Interactions between the 2.4 and 4.2 regions of S, the stress-specific
http://www.100md.com
《核酸研究医学期刊》
Laboratoire de Microbiologie et Génétique Moléculaire, UMR5100 CNRS–Université Toulouse III, 118, Route de Narbonne, 31062, Toulouse Cedex, France and 1 Unité des Régulations Transcriptionnelles, Institut Pasteur, 25 rue du Dr Roux, 75724 Paris cedex 15, France
*To whom correspondence should be addressed. Tel: +33 5 61 33 58 72; Fax: +33 5 61 33 58 86; Email: clg@ibcg.biotoul.fr
Present address:
Patricia Bordes, Department of Biological Sciences, Sir Alexander Fleming Building, Imperial College London, South Kensington Campus, London SW7 2AZ, UK
ABSTRACT
The s subunit of Escherichia coli RNA polymerase holoenzyme (ES) is a key factor of gene expression upon entry into stationary phase and in stressful conditions. The selectivity of promoter recognition by ES and the housekeeping E70 is as yet not clearly understood. We used a genetic approach to investigate the interaction of S with its target promoters. Starting with down-promoter variants of a S promoter target, osmEp, altered in the –10 or –35 elements, we isolated mutant forms of S suppressing the promoter defects. The activity of these suppressors on variants of osmEp and ficp, another target of S, indicated that S is able to interact with the same key features within a promoter sequence as 70. Indeed, (i) S can recognize the –35 element of some but not all its target promoters, through interactions with its 4.2 region; and (ii) amino acids within the 2.4 region participate in the recognition of the –10 element. More specifically, residues Q152 and E155 contribute to the strong preference of S for a C in position –13 and residue R299 can interact with the –31 nucleotide in the –35 element of the target promoters.
INTRODUCTION
The subunit of bacterial RNA polymerase (RNAP) is necessary for promoter recognition and transcription initiation. In Escherichia coli there are seven different subunits, which can associate with a single core RNAP. Switches in the use of factors allow the specific regulation of subsets of genes . Usually, each factor recognizes specific promoter sequences and, as a consequence, the different regulons they control do not overlap. However, this is not true for S, the rpoS-encoded master regulator of the transcriptional response to the entry into stationary phase and stress conditions (2–6). Strikingly, several promoters are recognized in vitro and in vivo by both ES and E70, the RNAP holoenzymes containing S and 70, respectively (4,7,8). Furthermore, the S- and 70-dependent promoters share almost identical optimal sequences (9). However, despite all these similarities, many genes of E.coli are specifically expressed under the control of S in vivo. Sequence comparison and genetic data established that nucleotides at positions –13/–14 in the promoters are important determinants of the selectivity of recognition by ES and E70 (8–11), but these nucleotides are not sufficient to account for the differential recognition of promoters by ES and E70 and the basis of factor selectivity is not completely understood.
Genetic studies have identified several regions in 70 necessary to initiate transcription by playing roles in RNAP holoenzyme assembly, promoter recognition or DNA opening. In particular, domains named 2.4 and 4.2 are involved in recognition of the –10 (TATAAT) and –35 (TTGACA) elements of the promoters, respectively (12–14). Moreover, the 2.5 region was shown to contact the TGX motif found 5' of the –10 element in promoters harboring a so-called extended –10 (TGXTATAAT) (15). These DNA recognition regions are particularly well conserved between S and 70 . Only partial structural information is available for E.coli 70 (17), but more complete data were reported recently for the RNAP of Thermus thermophilus (18) and Thermus aquaticus (19,20). In particular, crystal structures were determined for a fragment of A of T.aquaticus (the homolog of E.coli 70) complexed with a –35-mimicking DNA fragment and for RNAP holoenzyme complexed with an open promoter-mimicking fragment (19,20). These data confirmed the interactions of regions 2.4, 2.5 (renamed 3.0, from structural considerations) and 4.2 with –10, extended –10 and –35 elements of promoters, respectively. As yet, no structural information is available for S.
Figure 1. Domain organization of 70 and S. (A) Schematic of domain organization of 70 and S. The position of the two fragments of S subjected to mutagenesis is indicated. (B) Alignment of domains 2.4 of 70 and S. The region 2.4 of 70 is boxed. (C) Alignment of domains 4.2 of 70 and S. The region 4.2 of 70 is boxed.
We used the well-characterized promoter osmEp as a model to study promoter discrimination by ES. We demonstrated previously that osmEp is transcribed by both ES and E70, but with differential efficiencies (8,21). Transcription from osmEp during exponential phase is 70 dependent, whereas the induction of osmEp upon entry into stationary phase is controlled by S (8,21,22). Down-promoter mutations of osmEp have been isolated, and most of them reduced similarly the efficiency of transcription by ES and E70 (8). One exception was osmEp13T, changing the C at position –13 of osmEp –10 (CCAGGCT; Fig. 2) into a T. This mutation affected the transcription by ES more than by E70, and work with other promoters demonstrated that a C at this position is conserved within S-dependent promoters (9,10).
Figure 2. Wild-type and mutant promoters used in this work. Sequences of the wild-type osmEp and ficp are shown. The mutations used in this work are indicated above the sequence. –10 and –35 elements of the promoters are underlined. Nucleotides in bold correspond to the optimal sequence for ES (9).
In the present work, we used a genetic approach to investigate the mechanisms of promoter recognition by ES. We present evidence that mutations changing the amino acids Gln152 (Q152) and Glu155 (E155) of S improve specifically the recognition of osmEp variants altered in the –13 position within the –10 element, whilst modifications of Arg299 (R299) improve recognition of a promoter altered in the –35 element. When compared with results obtained with 70, these data show that both forms of holoenzyme use amino acids at similar positions to interact with DNA and that S can use contacts with the –35 region to recognize certain promoters. However, the effects of rpoS mutations on another target, ficp, suggest that ES can also recognize promoters through different types of interactions.
MATERIALS AND METHODS
Bacterial strains and culture conditions
The bacterial strains used in this study, all derived from E.coli K-12, are listed in Table 1. The plasmid pBADrpoS (a gift from Y.N.Zhou and S.Gottesman) is derived from the pBAD24 vector (23), and contains the rpoS open reading frame (ORF) cloned under the control of araBp. Cells were grown aerobically at 37°C in LB medium (24). MacConkey solid medium was supplemented with 1% lactose (Difco Laboratory). Ampicillin, kanamycin and tetracycline were used at concentrations of 100, 40 and 10 μg/ml, respectively.
Table 1. Bacterial strains
Genetic procedures
EcoRI DNA fragments carrying the mutated osmEp13A, osmEp13G and ficp12C promoters were constructed by two-step PCR amplification using overlapping mutagenic primers. After cleavage with EcoRI, these fragments were cloned in the recombination vector pOM41 (25). The resulting plasmids were introduced in strain pop3125 (26) and used to insert the corresponding promoters in front of the (malP-lac) fusion of pop3125 by homologous recombination, as described previously (27). rpoS-359::Tn10 or rpoS-359::Tek mutations were introduced by transduction with P1 stocks grown on strains RH90 or CLG141, respectively. Tek is a derivative of Tn10 with a kanamycin resistance cassette inserted into the tetA locus (28).
?-Galactosidase assays
?-Galactosidase activities were assayed as described by Miller (24), on cells in early stationary phase. The data shown are the average of at least two independent cultures, each measured in triplicate. The overall variation was <15%.
