Heterogeneity of RNA polymerase in Bacillus subtilis: evidence for an additional sigma factor in vegetative cells

Chromosomal location of a Bacillus subtilis DNA fragment uniquely transcribed by sigma-28-containing RNA polymerase

A fragment of the Bacillus subtilis chromosome containing a gene whose transcription is dependent on sigma-28-containing RNA polymerase has been genetically mapped by means of an integrable plasmid. This gene resides on the chromosome adjacent to the stage 0 sporulation locus spo0E between metC and citL. The gene was insertionally inactivated by cloning an internal EcoRI-HindIII fragment in the integrable plasmid pJH101 and by inserting the plasmid into the chromosome by transformation. Transformants bearing an inactivated gene were indifferent to this inactivation for both growth and sporulation.

Translational block to expression of the Escherichia coli Tn9-derived chloramphenicol-resistance gene in Bacillus subtilis

The Gram-negative product-encoding Tn9-derived chloramphenicol-resistance (Cmr) gene can be cloned but not phenotypically expressed in Bacillus subtilis. We show that, even when transcribed from B. subtilis promoters, the ribosomal binding site for the Cmr gene does not function well in B. subtilis. The Cmr gene product, chloramphenicol acetyltransferase (CmAcTase; acetyl-CoA:chloramphenicol 3-O-acetyltransferase, EC 2.3.1.28), is detected in B. subtilis when the promoters, ribosomal binding sites, and initiation codons of B. subtilis genes are fused to the Cmr gene. These gene fusions lead to the in vivo production of mRNAs containing B. subtilis translation start signals followed in an open reading frame by the translation start site normally used by Escherichia coli to initiate translation of Cmr mRNA. Both fusion and native CmAcTase proteins are produced in E. coli, but only fusion CmAcTase is produced in B. subtilis. We conclude that the absence of native CmAcTase in B. subtilis is due to inability of the E. coli ribosomal binding site to function well in B. subtilis. Since fusion CmAcTase polypeptides are produced in E. coli, we conclude that these particular B. subtilis regulatory elements function heterologously in E. coli. The absence of a suitable binding site on the Cmr gene for B. subtilis ribosomes is consistent with reports that many E. coli genes are not expressed in B. subtilis and that E. coli mRNA functions poorly in B. subtilis in vitro translation systems. The functioning of B. subtilis regulatory sequences in E. coli is consistent with in vivo and in vitro data showing the expression of B. subtilis genes in E. coli. To confirm the hypothesis that the large CmAcTase proteins are NH2-terminal fusions of native CmAcTase we partially determined the sequence of one CmAcTase fusion protein.

RNA polymerase of Myxococcus xanthus: purification and selective transcription in vitro with bacteriophage templates

DNA-dependent RNA polymerase from vegetative cells of the gram-negative, fruiting bacterium Myxococcus xanthus was purified more than 300-fold by a modified Burgess procedure (Lowe et al., Biochemistry 18:1344-1352, 1979), using Polymin P precipitation, 40 to 65% saturated ammonium sulfate fractional precipitation, double-stranded DNA cellulose chromatography, A5m gel filtration chromatography, and single-stranded DNA agarose chromatography. The last step separated the RNA polymerase into a core fraction and an enriched holoenzyme fraction. The core enzyme showed a subunit structure similar to that of the Escherichia coli polymerase, as follows: beta' and beta (145,000 and 140,000 daltons, respectively) and alpha (38,000 daltons). A comparison of the core enzyme and the holoenzyme implicated two polypeptides as possible sigma subunits. These polypeptides were closely related, as indicated by peptide analysis. M. xanthus RNA polymerase was capable of transcribing DNAs from E. coli phages T7, T4, and lambda, Bacillus subtilis phage phi 29, and M. xanthus phages Mx1, Mx4, and Mx8. Transcription of T7 and phi 29 DNAs was stimulated by KCl, whereas transcription of Mx1, Mx4, and Mx8 DNAs was inhibited by KCl. Magnesium ion dependence, rifampin and heparin sensitivities, and spermidine stimulation of M. xanthus RNA polymerase activity were similar to those found with E. coli RNA polymerase. The pH optimum of M. xanthus RNA polymerase activity was more basic than that of E. coli polymerase. M. xanthus RNA polymerase was capable of selective transcription in vitro when DNAs from phages T7 delta 111, phi 29, and Mx1 were used. The molecular weights of the resulting phage RNA transcripts made by M. xanthus RNA polymerase (as determined by agarose-acrylamide slab gel electrophoresis) were the same as the molecular weights of the transcripts synthesized by E. coli RNA polymerase. No discrete transcripts were detected as the in vitro RNA products of M. xanthus phage Mx4 and Mx8 DNA transcription. Southern transcript synthesized by M. xanthus RNA polymerase. Three transcripts (transcripts A, B, and C; molecular weights, 2.55 X 10(6), 1.95 X 10(6), and 1.56 X 10(6), respectively) were identified as in vitro RNA products of M. xanthus phage Mx1 DNA transcription when either E. coli or M. xanthus RNA polymerase was used. A Southern blot hybridization analysis indicated that the E. coli RNA polymerase and the M. xanthus RNA polymerase transcribe common SalI restriction fragments of Mx1 DNA.

