HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 282, Issue 25, 18046-18056, June 22, 2007

This Article Abstract Full Text (PDF) All Versions of this Article: 282/25/18046    most recent M610181200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Szalewska-Palasz, A. Articles by Shingler, V. Search for Related Content PubMed PubMed Citation Articles by Szalewska-Palasz, A. Articles by Shingler, V. Social Bookmarking                      What's this?

Properties of RNA Polymerase Bypass Mutants

IMPLICATIONS FOR THE ROLE OF ppGpp AND ITS CO-FACTOR DksA IN CONTROLLING TRANSCRIPTION DEPENDENT ON ^54*

Agnieszka Szalewska-Palasz, Linda U. M. Johansson, Lisandro M. D. Bernardo¹, Eleonore Skärfstad, Ewa Stec, Kristoffer Brännström, and Victoria Shingler²

From the Department of Molecular Biology, Umeå University, SE-901 87 Umeå, Sweden and the Department of Molecular Biology, University of Gdansk, 80822 Gdansk, Poland

Received for publication, October 31, 2006 , and in revised form, April 23, 2007.

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
The bacterial nutritional and stress alarmone ppGpp and its co-factor DksA directly bind RNA polymerase to regulate its activity at certain ⁷⁰-dependent promoters. A number of promoters that are dependent on alternative -factors function poorly in the absence of ppGpp. These include the Pseudomonas-derived ⁵⁴-dependent Po promoter and several other ⁵⁴-promoters, the transcription from which is essentially abolished in Escherichia coli devoid of ppGpp and DksA. However, ppGpp and DksA have no apparent effect on reconstituted in vitro ⁵⁴-transcription, which suggests an indirect mechanism of control. Here we report analysis of five hyper-suppressor mutants within the - and '-subunits of core RNA polymerase that allow high levels of transcription from the ⁵⁴-Po promoter in the absence of ppGpp. Using in vitro transcription and competition assays, we present evidence that these core RNA polymerase mutants are defective in one or both of two properties that could combine to explain their hyper-suppressor phenotypes: (i) modulation of competitive association with -factors to favor ⁵⁴-holoenzyme formation over that with ⁷⁰, and (ii) reduced innate stability of RNA polymerase-promoter complexes, which mimics the essential effects of ppGpp and DksA for negative regulation of stringent ⁷⁰-promoters. Both these properties of the mutant holoenzymes support a recently proposed mechanism for regulation of ⁵⁴-transcription that depends on the potent negative effects of ppGpp and DksA on transcription from powerful stringent ⁷⁰-promoters, and suggests that stringent regulation is a key mechanism by which the activity of alternative -factors is controlled to meet cellular requirements.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
The -subunit of bacterial holoenzyme RNA polymerases programs the multisubunit catalytic core enzyme (₂, , ', ) to recognize and specifically initiate transcription from promoter sequences. Escherichia coli core RNA polymerase has to accommodate seven different -factors. Six of these have sequence and functional similarity and belong to the ⁷⁰-family of proteins, while a seventh, ⁵⁴, is unrelated and targets RNA polymerase to the highly conserved and unusual –24, –12 class of promoters (consensus TGGCAC-N5-TTGCa/t (1)). Initiation of transcription involves multiple steps from initial closed complex formation at the promoter, through DNA melting to form the open complex, before reaching the final stage of promoter escape (reviewed in Ref. 2). The ⁵⁴-factor imposes kinetic constraints on transcriptional initiation by ⁵⁴-holoenzyme. Unlike ⁷⁰-holoenzyme, ⁵⁴-holoenzyme cannot undergo transition from the closed complex without the aid of a mechano-transcriptional activator that makes use of nucleotide hydrolysis to remodel the closed complex, thus allowing DNA melting (reviewed in Ref. 3).

There is now considerable evidence that the different -factors compete for limiting amounts of core RNA polymerase in the cell (reviewed in Ref. 4). For ⁵⁴, this was demonstrated by the dramatic enhancement of transcription from the ⁵⁴-Po promoter upon manipulation of the relative levels of household ⁷⁰, ⁵⁴, and the stationary and stress ^S-factor (5). Recent determination of the absolute levels of core RNA polymerase and -factors suggests that competition between -factors is likely to be even more extensive than previously anticipated (6). Changes in the composition of the different holoenzymes in the cell underlie many important bacterial adaptive processes undertaken to use available resources economically and to counteract stress conditions. In addition to alternative -factors having different affinities for core RNA polymerase, the levels of many are modulated in response to specific signals, growth conditions, and/or growth phases to allow conditional selective expression of different subsets of promoters. Growth phase and growth conditions also have a major effect on the efficiency of ⁵⁴-dependent transcription (7–9). However, as is the case for the household ⁷⁰, ⁵⁴ levels are more or less constant over the growth curve and under different growth conditions, and are maintained at 16–20% of the level of ⁷⁰ (10). Hence, preferential expression of ⁵⁴-promoters in early stationary phase and under starvation/stress conditions cannot be explained by elevated ⁵⁴ levels.

