Expression, Abundance, and RNA Polymerase Binding Properties of the δ Factor of Bacillus subtilis
1999; Elsevier BV; Volume: 274; Issue: 22 Linguagem: Inglês
10.1074/jbc.274.22.15953
ISSN1083-351X
AutoresFrancisco J. López de Saro, Noriko Yoshikawa, John D. Helmann,
Tópico(s)Legume Nitrogen Fixing Symbiosis
ResumoThe δ protein is a dispensable subunit ofBacillus subtilis RNA polymerase (RNAP) that has major effects on the biochemical properties of the purified enzyme. In the presence of δ, RNAP displays an increased specificity of transcription, a decreased affinity for nucleic acids, and an increased efficiency of RNA synthesis because of enhanced recycling. Despite these profound effects, a strain containing a deletion of the δ gene (rpoE) is viable and shows no major alterations in gene expression. Quantitative immunoblotting experiments demonstrate that δ is present in molar excess relative to RNAP in both vegetative cells and spores. Expression of rpoE initiates from a single, ςA-dependent promoter and is maximal in transition phase. A rpoE mutant strain has an altered morphology and is delayed in the exit from stationary phase. For biochemical analyses we have created derivatives of δ and ςA that can be radiolabeled with protein kinase A. Using electrophoretic mobility shift assays, we demonstrate that δ binds core RNAP with an apparent affinity of 2.5 × 106m−1, but we are unable to demonstrate the formation of a ternary complex containing core enzyme, δ, and ςA. The δ protein is a dispensable subunit ofBacillus subtilis RNA polymerase (RNAP) that has major effects on the biochemical properties of the purified enzyme. In the presence of δ, RNAP displays an increased specificity of transcription, a decreased affinity for nucleic acids, and an increased efficiency of RNA synthesis because of enhanced recycling. Despite these profound effects, a strain containing a deletion of the δ gene (rpoE) is viable and shows no major alterations in gene expression. Quantitative immunoblotting experiments demonstrate that δ is present in molar excess relative to RNAP in both vegetative cells and spores. Expression of rpoE initiates from a single, ςA-dependent promoter and is maximal in transition phase. A rpoE mutant strain has an altered morphology and is delayed in the exit from stationary phase. For biochemical analyses we have created derivatives of δ and ςA that can be radiolabeled with protein kinase A. Using electrophoretic mobility shift assays, we demonstrate that δ binds core RNAP with an apparent affinity of 2.5 × 106m−1, but we are unable to demonstrate the formation of a ternary complex containing core enzyme, δ, and ςA. RNA polymerase (RNAP) 1The abbreviations used are: RNAP, RNA polymerase; MOPS, 4-morpholinepropanesulfonic acid; PCR, polymerase chain reaction; PKA, protein kinase Afrom bacteria consists of a multisubunit enzyme complex that is the main target for the regulation of gene expression. The catalytic core of RNAP is composed of the subunits β, β′, and α and is sufficient to catalyze the polymerization of NTPs into RNA chains. In addition, a vast array of other proteins makes contacts with RNAP at one or more steps of the transcription cycle and modifies its activities in various ways (1Uptain S.M. Kane C.M. Chamberlin M.J. Annu. Rev. Biochem. 