Species specificity of promoter recognition by RNA polymerase and its transfer by the sigma factor

RNA polymerase holoenzyme from Micrococcus luteus synthesizes in vitro a run-off transcript of 85 nucleotides from a DNA fragment containing part of gene E of bacteriophage phi X174. This RNA starts with GTP as the 5' terminus 18 nucleotides downstream from the start of gene E on the viral (+)strand. Transcription does not occur when the fragment is cleaved 36 nucleotides upstream of the initiation site. No transcript is obtained with RNA polymerase core or holoenzyme from Escherichia coli. Other DNA fragments containing the three major E. coli promoters of phi X174 are transcribed by both enzymes although much less efficiently by M. luteus RNA polymerase. When subunit sigma in E. coli RNA polymerase is replaced by sigma from M. luteus the resulting hybrid enzyme actively transcribes the DNA fragment containing the inner region of gene D with formation of the same run-off transcript which is obtained with M. luteus holoenzyme. In the presence of sigma from E. coli this RNA is not synthesized. The hybrid enzyme also transcribes a DNA fragment containing the gene A promoter of phi X174 with even higher efficiency than RNA polymerase holoenzyme from E. coli.

An interaction between gramicidin and the sigma subunit of RNA polymerase

Gramicidin, a peptide antibiotic produced by Bacillus brevis, inhibits initiation of transcription by RNA polymerase (nucleosidetriphosphate:RNA nucleotidyltransferase, EC 2.7.7.6). We show here that the presence of gramicidin causes an increase in the rate of cleavage of the sigma subunit of Escherichia coli RNA polymerase by trypsin, although it does not alter the cleavage rate of any of the core subunits. Furthermore, whereas isolated sigma is cleaved much faster than is sigma in holoenzyme, gramicidin substantially decreases the trypsin cleavage rate of isolated sigma. Inhibition of RNA polymerase activity by gramicidin in consistent with a sigma-specific effect: the antibiotic is a strong inhibitor of transcription of T7 phage DNA, which requires sigma for activity, but it has little effect on transcription of sigma-independent templates, such as poly(dA-dT).poly)dA-dT) and calf thymus DNA. These results are discussed in light of the hypothesized role for gramicidin in the initiation of sporulation of B. brevis.

Nucleotide sequence of a Bacillus subtilis promoter recognized by Bacillus subtilis RNA polymerase containing sigma 37

We report the nucleotide sequence of the promoter for a Bacillus subtilis gene (spoVC) whose transcription is controlled by a 37,000 dalton species of B. subtilis sigma factor known as sigma 37 but not by the principal- sigma factor of 55,000 daltons (sigma 55). Using S1 nuclease mapping we show that the startpoint for sigma 37-directed transcription of the spoVC gene in vitro corresponded closely to the 5' terminus of in vivo synthesized spoVC RNA. The binding site for sigma 37-containing RNA polymerase extended from 43 bp to 51 bp (positions -43 to -51) upstream from the transcription startpoint to 22 bp (position +22) downstream from the startpoint. The nucleotide sequence of the spoVC promoter differed significantly from promoters whose recognition is controlled by sigma 55 but was similar to other sigma 37- controlled promoters in regions known to be important in promoter recognition. Our results are consistent with the hypothesis (Lee and Pero, J. Mol. Biol., in press) that sigma factors work by contacting specific bases in both the -35 and -10 regions of promoters.

Nucleotide sequences of two Bacillus subtilis promoters used by Bacillus subtilis sigma-28 RNA polymerase

RNA polymerase holoenzymes from many bacterial species share a common promoter recognition specificity since they use the same promoter sites on a variety of templates. These promoters generally include sequences homologous to the average sequences -TTGACA- and -TATAATA-, located -35 and -10 base pairs, respectively, upstream of the transcriptional state site. We have isolated a minor form of B. subtilis RNA polymerase in which the normal sigma subunit (sigma 55) is replaced by a smaller polypeptide (sigma 28) and which is highly specific for a class of promoter sites not used by the sigma 55-holoenzyme. Sequencing of two B. subtilis promoter sites used by the sigma 28-holoenzyme reveals identical sequences at -35 and -10 base pairs from the start site, which are -CTAAA- and -CCGATAT-, respectively. These results confirm that sigma subunit plays a major direct role in promoter sequence recognition, and support a model in which sigma interacts sequentially with -35 and -10 regions, respectively.