In its native Pseudomonas host, the ⁵⁴-Po promoter controls transcription of an operon required for mineralization of (methyl)phenols. Transcription from Po is strictly dependent on its cognate activator DmpR, which requires an aromatic effector as well as ATP for its activity (reviewed in Ref. 11). In both E. coli and Pseudomonas putida, transcription from the Po promoter in rich media is growth-phase-regulated. The swift increase in transcription at the transition from exponential to stationary phase is dependent on the bacterial alarmone ppGpp³ and its co-factor DksA (8, 12, 13). The unusual nucleotide ppGpp is produced in response to a variety of nutritional and physicochemical stresses (through the RelA and SpoT synthetases I and II), and is the mediator of the stringent response that encompasses down-regulation of transcription from stringent ⁷⁰-promoters, such as those for rRNA operons and tRNA genes, to balance translational capacity to reduced demand (reviewed in Ref. 14). Sensitivity of RNA polymerase to ppGpp requires the non-DNA-binding protein DksA, which together with ppGpp directly mediates negative and positive effects on the performance of RNA polymerase at certain ⁷⁰-promoters (15, 16). During rapid growth, transcription from the powerful stringent ⁷⁰-rRNA promoters occupies 60–70% of the transcriptional machinery (17). Thus, ppGpp/DksA mediated down-regulation of these powerful stringent ⁷⁰-promoters is likely to have global regulatory consequences through an increased pool of core RNA polymerase available for holoenzyme formation.

Down-regulation of transcription from stringent ⁷⁰-promoters by ppGpp and DksA involves destabilization of their intrinsically very short-lived competitor-resistant open promoter complexes, which are rate-limiting for transcription (15). Support for this mechanism comes from RNA polymerase suppressor mutants that likewise destabilize competitor-resistant complexes, and from promoter mutants with increased stability that become unresponsive to ppGpp (18–21). At positively regulated ⁷⁰-promoters, such as those involved in the biosynthesis and transport of some amino acids, dissociation of the open promoter complex is not rate-limiting. In these cases, ppGpp and DksA have been proposed to directly enhance transcription by decreasing the energy of a transition state intermediate(s) to accelerate the rate-limiting formation of open complexes, and to indirectly enhance this process by the reduced use of ⁷⁰-holoenzyme at the stringent ⁷⁰-promoters (16, 20).

Simultaneous lack of both ppGpp and DksA essentially abolishes detectable ⁵⁴-dependent transcription in E. coli without having any discernable direct effect on reconstituted in vitro ⁵⁴-transcription, either in the presence or the absence of competing ⁷⁰ (13). Previous analysis has identified mutations within the rpoBC genes encoding the - and '-subunits of core RNA polymerase and within rpoD encoding ⁷⁰, that all allow transcription from the ⁵⁴-Po promoter in E. coli in the absence of ppGpp (5). In this study we identified and analyzed the performance of five hyper-suppressor mutants within core RNA polymerase on ⁵⁴- and ⁷⁰-dependent transcription in vitro to gain further insight into the global regulatory role of ppGpp and DksA for ⁵⁴-dependent transcription. The in vitro properties of these mutant polymerases expand the known mechanism(s) of bypass mutants to include effects on both -factor competition for core RNA polymerase and on the ppGpp/DksA-sensitive step of transcription from stringent promoters. These results strongly support a model in which stringent regulation of ⁷⁰-promoters is a key mechanism by which the activity of ⁵⁴ and other alternative -factors is amplified to combat conditions of nutritional deficit and other stresses.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
Bacterial Strains and Plasmids—Parental strains and plasmids used in this work are listed in Table 1. Cultures were routinely grown at 30 °C in Luria broth (LB) (22) supplemented with appropriate antibiotics (carbenicillin, 100 µg/ml, tetracycline 15 µg/ml). Resistance to rifampicin was assessed on LB plates containing either 20 or 100 µg/ml rifampicin. Prototrophy was assessed over 3 days of growth on M9 minimal medium plates supplemented with 10 mM glucose and 100 µg/ml thiamine, and graded from no growth (–) to full prototrophy (+++) exhibited by ppGpp-proficient MG1655lac. The rpoBC alleles were transferred to defined genetic background using bacteriophage P1-mediated transduction employing the closely linked thiC39::Tn10 (Tc^R) marker. DNA sequencing was used to confirm co-transduction, and the entire DNA sequence of the rpoBC region of five strains was determined as previously described (5). The ⁷⁰-P_L in vitro transcription plasmid pVI901 was constructed by PCR amplification of the –68 to +2 region as an EcoRI-BamHI fragment, and cloning between these sites of pTE103. The fidelity of the DNA sequence of the resulting plasmid was confirmed by sequencing.

