerE36) (squares), and sigK mutant strains (triangles) containing promoter-lacZ fusions (cotX in panels A and B and gerE in panels C and D) were grown in DSM liquid medium. The mutant strains are EUKW9620 (K-E216A) (A), EUKW9703 (K-H225A) (B), EUKW9621 (K-E216A) (C), and EUKW9704 (K-H225A) (D). Samples were taken at the onset of stationary phase or sporulation (T0) and at 1-h intervals thereafter (T3 to T11) and assayed for -galactosidase accumulation. Two independent transductants for each of the above-mentioned strains were assayed for -galactosidase activity, and the averages are shown.

RNA polymerases containing ^K-E216A or ^K-H225Y have reduced transcription from the cotX promoter in vitro. In order to eliminate the possibility that the reduction in cotX expression in strains containing the substitutions in ^K was the result of some unknown indirect effect of the ^K mutations on the physiology of the cell, we examined the effects of two of these mutations in vitro with run-off transcription assays (Fig. 5 and 6). We isolated ^K RNA polymerase from EUKW9711 (wild-type ^K), EUKW9812 (E216A ^K), and EUKW9811 (H225Y ^K) cultures that were induced to sporulate. Cells were harvested approximately 7.5 h after the onset of sporulation, the time at which ^K is active. To ensure that we were isolating predominantly ^K RNA polymerase, these strains contain a mutation in spoIIIG, the gene that encodes the late forespore-specific ^G factor. The bofA::cat mutation is also present to bypass the forespore-specific signal that is normally required for ^K activation (9).

View larger version (39K): [in this window] [in a new window]   FIG. 5.   Comparison of cotX activation in vitro by wild-type and mutant K RNA polymerases. (A) An autoradiograph of radiolabelled transcripts that were subjected to electrophoresis on a 6% polyacrylamide gel is shown. DNA templates (1.5 µg of each template) were incubated with K RNA polymerase alone (lanes 1, 5, and 9) or with 0.25 µg (lanes 2, 6, and 10), 0.5 µg (lanes 3, 7, and 11), or 1 µg (lanes 4, 8, and 12) of partially purified GerE. Lanes 1 to 4, 5 to 8, and 9 to 12 contain RNA polymerase purified from EUKW9711 (wild type), EUKW9812 (K-E216A), and EUKW9811 (K-H225Y), respectively. The positions of run-off transcripts of the expected sizes, as judged by the migration of end-labeled 50-bp DNA ladder (Pharmacia), are indicated. (B) Quantification of the data shown in panel A. The cotX signal is standardized by division with the gerE signal from the same lane; the ratio is plotted against the concentration of GerE in the reaction mixture. EUKW9711 (circles), EUKW9812 (squares), and EUKW9811 (triangles).

View larger version (46K): [in this window] [in a new window]   FIG. 6.   In vitro transcription with limiting amounts of K polymerase. (A) An autoradiograph of radiolabelled transcripts that were subjected to electrophoresis on a 6% polyacrylamide gel is shown. One microgram of DNA template (pJZ6 for lanes 1 to 12 or pKW21 for lanes 13 to 15) was incubated with K RNA polymerase alone (lanes 1, 5, and 9) or with 0.5 µg of GerE (lanes 2 to 4, 6 to 8, and 10 to 12). Lanes 1 to 4, 5 to 8, and 9 to 12 contain RNA polymerase purified from EUKW9711 (wild type), EUKW9812 (K-E216A), and EUKW9811 (K-H225Y), respectively. Half of the amounts of RNA polymerase used in Fig. 5 was added in lanes 1, 4, 5, 8, and 9 and 12 to 15. One-eighth of the amounts was added in lanes 3, 7, and 11. A 1/32 fraction of the amounts was added in lanes 2, 6, and 10. The positions of run-off transcripts of the expected sizes, as judged by the migration of end-labeled 50-bp DNA ladder (Pharmacia), are indicated. (B) Quantification of the data shown in panel A. Data are represented as the amount of cotX signal made with RNA polymerase containing mutant K normalized against the gerE signal generated by using the equivalent amount of polymerase, divided by the equivalent expression for wild-type (wt) K. The ratio is plotted against the amount of polymerase in the reaction mixture. Error bars show the standard deviations for three independent experiments. EUKW9711 (wild type), EUKW9812 (E216A), and EUKW9811 (H225Y).