DNA manipulations
Isolation of DNA, digestion with restriction enzymes, ligation with T4 DNA ligase and transformation were carried out as described (29,30). Mutagenesis of rpoS was performed according to Diaz et al. (31). pBADrpoS DNA (20 ng) was lyophilized and incubated in 100 μl of 250 mM sodium acetate pH 4.3, 1 M sodium nitrite. Aliquots (25 μl) were removed after 30, 45, 60 and 90 s and the DNA was precipitated, washed and resuspended in 30 μl of H2O. A 5 ng aliquot of treated DNA was mixed with oligonucleotide couples rpoS1/rpoS2 (5'-GACTCAGCTTTACCTTGG-3'/5'-GAATCACC ACCCAGCGG-3') or rpoS3/rpoS4 (5'-CTTAACGAGCGCA TTACC-3'/5'-AATCTTCTCTCATCCGCC-3') and used for amplification of DNA fragments with Hot Tub DNA polymerase (Amersham Pharmacia Biotech) according to the manufacturer’s protocol. After cleavage with EagI + AccI or AccI + HindIII, the mutated rpoS fragments were substituted to the original fragment on pBADrpoS.
His6-tagged wild-type S and Q152P, R299G, R299H or R299S variants were constructed by PCR amplification using pBADrpoS as the template, and the oligonucleotides HisRpoS1 (5'-GGGAATTCACCATGCATCACCATCACC ATCACAGTCAGAATACGCTGGAAA-3') and HisRpoS2 (5'-CGACGCGCAAAATAAACTTC-3'). The amplified fragments were digested by EcoRI and EagI and introduced on the pBADrpoS* plasmids carrying the different variants of rpoS.
S purification and RNAP holoenzyme reconstitution
Strain CLG585 was transformed with pBAD-6HisrpoS, and ?-galactosidase assays demonstrated that the presence of the His6 tag had no effect on the activity of S. CLG585/pBAD-6HisrpoS was grown in Luria broth until OD600 of 0.6, and the production of tagged S was induced by addition of 0.02% arabinose. After 1 h at 30°C, the cells were harvested by centrifugation, and the resulting cell pellet stored at –80°C. The purification was performed with the QIAexpressionist kit (Qiagen, Chatsworth, CA). As a final step, elutes were concentrated by precipitation with ammonium sulfate and dialyzed against storage buffer . Analysis on electrophoresis gels indicated that the S preparations were 80% pure. Reconstitution of active holoenzymes was achieved by incubating 1 vol. of 5 μM core enzyme (Epicentre? Technologies) with 2 vols of S, S(Q152P), S(R299G), S(R299H) or S(R299S) at 10 μM for 20 min at 37°C (/core, approximately 4). The reconstituted holoenzymes were then diluted at room temperature in transcription buffer prior to their use for in vitro transcription experiments.
Single round in vitro transcription
Supercoiled plasmid templates (pJCD01osmEp+, pJCD01osmEp13T, pJCD01osmEp31C or pJCD01osmEp32G prepared from an overnight culture of a wild-type strain) were used for in vitro transcription assays as described previously (8,32). Transcripts were quantified on dried electrophoresis gels using a BAS-2000 PhosphorImager (Fuji) and TINA software (version 2.09).
RESULTS
Mutagenesis of rpoS
Random mutagenesis of rpoS was performed by PCR amplification in mutagenic conditions of two different fragments of the gene: either an internal part (codons 72–222) or its 3' end (codons 222–330). These two fragments encompass the entire regions 2 and 4 of S, respectively (Fig. 1A). The mutagenized DNA fragments were ligated into plasmid pBADrpoS, a vector that expresses S under the control of the araB promoter. After growth in rich medium without arabinose, this plasmid is able to produce S at a level roughly one-third of that found in a wild-type strain (33). Two independent mutant libraries were constructed by introducing the mutagenized internal or 3' fragment of rpoS in pBADrpoS.
Isolation of rpoS mutations suppressing down-promoter mutations of osmEp
The strain CLG585 carries a transcriptional lac fusion expressed under the control of the wild-type osmE promoter of E.coli (8). CLG591 and CLG619 carry the same lac fusion but expressed under the control of two down-promoter variants of osmEp: osmEp32G and osmEp13T, respectively . In addition, these strains carry an rpoS::Tek insertion that inactivates the chromosomal copy of rpoS (Table 1). When transformed with pBADrpoS+, CLG585 gave red colonies (Lac+) on MacConkey + lactose indicator plates. Substituting the cytosine at position –13 or the adenine at position –32 in osmEp resulted in both CLG591 and CLG619 transformed with pBADrpoS+ exhibiting white colonies (Lac–) on MacConkey + lactose. We used this phenotype to isolate mutations in rpoS able to suppress the osmE down-promoter mutations.
In a first set of experiments, the two rpoS mutant libraries were transformed into strain CLG619 (osmEp13T) and plated on MacConkey + lactose + ampicillin agar. Among approximately 50 000 white clones, 20 red colonies were identified. The plasmids isolated from these clones (pBADrpoS*) were transformed again in CLG619 and all gave the expected Lac+ phenotype. Sequence analysis showed that 16 plasmids contained single substitutions, which affected one of only three codons in rpoS, coding for Q152, E155 or R299 (Table 2). As shown in Figure 1, Q152 and E155 are in region 2.4 (involved in –10 element recognition in 70), and R299 is in region 4.2 (involved in –35 element recognition in 70). These amino acids correspond respectively to Q437, T440 and R584 in 70 (Fig. 1), three residues already identified as crucial for transcription initiation by 70 (13,14). The last four plasmids were not further studied because they contained two substitutions, including one affecting codons 152, 155 or 299 already isolated as a single substitution (Table 2).
Table 2. Isolation of rpoS mutants
In a second set of experiments, the two rpoS mutant libraries were transformed into strain CLG591 (osmEp32G) and plated on MacConkey + lactose + ampicillin plates. Among approximately 20 000 white clones from each library, eight red colonies were identified. All derived from the mutagenesis of the C-terminal fragment of rpoS and carried the same substitution at codon 299 of rpoS, changing Arg299 into a histidine (Table 2).
Western blot analysis showed that all the mutants isolated in the two sets of experiments contained very similar levels of S (data not shown), demonstrating that the phenotypic suppression of down-promoter mutations observed here was not a trivial consequence of higher amounts of the variants of S within the cells.
Substitutions at amino acids Q152 and E155 suppress preferentially mutations at the –13 position of osmEp
Plasmids pBADrpoS* carrying the mutations affecting Q152 and E155 (region 2.4 of S) were introduced into strains carrying different variants of osmEp (Fig. 2). ?-Galactosidase activity was measured after growth to early stationary phase in Luria broth in the absence of arabinose (Fig. 3A and B). Transcription of osmEp+ was not significantly affected by substitutions of Q152 or E155 in S. These substitutions were not able to restore the activity of promoter variants affected in the –35 element (osmEp30G, osmEp32G, osmEp34A and osmEp35C) or at position –7 in the –10 element (osmEp7C). In contrast, transcription of osmEp mutants carrying the modifications of position –13 was equivalent to that of the wild-type promoter with all the variants of Q152 and E155. The increase in activity obtained with osmEp13G was somehow smaller (between 2- and 3-fold) than with osmEp13A or osmEp13T (4- to 5-fold). However, we note that the activity measured here reflects the sum of ES- and E70-driven transcription and that a G in position –13 is the preferred nucleotide for transcription by E70 . Therefore, a smaller increase ratio for osmEp13G is probably due to a higher background of E70-driven transcription. Altogether, these results indicated that substitutions affecting Q152 and E155 suppress preferentially defects at position –13 in the –10 element of osmEp.
