Transcriptional Activation of the Bacillus subtilis spoIIG Promoter by the Response Regulator Spo0A Is Independent of the C-Terminal Domain of the RNA Polymerase Alpha Subunit

Abstract

In vitro transcription from the spoIIG promoter by Bacillus subtilis RNA polymerase reconstituted with wild-type alpha subunits and with C-terminal deletion mutants of the alpha subunit was equally stimulated by the response regulator Spo0A. Some differences in the structure of open complexes formed by RNA polymerase containing alpha subunit mutants were noted, although the wild-type and mutant polymerases appeared to use the same initiation mechanism.

The central regulator of the onset of Bacillus subtilis sporulation is the response regulator Spo0A, which both activates and represses transcription from operons expressed during late logarithmic growth and early stationary phase (8, 12, 30, 32). Promoters activated by Spo0A have 7-bp DNA sequences (0A boxes) which occur in tandem 5′ to the transcription start site (2, 30, 33). As with other members of the response regulator protein family, Spo0A is a two-domain protein whose activity is controlled by phosphorylation (9, 12, 17, 22, 23, 31). The phosphorylation of the N-terminal, receiver domain prevents inhibition of the transcription activation properties of the C-terminal, output domain (12, 23). The output domain of Spo0A (Spo0ABD) can be separated from the receiver domain via a single trypsin site in the connecting hinge region (11), and it has been shown that purified Spo0ABD stimulates transcription as effectively as does phosphorylated Spo0A (Spo0A∼P [4, 11, 25]).

One of our labs has recently used reconstituted B. subtilis RNA polymerases to examine the interaction of the bacteriophage φ29 transcription activator p4 with the RNA polymerase alpha subunit. These experiments showed that polymerase reconstituted with alpha subunits that have been shortened from the C terminus by 15 amino acids is no longer stimulated by p4 in vitro (19). The spoIIG operon encodes one of the first sporulation-specific sigma proteins, whose activities control the developmental process (8, 32). The spoIIG promoter is transcribed by RNA polymerase containing the ς^A subunit and is dependent on Spo0A∼P, both in vivo and in vitro (2, 4–6). The promoter region contains tandem 0A boxes at positions −56 to −35 and −95 to −70 relative to the start site of transcription (15, 16). While genetic (1, 29) evidence suggests that Spo0A interacts with the sigma subunit, other transcription regulators have been shown to interact with either the sigma or alpha subunits or with both (reviewed in reference 7). We thus sought evidence for the interaction of Spo0A and the polymerase alpha subunits at the spoIIG promoter.

RNA polymerase was reconstituted with the wild-type alpha subunit or with mutants with either the C-terminal 15 or 59 amino acids deleted (19). Comparison of the C-terminal sequences of the Escherichia coli alpha subunit with that of the B. subtilis alpha subunit indicates that the 59-amino-acid deletion is equivalent to a 73-amino-acid deletion of the E. coli alpha subunit. An assay with bacteriophage φ29 DNA (19, 20) indicated that the specific activities of the three reconstituted enzymes were similar. Equivalent amounts of the reconstituted polymerases were added separately to single-round in vitro transcription reactions containing Spo0ABD and a template carrying the spoIIG promoter (5). The products were separated by electrophoresis and quantitated by analysis with a PhosphorImager model SI and Imagequant software (version 1.0) (Fig. 1). The results demonstrated that the polymerases containing deletion mutants of the alpha subunit were stimulated by Spo0ABD with the same efficiency as was the polymerase reconstituted with wild-type alpha subunit. Thus, these results indicated that the C terminus of the alpha subunit was not required for RNA polymerase response to Spo0A∼P. We noted, however, that the maximum level of transcription by RNA polymerase reconstituted with the mutant alpha subunits was only 30% of that of the wild-type polymerase (data not shown), suggesting that the alpha subunit mutations might affect the interaction of RNA polymerase with the promoter DNA.

FIG. 1.

