Previous work has shown that expression of the Bacillus subtilis ans operon which codes for L-asparaginase and L-aspartase, is both increased and made insensitive to repression by NH4+ by the aspH1 mutation. In current work, the gene in which the aspH1 mutation resides has been identified and sequenced; this gene, termed ansR, is immediately upstream of, but transcribed in the opposite direction from, the ans operon. The promoter region of ansR contains -10 and -35 sequences similar to those recognized by RNA polymerase containing the major vegetative-cell sigma factor sigma A, and ansR appears to be monocistronic. The ansR gene codes for a 116-residue protein, but the aspH1 mutant allele has an additional guanine residue at codon 55, resulting in generation of a truncated polypeptide of only 58 residues. Insertional inactivation of ansR resulted in a phenotype identical to that of the aspH1 mutant. The predicted amino acid sequence of the ansR gene product (AnsR) was homologous to that of the repressor of B. subtilis prophage PBSX, and a helix-turn-helix motif, characteristic of many DNA-binding proteins, was present in the AnsR amino-terminal region. These results suggest that ansR codes for a repressor of the ans operon.

A number of Bacillus subtilis genes involved in NAD biosynthesis have been cloned and sequenced. One of the genes encodes a polypeptide homologous to Escherichia coli L-aspartate oxidase, and its mutation resulted in a nicotinic acid (Nic)-dependent phenotype; this gene was termed nadB. A second open reading frame (orf2) was found downstream of nadB, and an insertional plasmid separating orf2 and nadB also gave a Nic-dependent phenotype. This result suggests that orf2 may also be involved in NAD biosynthesis and that nadB and orf2 are in the same operon. Upstream of nadB was a third gene, transcribed in the opposite direction to that of nadB-orf2. The amino acid sequence derived from the third gene was quite similar to those derived from nifS genes of various nitrogen-fixing bacteria; therefore, the third gene was termed nifS. As with nadB and orf2, mutations in nifS also resulted in a Nic-dependent phenotype. The promoter regions of nadB and nifS overlapped each other and both contained -10 and -35 sequences which resemble those of E sigma A-type promoters. Transcription from both the nifS and nadB promoters, as well as expression of a nadB-lacZ fusion, was repressed by Nic. However, nadB transcription and nadB-lacZ expression were decreased, at most, only slightly by a deletion in nifS. The possible role of the nifS gene product in NAD biosynthesis is discussed.

The Bacillus subtilis divIVB1 mutation causes aberrant positioning of the septum during cell division, resulting in the formation of small, anucleate cells known as minicells. We report the cloning of the wild-type allele of divIVB1 and show that the mutation lies within a stretch of DNA containing two open reading frames whose predicted products are in part homologous to the products of the Escherichia coli minicell genes minC and minD. Just upstream of minC and minD, and in the same orientation, are three genes whose products are homologous to the products of the E. coli shape-determining genes mreB, mreC, and mreD. The B. subtilis mreB, mreC, and mreD genes are the site of a conditional mutation (rodB1) that causes the production of aberrantly shaped cells under restrictive conditions. Northern (RNA) hybridization experiments and disruption experiments based on the use of integrational plasmids indicate that the mre and min genes constitute a five-cistron operon. The possible involvement of min gene products in the switch from medial to polar placement of the septum during sporulation is discussed.