1997; 66: 117-172Crossref PubMed Scopus (400) Google Scholar). Such ancillary polypeptides modify RNAP during the promoter binding, initiation, elongation, and termination steps to effect changes in gene expression. The affinity of the various transcription factors for RNAP varies greatly during the transcription cycle; for example, ς factors have a high affinity for RNAP during the pre-initiation steps but very low affinity during elongation, whereas for NusA the opposite occurs (2Gill S.C. Weitzel S.E. von Hippel P.H. J. Mol. Biol. 1991; 220: 307-324Crossref PubMed Scopus (108) Google Scholar). In the Gram-positive bacterium Bacillus subtilis, purified preparations of RNAP are usually found associated with the 21.4-kDa δ factor (3Pero J. Nelson J. Fox T.D. Proc. Natl. Acad. Sci. U. S. A. 1975; 72: 1589-1593Crossref PubMed Scopus (69) Google Scholar, 4Lampe M. Binnie C. Schmidt R. Losick R. Gene (Amst.). 1988; 67: 13-20Crossref PubMed Scopus (25) Google Scholar). δ was found to be essential, together with a phage-encoded ς factor, to reconstitute selective transcription of the phage SP01 genome in vitro (3Pero J. Nelson J. Fox T.D. Proc. Natl. Acad. Sci. U. S. A. 1975; 72: 1589-1593Crossref PubMed Scopus (69) Google Scholar). Further biochemical characterization of δ activity has revealed complex effects on RNAP activity. For example, δ reduces the extent of contact between RNAP and promoter DNA, as detected by DNA footprinting assays, and antagonizes the formation of the open complex (5Achberger E.C. Hilton M.D. Whiteley H.R. Nucleic Acids Res. 1982; 10: 2893-2910Crossref PubMed Scopus (24) Google Scholar, 6Chen Y.-F. Helmann J.D. J. Mol. Biol. 1997; 267: 47-59Crossref PubMed Scopus (48) Google Scholar, 7Juang Y.L. Helmann J.D. J. Mol. Biol. 1994; 235: 1470-1488Crossref PubMed Scopus (143) Google Scholar, 8Juang Y.-L. Helmann J.D. Biochemistry. 1995; 34: 8465-8473Crossref PubMed Scopus (40) Google Scholar). These results suggest that RNAP binds to the promoter as a Eςδ complex. However, in solution ς and δ bind to RNAP with negative cooperativity (9Hyde E.I. Hilton M.D. Whiteley H.R. J. Biol. Chem. 1986; 261: 16565-16570Abstract Full Text PDF PubMed Google Scholar). In addition, δ can either inhibit or stimulate in vitrotranscription, depending on reaction conditions, and its stimulatory effects are clearly DNA template-dependent (7Juang Y.L. Helmann J.D. J. Mol. Biol. 1994; 235: 1470-1488Crossref PubMed Scopus (143) Google Scholar, 10Spiegelman G.B. Hiatt W.R. Whiteley H.R. J. Biol. Chem. 1978; 253: 1756-1765Abstract Full Text PDF PubMed Google Scholar, 11Dickel C.D. Burtis K.C. Doi R.H. Biochem. Biophys. Res. Commun. 1980; 95: 1789-1795Crossref PubMed Scopus (15) Google Scholar, 12Dobinson K.F. Spiegelman G.B. Biochemistry. 1987; 26: 8206-8213Crossref PubMed Scopus (39) Google Scholar, 13Juang Y.L. Helmann J.D. J. Mol. Biol. 1994; 239: 1-14Crossref PubMed Scopus (48) Google Scholar). δ has a very unusual tertiary structure with an amino-terminal half that is folded as a α/β type of structure and a carboxyl-terminal half that is unstructured in solution, as observed by circular dichroism analysis (14López de Saro F. Woody A.-Y.M. Helmann J.D. J. Mol. Biol. 1995; 252: 189-202Crossref PubMed Scopus (48) Google Scholar). The carboxyl-terminal region is highly acidic, consisting mostly of aspartate (37%), glutamate (34%), and hydrophobic residues (25%) and is essential for displacing RNA bound to RNA polymerase (14López de Saro F. Woody A.-Y.M. Helmann J.D. J. Mol. Biol. 1995; 252: 189-202Crossref PubMed Scopus (48) Google Scholar). This RNA displacement activity may explain our previous observation that δ enhances RNAP recycling in multiple cycle transcription reactions, thus stimulating overall transcription (13Juang Y.L. Helmann J.D. J. Mol. Biol. 1994; 239: 1-14Crossref PubMed Scopus (48) Google Scholar). These structural and functional studies suggest that δ acts by mimicking single-stranded RNA and points to a role of δ in the termination or recycling steps of transcription (14López de Saro F. Woody A.