Portal de Periódicos da CAPES

Multiple Mechanisms of Transcription Inhibition by ppGpp at the λp R Promoter

2002; Elsevier BV; Volume: 277; Issue: 46 Linguagem: Inglês

10.1074/jbc.m208768200

1083-351X

Katarzyna Potrykus, Grzegorz Węgrzyn, V. James Hernandez,

RNA Research and Splicing

General stress conditions in bacterial cells cause a global cellular response called the stringent response. The first event in this control is production of large amounts of a regulatory nucleotide, guanosine-3′,5′-(bis)pyrophospahte (ppGpp). It was proposed recently that ppGpp acts by decreasing stability of open complexes at promoters that make short-lived open complexes,e.g. the rRNA promoters. However, here we report that the bacteriophage λp R promoter, which forms long-lived open complexes, is inhibited by ppGpp in vitroas observed in vivo. We performed a systematic investigation of the ppGpp-specific inhibition of transcription initiation at λp R and found that ppGpp does decrease stability of open complexes at λp R, but only slightly. Likewise the equilbrium binding constant and rate of open complex formation by RNA polymerase at λp R are only slightly affected by ppGpp. The major effect of ppGpp-mediated inhibition is to decrease the rate of promoter escape. We conclude that ppGpp-mediated inhibition of transcription initiation is not restricted to promoters that make short-lived open complexes. Rather we conclude that the initial catalytic step of transcript formation is affected by ppGpp, specifically formation of the first phosphodiester bond is inhibited by ppGpp at λp R. General stress conditions in bacterial cells cause a global cellular response called the stringent response. The first event in this control is production of large amounts of a regulatory nucleotide, guanosine-3′,5′-(bis)pyrophospahte (ppGpp). It was proposed recently that ppGpp acts by decreasing stability of open complexes at promoters that make short-lived open complexes,e.g. the rRNA promoters. However, here we report that the bacteriophage λp R promoter, which forms long-lived open complexes, is inhibited by ppGpp in vitroas observed in vivo. We performed a systematic investigation of the ppGpp-specific inhibition of transcription initiation at λp R and found that ppGpp does decrease stability of open complexes at λp R, but only slightly. Likewise the equilbrium binding constant and rate of open complex formation by RNA polymerase at λp R are only slightly affected by ppGpp. The major effect of ppGpp-mediated inhibition is to decrease the rate of promoter escape. We conclude that ppGpp-mediated inhibition of transcription initiation is not restricted to promoters that make short-lived open complexes. Rather we conclude that the initial catalytic step of transcript formation is affected by ppGpp, specifically formation of the first phosphodiester bond is inhibited by ppGpp at λp R. The stringent response is a generalized adaptive response to nutritional deprivation and environmental stress (for a review, see Ref. 1Cashel M. Gentry D.R. Hernandez V.J. Vinella D. Neidhard F.C. Curtiss R. Ingraham J.L. Lin E.C.C. Low K.B. Magasanik B. Reznikoff W.S. Riley M. Schaechter M. Umbarger H.E. Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology. 2nd Ed. American Society for Microbiology, Washington, D. C.1996: 1458-1496Google Scholar). The production of a specific nucleotide, guanosine-5′,3′-(bis)pyrophosphate (ppGpp), 1The abbreviations used are: ppGpp, guanosine-3′,5′-(bis)pyrophospahte; RNAP, RNA polymerase 1The abbreviations used are: ppGpp, guanosine-3′,5′-(bis)pyrophospahte; RNAP, RNA polymeraseis the primary signaling and initiating event in the stringent response. This nucleotide interacts with RNA polymerase, most likely at the interface of the β and β′ subunits (2Chatterji D. Fujita N. Ishihama A. Genes Cells. 1998; 3: 279-287Crossref PubMed Scopus (139) Google Scholar, 3Toulokhonov I.I. Shulgina I. Hernandez V.J. J. Biol. Chem. 2001; 276: 1220-1225Abstract Full Text Full Text PDF PubMed Scopus (93) Google Scholar), which ultimately leads to the inhibition of synthesis of stable RNAs (rRNAs and tRNAs), and to the inhibition or activation of synthesis of specific mRNAs (for review, see Ref. 1Cashel M. Gentry D.R. Hernandez V.J. Vinella D. Neidhard F.C. Curtiss R. Ingraham J.L. Lin E.C.C. Low K.B. Magasanik B. Reznikoff W.S. Riley M. Schaechter M. Umbarger H.E. Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology. 2nd Ed. American Society for Microbiology, Washington, D. C.1996: 1458-1496Google Scholar). In Escherichia coli, ppGpp synthesis may be catalyzed by two enzymes: RelA and SpoT. The relA gene product, ppGpp synthetase I, is a ribosome-associated enzyme activated upon amino acid starvation due to disruption of the translation process via interaction of uncharged tRNAs with "hungry" codons at the ribosomal A site. The spoT gene product, ppGpp synthetase II, is responsible for ppGpp synthesis in a ribosome-independent manner during carbon source or energy limitation (4Hernandez V.J. Bremer H. J. Biol. Chem. 1991; 266: 5991-5999Abstract Full Text PDF PubMed Google Scholar, 5Xiao H. Kalman M. Ikehara K. Zemel S. Glaser G. Cashel M. J. Biol. Chem. 1991; 266: 5980-5990Abstract Full Text PDF PubMed Google Scholar, 6Gentry D.R. Cashel M. Mol. Microbiol. 