Figure 3. Effect of rpoS mutations affecting Q152 and E155 on transcription of wild-type or mutant osmE promoter in vivo. Strains carrying transcriptional lac fusions expressed under the control of the different variants of osmEp were transformed with the indicated derivatives of pBADrpoS. Overnight cultures of each strain were diluted 200-fold and cultures were grown aerobically for 8 h at 37°C in LB medium + ampicillin before ?-galactosidase activity was measured.
Substitution Q152P exerts similar effects in vitro and in vivo
An N-terminal polyhistidine tag was inserted in both the wild-type and Q152P variant of S, allowing their purification on Ni columns. After reconstitution of RNAP holoenzyme, we performed in vitro transcription experiments with supercoiled DNA of plasmids pJCD01-osmEp+ or pJCD01-osmEp13T . As shown in Figure 4, both forms of holoenzyme produced RNA I, a non-coding RNA transcribed from a promoter on the plasmid vector that harbors a C in position –13 (11). In addition, both forms of holoenzyme transcribed RNA species of 150 nt that are initiated at osmEp (8). Mutation osmEp13T resulted in a 2-fold decrease of the relative amount of osmE RNA versus RNA I, and the Q152P variant of S suppressed this defect, in agreement with the effect observed in vivo (Fig. 3). Therefore, the suppressor effect is probably due solely to an improved interaction between RNAP and the promoter.
Figure 4. Effect of Q152P modification of S on transcription of osmEp+ and osmEp13T in vitro. Supercoiled pJCD01osmEp+ or pJCD01osmEp13T were transcribed in vitro with ES or ES(Q152P). For each template, reactions were performed in triplicate with 45, 30 and 15 nM final concentration of RNAP from left to right, respectively. Autoradiography of a fixed and dried electrophoresis gel of the products is shown. The osmE' mRNA (150 nt) and vector-derived RNA I (107–108 nucleotides) are indicated. The amount of the two RNA species was quantitated with a Phosphor-Imager and the ratio osmE' RNA/RNA I for each lane is indicated in the histogram under the gel.
Substitutions at position R299 suppress preferentially a mutation at position –32 of osmEp in vivo
During our genetic screen, substitutions at position R299 were isolated in both sets of experiments, suppressing mutations of nucleotide –32 or –13. Suppression of a mutation of nucleotide –13 was unexpected, as R299 is located in domain 4.2, known to interact with the –35 element. When the four variants of R299 (R299C, R299G, R299S and R299H) were introduced into strains carrying different osmE mutant promoters, we observed that these mutations were able to increase transcription from wild-type osmEp 2- to 3-fold (Fig. 5). With promoter variants altered at positions –7, –13, –30, –34 or –35, we observed a similar increase in transcription, ranging from 2- to 5-fold. Therefore, transcription initiation at the osmE promoter seems globally more efficient when the residue R299 is modified, and this is probably why we could isolate such variants from an osmEp13T strain. However, when the different substitutions of R299 were introduced in a strain carrying osmEp32G, the increase in transcription was much stronger (from 9- to 18-fold, depending on the rpoS allele; Fig. 5). This allele-specific effect strongly suggests an interaction between the –35 element of the osmE promoter and the 4.2 region of S, which contains R299.
Figure 5. Effect of rpoS mutations affecting R299 on transcription of wild-type or mutant osmE promoters in vivo. Strains carrying transcriptional lac fusions expressed under the control of the different variants of osmEp were transformed with the indicated derivatives of pBADrpoS. Over night cultures of each strain were diluted 200-fold and cultures were grown aerobically for 8 h at 37°C in LB medium + ampicillin before ?-galactosidase activity was measured.
The effect of substitutions at R299 depends on the sequence in the –35 element
His6-tagged variants of S carrying modifications of R299 were used for in vitro transcription experiments with supercoiled DNA of plasmids pJCD01-osmEp+, pJCD01-osmEp13T, pJCD01-osmEp31C or pJCD01-osmEp32G. As compared with osmEp+, the mutation osmEp32G reduced 3-fold the in vitro transcription efficiency by wild-type S (Fig. 6A). Although the stimulation ratio was less pronounced than in vivo (Fig. 5), we observed that variants of S modified at R299 partly suppressed the down-promoter effect of osmEp32G in vitro (compare Fig. 6A with B, C and D). In contrast, the effect of R299 substitutions was very different when tested on osmEp31C, an up-promoter variant of osmEp (8) harboring a –35 element identical to the optimal sequence common for both E70 and ES (9). With wild-type S, osmEp31C behaved as an up-promoter mutation (4-fold stimulation, see Fig. 6A). However, when transcribed by holoenzymes reconstituted with R299 variants of S, osmEp31C showed equivalent or even slightly lower efficiency than osmEp+ (Fig. 6B, C and D). Thus, S exhibits a preference for a CG base pair at position –31 in osmEp, and this preference is abolished when R299 is changed into a histidine, a serine or a glycine.
Figure 6. Effect of modifications of R299 of S on transcription of osmEp wild-type and mutant derivatives. Supercoiled pJCD01osmEp+, pJCD01osmEp13T, pJCD01osmEp31C and pJCD01osmEp32G were transcribed in vitro with the indicated holoenzyme (45 nM final concentration). Autoradiography of a fixed and dried electrophoresis gel of the products is shown. osmE' mRNA and RNA I were quantitated with a PhosphorImager and the relative ratio osmE' RNA/RNA I (with the ratio obtained with pJCD01osmEp+ set as 1 for each form of holoenzyme) is shown in the histograms under the gels.
Effect of mutations in rpoS on transcription from ficp
The results observed with the osmE promoter suggest that S and 70 interact with this particular target in a similar manner, using domains 2.4 and 4.2 to recognize the –10 and –35 elements, respectively. To investigate whether these conclusions could be extended to other S-dependent promoters, we tested the effect of all the S variants obtained in this work on transcription from ficp, a promoter that exhibits a different organization. Indeed, ficp is completely dependent on S for its transcription (4,34) but, in contrast to osmEp, it has a very poor –35 element and a –10 element much closer to the consensus but without a C in position –12 (note that position –12 in ficp corresponds to position –13 in osmEp, Fig. 2). We first introduced a C in position –12 of ficp and, as shown in Figure 7, this resulted in a 2.5-fold increase in expression of a transcriptional fusion. This expression was abolished in an rpoS::Tn10 background, confirming that it is completely S-dependent (not shown). The variants of S modified at position 299 did not stimulate transcription from ficp. On the contrary, they resulted in an 2-fold reduction in expression of the ficp–lac fusion. Furthermore, the stimulation of expression due to the ficp12C mutation was still observed with all the variants of R299 of S. Therefore, these data indicate that R299 is not playing the same role in the recognition of ficp and of osmEp.