-Y.M. Helmann J.D. J. Mol. Biol. 1995; 252: 189-202Crossref PubMed Scopus (48) Google Scholar). Despite the clear effects of δ on in vitro transcription, initial attempts to demonstrate a role for δ in vivo were unsuccessful. δ mutant strains sporulate, lack auxotrophies, grow up to 49 °C, and plate phage SP01 normally (4Lampe M. Binnie C. Schmidt R. Losick R. Gene (Amst.). 1988; 67: 13-20Crossref PubMed Scopus (25) Google Scholar). In this study, we describe the expression of the rpoE locus, the phenotype of an rpoE deletion mutant, and preliminary biochemical analysis of the interactions between δ and RNAP. Restriction enzymes and T4 polynucleotide kinase were purchased from New England Biolabs, Inc.; avian myeloblastosis virus reverse transcriptase was from Promega Corp.; growth media were from Difco; and [α-32P]dATP and [35S]methionine were from NEN Life Science Products. Strains were grown with vigorous shaking at 37 °C in Luria broth (LB) (15Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1990Google Scholar) or in minimal medium (16Chen L. James L.P. Helmann J.D. J. Bacteriol. 1993; 175: 5428-5437Crossref PubMed Google Scholar), which is composed of 40 mm MOPS buffer, pH 7.4, 2% glucose (w/v), 2 mm potassium sulfate, pH 7.0, 10 μg/ml tryptophan, and 1× Bacillus salts (0.2 g of magnesium sulfate, 2 g of ammonium sulfate, 1 g of sodium citrate, and 1 g of potassium glutamate/liter). Antibiotic concentrations used were, for Escherichia coli, ampicillin at 100 μg/ml; forB. subtilis, erythromycin at 2 μg/ml; neomycin, 10 μg/ml; macrolides/lincomycin/streptogramin B with 1 μg/ml erythromycin and 25 μg/ml lincomycin. B. subtilis strains used were CU1065 (W168 trpC2 attSPβ), HB6002 (CU1065rpoE::lacZ, generated using the pTKlac vector), HB6005 (CU1065 PrpoE::cat-lacZ at the SPβ locus, generated using the pJPM122 vector (17Slack F.J. Mueller J.P. Sonenshein A.L. J. Bacteriol. 1993; 175: 4605-4614Crossref PubMed Google Scholar)), HB6010 (CU1065 ΔrpoE::cm, generated by transformation of CU1065 with RP17 DNA (14López de Saro F. Woody A.-Y.M. Helmann J.D. J. Mol. Biol. 1995; 252: 189-202Crossref PubMed Scopus (48) Google Scholar)), HB6012 (HB6002abrB::neo), and HB6013 (HB6002sinR::kan). The rpoE mutant used in this study is a deletion/insertion containing a chloramphenicol resistance cassette replacing the segment of the rpoE gene between codons 6 and 142 (4Lampe M. Binnie C. Schmidt R. Losick R. Gene (Amst.). 1988; 67: 13-20Crossref PubMed Scopus (25) Google Scholar). E. coli strains used were DH5α (supE44 ΔlacU169 φ80lacZΔM15 hsdR17 recA1endA1 gyrA96 thi-1 relA1) and, for protein overexpression, BL21/DE3 (BL21 with λDE3 (18Studier F.W. Rosenberg A.H. Dunn J.J. Dubendorff J.W. Methods Enzymol. 1990; 185: 60-89Crossref PubMed Scopus (6005) Google Scholar)). Plasmids used for the generation of lacZ fusions to the rpoE promoter region were pTKlac (19Kenney T.J. Moran C.P. J. Bacteriol. 1991; 173: 3282-3290Crossref PubMed Google Scholar) and pJPM122 (17Slack F.J. Mueller J.P. Sonenshein A.L. J. Bacteriol. 1993; 175: 4605-4614Crossref PubMed Google Scholar). The cloned rpoEgene was obtained originally from plasmid pCBδ11 (4Lampe M. Binnie C. Schmidt R. Losick R. Gene (Amst.). 1988; 67: 13-20Crossref PubMed Scopus (25) Google Scholar), a pBR322 derivative containing rpoE on a ∼2-kilobaseHindIII fragment. To generate a transcriptional fusion tolacZ at the rpoE locus, therpoE-containing HindIII fragment from pCBδ11 was cloned into pBSKII+ (Stratagene) such that therpoE gene was oriented toward the unique KpnI site. The resulting plasmid, pFL22, was digested with EcoRI and BglII, and the promoter-containing fragment was ligated to pTKlac and digested with EcoRI and BamHI to generate pFL23. Integration of pFL23 into the rpoE mutant strain RP3 (4Lampe M. Binnie C. Schmidt R. Losick R. Gene (Amst.). 