Western Blot Analysis—Crude extracts of cytosolic proteins, SDS-PAGE, electrotransfer, and Western blot analysis were essentially as described previously (23). Monoclonal antibodies against E. coli RNA polymerase subunits (, ', ⁷⁰, and ⁵⁴) were from Neoclone. Affinity-purified polyclonal rabbit antibodies against the N-terminal 232 residues of DmpR were as used in Ref. 13. Antibody-decorated bands were revealed using Amersham Biosciences polyvinylidene difluoride membrane and ECL-Plus reagents as directed by the manufacturer.

Nucleotides and Proteins—Deoxynucleotides were purchased from Roche Applied Science, and [-³²P]UTP from Amersham Biosciences and ICN Biomedicals, while ppGpp was synthesized and purified as previously described (24). Wild-type and mutant E. coli ⁷⁰ were purified essentially according to Fujita and Ishihama (25) as described in Ref. 5. E. coli ⁵⁴, IHF, and DmpR-His were purified as described (26, 27). N-terminally His-tagged E. coli DksA was a kind gift from A. Åberg (Umeå University, Sweden). Wild-type and mutant core RNA polymerases were purified according to a general protocol described in Burgess and Jendrisak (28), with modifications as in Ref. 29.

In Vitro Transcription—Transcription assays (final volume, 20 µl) were performed at 37 °C in transcription buffer containing 50 mM Tris-HCl pH 7.5, 50 mM KCl, 10 mM MgCl₂, 1 mM dithiothreitol, 0.1 mM EDTA, and 0.275 mg/ml bovine serum albumin essentially as described in Ref. 5. Briefly, core RNA polymerase was preincubated with the appropriate factor for 5 min to allow holoenzyme formation. Open complex formation (20 min) was initiated by the addition of supercoiled plasmid template. For ⁵⁴-Po transcription, IHF (10 nM), DmpR-His (50 nM), ATP (4 mM), and the aromatic effector 2-methylphenol (0.5 mM) were also included. Transcription was initiated by addition of NTPs (ATP, CTP, GTP to a final concentration of 0.4 mM, UTP at 0.06 mM, and 5µCi [-³²P]UTP). In assays that involved comparison with the ⁷⁰-rrnB P1 promoter, the CTP and GTP concentrations were reduced to 0.16 mM to ensure correct initiation from this promoter (15). Multiple round transcriptions were incubated for 5 min prior to addition of heparin (final concentration of 0.1 mg/ml) to prevent reinitiation. After incubation for a further 5 min, reactions were terminated by the addition of 5 µl of stop/load buffer (150 mM EDTA, 1.05 M NaCl, 7 M urea, 10% glycerol, 0.0375% xylene cyanol, and 0.0375% bromphenol blue). For single round transcription assays, heparin was added simultaneously with NTPs, and reactions were incubated for 10 min. Transcripts were analyzed on 4.5 or 6% polyacrylamide gel containing 7 M urea and quantified by phosphorimaging.

-Factor Competition Assays—In vitro competition assays between different -subunits for wild-type or mutant core RNA polymerases were performed by preincubating ⁷⁰ with the core enzyme for 5 min prior to addition of different concentrations of ⁵⁴. After a further incubation for 10 min, multiple round transcriptions were performed as described above.

RNA Polymerase-Promoter Complex Stability Assays—The stabilities of competitor-resistant complexes on specific promoters were assessed by preforming open complexes as described for the transcription assays above, after which complexes were challenged either with heparin (final concentration 0.1 mg/ml) or a 200-fold molar excess of double-stranded DNA. Single round transcription assays were used to measure the functional open complexes at the indicated times. For all promoters analyzed, challenges with double-stranded DNA gave longer half-lives than heparin challenges. To bring the half-lives into experimentally tractable time frames, a 60-bp double-stranded DNA fragment containing the full con promoter (30) was used for the highly unstable ⁷⁰-rrnB P1 promoter complex, and a 62-bp fragment encompassing the high affinity ⁵⁴-glnAp2 promoter (13) (sequence 5'-CCACCCGGGGCAATTTAAAAGTTGGCACAGATTTCGCTTTATCTTTTTTACGGCCAAGGCC-3', –24 –12 underlined) was used with the ⁵⁴-Po promoter. Heparin was used as the competitor for the very stable ⁷⁰-P_L promoter-RNA polymerase complex.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
RpoBC Mutants Suppress the Need for ppGpp for Efficient Transcription from the ⁵⁴-Po Promoter—Mutations in rpoB and rpoC capable of restoring transcription from Po in a ppGpp⁰ strain were previously isolated on rich media using a genetic system in which Po was placed in control of a promoterless tetracycline resistance gene (5). Among 41 of these strains, 31 different rpoB and rpoC alleles were identified (Table 2). The residues targeted by the mutations are highly conserved between E. coli and P. putida (data not shown), and Thermus aquaticus (Table 2) and Thermus thermophilus for which structures have been solved (31–34). With the exception of -G373S and -E374K, which map to the tip of the -claw, all the altered residues are located on the inner surface of the active site cleft in which single amino acid substitutions can affect more than one property of the RNA polymerase (35). As indicated in Table 2, some of these alleles target residues or adjacent residues that have been previously analyzed in the context of ppGpp suppression phenotypes or other attributes of RNA polymerase function (35), whereas others (to our knowledge) have not been reported previously.