Figure 7. Effect of the rpoS variants on transcription in vivo of the ficp+ and ficp12C promoters. Strains carrying transcriptional ficp+–lac or ficp12C–lac fusions were transformed with the indicated derivatives of pBADrpoS. Overnight cultures of each strain were diluted 200-fold and cultures were grown aerobically for 8 h at 37°C in LB medium + ampicillin before ?-galactosidase activity was measured.
Substitutions of Q152 in S were not observed to stimulate transcription from ficp. However, with the exception of Q152H, they abolish the preference for a C 5' of the –10 element, because the expression from wild-type ficp and ficp12C was equivalent. Similarly, modifications of E155 also abolished the preference for a C in –12 of ficp. Therefore, these data indicate that the 2.4 region of S interacts with the –10 region of ficp and, furthermore, they suggest that the amino acids Q152 and E155 are important for the specific recognition of the CG base pair 5' of the –10 element.
DISCUSSION
In this work, we used a genetic suppressor approach with variants of the osmE promoter altered either at position –13 (osmEp13T) or in the –35 element (osmEp32G), to investigate the mechanisms of promoter recognition by the stress-specific factor S of E.coli. We were able to isolate mutations in rpoS able to suppress both promoter defects and, strikingly, they affected only three codons of rpoS (Table 2). Because changes at each of the three positions appeared several times, it is likely that changes at these three codons are the only possible single modifications that fulfill the criteria of the screen. A BLAST search in the sequenced genomes of Gram-negative bacteria (http://www.ncbi.nlm.nih.gov/blast/Blast.cgi) allowed us to identify 21 clear homologs of E.coli S (not shown). All these homologs carry Q, E and R at the positions corresponding to Q152, E155 and R299 of S, emphasizing the importance of these residues for the functioning of S. We will discuss the role of these amino acids in promoter recognition in light of the available genetic and structural data.
Does S interact with –35 elements at its target promoters?
The –35 region is not essential in the strictly S-dependent fic promoter (34), and biochemical analysis indicated that S interacts only weakly with the –35 region of several promoters (35). Taken together with the poor sequence conservation in the –35 region of S targets (10,11,36), these observations suggested that a peculiarity of ES might be to recognize its targets without a strong and specific binding to the –35 element. Using synthetic promoters, Gaal et al. (9) demonstrated that a consensus –35 element increases the binding of ES and the efficiency of the promoter, suggesting that a –35 element can play a prominent role in the recognition of at least some promoters by ES. The data presented here demonstrate that an interaction with the –35 element is indeed important in vivo for the recognition of osmEp by ES. Therefore, S seems able to recognize promoters through different mechanisms, which do or do not involve strong interactions with a –35 element. Because most S-dependent promoters have a very poor –35 element (36), the case of osmEp may seem to be an exception. However, other S-dependent promoters having a good –35 element are known , and the identification of new targets of S may reveal additional members in this class of promoters. Interestingly, it is worth noting that this situation is not so different from that of 70 which can also recognize promoters without a –35 element, provided that they carry an extended –10 element (15,38).
How does S interact with the –35 elements at osmEp?
R299 of S is the homolog of R584 of 70 and R409 of A from T.aquaticus (Fig. 1). Genetic evidence suggested that R584 of 70 interacts with the CG base pair (TTGACA) in the –35 element (12,13), and structural data demonstrated that R409 of A donates two hydrogen bonds to the O6 and N5 of the guanine in the major groove (19). Wild-type S also exhibits a preference for a CG bp at position –31 (9), and we show here that R299 is responsible for this preference at the osmE promoter (Fig. 6). Therefore, our data strongly suggest that R299 of S can make direct, base-specific interactions with the guanine of the CG in position –31 of its target promoters. Such contacts are not possible between S and osmEp+ that harbors an AT base pair in position –31 (TTGAAA; Fig. 2). Furthermore, we have shown previously that this deviation is more deleterious for interaction with E70 than with ES and thereby contributes to the selectivity of recognition of osmEp (8). The structural data (structure coordinates 1KU7 in the Protein Data Bank) suggest that the replacement of the consensus G by a T on the template strand would place the methyl group on C5 of thymine in conflict with the arginine residues R584 of 70 or R299 of S. The four suppressor mutations isolated here substitute R299 by short side chain amino acids, and that should remove the conflict with –31T. We believe that the 2-fold stimulation of transcription from the wild-type and from most variants of osmEp (Fig. 5) can be explained by the elimination of this negative interaction. However, why the substitutions of R299 specifically suppress the effect of osmEp32G is not readily explained by the available structural data. Indeed, no base-specific contacts have been identified with the nucleotides in position –32 of the –35 element (19). One possibility is that such contacts do exist, but only transiently during the kinetic pathway of promoter recognition, and that they could not be seen in the crystal structures. Alternatively, the effect of osmEp32G could be due to a modification of the DNA helical structure and/or bending that could reinforce the negative effect of the repulsion between –31T and R299 of S, which would then explain the strong suppressor effect of the substitutions of R299 on osmEp32G. This hypothesis would be in agreement with the ability of suppressing osmEp32G by substituting R299 with four different amino acids that only have in common that they carry short side chains. An important role for the DNA structure would also be consistent with our previous observation that the efficiency of transcription of osmEp is modulated by supercoiling density (22).
Role of the 2.4 region of S in the recognition of the –10 element of promoters
Q152 and E155 of S are the homologs of Q437 and T440 of 70 and Q260 and N263 of A from T.aquaticus, respectively (Fig. 1). In the crystal structure of T.aquaticus RNAP holoenzyme complexed with an open promoter-mimicking DNA fragment, both Q260 and N263 are exposed on the same face of an amphipathic -helix and point to nucleotides in the major groove near –12 (20). Works based on suppressor analyses identified changes in these residues and led to the model that Q437 and T440 of 70 could interact directly with the base pair at position –12 (13,14). However, this conclusion has been challenged by both biochemical (39) and structural (20) data, and these residues may contribute to the recognition of nucleotides in –12 without direct contact but, for instance, via the repositioning of a nearby essential residue. In any case, it is striking that genetic screens for 70 suppressors of promoters mutated at position –12 or S suppressors of a promoter mutated at position –13 identified modifications affecting the same two residues. Recently, the analysis of a collection of rpoS mutations identified Q152 but not E155 as essential for the functioning of S (40). The same work concluded that 70 and S have a common global organization but several crucial amino acids differ between the two factors, suggesting that they may differ in the details of the promoter recognition process. Our work demonstrates that E155 is one of those residues important for promoter recognition that differ between 70 and S. Notably, the mutation leading to Q152R, that we identified here as a suppressor of osmEp13T (Table 2, Fig. 3), was found to abolish transcription from several S-dependent promoters (40). The origin of this difference is not known, but once again it may highlight the fact that different promoters could be recognized via different interactions, even within the –10 element, and it also emphasizes the importance of investigating the recognition of a variety of targets in order to fully understand the differential recognition of promoters by ES and E70.
Role of Q152 and E155 in S preference for a C at the –13 position of its target promoters
Previous work, also based on allele-specific suppressors, indicated that the amino acid K173 (corresponding to E458 in 70 and E281 in A of T.aquaticus) has a discriminatory role for the nucleotide at position –13 of S-dependent promoters (10). Our data show that Q152 and E155 are also important for this distinctive property of S, because modifications at both residues abolish the preference for a C at position –13 of osmEp (Fig. 3) and –12 of ficp (Fig. 6). Structural data in the Protein Data Bank] show that in A, Q260 and N263 on the one hand and E281 on the other hand belong to two -helices that form a clip facing the major groove in front of the positions –12/–13 on the template strand. Therefore, these residues are likely to participate in a structure involved in a direct interaction with the nucleotide in position –13 of promoters, in agreement with the genetic data.