1988; 67: 13-20Crossref PubMed Scopus (25) Google Scholar) generates strain HB6001 (rpoE-lacZ), whereas integration into the wild type (CU1065) generates HB6002(rpoE-lacZ). For generating the lacZtranscriptional fusion at the SPβ locus, a HindIII toSspI fragment of pFL20 (14López de Saro F. Woody A.-Y.M. Helmann J.D. J. Mol. Biol. 1995; 252: 189-202Crossref PubMed Scopus (48) Google Scholar) was cloned into pBSKII+ after digestion with HindIII and SmaI to generate pFL21. The promoter-containing fragment was excised by digestion withHindIII and BamHI and cloned into pJPM122 after digestion with the same enzymes to generate pFL60. pFL60 was first integrated into B. subtilis ZB307A (20Zuber P. Losick R. J. Bacteriol. 1987; 169: 2223-2230Crossref PubMed Google Scholar), and a transducing lysate was used to transfer of the PrpoE-cat-lacZ operon fusion into wild type (CU1065) and rpoE mutant (HB6010) backgrounds to generate strains HB6005 and HB6006, respectively. Overexpression of δ in B. subtilis was achieved using theP spac-containing plasmid, pDG148 (21Stragier P. Bonamy C. Karmazyn-Campelli C. Cell. 1988; 52: 697-704Abstract Full Text PDF PubMed Scopus (306) Google Scholar). TherpoE gene was removed from plasmid pFL32 (14López de Saro F. Woody A.-Y.M. Helmann J.D. J. Mol. Biol. 1995; 252: 189-202Crossref PubMed Scopus (48) Google Scholar) as anEcoRI to EcoRV fragment and cloned into pBSKII+ after digestion with EcoRI andSmaI to generate pFL24. Then the rpoE gene was removed from pFL24 as an XbaI fragment and cloned into pDG148 after digestion with XbaI to generate pFL25. RNA was purified using the RNeasy Total RNA kit purification system of Qiagen Inc. RNA from in vitro reactions was produced by mixing a polymerase chain reaction (PCR) fragment containing the rpoE promoter with 1 pmol ofBacillus RNA polymerase core enzyme supplemented with 10 pmol of purified ςA, 1× transcription buffer (40 mm Tris acetate, pH 8.0, 50 mm ammonium acetate, 20 mm potassium acetate, 4 mmmagnesium acetate, 0.1 mm dithiothreitol, and 0.04 mg/ml bovine serum albumin), and 0.1 mm NTPs, and incubating the mixture at 37 °C for 15 min. The "run-off" RNA was then purified using the Qiagen kit and precipitated with ethanol. A primer (5′-TCCTTTAGCTCTTCCTG-3′), centered at around 40 base pairs downstream of the translation start site, was labeled with [γ-32P]ATP and polynucleotide kinase (New England Biolabs, Inc.) following standard procedures (15Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1990Google Scholar). The labeled primer was then hybridized to the RNA in hybridization buffer (60 mm NaCl, 50 mm Tris-Cl, pH 8.0, and 10 mm dithiothreitol) by heating to 90 °C for 1 min and then cooling to 25 °C over 2 h. Next, nucleic acids were precipitated with ethanol in the presence of 0.3 m sodium acetate, pH 5.2, redissolved in extension buffer (50 mmTris-HCl, pH 8.3, 40 mm potassium chloride, 7 mm magnesium chloride, 1 mm dithiothreitol, and 0.1 mg/ml bovine serum albumin) supplemented with 10 mmdNTPs and 10 units of avian myeloblastosis virus reverse transcriptase, and incubated at 37 °C for 20 min. The resulting labeled DNA was then visualized by electrophoresis followed by autoradiography. To delete the 5′-untranslated region preceding the rpoE gene, the promoter region was amplified using the PCR and pFL24 as a template with primers NY1 (5′-GGCCCGAAGCTTTAACGGAAAACATCTCTCAGTCGG-3′) and NY2 (5′-CGGGAATTCCTTATACAAACCATACCTCTC-3′), and the resulting product was digested with HindIII and EcoRI (sites underlined). The 5′-end of the rpoE gene, together with the corresponding ribosome binding site, was amplified using primers NY3 (5′-GGCGAATTCTAGAAAGGGAGTGTCCGACCTTGGG-3′) and #107 (5′-GCGGATCCTACTTGACTGTCGGCTGAG-3′) and digested withEcoRI and BamHI. The two PCR products were cloned in a three-way ligation into pJPM122 digested withHindIII and BamHI to generate pNY50. Transformation of pNY50 (linearized with ScaI) into ZB307A generates HB6040. Transducing lysates, prepared from HB6040, were used to move the SPβ-borne PrpoEΔ-cat-lacZ operon fusions into CU1065 (generating HB6041) and the rpoE mutant strain HB6010 (generating HB6042). To assay gene expression, rich medium (2xYT) or minimal medium was inoculated from an overnight culture by 1:100 dilution, samples were taken, and β-galactosidase was assayed by the procedure of Miller (22Miller J.H. Experiments in Molecular Genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1972: 352Google Scholar). Immunoblots were carried out using rabbit anti-δ antibodies (14López de Saro F. Woody A.-Y.M. Helmann J.D. J. Mol. Biol. 1995; 252: 189-202Crossref PubMed Scopus (48) Google Scholar), goat horseradish peroxidase-coupled anti-rabbit secondary antibodies (Bio-Rad), and an enhanced chemiluminiscence detection system (NEN Life Science Products) following the instructions of the suppliers. Pure δ was obtained as described (14López de Saro F. Woody A.-Y.M. Helmann J.D. J. Mol. Biol. 1995; 252: 189-202Crossref PubMed Scopus (48) Google Scholar), and quantitation of its concentration was carried out using the Bradford reagent (Bio-Rad). Extracts of B. subtilis CU1065 cells were prepared by sonication of cells after a treatment with lysozyme and centrifugation at 12,000 rpm for 10 min. Spores of CU1065 and HB6010 were prepared by extensive washing, and extracts were prepared as described (23Nicholson W.L. Setlow P. Harwood C.R. Cutting S.M. Molecular Biological Methods for Bacillus. John Wiley & Sons, Inc., New York1990: 391-450Google Scholar). Derivatives of δ and ςA were constructed by the addition of a protein kinase A recognition motif to the amino terminus of the protein (24Kelman Z. Naktinis V. O'Donnell M. Methods Enzymol. 1995; 262: 430-442Crossref PubMed Scopus (47) Google Scholar). We used the PCR to place 11 additional codons at the 5′-end of the corresponding genes. Oligonucleotides used for PCR amplification of the δ gene and addition of the PKA sequence were a "PKA primer" (5′-AAAAGCTAGCCTGCGTCGTGCGTCCCTGGGTGATCAGGGTATCAAACAATATTCA-3′) and a 3′-rpoE primer (5′-CGCGGATCCCGACTATGAAAGTCAAGATCG-3′). The PKA primer encodes a PKA recognition site (ASLRRASLGDE) between a 5′-NheI site (singly underlined) and a BclI site (doubly underlined) used in later constructions. The PCR product was cloned first into pET11c (Novagen) as an NheI toBamHI fragment to generate pFL70. To generate δNPKA and δCPKA, we cloned therpoE gene from the pFL70 product as an XbaI toBamHI fragment into pBKSII+ (digested withXbaI and BamHI) to generate pFL80. For δNPKA we digested pFL80 at two BglII sites internal to the rpoE gene, filled the recessed termini using the Klenow enzyme and all four dNTPs, and ligated the resulting blunt termini with T4 ligase, thus creating a new BspDI site and an in-frame stop codon in plasmid pFL82 as described previously for unmodified rpoE (14López de Saro F. Woody A.-Y.M. Helmann J.D. J. Mol. Biol. 1995; 252: 189-202Crossref PubMed Scopus (48) Google Scholar). To create δCPKA, we digested pFL80 with BclI, partially digested withBglII, and ligated. A plasmid was identified with a deletion of the amino-terminal 109 amino acids of δ and was designated pFL85. For protein overproduction, each modified rpoE gene was removed as an XbaI to BamHI fragment and cloned into pET11c to generate plasmids pFL72 (δNPKA) and pFL75 (δCPKA). The modified rpoE genes were verified by DNA sequencing. For the construction of ςAPKA the following oligonucleotides were used as PCR primers in a reaction containing the cloned sigA gene as template: 5′-GGGCTCTAGACTGCCTCGTCGCTCCCTGGCTGATAAACAAACCCACGAC-3′ and 5′-CGCGGATCCTTAATCGTCCAGGAAGCTACG-3′. The resulting PCR product was digested with XbaI and BamHI (sites underlined) and ligated into pET11c digested with NheI andBamHI to generate pFL90. The sequence of the modifiedsigA gene was verified by automated DNA sequencing. Purification of δPKA and ςAPKAwas as described for the unmodified proteins (14López de Saro F. Woody A.