View this table: [in this window] [in a new window]   TABLE 2 rpoBC mutations isolated on the basis of restoration of 54-Po transcription

Five of the alleles (-R451C, -H551Y, -H1244L, -Q1264P, and '-L432R), like two previously analyzed -alleles (hatched bars, Fig. 1A), were found to give a hyper-suppressor phenotype resulting in >15-fold higher transcription levels than wild-type RNA polymerase in cells devoid of ppGpp. These five hyper-suppressor alleles also confer prototrophy to the otherwise auxotrophic ppGpp⁰ strain (Table 2) and they were chosen for more detailed analysis. DNA sequencing of the 10 kb rpoBC region confirmed that the mutation identified was the sole mutation present within this region of each of these five strains.

Western blot analysis of the expression levels of the RNA polymerase components in the five hyper-suppressor mutants revealed that and ⁵⁴ levels are all very similar to wild-type levels (Fig. 1B). The levels of ' and ⁷⁰ were found to be slightly higher in the ppGpp⁰ wild-type strain as compared with the ppGpp⁺ strain (Fig. 1B and Refs. 5, 38), and the levels of these proteins in the RpoBC mutant strains approximated those of the wild-type ppGpp⁺ strain. DmpR levels were found to vary somewhat between the different suppressor mutants, lying between those found in the ppGpp⁺ strain and the 2-fold lower levels found in the otherwise wild-type ppGpp⁰ strain. However, we have previously shown that production of additional DmpR in the ppGpp⁰ strain to levels that slightly exceed those in the wild-type do not influence the level of dependence on ppGpp (13). Thus, we conclude that the RpoBC suppressor phenotypes in vivo are primarily mediated through a mechanism that is independent of associated alterations in the levels of DmpR or the components of ⁵⁴-RNA polymerase.

RpoBC Mutants Also Suppress the Need for DksA for Efficient Transcription from the ⁵⁴-Po Promoter—Given the cooperativity of ppGpp and DksA in controlling both positive and negative effects at certain ⁷⁰-promoters (15, 16), we were also interested in determining the in vivo transcription from the ⁵⁴-Po promoter mediated by the mutant RpoBC proteins in the absence of DksA. To this end, the rpoBC alleles were introduced into wild-type E. coli MG1655 and isogenic ppGpp⁰ or DksA-null strains, and the activity of the luciferase transcriptional reporter plasmid pVI466 was monitored over the growth curve (Fig. 2). These strains do not have the additional lacX74 allele and, for reasons we do not currently understand, this slightly alters the fold effects of the different alleles (compare Figs. 1 and 2, circles).

All the rpoBC mutations detrimentally affected the growth rate of MG1655 and its ppGpp⁰ counterpart to a similar degree but affected the growth rates of DksA-null counterparts to different extents. Interestingly, in all cases transcription from the ⁵⁴-Po promoter was similar in the ppGpp⁺ and ppGpp⁰ counterparts (compare squares and circles in Fig. 2, B–F, noting change of scale), with the mutations resulting in 2–6-fold higher transcription from the ⁵⁴-Po promoter than in wild-type ppGpp⁺ cells. This corresponds to 20–60-fold enhanced transcription from ⁵⁴-Po by the mutant polymerases in ppGpp⁰ cells. Transcription from ⁵⁴-Po in strains lacking DksA was also enhanced by the rpoBC mutations, with transcription restored to between 0.5 and 2-fold the levels found in the wild-type ppGpp⁺ strain (triangles, Fig. 2, B–F). However, transcription from ⁵⁴-Po in the absence of DksA was found to be notably lower than that in the absence of ppGpp in all cases, suggesting a critical role for this protein in the normal functioning of the cell that is not fully compensated for by the mutations in core RNA polymerase. The essentially ppGpp-independent behavior of these mutants in vivo (Fig. 2) occurs despite the fact that the mutants remain co-responsive to ppGpp and DksA in vitro (see below). These results suggest that transcription from the ⁵⁴-Po promoter has reached a ceiling in each of these strains that is the maximal level that can be driven by each of the mutant polymerases.

View larger version (50K): [in this window] [in a new window]   FIGURE 1. Mutations within rpoBC bypass the requirement for ppGpp for efficient transcription from the 54-Po promoter. A, strains harboring the dmpR-Po-luxAB transcriptional reporter plasmid (pVI466) were grown and monitored for luciferase activity as described under "Experimental Procedures." Bars represent peak values observed after 6 h in the presence of the most potent aromatic effector of DmpR, and are given relative to the level observed in ppGpp+ MG1655lac (black bar, dashed line). Open and hatched bars indicate the transcription observed in ppGpp0 CF1693lac with the wild-type or the mutant RpoBC proteins as indicated. Open bars represent alleles originally selected on the basis of restoration of transcription from Po (Table 2), and hatched bars represent previously analyzed alleles isolated in the laboratory of M. Cashel on the basis of restoration of prototrophy (8). The results are the mean of two or three independent experiments in each case with standard deviations. The H551L strain was incorrectly assigned as H551Y in Ref. 5. B, immunoblot analysis of SDS-PAGE separated samples of 20 and 40 µg of soluble protein prepared from cells grown and harvested at 6hasin A, and analyzed using antibodies directed against the indicated proteins.