ACKNOWLEDGMENTS
We are grateful to F. Norel for the gift of anti-S antibodies, Y.N. Zhou and S. Gottesman for pBADrpoS and A.J. Carpousis for language improvements. Part of this work was supported by grants from the French Ministère de l’Enseignement Supérieur et de la Recherche (Programme de Recherche Fondamentale en Microbiologie, Maladies Infectieuses et Parasitaires) and from the Génop?le of Toulouse to C.G.
REFERENCES
Ishihama,A. (2000) Functional modulation of Escherichia coli RNA polymerase. Annu. Rev. Microbiol., 54, 499–518.
Mulvey,M.R. and Loewen,P.C. (1989) Nucleotide sequence of katF of Escherichia coli suggests KatF protein is a novel sigma transcription factor. Nucleic Acids Res., 17, 9979–9991.
Lange,R. and Hengge-Aronis,R. (1991) Identification of a central regulator of stationary-phase gene expression in Escherichia coli. Mol. Microbiol., 5, 49–59.
Tanaka,K., Takayanagi,Y., Fujita,N., Ishihama,A. and Takahashi,H. (1993) Heterogeneity of the principal factor in Escherichia coli: the rpoS gene product, 38, is a second principal factor of RNA polymerase in stationary-phase Escherichia coli. Proc. Natl Acad. Sci. USA, 90, 3511–3515.
Loewen,P.C. and Hengge-Aronis,R. (1994) The role of the sigma factor S (KatF) in bacterial global regulation. Annu. Rev. Microbiol., 48, 53–80.
Hengge-Aronis,R. (1996) Regulation of gene expression during entry into stationary phase. In Neidhardt,F.C. (ed.), Escherichia coli and Salmonella: Cellular and Molecular Biology. American Society for Microbiology Press, Washington, DC, pp. 1497–1512.
Altuvia,S., Almiron,M., Huisman,G., Kolter,R. and Storz,G. (1994) The dps promoter is activated by OxyR during growth and by IHF and S in stationary phase. Mol. Microbiol., 13, 265–272.
Bordes,P., Repoila,F., Kolb,A. and Gutierrez,C. (2000) Involvement of differential efficiency of transcription by Es and E70 RNA polymerase holoenzymes in growth phase regulation of the Escherichia coli osmE promoter. Mol. Microbiol., 35, 845–853.
Gaal,T., Ross,W., Estrem,S.T., Nguyen,L.H., Burgess,R.R. and Gourse,R.L. (2001) Promoter recognition and discrimination by ES RNA polymerase. Mol. Microbiol., 42, 939–954.
Becker,G. and Hengge-Aronis,R. (2001) What makes an Escherichia coli promoter S dependent? Role of the –13/–14 nucleotide promoter positions and region 2.5 of S. Mol. Microbiol., 39, 1153–1165.
Espinosa-Urgel,M., Chamizo,C. and Tormo,A. (1996) A consensus structure for S-dependent promoters. Mol. Microbiol., 21, 657–659.
Gardella,T., Moyle,H. and Susskind,M.M. (1989) A mutant Escherichia coli 70 subunit of RNA polymerase with altered promoter specificity. J. Mol. Biol., 206, 579–590.
Siegele,D.A., Hu,J.C., Walter,W.A. and Gross,C.A. (1989) Altered promoter recognition by mutant forms of the 70 subunit of Escherichia coli RNA polymerase. J. Mol. Biol., 206, 591–603.
Waldburger,C., Gardella,T., Wong,R. and Susskind,M.M. (1990) Changes in conserved region 2 of Escherichia coli 70 affecting promoter recognition. J. Mol. Biol., 215, 267–276.
Barne,K.A., Bown,J.A., Busby,S.J. and Minchin,S.D. (1997) Region 2.5 of the Escherichia coli RNA polymerase 70 subunit is responsible for the recognition of the ‘extended-10’ motif at promoters. EMBO J., 16, 4034–4040.
Lonetto,M., Gribskov,M. and Gross,C.A. (1992) The 70 family: sequence conservation and evolutionary relationships. J. Bacteriol., 174, 3843–3849.
Malhotra,A., Severinova,E. and Darst,S.A. (1996) Crystal structure of a 70 subunit fragment from E.coli RNA polymerase. Cell, 87, 127–136.
Vassylyev,D.G., Sekine,S., Laptenko,O., Lee,J., Vassylyeva,M.N., Borukhov,S. and Yokoyama,S. (2002) Crystal structure of a bacterial RNA polymerase holoenzyme at 2.6 ? resolution. Nature, 417, 712–719.
Campbell,E.A., Muzzin,O., Chlenov,M., Sun,J.L., Olson,C.A., Weinman,O., Trester-Zedlitz,M.L. and Darst,S.A. (2002) Structure of the bacterial RNA polymerase promoter specificity subunit. Mol. Cell, 9, 527–539.
Murakami,K.S., Masuda,S., Campbell,E.A., Muzzin,O. and Darst,S.A. (2002) Structural basis of transcription initiation: an RNA polymerase holoenzyme–DNA complex. Science, 296, 1285–1290.
Conter,A., Menchon,C. and Gutierrez,C. (1997) Role of DNA supercoiling and RpoS sigma factor in the osmotic and growth phase-dependent induction of the gene osmE of Escherichia coli K12. J. Mol. Biol., 273, 75–83.
Bordes,P., Conter,A., Morales,V., Bouvier,J., Kolb,A. and Gutierrez,C. (2003) DNA supercoiling contributes to disconnect S accumulation from S-dependent transcription in Escherichia coli. Mol. Microbiol., 48, 561–571.
Guzman,L.M., Belin,D., Carson,M.J. and Beckwith,J. (1995) Tight regulation, modulation and high-level expression by vectors containing the arabinose PBAD promoter. J. Bacteriol., 177, 4121–4130.
Miller,J.H. (1992) A Short Course in Molecular Genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
Vidal-Ingigliardi,D. and Raibaud,O. (1985) A convenient technique to compare the efficiency of promoters in Escherichia coli. Nucleic Acids Res., 13, 5919–5926.
Debarbouillé,M., Shuman,H.A., Silhavy,T.J. and Schwartz,M. (1978) Dominant constitutive mutations in malT, the positive regulator gene of the maltose regulon in Escherichia coli. J. Mol. Biol., 124, 359–371.
Gutierrez,C. and Devedjian,J.C. (1991) Osmotic induction of gene osmC expression in Escherichia coli K12. J. Mol. Biol., 220, 959–973.
Fran?ois,V., Louarn,J., Patte,J. and Louarn,J.-M. (1987) A system for in vivo selection of genomic rearrangements with predetermined endpoints in Escherichia coli using modified Tn10 transposons. Gene, 56, 99–108.
Sambrook,J., Fritsch,E.F. and Maniatis,T. (1989) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
Silhavy,T.J., Berman,M. and Enquist,L. (1984) Experiments with Gene Fusions. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
Diaz,J.J., Rhoads,D.D. and Roufa,D.J. (1991) PCR-mediated chemical mutagenesis of cloned duplex DNAs. Biotechniques, 11, 204–211.