1996; 19: 1373-1384Crossref PubMed Scopus (141) Google Scholar, 7Murray K.D. Bremer H. J. Mol. Biol. 1996; 259: 41-57Crossref PubMed Scopus (135) Google Scholar). SpoT is a bifunctional protein capable of both ppGpp synthesis and hydrolysis and is the only major intracellular ppGpp degradase. Knock-out mutants in the relA gene do not contain functional ppGpp synthetase I, and do not produce ppGpp during amino acid starvation due to the fact that ppGpp synthetase II is inactive under these conditions (7Murray K.D. Bremer H. J. Mol. Biol. 1996; 259: 41-57Crossref PubMed Scopus (135) Google Scholar). This leads to a decrease in ppGpp level following amino acid-starvation ofrelA cells, such a phenotype is called the "relaxed" response. Although inhibition of synthesis of stable RNAs during the stringent response is very effective, the molecular mechanism underlying this phenomenon remains unclear. Travers (8Travers A. Nucleic Acids Res. 1984; 12: 2605-2618Crossref PubMed Scopus (101) Google Scholar) suggested that the nucleotide sequence located between the −10 consensus hexamer and +1 nucleotide position, which he termed the "discriminator," determines sensitivity of a promoter to ppGpp. He proposed that during the stringent response, promoters with a GC-rich discriminator are negatively regulated, while those with AT-rich discriminators were positively regulated. However, extensive mutational analyses revealed that the sequence identity of this region is not sufficient to explain the regulatory response of a promoter during the stringent response (for a review, see Ref. 1Cashel M. Gentry D.R. Hernandez V.J. Vinella D. Neidhard F.C. Curtiss R. Ingraham J.L. Lin E.C.C. Low K.B. Magasanik B. Reznikoff W.S. Riley M. Schaechter M. Umbarger H.E. Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology. 2nd Ed. American Society for Microbiology, Washington, D. C.1996: 1458-1496Google Scholar). Other authors (9Kingston R.E. Nierman W.C. Chamberlin M.J. J. Biol. Chem. 1981; 256: 2787-2797Abstract Full Text PDF PubMed Google Scholar, 10Vogel U. Jensen K.F. J. Biol. Chem. 1995; 270: 18335-18340Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar, 11Vogel U. Jensen K.F. J. Biol. Chem. 1994; 269: 16236-16241Abstract Full Text PDF PubMed Google Scholar) have demonstrated that ppGpp decreases the rate of transcription elongation in E. coli cells. However, assuming that this is the sole effect of ppGpp on transcription it would be hard to explain why some promoters display differential sensitivity to ppGpp with some being highly sensitive and others being relatively insensitive. Recently, Barkeret al. (12Barker M.M. Gaal T. Gourse R.L. J. Mol. Biol. 2001; 305: 689-702Crossref PubMed Scopus (177) Google Scholar) demonstrated that ppGpp decreased the stability of open complexes at all promoters they investigated. Open complexes at the major promoters located upstream of genes coding for rRNAs,rrn P1 promoters, are extremely unstable. It was proposed that ppGpp inhibits transcription preferentially from these promoters by further destabilizing these intrinsically unstable open complexes and, furthermore, that ppGpp-mediated transcription inhibition is generalized to affect all similarly behaving promoters. The hypothesis described above could explain differences between various promoters in their sensitivities to ppGpp. Namely, the open complex stability of a specific promoter may dictate its sensitivity to ppGpp with promoters forming stable open promoter complexes displaying resisitance to ppGpp. Contrary to this hypothesis, a ppGpp-dependent inhibition of the λp R promoter in vivo was shown (13Szalewska-Palasz A. Wegrzyn A. Herman A. Wegrzyn G. EMBO J. 1994; 13: 5779-5785Crossref PubMed Scopus (57) Google Scholar, 14Szalewska-Palasz A. Wegrzyn G. Acta Biochim. Pol. 1995; 42: 233-239Crossref PubMed Scopus (14) Google Scholar, 15Wrobel B. Murphy H. Cashel M. Wegrzyn G. Mol. Gen. Genet. 1998; 257: 490-495Crossref PubMed Scopus (15) Google Scholar, 16Slominska M. Neubauer P. Wegrzyn G. Virology. 1999; 262: 431-441Crossref PubMed Scopus (35) Google Scholar), while λp R has been demonstrated to form very stable open complexes in vitro (17McKane M. Gussin G.N. J. Mol. Biol. 2000; 299: 337-349Crossref PubMed Scopus (14) Google Scholar). Our principle aim in the these studies was to elucidate the mechanism of λp R promoter inhibition by ppGpp to: 1) extend the current general mechanistic picture of ppGpp-mediated transcription inhibition and 2) to verify our previous in vivo observations with respect to ppGpp inhibition at λp R, in vitro. Toward these goals we first sought to establish in vitro conditions that would reproduce the level of transcription inhibition of the λp R promoter by ppGpp observed in vivo. Once optimization of in vitro conditions for ppGpp inhibition of λp R were established, we proceeded to investigate the specific mechanism of ppGpp inhibition at λp R. The λp R promoter region (−75 to +35 bp) was generated by PCR using wild-type λ DNA as template and the following primers λpR3 (5′-CCCGGGTTAAGCGTTGTTCCA) and λpR4 (5′-CGAATTCACCGCAGGGATAAA-3′). Following digestion withEcoRI and SmaI, the DNA fragment was then