To compare ⁷⁰-dependent transcription relative to ⁵⁴-dependent transcription, and to compensate for possible differences in specific activities of different mutant core preparations, in Fig. 3E, the RpoB mutant proteins were found to mediate 1.4–1.6-fold higher transcription with ⁵⁴ than with ⁷⁰, while the RpoC mutant ('-L432R) had little effect. We also performed similar titrations with graded concentrations of DmpR in reaction mixes to detect any potential influence of the RpoBC mutations on interaction with this obligatory activator for the ⁵⁴-Po promoter. However, no differences from the wild-type were observed (data not shown).

The results described above suggest that wild-type core, which has evolved to accommodate a number of different -factors, is suboptimal for ⁵⁴-dependent transcription. Nevertheless, the 1.1–1.6-fold effects of these mutations on relative ⁵⁴/⁷⁰ transcription in vitro (Fig. 3E) appeared insufficient to account for their 20–60-fold effects observed in cells devoid of ppGpp (Fig. 2). Thus, we sought to test properties of the core RNA polymerases that could account for their dramatic phenotypes in vivo.

View larger version (40K): [in this window] [in a new window]   FIGURE 2. rpoBC alleles bypass the requirement for ppGpp and DksA. Growth profiles (closed symbols) and transcriptional profiles (open symbols) of strains harboring pVI466 (dmpR-Po-luxAB). Strains possess either wild-type (wt) alleles or the indicated rpoBC alleles in ppGpp+ DksA+ MG1655 (squares), or its otherwise isogenic ppGpp0 (CF1693, circles) and DksA-null (RK201, triangles) counterparts. The data are the average of two independent experiments with standard deviations. The data for rpoBC-wt derivatives in A have been published previously (13), and are shown for comparison. Note the change of scale for the rpoBC mutations in B–F, in which the dashed line indicates the maximal wild-type DksA+ ppGpp+ level.

View larger version (30K): [in this window] [in a new window]   FIGURE 3. Relative 54- and 70-dependent in vitro transcription by wild-type and RpoBC mutant RNA polymerases. A and C, show multiple round in vitro titrations of 10 nM core with increasing concentrations of 54 on the Po promoter (pVI695; 2 nM)(A) and 70 on the PL promoter (pVI901, 1 nM)(C). Values were normalized to wild-type core polymerase with 80 nM -factor set as 1. B and D show the corresponding relative levels of transcripts with 80 nM of the appropriate -factor and 10 nM core RNA polymerase with the indicated mutations as listed in E. E, ratio of 54 to 70 transcripts with the indicated core RNA polymerases as determined from B and D. Data are the average of three independent experiments with standard errors.

View larger version (28K): [in this window] [in a new window]   FIGURE 4. In vitro competition between 70 and 54 for wild-type and mutant core RNA polymerases. A, multiple round in vitro competition assay with 10 nM core RNA polymerase preincubated with 20 nM 70 and then challenged with increasing concentrations of 54 in the presence of 11 nM pVI695 (54-Po) and 11 nM pVI901 (70-PL). For each RNA polymerase, 100% 70-PL transcription corresponds to that observed in the absence of 54, while 100% 54-Po transcription corresponds to that observed with 10 nM core and 80 nM 54 in the absence of 70. The inset shows the specificity of the transcripts from 54-Po and 70-PL for their respective holoenzymes. Mixtures contained the indicated -factor(s) and other necessary components for both 54- and 70-dependent transcription. B, open bars are the percent 54-Po transcripts found under the conditions described for A, with wild-type or mutant core RNA polymerases preincubated with 70 and challenged with 160 nM 54. The hatched bar shows the results from an equivalent experiment with wild-type core, which was preincubated with a mutant 70-40Y protein [RpoD-DSA (536–538)]. Data represent the mean ± S.D. from two independent experiments.

The RpoBC Mutant Enzymes are Co-responsive to ppGpp and DksA—None of the residues altered in the five RpoBC mutants are directly involved in the binding of ppGpp close to the active site, in either of its two orientations observed in co-crystals of T. thermophilus RNA polymerase and ppGpp (33). DksA and ppGpp have little or no effect on reconstituted in vitro transcription from ⁵⁴-Po under conditions that recapitulate known positive and negative effects at ⁷⁰-dependent promoters (Fig. 5A and Ref. 13). Similarly, in vitro transcript levels from the ⁵⁴-Po promoter with the RpoBC mutants in the absence or presence of ppGpp (200 µM) and DksA (200 nM) are the same (data not shown). Thus, to assess the responses of the mutant RNA polymerases to ppGpp and DksA, we took advantage of their negative co-regulation of transcription from the stringent ⁷⁰-rrnB P1 promoter (Fig. 5A and Ref. 15).

