The alternative transcription factor sigma B of Bacillus subtilis is activated during the stationary growth phase by a regulatory network responsive to stationary-phase signals. On the basis of the results reported here, we propose that sigma B controls a general stress regulon that is induced when cells encounter a variety of growth-limiting conditions. Expression of genes controlled by sigma B, including the ctc gene and the sigB operon that codes for sigma B and its associated regulatory proteins, was dramatically induced in both the exponential and stationary phases by environmental challenges known to elicit a general stress response. After cells were subjected to salt stress, the increased expression of lacZ transcriptional fusions to the ctc and sigB genes was entirely dependent on sigma B, and primer extension experiments confirmed that the sigma B-dependent transcriptional start site was used during salt induction of sigB operon expression. Western blotting (immunoblotting) experiments measuring the levels of sigma B protein indicated that ethanol addition and heat stress also induced sigma B activity during logarithmic growth. Salt and ethanol induction during logarithmic growth required RsbV, the positive regulator of sigma B activity that is normally necessary for activity in stationary-phase cells. However, heat induction of sigma B activity was largely independent of RsbV, indicating that there are two distinct pathways by which these environmental signals are conveyed to the transcriptional apparatus.

Alternative transcription factor sigma B of Bacillus subtilis controls a stationary-phase regulon induced under growth conditions that do not favor sporulation. Little is known about the metabolic signals and protein factors regulating the activity of sigma B. The operon containing the sigma B structural gene has the gene order orfV-orfW-sigB-rsbX, and operon expression is autoregulated positively by sigma B and negatively by the rsbX product (rsbX = regulator of sigma B). To establish the roles of the orfV and orfW products, orfV and orfW null and missense mutations were constructed and tested for their effects on expression of the sigma B-dependent genes ctc and csbA. These mutations were tested in two contexts: in the first, the sigB operon was under control of its wild-type, sigma B-dependent promoter, and in the second, the sigB operon promoter was replaced by the inducible Pspac promoter. The principal findings are that (i) the orfV (now called rsbV) product is a positive regulator of sigma B-dependent gene expression; (ii) the orfW (now called rsbW) product is a negative regultor of such expression; (iii) sigma B is inactive during logarithmic growth unless the rsbW product is absent; (iv) the rsbX, rsbV, and rsbW products have a hierarchical order of action; and (v) both the rsbV and rsbW products appear to regulate sigma B activity posttranslationally. There are likely to be at least two routes by which information can enter the system to regulate sigma B: via the rsbX product, and via the rsbV and rsbW products.

Transcription factor sigma B of Bacillus subtilis is active during the stationary growth phase, but its physiological role remains unknown. Understanding the function and regulation of genes controlled by sigma B (csb genes) should provide important clues to sigma B function in stationary-phase cells. To this end, we used a genetic approach to identify six new csb genes. This strategy relies on two elements: (i) random transcriptional fusions between the Escherichia coli lacZ gene and genes on the B. subtilis chromosome, generated in vivo with transposon Tn917lacZ, and (ii) a plate transformation technique to introduce a null sigB mutation into the fusion-bearing recipients directly on indicator plates. This strategy allowed the comparison of fusion expression in strains that were isogenic save for the presence or absence of a functional sigma B protein. Beginning with 1,400 active fusions, we identified 11 that were wholly or partly controlled by sigma B. These fusions mapped to six different loci that exhibit substantial contrasts in their patterns of expression in the logarithmic and stationary growth phases, suggesting that they participate in diverse cellular functions. However, for all six loci, the sigma B-dependent component of their expression was manifest largely in the stationary phase. The high frequency of six independent csb loci detected in a random collection of 1,400 fusions screened, the fact that four of the six new loci were defined by a single fusion, and the absence of the previously identified ctc and csbA genes in the present collection strongly suggest that sigma B controls a large stationary-phase regulon.