Spo0A∼P stimulation of transcription by RNA polymerase containing wild-type or mutant alpha subunits. Reconstituted RNA polymerase (prepared as described previously [19]) was incubated with a DNA fragment containing the spoIIG promoter, ATP and [α-³²P]GTP, and increasing inputs of Spo0A∼P that had been phosphorylated in vitro with purified phosphorelay proteins (5). After 2 min, a mixture of heparin, UTP, and CTP was added to permit a single round of transcript elongation as described earlier (5). Reaction products were separated by electrophoresis, and full-length transcripts were quantitated with a Molecular Dynamics PhosphorImager and Imagequant (version 1.0) software. Transcription is reported as fold stimulation over that with no Spo0A∼P added for each polymerase preparation: triangles, wild-type alpha subunit; open circles, alpha subunit truncated by 15 amino acids; closed circles, alpha subunit truncated by 59 amino acids.

We have recently shown that during transcription initiation, Spo0ABD and the polymerase separate the DNA strands at the spoIIG promoter in a stepwise manner. The first step melts the −10 region, extending to −3. Addition of nucleotides leads to denaturation of the +1 site and sequences downstream (25). Mutations of other regulators have suggested that there might be alternate pathways of transcription stimulation (18), so we tested whether the melting step was the same for the RNA polymerase reconstituted with the three different alpha subunits. We investigated the DNA denaturation, using KMnO₄, which preferentially reacts with thymines on single-strand DNA, allowing them to be selectively cleaved (28).

A PvuII/BamHI DNA fragment containing the spoIIG promoter was end labeled at the BamHI site on the template strand. The labeled DNA was incubated with the reconstituted polymerases and Spo0ABD with various combinations of nucleotides. The sequence of the spoIIG promoter allows for synthesis of an ApA dimer in the presence of ATP, a trimer when the dinucleotide ApA and GTP are added, and an 11-mer transcript when ATP and GTP are added. KMnO₄ was added for 3 min and, after cleavage at modified thymines, the fragments were resolved by electrophoresis (Fig. 2). Complexes formed by all three forms of the polymerase plus Spo0ABD without initiating nucleotides contained a denatured region which included the thymines at −13 to −3. Addition of ApA and GTP led to further denaturation of the +1 and +2 sites, and inclusion of ATP and GTP induced denaturation between −13 and +13. The overall patterns of denaturation induced by all three enzymes were similar to each other, in agreement with the transcription results, and to the pattern we reported earlier for wild-type RNA polymerase (25). Thus, the three forms of polymerase use the same initiation mechanism.

FIG. 2.

Sensitivity of the spoIIG promoter region to KMnO₄ induced by Spo0ABD plus RNA polymerase containing wild-type or mutant alpha subunits. A PvuII/BamHI DNA fragment of plasmid pUCIIGtrpA, labeled with T4 polynucleotide kinase at the BamHI site (135 bp 3′ to the spoIIG promoter [6]), was incubated with Spo0ABD (the C-terminal fragment of Spo0A [400 nM] [11]), RNA polymerase reconstituted with wild-type or mutant alpha subunits (40 nM), and combinations of nucleotides. After 2 min, the KMnO₄ was added, and the samples were processed to detect modified thymine residues (28). After cleavage with piperidine, the DNA fragments were separated by electrophoresis on denaturing polyacrylamide gels and detected by exposure to X-ray film. The form of the alpha subunit (WT, wild type; −15, C-terminal 15 amino acids deleted; −59, C-terminal 59 amino acids deleted) used to reconstitute the polymerase (RNAP) is indicated above the panels, and positions of the bases in the DNA relative to the start site of transcription (+1) are indicated on the left.

Since some differences in the levels of thymine sensitivity could be seen, we compared the sensitivities of thymines, relative to the +1 site, under ATP and GTP initiation conditions. Complexes containing wild-type RNA polymerase displayed approximately equal degrees of sensitivity for all thymines between positions −13 and −3. Complexes formed with RNA polymerase lacking the C-terminal 15 amino acids of the alpha subunit displayed a pattern of thymine sensitivity similar to that of wild-type RNA polymerase with the exception of the −3 and −4 positions, which were less reactive (1.4- and 2-fold, respectively). Complexes formed with RNA polymerase lacking the C-terminal 59 amino acids of the alpha subunit showed reduced thymine reactivity at positions −13 and −4 (1.8- and 4.5-fold, respectively) and enhanced reactivity at positions −11 and −7 (1.4- and 2.8-fold, respectively). Lack of denaturation of the −3 and −4 positions has previously been demonstrated to have a significant deleterious effect on transcription initiation (25), suggesting that the C terminus of the alpha subunit may contribute to open complex formation.