cloned into pHG86 (18Giladi H. Igarashi K. Ishihama A. Oppenheim A.B. J. Mol. Biol. 1992; 227: 985-990Crossref PubMed Scopus (38) Google Scholar), previously cut with the same enzymes. The resulting plasmid, called pKP-70pR, served as the template in the PCR reactions where primers pR4 and pUC1 (5′-GTTTTCCCAGTCACGAC-3′; pUC19 universal primer) were used. The 268-bp PCR fragment thus obtained was used in the in vitro transcription reactions (the full-length fragment produced in such reactions was 198 nucleotides long). The λpL promoter-containing region was generated by PCR using wild-type λ DNA as the template and the following primers: pL1, 5′-GAATTCCCATACATTAGTGAGTTGA-3′ and pL2, 5′-GGATTCTGATTGCTGCCTTGA-3′. The resulting fragment used in thein vitro transcription experiments was 1205 bp long, and the transcript initiating from the λpL promoter was 660 nucleotides long. Deoxynucleotides used in the PCR reactions as well as the restriction enzymes were purchased from MBI Fermentas. Taq polymerase was obtained fromFINNZYMES, T4 DNA ligase from Invitrogen, andE. coli RNA polymerase holoenzyme (RNAP) from EpiCentre Technologies. Nucleotides used in in vitro transcription experiments were from Roche Molecular Biochemicals. ppGpp was a generous gift of Dr. Mike Cashel (National Institutes of Health). ApU, used in the abortive initiation transcriptions, and [α-32P]UTP were purchased from ICN Biomedicals. In vitrotranscription assays were performed using 20 nmλpR template and 60–140 nm RNAPs indicated in the following transcription buffer: 50 mm Tris-HCl (pH 8.0), 10 mm MgCl2, 10 mmβ-mercaptoethanol, 10 μg/ml bovine serum albumin, 140 mm KCl (unless indicated otherwise), 100 μmATP, GTP, and CTP, and 10 μm UTP (10 μCi/reaction [α-32P]UTP) and either 250 μm GDP or 250 μm ppGpp. The final reaction volume was 20 μl. RNA polymerase was preincubated with GDP or ppGpp for 7 min at room temperature prior to mixing with KCl, followed by 2-min incubation at 37 °C, and reactions were started by mixing RNAP with λ promoter templates and NTP substrates at 37 °C. Transcription was terminated after 10 min by addition of 100 μl of stop solution (30 mm EDTA, 300 mm sodium acetate (pH 5.5), 1 mg/ml glycogen) and followed by precipitation with 370 μl of ice-cold absolute ethanol for 1 h at −20 °C. Precipitates were collected by centrifugation and resuspended in 20 μl of formamide loading buffer (66% formamide, 1 mmEDTA, 0.05% bromphenol blue, 1× Tris borate-EDTA buffer). 10 μl of RNA re-suspensions were resolved on 7 m urea, 6% polyacrylamide gels (acrylamide to bis ratio 19:1). Gels were visualized and quantified by phosphorimaging on a Bio-Rad Molecular Imaging system. RNAP (60 nm) and DNA template (20 nm) were incubated for 7 min at room temperature in transcription buffer with 250 μm GDP or 250 μm ppGpp, followed by incubation for 30 min at 37 °C with 140 mm KCl. After heparin (100 μg/ml final) addition, aliquots (10 μl) were removed to a tube containing 10 μl of an NTPs mix (final concentrations: 100 μm ATP, GTP, and CTP, 10 μm UTP (10 μCi/reaction [α-32P]UTP), and 100 μg/ml heparin) in transcription buffer, at the indicated times. Reactions were chased out, stopped after 7 min, and analyzed as described above. Determination of KB and kf—K B and k fwere determined by τ plots analysis, as described previously (17McKane M. Gussin G.N. J. Mol. Biol. 2000; 299: 337-349Crossref PubMed Scopus (14) Google Scholar,19Goodrich J.A. McClure W.R. J. Mol. Biol. 1992; 224: 15-29Crossref PubMed Scopus (46) Google Scholar). DNA (1 nm) and RNAP (7.5–90 nm) were combined in transcription buffer containing 50 mm Tris-HCl (pH 8.0), 10 mm MgCl2, 10 mmβ-mercaptoethanol, 10 μg/ml bovine serum albumin, 140 mm KCl, 500 μm ApU, 50 μm GTP, 5 μm UTP (15 μCi of [α-32P]UTP), and 250 μm GDP or 250 μm ppGpp. Prior to mixing, RNAP was preincubated with GDP or ppGpp for 7 min at room temperature in a transcription buffer without KCl. At indicated times (0, 0.25, 0.5, 1, 2, 4, 8, and 16 min), 5-μl aliquots were removed and added to 5 μl of stop solution (7 m urea, 0.1m EDTA, 0.4% SDS, 40 mm Tris-HCl (pH 8.0), 0.5% (w/v) bromphenol blue, 0.5% (w/v) xylene cyanol) to terminate the reaction. The control reactions were started with the addition of NTPs to the tube containing RNAP prebound to DNA. Reactions were analyzed on 7 m urea, 20% polyacrylamide gels (acrylamide to bis ratio 59:1), with 0.5 × TBE top buffer and 1.2 × TBE bottom buffer. The amount of product (ApUpGpU) was quantitated by phosphorimaging on a Bio-Rad Molecular Imaging system. To determine τobs (lag time observed) for each RNAP concentration, a best fit line was determined through the experimental curve parallel to the control curve (control curve reflected the rates of abortive initiation in control reactions initiated with nucleotides) (20McClure W.R. Proc. Natl. Acad. Sci. U. S. A. 1980; 77: 5634-5638Crossref PubMed Scopus (270) Google Scholar). To determine KB and kf, SigmaPlot (Jandel Scientific) was used to perform a least square fit of the data to Equation 1 (20McClure W.R. Proc. Natl. Acad. Sci. U. S. A. 1980; 77: 5634-5638Crossref PubMed Scopus (270) Google Scholar), τobs=1/kf+1/KBkf[RNAP](Eq. 1) where τobs is the lag time observed,k f is the isomerization constant,K B is the binding constant of the RNAP to the promoter, and [RNAP] is the RNA polymerase concentration. RNAP (60 nm) was incubated with 20 nm DNA template and 250 μmGDP or 250 μm ppGpp at room temperature for 7 min in the transcription buffer, followed by the addition of KCl to 140 mm and incubation for 30 min at 37 °C. RNA sythesis was started by addition of NTPs (final concentrations: 100 μmATP, GTP, and CTP and 10 μm UTP (10 μCi/reaction [α-32P]UTP)) and heparin (100 μg/ml). 