DNA in spores of Bacillus and Clostridium species is associated with small, acid-soluble proteins (SASP) of the alpha/beta type; the presence of these proteins is a major factor in causing spore resistance to UV light, alpha/beta-type SASP did not bind to single-stranded DNA, single- or double-stranded RNA, or DNA-RNA hybrids in vitro. However, these proteins bound a variety of double-stranded DNAs and conferred protection against DNase cleavage. The binding of alpha/beta-type SASP to DNA saturated at a protein/DNA ratio (wt/wt) of 4:1 to 5:1, which is approximately 1 SASP per 4 bp. alpha/beta-type SASP-DNA interaction did not require divalent cations, was independent of pH between 6 and 8, and, for some SASP-DNA pairs, was relatively insensitive to salt up to 0.3 M. The relative affinity of alpha/beta-type SASP for different DNAs was poly(dG).poly(dC) greater than poly(dG-dC).poly(dG-dC) greater than plasmid pUC19 greater than poly(dA-dT).poly(dA-dT), with poly(dA).poly(dT) giving no detectable binding. This order in alpha/beta-type SASP-DNA affinities parallels the facility with which the DNAs adopt an A-like conformation, the conformation in alpha/beta-type SASP-DNA complexes. An oligo(dG).oligo(dC) of 12 bp was bound by alpha/beta-type SASP. While a 26-bp oligo(dG).oligo(dC) bound more tightly than the 12-mer, there was no significant increase in affinity for alpha/beta-type SASP with further increase in size of oligo(dG).oligo(dC). In contrast, binding of alpha/beta-type SASP to oligo(dA-dT).oligo(dA-dT) was minimal up to at least a 70-mer, and binding to poly(dA-dT).poly(dA-dT) was very cooperative. In addition to blocking DNase digestion, binding of alpha/beta-type SASP to DNA blocked (i) cleavage of the DNA backbone by hydroxyl radicals and orthophenanthroline-Cu2+, (ii) DNA cleavage by restriction enzymes, in particular those with specificity for GC-rich sequences; and (iii) in vitro transcription of some but not all genes. However, methylation of dG residues by dimethyl sulfate was not affected by alpha/beta-type SASP binding.

Translational lacZ fusions to forespore genes of Bacillus subtilis were not expressed in spoIIAC (sigma F) or spoIIIE mutants when the lacZ fusions were integrated at the loci of the same genes or at the SP beta locus. However, some of these genes, including gerA, gpr, spoIIIG (sigma G), and sspE, were expressed in spoIIIE mutants and spoIIIE spoIIIG double mutants (but not in spoIIAC mutants) when the lacZ fusions were integrated at the amyE locus. When tested, the beta-galactosidase made in these mutants was found only in the forespore, and the 5' ends of the mRNAs produced in these mutants were identical to those in a Spo+ background. Analysis of the in vitro transcription of forespore genes by RNA polymerase containing sigma F (E sigma F) revealed a direct correlation between good in vitro transcription by E sigma F and expression at the amyE locus in spoIIIE mutants. This result suggests that forespore genes are transcribed by E sigma F in spoIIIE and spoIIIE spoIIIG mutants. Comparison of the promoter regions of genes transcribed well and poorly by E sigma F in vivo and in vitro showed that good transcription by E sigma F was correlated with G residues at positions -15 and -16, a purine residue at position -13, and a T residue at position -7 relative to the start site of transcription. The importance of these residues in sigma F recognition was confirmed by analysis of the E sigma F-dependent transcription in vivo and in vitro of mutant ssp genes.

L-Aspartase was purified from Bacillus subtilis, its N-terminal amino acid sequence was determined to construct a probe for the aspartase gene, and the gene (termed ansB) was cloned and sequenced. A second gene (termed ansA) was found upstream of the ansB gene and coded for L-asparaginase. These two genes were in an operon designated the ans operon, which is 80% cotransformed with the previously mapped aspH1 mutation at 215 degrees. Primer extension analysis of in vivo ans mRNA revealed two transcription start sites, depending on the growth medium. In wild-type cells in log-phase growth in 2x YT medium (tryptone-yeast extract rich medium), the ans transcript began at -67 relative to the translation start site, while cells in log-phase growth or sporulating (t1 to t4) in 2x SG medium (glucose nutrient broth-based moderately rich medium) had an ans transcript which began at -73. The level of the -67 transcript was greatly increased in an aspH mutant grown in 2x YT medium; the -67 transcript also predominated when this mutant was grown in 2x SG medium, although the -73 transcript was also present. In vitro transcription of the ans operon by RNA polymerase from log-phase cells grown in 2x YT medium and log-phase or sporulating cells grown in 2x SG medium yielded only the -67 transcript. Depending on the growth medium, the levels of asparaginase and aspartase were from 2- to 40-fold higher in an aspH mutant than in wild-type cells, and evidence was obtained indicating that the gene defined by the aspH1 mutation codes for a trans-acting transcriptional regulatory factor. In wild-type cells grown in 2x SG medium, the levels of both aspartase and asparaginase decreased significantly by t0 of sporulation but then showed a small increase, which was mirrored by changes in the level of beta-galactosidase from an ansB-lacZ fusion. The increase in the activities of ans operon enzymes between t2 and t5 of sporulation was found primarily in the forespore, and the great majority of the increased was found in the mature spore. However, throughout sporulation the only ans transcript detected was the -73 form, and no sporulation-specific RNA polymerase tested yielded a -73 transcript in vitro.