View larger version (45K): [in this window] [in a new window]   FIGURE 5. In vitro transcription in the presence and absence of ppGpp and DksA. A, multiple round in vitro transcription with wild-type core RNA polymerase at 5 nM, 40 nM -factor, 1 nM DNA template and increasing concentrations of ppGpp in the presence of 200 nM DksA. 54-dependent transcription from the Po promoter was monitored using pVI695, while 70-dependent transcription from the rrnB P1 promoter was monitored using pRLG6214. Transcription in the absence of ppGpp has been set to 1 in each case. B, single round in vitro 70-transcription with wild-type and mutant holoenzymes (10 nM core and 80 nM 70) on 5 nM of the rrnB P1 DNA template pRLG6214. Transcription with the wild-type enzyme has been set as 1. C, multiple round in vitro transcription using the indicated wild-type or mutant core enzyme at 10 nM, 80 nM 70, 5 nM pRLG6214, and either DksA storage buffer (no additions), ppGpp (200 µM) and/or DksA (200 nM). Transcription in the absence of any addition was set as 100% for each core enzyme. Data in A and B are the average of two independent experiments with standard deviation, data in C are the average of three to five independent experiments with standard errors.

RpoBC Mutants Are Innately Defective in the Stability of Their Competitor-resistant Complexes—The molecular mechanism(s) by which ppGpp and DksA directly affect transcriptional initiation from ⁷⁰-promoters has not been fully resolved. However, both of these regulatory molecules reduce the lifetime of competitor-resistant RNA polymerase complexes at all ⁷⁰-promoters so far tested. The lifetimes of the competitor-resistant promoter complexes are critical for regulation of stringent ⁷⁰-promoters such as the rrnB P1 promoter, since this step is rate-limiting for transcription. Given the apparent critical role of competitor-resistant complexes for transcription from stringent promoters, we first determined the lifetimes of complexes in the presence and absence of ppGpp and/or DksA at the three promoters employed in this study. In these experiments, preformed RNA polymerase-promoter complexes were challenged with either competitor DNA for ⁵⁴-Po (Fig. 6A) and ⁷⁰-rrnB P1 (Fig. 6B), or heparin for ⁷⁰-P_L (Fig. 6C), and the levels of functional complexes were followed over time. Consistent with previous findings, ppGpp and DksA were found to cooperate in reducing the lifetime of competitor-resistant complexes at both the ⁷⁰-dependent rrnB P1 and P_L promoters (Fig. 6, B and C). Likewise, ppGpp and DksA were found to cooperate in reducing the lifetime of competitor-resistant complexes at the ⁵⁴-Po promoter (Fig. 6A). This latter result contrasts with previous data in which the cooperative effects of ppGpp and DksA were not detectable if heparin was used as the competitor (13). It should be emphasized, however, that reduced open complex stability will only have regulatory consequences on transcription from negatively regulated promoters that have open complex stability as the rate-limiting step, and that ppGpp and DksA have little if any direct influence on transcription from the ⁵⁴-Po promoter (Fig. 5A and Ref. 13).

View larger version (32K): [in this window] [in a new window]   FIGURE 6. The effects of ppGpp and DksA on promoter half-lives with wild-type or mutant RNA polymerases. The lifetimes of competitor-resistant complexes generated with 10 nM core RNA polymerase and 80 nM -factor. A–C, time courses in the presence of no additions (DksA storage buffer, squares), 200 µM ppGpp (triangles), 200 nM DksA (circles), or both (open diamonds), are shown with the values found at 20 s after competitor addition in the absence of ppGpp and DksA set as 100%. Templates used: A, 54-Po (2 nM pVI695, DNA competitor); B, 70-rrnB P1 (7 nM pRLG6214, DNA competitor); C, 70-PL, (11 nM pVI901, and heparin as competitor); note difference in time scale for the three promoters. D–F, assays were performed as for A–C using wild-type or mutant core enzymes in the absence of ppGpp and DksA. Only data at a single time point (as indicated on the Y-axes) after addition of the competitor is shown as bars. Bars 1–6 in E and F are as indicated in D. Dashed lines indicate the corresponding level of the wild-type holoenzyme in the presence of ppGpp and DksA. Data are the average of three to five independent experiments with standard errors.

Similar to the data for competitive association with -factors, there is no apparent direct correlation between the innate stabilities of the mutant RNA polymerase promoter complexes on rrnB P1 (Fig. 6E) and their in vivo phenotypes (Fig. 2). However, it is notable that the two mutants most affected in open complex stability on rrnB P1 ('-L432R and -R451C) are the ones least affected in interaction with -factors, while -Q1264P that has no apparent effect on rrnB P1 open complex stability is among the three mutants most affected in their interaction with -factors (Figs. 6E and 4). The remaining two mutants (-H551Y, -H1244L) show marked defects in both properties. As considered in greater detail in the discussion, reduced intrinsic lifetimes at the powerful stringent promoters and defects in -factor competition for limiting core, would likely combine to facilitate ⁵⁴-holoenzyme formation and thus ⁵⁴-dependent transcription in vivo.

