Identification of a Second Region of the Spo0A Response Regulator of Bacillus subtilis Required for Transcription Activation

Abstract

Deletion of the 10 C-terminal amino acids of the Bacillus subtilis response regulator Spo0A or valine substitution at D258 and L260 resulted in a sporulation-negative phenotype and loss of in vivo activation of the spoIIG and spoIIA operon promoters. Repression of the abrB promoter was not affected by the mutations. In combination with the previously characterized mutation (A257V), the results identify amino acids at positions 257, 258, and 260 as being required for transcription activation by Spo0A.

The Bacillus subtilis response regulator Spo0A stimulates transcription from a variety of stationary-phase and sporulation-specific promoters (7, 15, 23). Stimulation by Spo0A is mediated by the C-terminal domain, whose activity is blocked until the N terminus has been phosphorylated (1, 11, 15, 23). Spo0A is unusual in that it activates transcription from promoters transcribed by RNA polymerase holoenzyme containing either the ς^A or ς^H sigma factor (reviewed in reference 23).

One region of Spo0A that is required for transcription activation is between amino acids 227 and 240 in the C-terminal domain. Mutations in this region block stimulation of ς^A-dependent promoters, and this region has been proposed as a site of contact with the ς^A subunit (2, 4, 5, 12, 22). Spo0A-ς contact is supported by identification of mutations in both ς^A and ς^H that prevent transcription from Spo0A-dependent promoters but have no effect on transcription from Spo0A-independent promoters (2, 4, 22). Mutations in the ς^A contact region do not affect transcription from ς^H-dependent promoters, suggesting that Spo0A may have a separate contact region for ς^H (4, 12) or that it may activate transcription via different mechanisms at ς^A- and ς^H-dependent promoters.

Deletion of the C-terminal 15 residues of Spo0A generates a mutant blocked at stage 0 of sporulation (8). Substitution of either valine or glutamic acid for the alanine at position 257, which is the 11th amino acid from the C terminus, causes a sporulation-deficient phenotype and abolishes transcription stimulation of the ς^H-dependent spoIIA operon promoter, although similar effects on ς^A-dependent promoters have not been reported (3, 9, 17, 20). The A257V mutation does not prevent in vivo repression of the abrB promoter by Spo0A, so the C terminus appears to be involved in transcription activation (20).

We further investigated the extreme C terminus of Spo0A by creating deletion and point mutants. All mutations were introduced into the spo0A gene by PCR amplification using an upstream primer, mutagenic primers designed to anneal at the end of the coding sequence of the spo0A gene, and plasmid pKK0A (11) as the template. The mutated products were cloned into pGEM-T (Promega) and the sequences were verified (21). Plasmid DNA from each clone was cut at unique SphI and SstI sites to release the fragment containing the spo0A gene fragment, which was then cloned into an integrative vector, pJM103 (18), that had been digested with the same enzymes. To clone the three previously known mutants, the spo0A gene from strains carrying the alleles spo0A9V (A257V), spo0A153 (A257E), and spo0AΔ15 (resulting in deletion of the terminal 15 amino acids) (obtained from J. A. Hoch, Scripps Institute, San Diego, Calif.) was amplified and cloned into pGEM-T. The amino acid sequence from position 251 to the C terminus of each mutant studied is shown in Table 1.

TABLE 1.

Amino acid sequences of C terminus mutants of Spo0A

Protein	Amino acid sequencea	Reference or source
Spo0A (WTb)	EFIAMVADKLRLEHKAS	9
Spo0AΔ10	EFIAMVA	This study
Spo0AΔ15	EF	9
Spo0AA275V (spo0A9V)	EFIAMVVDKLRLEHKAS	9
Spo0AA257E (spo0A153)	EFIAMVEDKLRLEHKAS	9
Spo0AD258V	EFIAMVAVKLRLEHKAS	This study
Spo0AK259V	EFIAMVADVLRLEHKAS	This study
Spo0AL260V	EFIAMVADKVRLEHKANH	This study
Spo0AR261V	EFIAMVADKLVLEHKAS	This study
Spo0AL262V	EFIAMVADKLRVEHKAS	This study
Spo0AE263V	EFIAMVADKLRLVHKAS	This study
Spo0AH264V	EFIAMVADKLRLEVKAS	This study
Spo0AK265V	EFIAMVADKLRLEHVAS	This study
Spo0AA266V	EFIAMVADKLRLEHKVS	This study
Spo0AS267V	EFIAMVADKLRLEHKAV	This study

^aThe position of the substitution is underlined. 