Bordes,P., Bouvier,J., Conter,A., Kolb,A. and Gutierrez,C. (2002) Transient repressor effect of Fis on the growth phase-regulated osmE promoter of Escherichia coli K12. Mol. Gen. Genet., 268, 206–213.
Becker,G., Klauck,E. and Hengge-Aronis,R. (1999) Regulation of RpoS proteolysis in Escherichia coli: the response regulator RssB is a recognition factor that interacts with the turnover element in RpoS. Proc. Natl Acad. Sci. USA, 96, 6439–6444.
Hiratsu,K., Shinagawa,H. and Makino,K. (1995) Mode of promoter recognition by the Escherichia coli RNA polymerase holoenzyme containing the S subunit: identification of the recognition sequence of the fic promoter. Mol. Microbiol., 18, 841–850.
Colland,F., Fujita,N., Kotlarz,D., Bown,J.A., Meares,C.F., Ishihama,A. and Kolb,A. (1999) Positioning of S, the stationary phase factor, in Escherichia coli RNA polymerase–promoter open complexes. EMBO J., 18, 4049–4059.
Lee,S.J. and Gralla,J.D. (2001) Sigma38 (rpoS) RNA polymerase promoter engagement via –10 region nucleotides. J. Biol. Chem., 276, 30064–30071.
Lacour,S., Kolb,A., Boris Zehnder,A.J. and Landini,P. (2002) Mechanism of specific recognition of the aidB promoter by S-RNA polymerase. Biochem. Biophys. Res. Commun., 292, 922–930.
Kumar,A., Malloch,R.A., Fujita,N., Smillie,D.A., Ishihama,A. and Hayward,R.S. (1993) The minus 35-recognition region of Escherichia coli 70 is inessential for initiation of transcription at an ‘extended minus 10’ promoter. J. Mol. Biol., 232, 406–418.
Fenton,M.S., Lee,S.J. and Gralla,J.D. (2000) Escherichia coli promoter opening and –10 recognition: mutational analysis of 70. EMBO J., 19, 1130–1137.
Lee,S.J. and Gralla,J.D. (2002) Promoter use by 38 (rpoS) RNA polymerase. Amino acid clusters for DNA binding and isomerization. J. Biol. Chem., 277, 47420–47427.
Casadaban,M.J. (1976) Transposition and fusion of the lac genes to selected promoters in Escherichia coli using bacteriophage lambda and Mu. J. Mol. Biol., 104, 541–555.(Claire Checroun, Patricia Bordes, Olivie)
*To whom correspondence should be addressed. Tel: +33 5 61 33 58 72; Fax: +33 5 61 33 58 86; Email: clg@ibcg.biotoul.fr
Present address:
Patricia Bordes, Department of Biological Sciences, Sir Alexander Fleming Building, Imperial College London, South Kensington Campus, London SW7 2AZ, UK
ABSTRACT
The s subunit of Escherichia coli RNA polymerase holoenzyme (ES) is a key factor of gene expression upon entry into stationary phase and in stressful conditions. The selectivity of promoter recognition by ES and the housekeeping E70 is as yet not clearly understood. We used a genetic approach to investigate the interaction of S with its target promoters. Starting with down-promoter variants of a S promoter target, osmEp, altered in the –10 or –35 elements, we isolated mutant forms of S suppressing the promoter defects. The activity of these suppressors on variants of osmEp and ficp, another target of S, indicated that S is able to interact with the same key features within a promoter sequence as 70. Indeed, (i) S can recognize the –35 element of some but not all its target promoters, through interactions with its 4.2 region; and (ii) amino acids within the 2.4 region participate in the recognition of the –10 element. More specifically, residues Q152 and E155 contribute to the strong preference of S for a C in position –13 and residue R299 can interact with the –31 nucleotide in the –35 element of the target promoters.
INTRODUCTION
The subunit of bacterial RNA polymerase (RNAP) is necessary for promoter recognition and transcription initiation. In Escherichia coli there are seven different subunits, which can associate with a single core RNAP. Switches in the use of factors allow the specific regulation of subsets of genes . Usually, each factor recognizes specific promoter sequences and, as a consequence, the different regulons they control do not overlap. However, this is not true for S, the rpoS-encoded master regulator of the transcriptional response to the entry into stationary phase and stress conditions (2–6). Strikingly, several promoters are recognized in vitro and in vivo by both ES and E70, the RNAP holoenzymes containing S and 70, respectively (4,7,8). Furthermore, the S- and 70-dependent promoters share almost identical optimal sequences (9). However, despite all these similarities, many genes of E.coli are specifically expressed under the control of S in vivo. Sequence comparison and genetic data established that nucleotides at positions –13/–14 in the promoters are important determinants of the selectivity of recognition by ES and E70 (8–11), but these nucleotides are not sufficient to account for the differential recognition of promoters by ES and E70 and the basis of factor selectivity is not completely understood.
Genetic studies have identified several regions in 70 necessary to initiate transcription by playing roles in RNAP holoenzyme assembly, promoter recognition or DNA opening. In particular, domains named 2.4 and 4.2 are involved in recognition of the –10 (TATAAT) and –35 (TTGACA) elements of the promoters, respectively (12–14). Moreover, the 2.5 region was shown to contact the TGX motif found 5' of the –10 element in promoters harboring a so-called extended –10 (TGXTATAAT) (15). These DNA recognition regions are particularly well conserved between S and 70 . Only partial structural information is available for E.coli 70 (17), but more complete data were reported recently for the RNAP of Thermus thermophilus (18) and Thermus aquaticus (19,20). In particular, crystal structures were determined for a fragment of A of T.aquaticus (the homolog of E.coli 70) complexed with a –35-mimicking DNA fragment and for RNAP holoenzyme complexed with an open promoter-mimicking fragment (19,20). These data confirmed the interactions of regions 2.4, 2.5 (renamed 3.0, from structural considerations) and 4.2 with –10, extended –10 and –35 elements of promoters, respectively. As yet, no structural information is available for S.
Figure 1. Domain organization of 70 and S. (A) Schematic of domain organization of 70 and S. The position of the two fragments of S subjected to mutagenesis is indicated. (B) Alignment of domains 2.4 of 70 and S. The region 2.4 of 70 is boxed. (C) Alignment of domains 4.2 of 70 and S. The region 4.2 of 70 is boxed.
We used the well-characterized promoter osmEp as a model to study promoter discrimination by ES. We demonstrated previously that osmEp is transcribed by both ES and E70, but with differential efficiencies (8,21). Transcription from osmEp during exponential phase is 70 dependent, whereas the induction of osmEp upon entry into stationary phase is controlled by S (8,21,22). Down-promoter mutations of osmEp have been isolated, and most of them reduced similarly the efficiency of transcription by ES and E70 (8). One exception was osmEp13T, changing the C at position –13 of osmEp –10 (CCAGGCT; Fig. 2) into a T. This mutation affected the transcription by ES more than by E70, and work with other promoters demonstrated that a C at this position is conserved within S-dependent promoters (9,10).
Figure 2. Wild-type and mutant promoters used in this work. Sequences of the wild-type osmEp and ficp are shown. The mutations used in this work are indicated above the sequence. –10 and –35 elements of the promoters are underlined. Nucleotides in bold correspond to the optimal sequence for ES (9).