As a third test of the reconstituted RNA polymerases, we examined complexes containing the mutant polymerases Spo0ABD and the spoIIG promoter by electrophoretic mobility shift assays (Fig. 3). For each of the three enzymes, the presence of Spo0ABD stimulated the amount of complexes formed, and the inclusion of the initiation nucleotides ATP and GTP further increased the level of complexes. The levels of complexes in the presence of Spo0ABD and ATP and GTP were roughly proportional to the final levels of transcription which were detected in the experiment shown in Fig. 1.

FIG. 3.

Electrophoretic mobility shift assay of complexes formed by RNA polymerase containing mutant and wild-type alpha subunits. The PuvII/BamHI DNA fragment described in the legend to Fig. 2 was incubated with Spo0ABD (400 nM), RNA polymerase (RNAP) reconstituted with wild-type (WTα) or mutant (−15α or −59α) alpha subunits (40 nM), and ATP plus GTP in transcription initiation buffer (6). After 3 min at 37°C, the samples were electrophoresed on a 5% nondenaturing polyacrylamide gel in 1× Tris-acetate-EDTA buffer (27). The gel was dried and exposed to X-ray film. FD, free DNA.

The complex formed by RNA polymerase reconstituted with the wild-type alpha subunit plus Spo0ABD and initiating nucleotides (ATP and GTP) had a mobility that was detectably lower than those formed by the polymerase containing the deletion mutant alpha subunits. As demonstrated for the E. coli alpha subunit (24, 26), the alpha subunit of B. subtilis probably interacts with AT-rich UP elements of promoters (10). There is a 5-bp AT-rich region directly 5′ to the site 2 0A boxes (positions −57 to −61 of the spoIIG promoter). Interaction of the C terminus of the alpha subunit with this region might stabilize a bend in the DNA template. Since DNA molecules with static bends migrate more slowly than linear molecules, this could explain the lower mobility of the complex formed with wild-type polymerase compared to those of complexes formed with polymerase containing the mutant alpha subunits. The wild-type alpha subunit interaction could stabilize the initiated complex, raising the levels of complexes and thus supporting an intrinsic role for the alpha subunit in transcription initiation.

The data in Fig. 3 also indicate that compared to wild type, polymerase reconstituted with mutant alpha subunits appeared to form a higher level of complexes with the spoIIG promoters on their own. This difference supports the notion that the alpha subunit affects the interaction of the polymerase with the spoIIG promoter. While we do not have direct evidence, it is possible that the initial complexes formed by the enzymes with deletion mutant alpha subunits differ because the mutant alpha subunits do not interact with the AT sequence. While these binary complexes appear to be more stable and can still respond to Spo0ABD, they must have a lower rate of initiation.

The phage φ29 transcription regulator p4 interacts with the alpha subunit to activate transcription at the φ29 A3 promoter, since deletion of the C-terminal 15 amino acids of the alpha subunit blocks p4 activation (19). p4 binds to DNA as a pair of dimers at consensus sequences separated by 15 bp. This arrangement allows p4 to both bend the DNA and directly interact with the alpha subunit (3, 19, 20). The major effect of p4 on transcription from the A3 promoter is stabilization of polymerase binding to the promoter (21). This contrasts with Spo0A∼P stimulation of spoIIG, since Spo0A∼P does not affect polymerase binding but binds to the polymerase-DNA complex, stimulating an isomerization, which increases the rate of initiation by up to 50-fold. Since Spo0ABD stimulation of in vitro transcription by the mutant polymerases was not reduced by the alpha subunit deletions, we conclude that Spo0A∼P stimulation of the spoIIG promoter does not require interaction with the alpha subunit of the polymerase. This conclusion is supported by genetic results indicating that the primary interaction of Spo0A at the spoIIG promoter is with the sigma subunit (1, 29). However, the overall level of transcription and the apparent differences in the electrophoretic mobilities of initiated complexes indicate that the C terminus of the alpha subunit is required for optimal interaction with the spoIIG promoter.