Transcription factor rB of Bacillus subtilis controls a large stationary-phase regulon, but in no case has the physiological function of any gene in this regulon been identified. Here we show that transcription of gtaB is partly dependent on &B in vivo and thatgtaB encodes UDP-glucose pyrophosphorylase. ThegtaB reading frame was initially identified by a &B-dependent Tn9l7lacZ fusion, csb42. We cloned the region surrounding the csb42 insertion, identified the reading frame containing the transposon, and found that this frame encoded a predicted 292-residue product that shared 45% identical residues with the UDP-glucose pyrophosphorylase of Acetobacter xylinum. The identified reading frame appeared to lie in a monocistronic transcriptional unit. Primer extension and promoter activity experiments identified tandem promoters, one &rB dependent and the other oM' independent, immediately upstream from the proposed coding region. A sequence resembling a factor-independent terminator closely followed the coding region. By polymerase chain reaction amplification of a B. subtilis genomic library carried in yeast artificial chromosomes, we located the UDP-glucose pyrophosphorylase coding region near gtaB, mutations in which confer phage resistance due to decreased glycosylation of cell wall teichoic acids. Restriction mapping showed that the coding region overlapped the known location ofgtaB. Sequence analysis of a strain carrying the gtaB290 allele found an alteration that would change the proposed initiation codon from AUG to AUA, and an insertion-deletion mutation in this frame conferred phage resistance indistinguishable from that elicited by the gtaB290 mutation. We conclude thatgtaB encodes UDP-glucose pyrophosphorylase and is partly controlled by &rB. Because fled by one such fusion, csb42, is transcribed from a edependent promoter in vivo and encodes UDP-glucose pyrophosphorylase. We further demonstrate that the csb42 fusion is an allele of gtaB, mutations in which confer phage resistance as a result of their decreased glycosylation of cell wall teichoic acids (48). From the known role of UDPglucose pyrophosphorylase in stationary-phase survival of enteric bacteria (15, 19), we speculate that some of the genes controlled by might respond to stress in the stationary growth phase. MATERIALS AND METHODSBacteria, phage, and genetic methods. Escherichia coli DH5a (Bethesda Research Laboratories) was the host for all plasmid constructions. B. subtilis strains used are shown in Table 1. B. subtilis PB2 and its derivatives were recipients for natural transformations (13) with linear and plasmid DNA, both for strain constructions and for genetic crosses to map the chromosomal locus of the csb42 fusion. We made the gtaBA1::ery insertion-deletion mutation in strain PB319 by removing the 622-nucleotide (nt) SnaBI-BglII fragment from within the coding region identified by csb42 and replacing it with a 1.4-kb fragment carrying the macrolide-lincosamide-streptogramin B (ery) resistance gene from pE194 (21). Transformation sele...

Chimeric molecules of the cAMP-dependent protein kinase (PKA) holoenzyme (R 2 C 2 ) and of a ⌬ 1-91 RC dimer were reconstituted using deuterated regulatory (R) and protiated catalytic (C) subunits. Small angle scattering with contrast variation has revealed the shapes and dispositions of R and C in the reconstituted complexes, leading to low resolution models for both forms. The crystal structures of C and a truncation mutant of R fit well within the molecular boundaries of the RC dimer model. The area of interaction between R and C is small, seemingly poised for dissociation upon a conformational transition within R induced by cAMP binding. Within the RC dimer, C has a "closed" conformation similar to that seen for C with a bound pseudosubstrate peptide. The model for the PKA holoenzyme has an extended dumbbell shape. The interconnecting bar is formed from the dimerization domains of the R subunits, arranged in an antiparallel configuration, while each lobe contains the cAMP-binding domains of one R interacting with one C. Our studies suggest that the PKA structure may be flexible via a hinge movement of each dumbbell lobe with respect to the dimerization domain. Sequence comparisons suggest that this hinge might be a property of the R II PKA isoforms.Protein phosphorylation is one of the most important mechanisms for the regulation of biochemical function in eukaryotic cells. It is catalyzed by a family of enzymes, the protein kinases, of which several hundred have been identified. The cAMP-dependent protein kinase (PKA) 1 was one of the earliest kinases to be discovered, and it serves as a prototype for understanding kinase structure-function relationships and regulatory mechanisms (1, 2). In the absence of cAMP, PKA is an inactive tetramer (R 2 C 2 ) with two identical regulatory (R) and two identical catalytic (C) subunits. The two R subunits are homodimerized at their amino-terminal ends (3, 4), and, physiologically, the R subunit appears always to exist as a dimer. The R subunit also has two in tandem cAMP-binding sites and a pseudosubstrate autoinhibitory domain that binds to C and inhibits catalysis in the absence of cAMP. Upon binding cAMP, the PKA holoenzyme dissociates into an R 2 homodimer and two active C subunits (5-7). Whether dissociation is absolutely required for activation, however, remains in question (8, 9). Saturation of both cAMP-binding sites on each R is required for activation.There are three isoforms of C (C␣, C␤, and C␥) and two major isoforms of R (R I and R II ) that are further distinguished into subforms (␣ and ␤) (10). The physiological importance of these isozyme variations is not fully understood, but anchoring proteins (AKAPs) for R II give it a unique cellular distribution (11,12). R I and R II show sequence homology in their cAMP-binding and pseudosubstrate domains but differ extensively in their dimerization domains as well as in the sequence connecting the dimerization and pseudosubstrate domains.Structural data have been obtained for the individual PKA subunits, but info...