In the present work, we used a genetic approach to investigate the mechanisms of promoter recognition by ES. We present evidence that mutations changing the amino acids Gln152 (Q152) and Glu155 (E155) of S improve specifically the recognition of osmEp variants altered in the –13 position within the –10 element, whilst modifications of Arg299 (R299) improve recognition of a promoter altered in the –35 element. When compared with results obtained with 70, these data show that both forms of holoenzyme use amino acids at similar positions to interact with DNA and that S can use contacts with the –35 region to recognize certain promoters. However, the effects of rpoS mutations on another target, ficp, suggest that ES can also recognize promoters through different types of interactions.
MATERIALS AND METHODS
Bacterial strains and culture conditions
The bacterial strains used in this study, all derived from E.coli K-12, are listed in Table 1. The plasmid pBADrpoS (a gift from Y.N.Zhou and S.Gottesman) is derived from the pBAD24 vector (23), and contains the rpoS open reading frame (ORF) cloned under the control of araBp. Cells were grown aerobically at 37°C in LB medium (24). MacConkey solid medium was supplemented with 1% lactose (Difco Laboratory). Ampicillin, kanamycin and tetracycline were used at concentrations of 100, 40 and 10 μg/ml, respectively.
Table 1. Bacterial strains
Genetic procedures
EcoRI DNA fragments carrying the mutated osmEp13A, osmEp13G and ficp12C promoters were constructed by two-step PCR amplification using overlapping mutagenic primers. After cleavage with EcoRI, these fragments were cloned in the recombination vector pOM41 (25). The resulting plasmids were introduced in strain pop3125 (26) and used to insert the corresponding promoters in front of the (malP-lac) fusion of pop3125 by homologous recombination, as described previously (27). rpoS-359::Tn10 or rpoS-359::Tek mutations were introduced by transduction with P1 stocks grown on strains RH90 or CLG141, respectively. Tek is a derivative of Tn10 with a kanamycin resistance cassette inserted into the tetA locus (28).
?-Galactosidase assays
?-Galactosidase activities were assayed as described by Miller (24), on cells in early stationary phase. The data shown are the average of at least two independent cultures, each measured in triplicate. The overall variation was <15%.
DNA manipulations
Isolation of DNA, digestion with restriction enzymes, ligation with T4 DNA ligase and transformation were carried out as described (29,30). Mutagenesis of rpoS was performed according to Diaz et al. (31). pBADrpoS DNA (20 ng) was lyophilized and incubated in 100 μl of 250 mM sodium acetate pH 4.3, 1 M sodium nitrite. Aliquots (25 μl) were removed after 30, 45, 60 and 90 s and the DNA was precipitated, washed and resuspended in 30 μl of H2O. A 5 ng aliquot of treated DNA was mixed with oligonucleotide couples rpoS1/rpoS2 (5'-GACTCAGCTTTACCTTGG-3'/5'-GAATCACC ACCCAGCGG-3') or rpoS3/rpoS4 (5'-CTTAACGAGCGCA TTACC-3'/5'-AATCTTCTCTCATCCGCC-3') and used for amplification of DNA fragments with Hot Tub DNA polymerase (Amersham Pharmacia Biotech) according to the manufacturer’s protocol. After cleavage with EagI + AccI or AccI + HindIII, the mutated rpoS fragments were substituted to the original fragment on pBADrpoS.
His6-tagged wild-type S and Q152P, R299G, R299H or R299S variants were constructed by PCR amplification using pBADrpoS as the template, and the oligonucleotides HisRpoS1 (5'-GGGAATTCACCATGCATCACCATCACC ATCACAGTCAGAATACGCTGGAAA-3') and HisRpoS2 (5'-CGACGCGCAAAATAAACTTC-3'). The amplified fragments were digested by EcoRI and EagI and introduced on the pBADrpoS* plasmids carrying the different variants of rpoS.
S purification and RNAP holoenzyme reconstitution
Strain CLG585 was transformed with pBAD-6HisrpoS, and ?-galactosidase assays demonstrated that the presence of the His6 tag had no effect on the activity of S. CLG585/pBAD-6HisrpoS was grown in Luria broth until OD600 of 0.6, and the production of tagged S was induced by addition of 0.02% arabinose. After 1 h at 30°C, the cells were harvested by centrifugation, and the resulting cell pellet stored at –80°C. The purification was performed with the QIAexpressionist kit (Qiagen, Chatsworth, CA). As a final step, elutes were concentrated by precipitation with ammonium sulfate and dialyzed against storage buffer . Analysis on electrophoresis gels indicated that the S preparations were 80% pure. Reconstitution of active holoenzymes was achieved by incubating 1 vol. of 5 μM core enzyme (Epicentre? Technologies) with 2 vols of S, S(Q152P), S(R299G), S(R299H) or S(R299S) at 10 μM for 20 min at 37°C (/core, approximately 4). The reconstituted holoenzymes were then diluted at room temperature in transcription buffer prior to their use for in vitro transcription experiments.
Single round in vitro transcription
Supercoiled plasmid templates (pJCD01osmEp+, pJCD01osmEp13T, pJCD01osmEp31C or pJCD01osmEp32G prepared from an overnight culture of a wild-type strain) were used for in vitro transcription assays as described previously (8,32). Transcripts were quantified on dried electrophoresis gels using a BAS-2000 PhosphorImager (Fuji) and TINA software (version 2.09).
RESULTS
Mutagenesis of rpoS
Random mutagenesis of rpoS was performed by PCR amplification in mutagenic conditions of two different fragments of the gene: either an internal part (codons 72–222) or its 3' end (codons 222–330). These two fragments encompass the entire regions 2 and 4 of S, respectively (Fig. 1A). The mutagenized DNA fragments were ligated into plasmid pBADrpoS, a vector that expresses S under the control of the araB promoter. After growth in rich medium without arabinose, this plasmid is able to produce S at a level roughly one-third of that found in a wild-type strain (33). Two independent mutant libraries were constructed by introducing the mutagenized internal or 3' fragment of rpoS in pBADrpoS.
Isolation of rpoS mutations suppressing down-promoter mutations of osmEp
The strain CLG585 carries a transcriptional lac fusion expressed under the control of the wild-type osmE promoter of E.coli (8). CLG591 and CLG619 carry the same lac fusion but expressed under the control of two down-promoter variants of osmEp: osmEp32G and osmEp13T, respectively . In addition, these strains carry an rpoS::Tek insertion that inactivates the chromosomal copy of rpoS (Table 1). When transformed with pBADrpoS+, CLG585 gave red colonies (Lac+) on MacConkey + lactose indicator plates. Substituting the cytosine at position –13 or the adenine at position –32 in osmEp resulted in both CLG591 and CLG619 transformed with pBADrpoS+ exhibiting white colonies (Lac–) on MacConkey + lactose. We used this phenotype to isolate mutations in rpoS able to suppress the osmE down-promoter mutations.
In a first set of experiments, the two rpoS mutant libraries were transformed into strain CLG619 (osmEp13T) and plated on MacConkey + lactose + ampicillin agar. Among approximately 50 000 white clones, 20 red colonies were identified. The plasmids isolated from these clones (pBADrpoS*) were transformed again in CLG619 and all gave the expected Lac+ phenotype. Sequence analysis showed that 16 plasmids contained single substitutions, which affected one of only three codons in rpoS, coding for Q152, E155 or R299 (Table 2). As shown in Figure 1, Q152 and E155 are in region 2.4 (involved in –10 element recognition in 70), and R299 is in region 4.2 (involved in –35 element recognition in 70). These amino acids correspond respectively to Q437, T440 and R584 in 70 (Fig. 1), three residues already identified as crucial for transcription initiation by 70 (13,14). The last four plasmids were not further studied because they contained two substitutions, including one affecting codons 152, 155 or 299 already isolated as a single substitution (Table 2).