E216 and H225 of ^K are not required for cotD promoter activity. GerE also stimulates the activity of the cotD promoter (24, 54, 55). In this case GerE binds to sites centered at 25.5 and 47.5 (24); therefore, GerE may interact with a different surface of the ^K RNA polymerase than the surface required for stimulation of cotX promoter activity. To determine whether positions 216 and 225 in ^K are required for GerE stimulation of cotD promoter activity, we transduced strains containing the mutant sigK alleles with an SP phage that carried the cotD-lacZ promoter fusion. We assessed the effect of the alanine substitutions on cotD promoter utilization by monitoring -galactosidase accumulation in the cultures during sporulation (Fig. 7). In contrast to their effect on cotX promoter activity (Fig. 4A and B), the E216A and H225A substitutions in ^K had little effect on cotD promoter activity (Fig. 7). Evidently, E216 and H225 in ^K are not necessary for cotD promoter activity. Therefore if, as argued below, H225 of ^K is a site required for interaction with GerE at the cotX promoter, GerE may stimulate cotD promoter activity by a different interaction with ^K RNA polymerase.

View larger version (14K): [in this window] [in a new window]   FIG. 7.   Effect of alanine substitutions in sigK on transcription from the cotD promoter. B. subtilis MO1100 (wild type) (circles), VO558 (gerE36) (squares), and mutant strains (triangles) containing cotD-lacZ were grown in DSM liquid medium. The mutant strains are EUKW9622 (E216A) (A) and EUKW9705 (H225A) (B). Samples were taken at the onset of stationary phase or sporulation (T0) and at 1-h intervals thereafter (T3 to T10) and assayed for -galactosidase accumulation. Two independent transductants for each of the above-mentioned strains were assayed for -galactosidase activity, and the averages are shown.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

We have found that amino acid substitutions at positions 216 and 225 in ^K decrease cotX promoter activation without affecting expression from the gerE promoter. Both the originally isolated substitutions and the alanine substitutions at these positions have similar effects; therefore, it seems unlikely that all of these substitutions change the specificity of promoter recognition such that the mutant sigmas could not interact with cotX promoter DNA but were able to interact productively with gerE promoter DNA. Because the alanine substitutions reduce cotX promoter activity, we infer that the side chains of the lysine substituted at position 216 and the tyrosine substituted at position 225 were not interfering with activation but, rather, that E216 and H225 of ^K are required for maximum cotX promoter activity. In vitro transcription results show that the decrease in cotX promoter activity is not a result of some unknown change in the physiology of the cell. In addition, the substitutions at positions 216 and 225 in ^K do not grossly alter the structure of ^K, because the mutant polymerases were still able to use the gerE promoter in vitro. RNA polymerase containing the E216A-substituted ^K was stimulated efficiently in vitro by GerE to use the cotX promoter, but its activity on the gerE promoter in the absence of GerE was lower than that of wild-type polymerase. Therefore, the E216A substitution may have affected interaction of ^K with cotX promoter DNA. In contrast, GerE stimulated use of the cotX promoter in vitro by RNA polymerase containing the H225Y-substituted ^K less efficiently than by wild-type ^K RNA polymerase. Therefore, H225 in ^K is probably involved in a ^K-GerE interaction at the cotX promoter. We believe that it is unlikely, but we cannot eliminate the possibility that H225 is also involved in an interaction with cotX promoter DNA.

We suggest that interaction of GerE with the region of ^K near position 225 is important for cotX promoter activity, but we do not exclude the possibility that additional interactions between GerE and RNA polymerase may also be required for stimulation of cotX promoter activity. The amino acid substitutions at positions 216 and 225 in ^K cause a decrease in cotX promoter activity but do not abolish cotX transcription completely. This raises the possibility that GerE makes more than one contact with RNA polymerase at this promoter. E. coli catabolite gene activator protein (CAP) is an example of a transcriptional activator that makes several contacts with RNA polymerase. At class II CAP-dependent promoters, CAP binds as a dimer to sequences that overlap the 35 region of the promoter. The upstream subunit of the dimer contacts the C-terminal domain of the  subunit, while the downstream subunit interacts with the N-terminal domain of the  subunit (7, 42, 56). Similarly, E. coli FNR protein also has more than one interaction with RNA polymerase when it binds as a dimer to sites that overlap the 35 region. In this case, the upstream FNR molecule interacts with the C-terminal domain of  and the downstream molecule contacts the ⁷⁰ subunit of RNA polymerase (36).

In the cotX promoter, there are two adjacent binding sites, one of which overlaps the 35 region (53). Our results support a model in which GerE activates transcription from the cotX promoter by interacting with RNA polymerase specifically through the ^K subunit. It is not clear, however, which GerE molecule is involved in this interaction, but from the studies of other activators, it would be expected to be between ^K and the GerE molecule bound to the downstream site which overlaps the 35 region. This contact may not be the only interaction between GerE and the RNA polymerase at cotX, because the mutations do not abolish cotX promoter activity. Therefore, the upstream GerE molecule may interact with the C-terminal domain of the  subunit. This additional interaction could account for the residual activity from the cotX-lacZ promoter fusion in the mutant strains. Moreover, the H225A substitution in ^K did not interfere with the activation of another promoter that is stimulated by GerE, cotD. Therefore, despite its small size, GerE appears to be capable of activating promoters in more than one way, presumably by interacting with different surfaces of ^K RNA polymerase.