-Y.M. Helmann J.D. J. Mol. Biol. 1995; 252: 189-202Crossref PubMed Scopus (48) Google Scholar, 25Chang B.Y. Doi R.H. J. Bacteriol. 1990; 172: 3257-3263Crossref PubMed Google Scholar). Briefly, δNPKA was purified from E. coli BL21/DE3 cells by passage of an extract of isopropyl-1-thio-β-d-galactopyranoside-induced cells through an S-Sepharose chromatography column previously equilibrated with buffer A (50 mm HEPES, pH 8.2, 2 mm EDTA, and 5% glycerol in water). After washing and elution of δNPKA with buffer A supplemented with 200 mmNaCl, the protein solution was precipitated with ammonium sulfate (60% saturation) and centrifuged. Further purification of δNPKA was carried out by the passage of the protein through fast protein liquid chromatography Superdex 70 (size exclusion chromatography) and Mono-S (ion exchange chromatography) columns. PKA derivatives were labeled with [γ-32P]ATP and protein kinase A (New England Biolabs) as described (24Kelman Z. Naktinis V. O'Donnell M. Methods Enzymol. 1995; 262: 430-442Crossref PubMed Scopus (47) Google Scholar). The rpoEpromoter region contains a predicted ςA-type promoter element located ∼90 base pairs upstream of the translational start site (Fig. 1 A). Primer extension analysis (Fig. 1 B) with RNA purified from vegetatively growing cells and from in vitro transcription assays using B. subtilis RNA polymerase containing ςA shows that transcription starts with approximately equal frequency at either of two purine residues located 84 and 86 base pairs from the translation start site. The putative −10 region of the promoter conforms to a ςA consensus promoter (26Helmann J.D. Nucleic Acids Res. 1995; 23: 2351-2360Crossref PubMed Scopus (333) Google Scholar), but there is weak if any similarity in the −35 region. The level of expression of rpoE was monitored by β-galactosidase assays of cells containing rpoE-lacZ transcriptional fusions. When measured at the rpoE locus (strain HB6002), expression reaches a maximum of 60–80 Miller units at the transition between logarithmic and stationary phase (Fig.2). Other strains in which the fusion was placed at the SPβ locus, either in wild type or in an rpoEmutant, produced essentially identical results (data not shown). Folding of the untranslated RNA molecule with the Mulfold program (27Walter A.E. Turner D.H. Kim J. Lyttle M.H. Muller P. Mathews D.H. Zucker M. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 9218-9222Crossref PubMed Scopus (428) Google Scholar) shows that this region contains several inverted repeats predicted to form stable stem loops (estimated ΔG of −23.3 and −22.7 kcal/mol at 25 °C and 1 m NaCl) (Fig. 1 A). To investigate the possible regulatory function of these sequences, a deletion mutation was engineered and placed in wild type (HB6041) orrpoE mutant (HB6042) cells. β-galactosidase analysis of these two strains showed that rpoE expression is increased 2-fold in the absence of these sequences. However, as noted for the wild-type leader region, the presence or absence of δ had no detectable effect on expression. In vitro transcription from the rpoE promoter showed that small RNA fragments about 50 nucleotides long are produced when