20-μl aliquots were removed at the indicated times, and the reactions were stopped and analyzed as described above. RNA products were resolved on a 7 m urea, 6% polyacrylamide gels (acrylamide to bis ratio 19:1) or, to visualize the abortive transcription products, on a 7 m urea, 25% polyacrylamide gels (acrylamide to bis ratio 12:1). RNAP (60 nm) was incubated with 250 μm GDP or 250 μm ppGpp in the transcription buffer for 7 min at room temperature followed by addition of KCl to 140 mm. Next, RNAP was added to NTP substrates (final concentrations: 250 μm UTP, 250 μm CTP, 20 μm GTP (10 μCi/reaction [α-32P]GTP), 15 μm ATP, and 0–500 μm ApU initiator) and λpR template (20 nm) to initiate the reaction. After 10-min incubation at 37 °C, the reactions were stopped and analyzed as described above. Using Northern blot analysis and analysis of activity of a gene fusion consisting of the lacZ gene under control of the λp R promoter, it was previously demonstrated that the bacteriophage λp R promoter activity is significantly decreased during the stringent, but not relaxed, response of E. coli (13Szalewska-Palasz A. Wegrzyn A. Herman A. Wegrzyn G. EMBO J. 1994; 13: 5779-5785Crossref PubMed Scopus (57) Google Scholar, 14Szalewska-Palasz A. Wegrzyn G. Acta Biochim. Pol. 1995; 42: 233-239Crossref PubMed Scopus (14) Google Scholar, 16Slominska M. Neubauer P. Wegrzyn G. Virology. 1999; 262: 431-441Crossref PubMed Scopus (35) Google Scholar). Genetic studies strongly suggested that this inhibition is due to direct rather than indirect effects of ppGpp on the reporter gene expression (15Wrobel B. Murphy H. Cashel M. Wegrzyn G. Mol. Gen. Genet. 1998; 257: 490-495Crossref PubMed Scopus (15) Google Scholar). Collectively, these studies revealed 5–10 times less λp R-initiated transcripts in amino acid-starved wild-type bacteria relative to normal growth conditions. Decreased levels of transcripts derived from the λp R promoter could not be explained by lower rates of transcription elongation (according to the model proposed by Vogel and Jensen (10Vogel U. Jensen K.F. J. Biol. Chem. 1995; 270: 18335-18340Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar, 11Vogel U. Jensen K.F. J. Biol. Chem. 1994; 269: 16236-16241Abstract Full Text PDF PubMed Google Scholar)), since replacement of λp R with the p lacpromoter, known to be independent of ppGpp, in the constructs used for measurement of transcription efficiency resulted in complete restoration of transcript abundance during the stringent response to levels comparable with those observed under normal growth conditions (13Szalewska-Palasz A. Wegrzyn A. Herman A. Wegrzyn G. EMBO J. 1994; 13: 5779-5785Crossref PubMed Scopus (57) Google Scholar). Despite these many observations implicating ppGpp-dependent inhibition of λp R, direct evidence of ppGpp inhibition of λp R has been lacking. Thus, we initiated in vitro transcription assays, first to verify, then to investigate, ppGpp-mediated inhibition of λp R. We initiated these investigations by accessing the effect of ppGpp on in vitro transcription of λp R at different KCl concentrations based on previous observations that the ppGpp-specific transcription inhibitory effects are profoundly salt-sensitive (12Barker M.M. Gaal T. Gourse R.L. J. Mol. Biol. 2001; 305: 689-702Crossref PubMed Scopus (177) Google Scholar, 21Ohlsen K.L. Gralla J.D. Mol. Microbiol. 1992; 16: 2243-2251Crossref Scopus (47) Google Scholar). We found that levels of transcripts derived from a λp R promoter template in multiple-round in vitro transcription assays were severalfold (5-fold on average) lower in the presence of ppGpp than in control experiments in which GDP was used at the same concentrations as control (Fig. 1). The degree of ppGpp-mediated inhibition of λp R was observed to be highly salt-dependent, as has been previously reported for the rrnB P1 promoter (12Barker M.M. Gaal T. Gourse R.L. J. Mol. Biol. 2001; 305: 689-702Crossref PubMed Scopus (177) Google Scholar, 21Ohlsen K.L. Gralla J.D. Mol. Microbiol. 1992; 16: 2243-2251Crossref Scopus (47) Google Scholar). The greatest degree of inhibition was observed between 140 and 160 mm KCl (Fig. 1). In all further experiments we used KCl at a final concentration of 140 mm. In addition, we found that to observe ppGpp inhibition of λp R, ppGpp had to be preincubated with RNAP prior to addition of KCl; thus, apparently, the initial interaction of ppGpp itself with RNAP is salt-sensitive (data not shown). The ppGpp-mediated inhibition of λp R was promoter-specific as the λp L promoter appeared to be insensitive to ppGpp under the same experimental conditions (Figs. 1 and2). These results are in good agreement with both in vivo (16Slominska M. Neubauer P. Wegrzyn G. Virology. 