The Bacillus subtilis spoIIIG gene codes for a sigma factor termed sigma G which directs transcription of genes expressed only in the forespore compartment of the sporulating cell. Use of spoIIIG-lacZ transcriptional fusions showed that spoIIIG is cotranscribed with the spoIIG operon beginning at t0.5-1 of sporulation. However, this large mRNA produced little if any sigma G, and transferring the spoIIIG gene without the spoIIG promoter into the amyE locus resulted in a Spo+ phenotype. Significant translation of spoIIIG began at t2.5-3 with use of an mRNA whose 5' end is just upstream of the spoIIIG coding sequence. Synthesis of this spoIIIG-specific mRNA was not abolished by a deletion in spoIIIG itself. Similar results were obtained when a spoIIIG-lacZ translational fusion lacking the spoIIG promoter was integrated at the amyE locus. These data suggest that synthesis of sigma G is dependent neither on transcription from the spoIIG promoter nor on sigma G itself but can be due to another transcription factor. This transcription factor may be sigma F, the product of the spoIIAC locus, since a spoIIAC mutation blocked spoIIIG expression, and sequences upstream of the 5' end of the spoIIIG-specific mRNA agree well with the recognition sequence for sigma F. RNA polymerase containing sigma F (E sigma F) initiated transcription in vitro on a spoIIIG template at the 5' end found in vivo, as did E sigma G. However, E sigma F showed a greater than 20-fold preference for spoIIIG over a known sigma G-dependent gene compared with the activity of E sigma G.

During sporulation of Bacillus subtilis, two identical genomes segregate in two compartments, the forespore and mother cell. These genomes are expressed differentially, with some genes such as sspE turned on only in the forespore. In vitro transcription of sspE was obtained only with RNA polymerase extracted from sporulating cells. Fractionation of factors associated with this enzyme and reconstitution with core RNA polymerase from vegetative cells generated an enzyme accurately transcribing sspE in vitro and led to purification of a polypeptide with the amino-terminal sequence of the spoIIIG product. Inactivation of spoIIIG abolished expression of sspE and five other forespore-specific genes, whereas synthesis of the spoIIIG product in vegetative cells rapidly turned these genes on. Therefore, spoIIIG encodes a sigma-factor, sigma G, which controls the expression of multiple genes in the forespore compartment.

During sporulation in Bacillus subtilis, expression of the genes sspA, sspB, sspC, sspD, and sspE, which encode a family of small, acid-soluble spore proteins, as well as of the spoVA and gdh operons is transcriptionally activated at stage III of sporulation only in the forespore compartment. Transcription of these genes is mediated by RNA polymerase containing sigma G (E sigma G), the product of the sigG gene, which is itself expressed at stage III in the developing forespore. We have determined the 5' ends of transcripts generated both in vivo and in vitro by the action of E sigma G on various genes of B. subtilis and other bacilli. The 5' ends of the in vivo and in vitro mRNAs were found to coincide and were therefore considered to define the transcription initiation sites for the genes examined. We identified highly homologous DNA sequences centered at 35 and 10 base pairs preceding the transcriptional start sites of the genes examined. Consequently, we propose that these sequences define a class of promoters recognized only by E sigma G which allow transcription of genes expressed uniquely at stage III in the developing forespore.

The nucleotide sequences of the single genes coding for the B-type small, acid-soluble spore proteins (SASP) of Bacillus cereus, B. stearothermophilus, and "Thermoactinomyces thalpophilus" were determined, and the amino acid sequences of all B-type SASP were compared. While this type of SASP showed significant sequence conservation around the two spore protease cleavage sites, alignment of these sequences required the introduction of gaps, and even then only 19 of the residues were conserved exactly in all five proteins. However, all five B-type SASP did contain a large (27 to 35-residue), rather well-conserved amino acid sequence repeat, and four of the five proteins had well-conserved regions of 14 to 17 amino acids which appeared three times.