^bWT, wild type. 

Plasmids carrying the mutated spo0A genes were used to transform JH16304, a strain constructed from strain JH642, with a spoIIG::lacZ fusion integrated into the amyE gene by using plasmid pDH32. Transformants resulting from Campbell-type recombination between the plasmid-borne spo0A gene and the chromosomal allele were selected, and 10 representatives from each transformation were examined for their ability to sporulate (6). We reasoned that if A257 was the only critical amino acid within the last 15 amino acids of the sequence, deletion of the 10 amino acids C terminal to A257 would not affect sporulation. As shown in Table 2, this was not the case for the DR2004 mutant, so we extended the analysis by carrying out valine-scanning mutagenesis of the 10C-terminal amino acids. Of the valine substitution mutants, DR2006 and DR2008, which carry the spo0AD258V and spo0AL260V alleles, respectively, had sporulation frequencies of <0.1% (Table 2). The new mutants, along with the three previously identified mutants with Spo⁻ phenotypes, were analyzed for expression of the spoIIG::lacZ promoter fusion (Fig. 1).

TABLE 2.

Effects of Spo0A mutations on spore formation

Strain	Protein	No. of cells/ml	No. of spores/ml	Sporulation frequency (%)a
JH16304	Spo0A	4.0 × 108	2.8 × 108	70
DR2001	Spo0AA275V	4.0 × 108	4.0 × 104	<0.1
DR2002	Spo0AA257E	2.0 × 108	<1	<0.1
DR2004	Spo0AΔ10	4.0 × 107	100	<0.1
DR2005	Spo0AΔ15	2.2 × 107	<1	<0.1
DR2006	Spo0AD258V	4.0 × 108	<1	<0.1
DR2007	Spo0AK259V	4.0 × 108	3.0 × 108	75
DR2008	Spo0AL260V	2.0 × 107	5.0 × 102	<0.1
DR2009	Spo0AR261V	4.0 × 108	4.0 × 108	100
DR2010	Spo0AL262V	5.0 × 108	3.0 × 108	60
DR2011	Spo0AE263V	5.0 × 108	4.0 × 108	80
DR2012	Spo0AH264V	3.7 × 108	3.0 × 108	81
DR2013	Spo0AK265V	5.0 × 108	3.0 × 108	60
DR2014	Spo0AA266V	3.8 × 108	2.7 × 108	71
DR2015	Spo0AS267V	5.0 × 108	1.5 × 108	30

^aCells were grown in Schaeffer sporulation medium (14) and sporulation frequency was determined as described in reference 6. 

FIG. 1.

Effects of spo0A mutations on expression of spoIIG::lacZ transcriptional fusion. The indicated strains were grown in Schaeffer sporulation medium (14). Samples were collected at 1-h intervals from mid-log (T₋₂) into stationary phase (T₁ to T₄) and assayed for β-galactosidase activity (8). T₀ indicates the end of exponential growth.

The strain carrying the wild-type spo0A gene showed stimulation of the spoIIG promoter beginning at 1 h after the end of log phase (T₁) and reaching a maximum at T₃. Deletion of either the 10 or the 15 C-terminal amino acids of Spo0A (spo0AΔ10 and spo0AΔ15) resulted in a reduction of spoIIG promoter activity to 14 and 10% of the wild-type level, respectively. Mutants DR2006 (D258V) and DR2008 (L260V) and the previously known mutants DR2001 (A257V) and DR2002 (A257E) showed less than 10% of wild-type expression of the promoter, a level similar to that in a spo0A null strain (5, 12).

To test whether the mutants that could not activate the spoIIG promoter would activate a ς^H-dependent promoter, we transformed the plasmids containing the Spo0A mutations into JH16302, which carries a spoIIA::lacZ fusion (obtained from M. Perego, Scripps Institute). Cultures of cells were grown to stationary phase and the level of β-galactosidase activity was measured. The results (Fig. 2) showed that, like the A257V mutation (20), the D258V and L260V mutations did not activate the spoIIA promoter.