⁷⁰ Mutants Reduce the Lifetime of RNA Polymerase-rrnB P1 Promoter Complexes—In addition to the RpoBC mutants discussed above, four ⁷⁰ mutants have previously been analyzed with respect to their ability to bypass the need for ppGpp and DksA for efficient ⁵⁴-Po transcription in vivo (5, 13). Two of these mutants, ^70-35D [RpoD-Y571H] and ^70-40Y [RpoD-DSA (536–538)], were isolated in the same genetic screen as the RpoBC mutants reported here (5), while two others [^70-P504L and ^70-S506F] were isolated on the basis of their ability to restore prototrophy to ppGpp⁰ E. coli (38). Since holoenzyme RNA polymerases utilizing ⁷⁰ or ⁵⁴ do not show any promoter cross-recognition, the defects of these ⁷⁰ mutants in competing against ⁵⁴ for limiting core RNA polymerase have been the focus of attention. In the light of our finding with the RpoBC mutants described above, we tested the effects of the four ⁷⁰ mutants on the stabilities of competitor-resistant complexes at the stringent ⁷⁰-rrnB P1 promoter. As shown in Fig. 7, three of the mutants were found to exhibit markedly diminished lifetimes (with that for ^70-40Y being shorter than for ^70-P504L and ^70-S506F), while the lifetime for the fourth, ^70-35D, was similar to wild-type. The hierarchy of defects in complex stability at the rrnB P1 promoter is the same as their hierarchy of defects in competing against ⁵⁴ for core RNA polymerase, namely ^70-40Y > ^70-P504L and ^70-S506F >^70-35D. This hierarchy also reflects their suppressor phenotype in a ppGpp⁰ strain, with ^70-40Y restoring ⁵⁴-Po transcription to 1.5-fold that of the ⁷⁰ wild-type ppGpp⁺ strain, ^70-P504L and ^70-S506F to approximately wild-type levels, with ^70-35D having only very minor effects (5). Thus, like some of the RpoBC mutants, the ⁷⁰ suppressor mutants have defects in both (i) competitive association of core with -factors and (ii) stability of the ppGpp/DksA-susceptible RNA polymerase complexes at the stringent ⁷⁰-rrnB P1 promoter.

View larger version (25K): [in this window] [in a new window]   FIGURE 7. The half-lives of rrnB P1 complexes with mutant 70-holoenzymes. Experiments were performed as in Fig. 6B except that wild-type core was used with different forms of 70: wild-type (squares), 70-35D [RpoD-Y571H] (triangles), 70-P504L (inverted-triangles), 70-S506F (diamonds), or 70-40Y [RpoD-DSA (536–538)] (circles). The dashed line shows the time course for the transcript from the 70-PRNA1 promoter that is also present on pRLG6214. Data are the mean of two independent experiments with standard deviations.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Interaction of core RNA polymerase with -factors involves a very extensive interface including interactions of the three globular domains on the surface of the polymerase as well as with the extensive 3.2-loop of the -factors that penetrates deeply into the active site cleft (32). Thus, active site cleft mutants have the potential to directly or indirectly alter association with -factors and alter other steps in the transcriptional initiation pathway to affect their in vivo phenotypes. Our analysis of the relative performance of mutant holoenzymes in transcription from promoters that are independent of ppGpp and DksA in vitro (⁵⁴-Po and the ⁷⁰-P_L and P_RNA1 promoters) has shown that they only mediate 1.1–1.6-fold enhancement of ⁵⁴-transcription relative to that of ⁷⁰ in vitro (Fig. 3), which is insufficient to account for their in vivo phenotypes. However, we found that the and ' mutants are altered in one or both of two properties, namely, in their interaction with -factors to favor association with ⁵⁴ under competition for limiting core RNA polymerase (Fig. 4), and reduced innate stabilities of competitor-resistant promoter complexes (Fig. 6). This latter property is critical for ppGpp- and DksA-mediated down-regulation of stringent ⁷⁰-promoters, but would have no regulatory consequences on transcription from promoters in which this is not the rate-limiting step. Unlike ⁷⁰ mutants, with which the magnitude of their in vivo bypass phenotypes follows that of their defects in associating with core RNA polymerase and in open complex stability on rrnB P1 (Fig. 7, Ref. 5), the and ' mutants are differentially affected in these two properties. Two of the mutants ('-L432R and -R451C) are primarily defective in open complex stability, one of the mutants is only altered in -factor interaction (-Q1264P), while two are markedly altered in both properties (-H551Y, -H1244L) (Figs. 4 and 6). How these effects on RNA polymerase function can combine to cause their in vivo hyper-suppressor phenotypes is discussed below in the context of a recently proposed model for indirect (passive) control of ⁵⁴-dependent transcription that operates through the regulation of ⁷⁰-stringent promoters (13).