1999; 262: 431-441Crossref PubMed Scopus (35) Google Scholar) and in vitro (12Barker M.M. Gaal T. Gourse R.L. J. Mol. Biol. 2001; 305: 689-702Crossref PubMed Scopus (177) Google Scholar) studies that demonstrated that λp L is not inhibited by ppGpp. Moreover, since these results indicate no significant influence of ppGpp on the amounts of p L-derived transcripts, the decreased levels of p R-derived transcripts cannot be trivially ascribed to lower rates of transcription elongation but are promoter-specific. The level of inhibition of the p R promoter was dose-dependent on ppGpp concentration, but above a concentration of 200 μm the efficacy of ppGpp-dependent inhibition was roughly constant (Fig. 2). In all further studies a saturating level of 250 μm ppGpp or GDP (in control experiments) was used. The degree of inhibition we observed in vitro (about 5-fold) was in a good correlation with in vivo measurements that indicated 5–10-fold decreased levels of p R-derived transcripts under conditions of increased ppGpp levels in vivo (13Szalewska-Palasz A. Wegrzyn A. Herman A. Wegrzyn G. EMBO J. 1994; 13: 5779-5785Crossref PubMed Scopus (57) Google Scholar, 15Wrobel B. Murphy H. Cashel M. Wegrzyn G. Mol. Gen. Genet. 1998; 257: 490-495Crossref PubMed Scopus (15) Google Scholar). Hence, we conclude that inhibition of transcription from thep R promoter by ppGpp is direct. Next, we focused on elucidation of the specific influence of ppGpp on specific steps of the transcription initiation process at λp R. Barker et al. (12Barker M.M. Gaal T. Gourse R.L. J. Mol. Biol. 2001; 305: 689-702Crossref PubMed Scopus (177) Google Scholar) reported that ppGpp decreases stability of open complexes on all promoters tested; however, only in the case of promoters, which form intrinsically short-lived open complexes, the observed ppGpp-dependent destabilization of open complexes resulted in inhibition of promoter activity. Based on our observations of ppGpp-specific p R promoter inhibition (Figs. 1and 2) we also investigated effects of ppGpp on the stability of open complexes formed at p R. RNA polymerase-promoter complexes were preformed on DNA templates in the presence of RNAP pretreated with either GDP (control experiments) or ppGpp. This was followed by addition of heparin (to prevent re-association of RNA polymerase) and subsequent chase with nucleotides (NTPs) at different times to determine the relative fraction of complexes remaining by quantifying levels of p R-specific transcripts. In agreement with the observations of Barker et al. (12Barker M.M. Gaal T. Gourse R.L. J. Mol. Biol. 2001; 305: 689-702Crossref PubMed Scopus (177) Google Scholar), we found that ppGpp decreased stability of open complexes formed at p R about 2-fold (Fig.3). In GDP control experiments we confirmed the previous observations (17McKane M. Gussin G.N. J. Mol. Biol. 2000; 299: 337-349Crossref PubMed Scopus (14) Google Scholar) that p Rmakes long-lived open complexes (Fig. 3). We estimated the average lifetime of open complexes at p R to be ∼1 h compared with half-lives of less than a minute for rrnB andrrnD P1 open complexes (12Barker M.M. Gaal T. Gourse R.L. J. Mol. Biol. 2001; 305: 689-702Crossref PubMed Scopus (177) Google Scholar). The 2-fold destabilization of RNAP·p R open complexes in the presence of ppGpp, however, appears not to account for the observed 5-fold decrease in the activity of the p R promoter. Since the stability of open complexes formed at λp R is not dramatically affected in the presence of ppGpp, at least not to a level which can explain the full level of inhibition observed (Figs. 1 and 2), we next determined the apparent equilibrium binding constant of RNAP (KB) and the rate constant of open complex formation (kf) atp R in the presence of either GDP (control experiments) or ppGpp. This was accomplished by performing τ plot analyses using in vitro abortive transcription assays based on the classical method described by McClure (20McClure W.R. Proc. Natl. Acad. Sci. U. S. A. 1980; 77: 5634-5638Crossref PubMed Scopus (270) Google Scholar) (see "Experimental Procedures" for details). The results presented in Fig. 4 and TableI indicate that ppGpp has only a slight effect on either KB and kf at thep R promoter. Therefore, changes in either of these parameters appears not to be responsible for the observed decrease in transcription efficiency from p R by ppGpp.Table ISummary of kinetic constants for the λpR promoter in the presence of GDP or ppGppNucleotideτobsk fK BK B k fs10−2 s−1106m−1104m−1s−1GDP22.6 ± 4.834.42 ± 0.6430.26 ± 5.18135.85 ± 22.09ppGpp16.6 ± 2.056.02 ± 0.3824.59 ± 7.01148.06 ± 9.36Values for τobs, k f, andK B k f were calculated from τ plots (Fig. 4) as outlined under "Experimental Procedures." S.E. values were determined from residuals of the least square fit to Equation 1. Open table in a new tab Values for τobs, k f, andK B k f were calculated from τ plots (Fig. 4) as outlined under "Experimental Procedures." S.E. values were determined from residuals of the least square fit to Equation 1. Next we sought to investigate possible effects of ppGpp on pR promoter clearance also termed "promoter escape." Following formation of RNAP-λpR binary complexes, single-round transcription assays were