FIG. 2.

Effects of spo0A mutations on expression of spoIIA::lacZ transcriptional fusion. The strains were grown in Schaeffer sporulation medium (14). Samples were collected at 1-h intervals from mid-log (T₋₂) into stationary phase (T₁ to T₄) and assayed for β-galactosidase activity (8). T₀ indicates the end of exponential growth. Symbols: open circles, wild-type Spo0A; pluses, Spo0AA257V; squares, Spo0AD258V; triangles, Spo0AL260V.

The possibility that the valine substitutions destabilized the Spo0A protein was tested by monitoring the activity of the abrB promoter, which is repressed by Spo0A-P (19, 20, 24, 25). The spoIIG::lacZ fusion in JH642, DM2001, DR2006, and DR2008 was replaced by transforming the strains with DNA from strain JH12604 (obtained from M. Perego, Scripps Institute), which carries an abrB::lacZ fusion integrated into the amyE locus, selecting for spectinomycin resistance, which is associated with the fusion in this strain. Cultures of the transformants were grown and the level of β-galactosidase activity was determined. The results (Fig. 3) showed that the abrB promoter was repressed with the same kinetics in both the wild-type and mutant strains. Thus, the mutations did not affect the stability of the Spo0A protein, and because abrB repression requires phosphorylation, the data implied that phosphorylation of the mutant proteins was normal. We concluded that amino acids A257, D258, and L260 represent a second region that, in addition to the residues between 227 and 240, is required for transcription activation by Spo0A.

FIG. 3.

Effects of spo0A mutations on expression of abrB::lacZ transcriptional fusion. The strains were grown in Schaeffer sporulation medium (14). Samples were collected at 1-h intervals from mid-log (T₋₂) into stationary phase (T₁ to T₄) and assayed for β-galactosidase activity (8). T₀ indicates the end of exponential growth. Symbols: open circles, wild-type Spo0A; triangles, Spo0AA257V; filled squares, Spo0AD258V; open squares, Spo0AL260V.

We modified the classical alanine-scanning mutagenesis technique (26) to probe the extreme C-terminal residues of Spo0A because the target region contained several alanine residues, and one valine substitution mutation, at position 257, had already been isolated (20). Next to alanine, valine is the most suitable amino acid for negating electrostatic effects while minimizing additional steric effects. Four of the 10 C-terminal amino acids of Spo0A have positively charged side chains. Since none of the valine substitutions at these residues affected Spo0A activity, we concluded that the C terminus was not a “positive charge patch” needed for transcription activation.

The A257V, D258V, and L260V mutations affected both ς^A- and ς^H-dependent transcription activation. The isolation of two intragenic suppressors of A257V (20), H162R (suv4) and L174F (suv3), suggests that A257, and, by extension, D258 and L260 could be involved in maintaining the activated structure of Spo0A. A similar role has been assigned to the residues in the extreme C terminus of OmpR, which interacts with central amino acids to create a compact hydrophobic structure (16).

Loss of spoIIA activation in the A257V mutant has been interpreted as an indication that this region is needed for specific interaction with the ς^H subunit of RNA polymerase (20). The hypothesis that Spo0A-P contacts ς^H and ς^A with different subdomains is attractive, since mutations in the ς^A contact region do not affect activation of ς^H-dependent promoters (4, 12, 14, 22). However, while ς^H mutants are known that reduce transcription from Spo0A-dependent but not Spo0A-independent promoters (2, 4, 22), no Spo0A mutants are known that block activation of ς^H-dependent promoters but not ς^A-dependent promoters. Furthermore, the available data suggest that the Spo0A binding sites (0A boxes) that are critical for activation of the spoIIA promoter are located further upstream than are the 0A boxes needed for spoIIG activation, and they also suggest that the orientation of the 0A boxes upstream of the spoIIA promoter is inverted relative to the orientation of the 0A boxes at the spoIIG promoter (23, 27). These factors lead to the possibility that the mechanism of Spo0A activation at ς^H-dependent promoters is different than the mechanism at ς^A-dependent promoters. The mutations identified in this study are consistent with this view, although a more general role for these residues in maintaining the structure of the protein cannot be ruled out.