The passive model for control of ⁵⁴-dependent transcription operates through predicted global regulatory consequences of the negative action of ppGpp and DksA at the seven powerful stringent ⁷⁰-rRNA operon promoters of E. coli. During intermediate-to-rapid growth in rich media, the activity of these promoters sequester 60–70% of the transcriptional machinery (17). Thus, under these conditions in which ppGpp levels are low, much of the core RNA polymerase is occupied in production of the abundant transcripts from these multiple powerful promoters, leaving little free core available for association with the constant levels of ⁵⁴ and ⁷⁰. This in turn would lead to low levels of ⁵⁴-holoenzyme, and consequent low promoter occupancy and output from cognate ⁵⁴-promoters. However, under slow growth and stress conditions that elicit high levels of ppGpp, the potent down-regulation of transcription from the ⁷⁰-rRNA operon promoters in response to ppGpp and DksA would lead to higher levels of core being available for association with -factors. As a result, ⁵⁴-holoenzyme levels would increase, leading to enhanced ⁵⁴-promoter occupancy and output. Within this model, both core and ⁷⁰-suppressor mutants would bypass the requirement for ppGpp in two ways: 1) by directly reducing transcription from the powerful stringent ⁷⁰-rRNA operon promoters to elevate the levels of core available for association with other -factors, and 2) by altering core to -factor association to favor the formation of ⁵⁴-holoenzyme over that of ⁷⁰-holoenzyme. It is worth noting that ⁷⁰ mutants disfavor interaction between ⁷⁰ and the core enzyme. Core mutants, on the other hand, could either enhance association with ⁵⁴ and/or disfavor association with ⁷⁰ and probably also other ⁷⁰-family alternative -factors. In either case, this would lead to a competitive advantage of ⁵⁴ over all other -factors for core RNA polymerase that may contribute to the hyper-suppressor phenotype of core mutants. In addition, in the case of the core RNA polymerase mutants, these effects on holoenzyme levels would further amplify the enhanced output from the ⁵⁴-Po promoter that is observed in vitro (Fig. 3) with these mutant holoenzymes. Although not specifically tested, these properties of the mutant polymerases would alter translational capacity within the cell, which probably underlies the reduced growth rates observed in the mutant strains (Fig. 2).

The passive model for the action of ppGpp and DksA as outlined above would predict that low affinity promoters that have ⁵⁴-holoenzyme binding as a rate-limiting step would be more susceptible to loss of these regulatory molecules than their high-affinity counterparts. Consistent with this prediction, we have previously shown that lack of ppGpp and DksA have greater effect on low affinity ⁵⁴-promoters than on high affinity promoters (13). Recent determinations of absolute levels of core RNA polymerase and -factors have clarified that -factor numbers far exceed those of core, and it follows that conditions within the cell are such that competition for limiting core will most likely be severe (6). The mathematical modeling of equilibrium binding performed in that study also highlights the role of nonspecific DNA binding by RNA polymerase in -factor competition and redistribution of RNA polymerase, to have effects that would have greatest impact on low affinity promoters. All the mutant holoenzymes tested in the present study are affected in their DNA binding properties as assessed by open complex stability assays. Hence, although purely speculative, it is plausible that binding of ppGpp and DksA (and mutant RNA polymerases that mimic their effects) also alters nonspecific DNA binding. Such alterations might contribute to in vivo suppression phenotypes of mutant RNA polymerases.

The mechanism outlined above for ppGpp/DksA-mediated passive regulation of ⁵⁴-dependent transcription is also pertinent to the functioning of other alternative -factors. In addition to ⁵⁴, DksA and/or ppGpp have been shown to be important for transcription from promoters dependent on ^S, ^H, and ^E that control processes critical for stress survival (39, 40). In the case of ^S, mutant RNA polymerases that mimic the effects of ppGpp have also been shown to bypass the requirement for ppGpp in vivo by a mechanism that is independent of ppGpp effects on ^S levels (39, 41). Although the direct effects of ppGpp in synergy with DksA have not been tested at these ^S-, ^H-, and ^E-dependent promoters, it appears likely that passive regulation of the levels of the cognate holoenzymes would at least in part contribute to their regulation and thus to the pleiotropic roles of ppGpp and DksA in combating nutritional deprivation and other stress conditions.

	FOOTNOTES

 
^* This work was supported in part by grants from the Swedish Foundation for Strategic Research and the Swedish Research Council (to V. S.) and the Polish Ministry of Education and Science (Grant 2 PO4A 034 28, to A. S.-P.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

¹ Recipient of a student fellowship from the Foundation for Science and Technology (FCT, Portugal).

² To whom correspondence should be addressed. Tel.: 46-90-785-2534; Fax: 46-90-771420; E-mail: victoria.shingler{at}molbiol.umu.se.

³ The abbreviations used are: ppGpp, guanosine bispyrophosphate; Cb, carbenicillin; Cm, chloramphenicol; Km, kanamycin; Tc, tetracycline; AA, amino acid; T. Aq, T. aquaticus; OPC, open complex; iNTP, initiating nucleotide; wt, wild type.