these sequences are present, indicating that these stem loops might function as transcriptional terminators. As with the in vivo result, the presence or absence of δ in the transcription reaction did not make a difference (data not shown). Thus, expression of rpoE is unaffected by δ, and we find no apparent autoregulation of this locus. Quantitative immunoblots were performed to estimate the abundance of δ protein. The antiserum used for these studies is highly specific and reacts with a single 21-kDa band present in wild-type cells but absent from rpoE mutant strains (14López de Saro F. Woody A.-Y.M. Helmann J.D. J. Mol. Biol. 1995; 252: 189-202Crossref PubMed Scopus (48) Google Scholar) (Fig.3 B, lane 1). When 10 μg of cell lysate is analyzed, the amount of δ detected remains constant during growth at between 10 and 50 ng (Fig. 3 A, lanes 1–3) as judged by comparison with the pure protein. We therefore estimate that δ represents 0.3 ± 0.1% by weight of soluble cell protein. Because RNA polymerase (Mr = 337,000) represents 1% by weight of soluble cell protein (28Burgess R.R. Jendrisak J.J. Biochemistry. 1975; 14: 4634-4638Crossref PubMed Scopus (843) Google Scholar), we conclude that δ is an abundant protein (∼104molecules/cell) present in an approximately 5:1 molar excess relative to RNAP. Purified spores also contain similar levels of δ (Fig.3 B, lane 2). The fact that therpoE gene seems to be up-regulated in the transition state between the logarithmic and stationary phase (Fig. 2) might be indicative of a role of δ in stationary phase phenomena. To investigate the consequences of an rpoE disruption on cell growth, we chose a mutation (RP17) in which most of the rpoEcoding sequence is deleted (4Lampe M. Binnie C. Schmidt R. Losick R. Gene (Amst.). 1988; 67: 13-20Crossref PubMed Scopus (25) Google Scholar). We find that rpoE mutants generate a striking colony morphology when incubated for extended times (3–4 days) on Luria agar at room temperature; colonies of HB6010 show a rough edge with a fractal geometry (Fig.4 B) in contrast to the isogenic parental strain, which has smooth edged colonies. In addition, after dilution of 24-h-old cells into fresh medium (1:100), the cells of HB6010 are markedly more elongated and tend to occur in aggregates, as seen under the light microscope (Fig. 4 D), and this phenotype persists for several generations. HB6010 also shows a reproducible 30-min delay in entrance into the logarithmic phase (Fig.5), as compared with CU1065, in either minimal medium or LB growth medium.Figure 5An rpoE mutant has a lengthened lag phase. Growth curves of strains CU1065 (♦) and HB6010 (▪) when diluted 100:1 in fresh liquid LB medium after a 24 h incubation at 37 °C with vigorous shaking. Viable cell counts (colony forming units) demonstrated that viability is identical for both strains even after prolonged incubation. The average difference of time of entrance into logarithmic growth between CU1065 and HB6010 is 30 min.View Large Image Figure ViewerDownload Hi-res image Download (PPT) In contrast with these results, most other attempts to define a phenotype for the rpoE mutant strain have not been successful. HB6010, a rpoE deletion mutant, sporulates with the same efficiency as CU1065, a wild-type strain, and the purified spores are equally sensitive to UV irradiation (data not shown). Growth rate in Luria broth, sporulation medium, or minimal medium is also indistiguishable between the two strains, as originally found by Lampeet al. (4Lampe M. Binnie C. Schmidt R. Losick R. Gene (Amst.). 