performed by the simultaneous addition of heparin and NTP substrates. The rate of accumulation of full-length products was determined with increasing time in the presence or absence of ppGpp (Fig. 5). In the presence of ppGpp there was observed a significant lag in formation of full-length transcripts compared with the GDP control (approximately a 4-fold difference; lasting 30 s in the presence of GDP and 2 min in the presence of ppGpp (see inset Fig. 5). Therefore, escape of the RNAP from the pR promoter is impaired in the presence of ppGpp. Surprisingly, we did not observe changes in the amounts of abortive transcripts accumulating under these conditions (data not shown). In addition, this observed lag in promoter clearance at λpR is sufficient to account for the observed 5-fold inhibition (80% inhibition) observed in multiple-round transcription assays (Figs. 1 and 2). Specifically, over the duration of a 10-min multiple round assay with a respective clearance time of 30 s or 2 min in the presence of GDP or ppGpp, respectively, 20 versus5 rounds of transcription would occur leading to an apparent observed 75% inhibition. The results of the promoter clearance assay (Fig. 5) was in conflict with observations that there was little or no observable alteration in the apparent rate of open complex formation determined by the McClure abortive transcription assay method (Fig. 4). In the experiments presented in Fig. 4, determination of apparent kinetic constants, KB and kf, relied on measurement of the rate of formation of the initial transcribed tetramer RNA species from λpR, ApUpGpU (see "Experimental Procedures"). However, no ppGpp-induced effect was observed in the kinetics of formation of this small RNA, which would have led to an overestimation of the rate of open complex formation, and apparent erroneous indication of a ppGpp-dependent effect on the rate of open complex formation, which it did not (Fig. 4; Table I). Despite this, a ppGpp-induced slower rate of promoter escape was indeed evident (Fig.5). Thus, we hypothesized that the lower rate of promoter escape is not evident during early steps in the RNA polymerization process either 1) because the slow step in promoter escape occurs after the RNAP has advanced past the stage of formation of the tetramer RNA or 2) that the use of a dinucleotide in the McClure abortive assay method (see "Experimental Procedures") to produce the tetramer RNA caused the initial transcription process to be resistant to the presence of ppGpp. To test the latter of these possibilities we investigated whether programming transcription initiation with the initiating dinucleotide (corresponding to +1 and +2 positions), ApU, would have an effect on ppGpp-dependent inhibition of λpR. For these experiments, we used a low concentration of ATP (15 μm), which was not limiting for elongation (data not shown) but reduced competition for initiation by ApU, and increasing concentrations of ApU initiator, from 0 to 500 μm in the presence or absence of ppGpp. Results shown in Fig. 6demonstrate that as initiation is increasingly driven by the dinucleotide ApU, transcription from λpR becomes completely resistant to the inhibitory effects of ppGpp. That is, at 0 concentration of ApU when initiation is being driven solely by ATP, the 80% ppGpp inhibitory effect is evident (compared with the GDP control), but at increased concentrations of ApU when initiation is increasingly driven by the dinucleotide initiator, less difference in the level of transcript accumulation is observed, and ppGpp inhibition is progressively diminished and is no longer apparent at a concentration above 50 μm ApU (Fig. 6). This effect is specific for the dinucleotide initiator, since increasing the concentration of the initiating nucleotide ATP from 0 to 500 μm, using a nonspecific dinucleotide (ApC) or using a dinucleotide corresponding to the −1 and +1 positions (CpA), still displayed ppGpp-specific inhibition (data not shown). ppGpp is a stress-induced small molecular weight effector of RNAP being produced in large amounts in response to a variety of environmental stresses, for example, the onset of amino acid starvation in bacterial cells (for a review, see Ref. 1Cashel M. Gentry D.R. Hernandez V.J. Vinella D. Neidhard F.C. Curtiss R. Ingraham J.L. Lin E.C.C. Low K.B. Magasanik B. Reznikoff W.S. Riley M. Schaechter M. Umbarger H.E. Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology. 2nd Ed. American Society for Microbiology, Washington, D. C.1996: 1458-1496Google Scholar). The production of ppGpp leads directly to global changes in bacterial physiology, called the stringent response. ppGpp binds directly and allosterically to RNA polymerase (2Chatterji D. Fujita N. Ishihama A. Genes Cells. 1998; 3: 279-287Crossref PubMed Scopus (139) Google Scholar, 3Toulokhonov I.I. Shulgina I. Hernandez V.J. J. Biol. Chem. 2001; 276: 1220-1225Abstract Full Text Full Text PDF PubMed Scopus (93) Google Scholar, 22Reddy P.S. Raghaven A. Chatterji D. Mol. Microbiol. 