1988; 67: 13-20Crossref PubMed Scopus (25) Google Scholar). We did not detect any differences in motility or in resistance to H2O2 between CU1065 and HB6010 (data not shown). Double mutations of rpoE with other known regulators of transition state phenomena, such as abrB orsinR, did not result in phenotypes distinguishable from the phenotypes because of these mutations alone (data not shown). Finally, induction of δ from the isopropyl-1-thio-β-d-galactopyranoside inducibleP spac promoter (in pFL25) led to the overproduction of δ protein as observed by immunoblot analysis but did not produce any appreciable changes in the growth rate or viability nor did it have any major effects on protein composition of the cell, as judged by a [35S]methionine pulse-chase experiment followed by SDS-polyacrylamide gel electrophoresis (data not shown). For use in biochemical studies, we placed a 9-amino acid recognition site for protein kinase A at the amino terminus of δ to generate δPKA. Overexpression and purification of δPKA resulted in a protein that displaced nucleic acids as efficiently as wild-type δ (14López de Saro F. Woody A.-Y.M. Helmann J.D. J. Mol. Biol. 1995; 252: 189-202Crossref PubMed Scopus (48) Google Scholar). We used 4% native polyacrylamide gel electrophoresis to study the interaction of δPKA with RNAP. As expected for a small protein with a −49 net charge, δPKA migrated very fast, but its mobility was greatly reduced in the presence of increasing amounts of core RNAP (Fig. 6). Interestingly, two very closely spaced bands are observed suggesting that two distinct forms of RNAP-δ complex are present. Similar results are seen with E. coli RNAP (data not shown). Under these conditions 0.28 μm core enzyme binds less than 50% of the labeled δ, whereas 0.54 μm binds nearly all of the labeled δ. Because δ is present at very low levels (2 nm) in these reactions, we can estimate the apparent dissociation constant as the concentration of RNAP required for half-maximal binding (assuming all RNAP can bind δ). By interpolation, this is ∼0.4 μm, corresponding to an association constant of 2.5 × 106m−1. The presence of a single complex upon the addition of RNAP is consistent with a stoichiometry of 1:1 as previously observed (9Hyde E.I. Hilton M.D. Whiteley H.R. J. Biol. Chem. 1986; 261: 16565-16570Abstract Full Text PDF PubMed Google Scholar, 14López de Saro F. Woody A.-Y.M. Helmann J.D. J. Mol. Biol. 1995; 252: 189-202Crossref PubMed Scopus (48) Google Scholar). We used labeled δPKA and ςPKA to investigate the interactions of δ and ς with RNAP in the presence and absence of the strong trnS promoter (Fig.7). As before, the addition of core RNAP (lane 2) leads to the appearance of a slower mobility species (Eδ). To see whether we could detect formation of a Eςδ complex, we added an excess of unlabeled ςA. However, under these conditions, there was no "supershift" detected, suggesting that ςA was unable to bind the Eδ complex (lane 4). However, δ is displaced from the core by the presence of DNA (lanes 3 and 5). In a parallel series of experiments using radiolabeled ςAPKA, we observed the formation of a slow mobility complex corresponding to promoter-bound EςA(lanes 7 and 9) (data not shown), and this complex is stable to the addition of excess cold δ. We again failed to detect a δ-dependent supershift, suggesting that, under the conditions tested, binding favors complexes of Eδ or Eς·DNA but not Eδ·DNA or Eςδ·DNA. Previous analyses indicated that δ, a stoichiometric component of RNAP purified from B. subtilis, is dispensable for growth. Indeed, a rpoE mutant had no apparent phenotype (4Lampe M. Binnie C. Schmidt R. Losick R. Gene (Amst.). 1988; 67: 13-20Crossref PubMed Scopus (25) Google Scholar). Nevertheless, δ has large effects on the biochemical properties of RNAP (3Pero J. Nelson J. Fox T.D. Proc. Natl. Acad. Sci. U. S. A. 1975; 72: 1589-
Referência(s)