1995; 15: 255-265Crossref PubMed Scopus (77) Google Scholar), resulting in direct transcriptional inhibition of those promoters responsible for synthesis of stable ribosomal and transfer RNAs. However, it is intriguing that the modification of RNA polymerase by ppGpp results in inhibition of transcription from some promoters, whereas other promoters are insensitive to this nucleotide or even stimulated during the stringent response (see Ref. 1Cashel M. Gentry D.R. Hernandez V.J. Vinella D. Neidhard F.C. Curtiss R. Ingraham J.L. Lin E.C.C. Low K.B. Magasanik B. Reznikoff W.S. Riley M. Schaechter M. Umbarger H.E. Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology. 2nd Ed. American Society for Microbiology, Washington, D. C.1996: 1458-1496Google Scholar for review). The most extensively investigated promoters under negative ppGpp regulation are the major rRNA promoters, designated the rrnP1 promoters. Recently it was proposed that ppGpp decreases stability of open complexes at all promoters but inhibits transcription directly only from promoters that form intrinsically short-lived open complexes, as is exemplified by the rrnB and rrnD P1 promoters (12Barker M.M. Gaal T. Gourse R.L. J. Mol. Biol. 2001; 305: 689-702Crossref PubMed Scopus (177) Google Scholar). The inhibitory effect of ppGpp on transcription from these promoters appears to be primarily through the destabilization of the RNAP·rrn P1 binary open complexes which are intrinsically short lived. The mechanism of ppGpp-mediated inhibition of promoters forming relatively stable open complexes, however, remained obscure. In fact, previous in vivo data (13Szalewska-Palasz A. Wegrzyn A. Herman A. Wegrzyn G. EMBO J. 1994; 13: 5779-5785Crossref PubMed Scopus (57) Google Scholar, 14Szalewska-Palasz A. Wegrzyn G. Acta Biochim. Pol. 1995; 42: 233-239Crossref PubMed Scopus (14) Google Scholar, 15Wrobel B. Murphy H. Cashel M. Wegrzyn G. Mol. Gen. Genet. 1998; 257: 490-495Crossref PubMed Scopus (15) Google Scholar, 16Slominska M. Neubauer P. Wegrzyn G. Virology. 1999; 262: 431-441Crossref PubMed Scopus (35) Google Scholar) indicated that the bacteriophage λp R, shown to form long-lived open complexes in vitro (17McKane M. Gussin G.N. J. Mol. Biol. 2000; 299: 337-349Crossref PubMed Scopus (14) Google Scholar), is also inhibited by ppGpp. Here, using in vitro transcription assays we confirmed previous in vivo results that ppGpp inhibits transcription from the λp R promoter directly (Figs. 1 and2). We then further investigated the specific ppGpp-dependant mechanism of λp R inhibition. We found that ppGpp decreases stability of open complexes at λp Rabout 2-fold (Fig. 3), but since overall inhibition was ∼80% (Fig.2), this 2-fold destabilization cannot account solely for the observed ppGpp-mediated inhibition of λp R. In addition, neither the equilibrium binding constant, K B, nor the rate constant of open complex formation, kf, at λp R were significantly influenced by ppGpp (Fig. 4, Table I). In fact, we found that a decrease in the rate of promoter escape of RNAP from λp R in the presence of ppGpp (Fig. 5) can nearly account for the full inhibition by ppGpp observed on λp R (75% inhibition). However, it seems that ppGpp can affect several steps in the transcription process each to varying degrees. Thus, it is likely that the cumulative effects of inhibition at perhaps each of these steps is what ultimately leads to the total 80% inhibitory effect of ppGpp on λp R. Our results clearly indicate that ppGpp-mediated inhibition of transcription initiation is not restricted to promoters that make short-lived open complexes, as proposed by Barker et al.(12Barker M.M. Gaal T. Gourse R.L. J. Mol. Biol. 2001; 305: 689-702Crossref PubMed Scopus (177) Google Scholar). Clearly, the general mechanism of ppGpp inhibitory action is more complicated than solely decreasing the stability of open complexes. On the other hand, it is likely that the 2–3-fold ppGpp-mediated decrease in half-lives of open complexes formed at rrn P1 promoters, which are unstable even under normal conditions, is sufficient to account for the inhibition observed at these promoters (12Barker M.M. Gaal T. Gourse R.L. J. Mol. Biol. 2001; 305: 689-702Crossref PubMed Scopus (177) Google Scholar). Therefore, the model of ppGpp-mediated regulation of transcription initiation proposed by Barker et al. (12Barker M.M. Gaal T. Gourse R.L. J. Mol. Biol. 2001; 305: 689-702Crossref PubMed Scopus (177) Google Scholar), while it appears to be valid for rrn promoters, is insufficient to explain the mechanism of ppGpp inhibition of activities of promoters exemplified in this report by λp R. Unexpectedly, we found that when "programming" transcription initiation from the natural λp R start site using the dinucleotide initiator, corresponding to +1 and +2 positions (ApU), the ppGpp-mediated inhibition of λp Rwas abrogated (Fig. 6). Our testing of this possibility was an attempt to reconcile differences observed in the transcription of abortive RNA during τ plot analyses (Fig. 4) and of the observed effect on the rate of promoter escape (Fig. 5), found in the presence of ppGpp. In the former assay, RNA synthesis was performed using the dinucleotide initiator ApU, while in the latter, initiation was via the normal mode of RNA synthesis. The ppGpp inhibitory effect being evident in the latter case but not the former, we elected to access the impact of the dinucleotide initiator on ppGpp action. The simplest interpretation of this result, as well as previous results presented here, is that ppGpp-mediated inhibition at λp Rprimarily slows the formation of the first phosphodiester bond, leading to a low rate of initial transcription resulting ultimately in a low overall rate of promoter escape. When transcription is initiated with the ApU dinucleotide, but not with CpA (corresponding to −1 and +1 positions), the slow step in initial transcription caused by ppGpp is effectively "by-passed" and leads to apparent resistance of λp R to ppGpp inhibitory effects. Likewise, no effect was observed when a nonspecific dinucleotide, ApC, was used. This new finding regarding the mode of action of ppGpp likely has implications for the regulation of other ppGpp regulated promoters like the rrn P1 promoters. It has been previously noted that intrinsically unstable rrnB P1 promoter open complexes as discussed above can be stabilized by the addition of the initiating +1 and +2 nucleotides, ATP and CTP, due apparently to dinucleotide formation (23Gourse R.L. Nucleic Acids Res. 1988; 16: 9789-9809Crossref PubMed Scopus (99) Google Scholar). Thus, it appears that the presence of the initiating dinucleotide can stabilize the binary complex of RNAP and therrnB P1 promoter, and it is the intrinsic instability of therrnB P1 open complex that is the kinetic target of ppGpp action (12Barker M.M. Gaal T. Gourse R.L. J. Mol. Biol. 2001; 305: 689-702Crossref PubMed Scopus (177) Google Scholar). Given these facts and our current observations, it is tempting to speculate that the dinucleotide initiator of therrnB P1, ApC, would likewise stabilize the rrnBP1 open complex and lead to ppGpp resistance as was observed here for λp R. If this were true, it would cast the ppGpp inhibition of the rrnB P1 promoter in vivoin a new light, since it is difficult to envision how ppGpp would lead to inhibition of rrn promoters in vivo due to high intracellular concentrations of ATP and CTP in cells that remain well over the K m for binding to RNAP under all conditions of growth (24Petersen C. Moller L.B. J. Biol. Chem. 2000; 275: 3931-3935Abstract Full Text Full Text PDF PubMed Scopus (61) Google Scholar) and thus should allow formation of dinucleotides at RNAP·rrn promoter complexes and thus stabilization of rrn open complexes. If, however, the primary effect of ppGpp were to slow the formation of the dinucleotide by kinetically inhibiting or slowing the formation of the first phosphodiester bond, this could in turn slow formation of a stable RNAP·rrnB P1 binary complex in vivo resulting in accumulation of closed promoter complexes and rapid RNAP promoter dissociation. Currently studies of ppGpp inhibition on therrnB P1 promoter in the presence of initiating dinucleotide are ongoing and will be reported elsewhere. Bacteriophage λ, as a temperate virus, after infection ofE. coli must make a decision whether to lyse the host cell or to form a lysogen (for recent reviews, see Refs. 25Taylor K. Wegrzyn G. Busby S.J.W. Thomas C.M. Brown N.L. Molecular Microbiology. Springer-Verlag, Berlin1998: 81-97Crossref Google Scholar and 26Wegrzyn G. Wegrzyn A. Baranska S. Czyz A. Recent Res. Dev. Virol. 2001; 3: 375-386Google Scholar). This is a crucial decision for the phage, which should ensure its effective propagation in bacteria growing under different conditions. Nutrition conditions are among the most important factors influencing the lysis versus lysogenization decision of λ phage. Since ppGpp synthesis reflects availability of amino acids and carbon source in cells, one might assume that this nucleotide should be involved in the regulation of phage λ development. In fact, in vivostudies revealed that ppGpp is a major factor influencing the lysisversus lysogenization decision (16Slominska M. Neubauer P. Wegrzyn G. Virology. 1999; 262: 431-441Crossref PubMed Scopus (35) Google Scholar). The highest efficiency of lysogenization was observed at ppGpp concentrations moderately increased relative to those measured in wild-type bacteria cultivated under standard laboratory conditions. It was demonstrated that ppGpp regulates activities of several λ promoters, including those involved in the lysis versus lysogenization decision (16Slominska M. Neubauer P. Wegrzyn G. Virology. 1999; 262: 431-441Crossref PubMed Scopus (35) Google Scholar). Activity of the p R promoter is crucial for effective lystic development of the phage. Therefore, the role that ppGpp-mediated inhibition at λp R plays in the context of overall λ phage physiology appears to be as an environmental "sensor" system, integral to the λ genetic regulatory network, which surveys the physiological state of the host cell at the moment of infection and appropriately dictates the lysisversus lysogeny decision. Although thep L promoter, which is not inhibited by ppGpp, is also active during the lytic development of phage λ, it also directs expression of a gene required for efficient lysogenization, namelycIII. This may be a physiological explanation of different response to ppGpp of the two λ promoters (p Rand p L). We are indepted to Dr. Mike Cashel for his advice and intellectual support of this investigation as well as providing high quality purified ppGpp for in vitrotranscription analyses.