Table 2. Isolation of rpoS mutants
In a second set of experiments, the two rpoS mutant libraries were transformed into strain CLG591 (osmEp32G) and plated on MacConkey + lactose + ampicillin plates. Among approximately 20 000 white clones from each library, eight red colonies were identified. All derived from the mutagenesis of the C-terminal fragment of rpoS and carried the same substitution at codon 299 of rpoS, changing Arg299 into a histidine (Table 2).
Western blot analysis showed that all the mutants isolated in the two sets of experiments contained very similar levels of S (data not shown), demonstrating that the phenotypic suppression of down-promoter mutations observed here was not a trivial consequence of higher amounts of the variants of S within the cells.
Substitutions at amino acids Q152 and E155 suppress preferentially mutations at the –13 position of osmEp
Plasmids pBADrpoS* carrying the mutations affecting Q152 and E155 (region 2.4 of S) were introduced into strains carrying different variants of osmEp (Fig. 2). ?-Galactosidase activity was measured after growth to early stationary phase in Luria broth in the absence of arabinose (Fig. 3A and B). Transcription of osmEp+ was not significantly affected by substitutions of Q152 or E155 in S. These substitutions were not able to restore the activity of promoter variants affected in the –35 element (osmEp30G, osmEp32G, osmEp34A and osmEp35C) or at position –7 in the –10 element (osmEp7C). In contrast, transcription of osmEp mutants carrying the modifications of position –13 was equivalent to that of the wild-type promoter with all the variants of Q152 and E155. The increase in activity obtained with osmEp13G was somehow smaller (between 2- and 3-fold) than with osmEp13A or osmEp13T (4- to 5-fold). However, we note that the activity measured here reflects the sum of ES- and E70-driven transcription and that a G in position –13 is the preferred nucleotide for transcription by E70 . Therefore, a smaller increase ratio for osmEp13G is probably due to a higher background of E70-driven transcription. Altogether, these results indicated that substitutions affecting Q152 and E155 suppress preferentially defects at position –13 in the –10 element of osmEp.
Figure 3. Effect of rpoS mutations affecting Q152 and E155 on transcription of wild-type or mutant osmE promoter in vivo. Strains carrying transcriptional lac fusions expressed under the control of the different variants of osmEp were transformed with the indicated derivatives of pBADrpoS. Overnight cultures of each strain were diluted 200-fold and cultures were grown aerobically for 8 h at 37°C in LB medium + ampicillin before ?-galactosidase activity was measured.
Substitution Q152P exerts similar effects in vitro and in vivo
An N-terminal polyhistidine tag was inserted in both the wild-type and Q152P variant of S, allowing their purification on Ni columns. After reconstitution of RNAP holoenzyme, we performed in vitro transcription experiments with supercoiled DNA of plasmids pJCD01-osmEp+ or pJCD01-osmEp13T . As shown in Figure 4, both forms of holoenzyme produced RNA I, a non-coding RNA transcribed from a promoter on the plasmid vector that harbors a C in position –13 (11). In addition, both forms of holoenzyme transcribed RNA species of 150 nt that are initiated at osmEp (8). Mutation osmEp13T resulted in a 2-fold decrease of the relative amount of osmE RNA versus RNA I, and the Q152P variant of S suppressed this defect, in agreement with the effect observed in vivo (Fig. 3). Therefore, the suppressor effect is probably due solely to an improved interaction between RNAP and the promoter.
Figure 4. Effect of Q152P modification of S on transcription of osmEp+ and osmEp13T in vitro. Supercoiled pJCD01osmEp+ or pJCD01osmEp13T were transcribed in vitro with ES or ES(Q152P). For each template, reactions were performed in triplicate with 45, 30 and 15 nM final concentration of RNAP from left to right, respectively. Autoradiography of a fixed and dried electrophoresis gel of the products is shown. The osmE' mRNA (150 nt) and vector-derived RNA I (107–108 nucleotides) are indicated. The amount of the two RNA species was quantitated with a Phosphor-Imager and the ratio osmE' RNA/RNA I for each lane is indicated in the histogram under the gel.
Substitutions at position R299 suppress preferentially a mutation at position –32 of osmEp in vivo
During our genetic screen, substitutions at position R299 were isolated in both sets of experiments, suppressing mutations of nucleotide –32 or –13. Suppression of a mutation of nucleotide –13 was unexpected, as R299 is located in domain 4.2, known to interact with the –35 element. When the four variants of R299 (R299C, R299G, R299S and R299H) were introduced into strains carrying different osmE mutant promoters, we observed that these mutations were able to increase transcription from wild-type osmEp 2- to 3-fold (Fig. 5). With promoter variants altered at positions –7, –13, –30, –34 or –35, we observed a similar increase in transcription, ranging from 2- to 5-fold. Therefore, transcription initiation at the osmE promoter seems globally more efficient when the residue R299 is modified, and this is probably why we could isolate such variants from an osmEp13T strain. However, when the different substitutions of R299 were introduced in a strain carrying osmEp32G, the increase in transcription was much stronger (from 9- to 18-fold, depending on the rpoS allele; Fig. 5). This allele-specific effect strongly suggests an interaction between the –35 element of the osmE promoter and the 4.2 region of S, which contains R299.
Figure 5. Effect of rpoS mutations affecting R299 on transcription of wild-type or mutant osmE promoters in vivo. Strains carrying transcriptional lac fusions expressed under the control of the different variants of osmEp were transformed with the indicated derivatives of pBADrpoS. Over night cultures of each strain were diluted 200-fold and cultures were grown aerobically for 8 h at 37°C in LB medium + ampicillin before ?-galactosidase activity was measured.
The effect of substitutions at R299 depends on the sequence in the –35 element
His6-tagged variants of S carrying modifications of R299 were used for in vitro transcription experiments with supercoiled DNA of plasmids pJCD01-osmEp+, pJCD01-osmEp13T, pJCD01-osmEp31C or pJCD01-osmEp32G. As compared with osmEp+, the mutation osmEp32G reduced 3-fold the in vitro transcription efficiency by wild-type S (Fig. 6A). Although the stimulation ratio was less pronounced than in vivo (Fig. 5), we observed that variants of S modified at R299 partly suppressed the down-promoter effect of osmEp32G in vitro (compare Fig. 6A with B, C and D). In contrast, the effect of R299 substitutions was very different when tested on osmEp31C, an up-promoter variant of osmEp (8) harboring a –35 element identical to the optimal sequence common for both E70 and ES (9). With wild-type S, osmEp31C behaved as an up-promoter mutation (4-fold stimulation, see Fig. 6A). However, when transcribed by holoenzymes reconstituted with R299 variants of S, osmEp31C showed equivalent or even slightly lower efficiency than osmEp+ (Fig. 6B, C and D). Thus, S exhibits a preference for a CG base pair at position –31 in osmEp, and this preference is abolished when R299 is changed into a histidine, a serine or a glycine.
Figure 6. Effect of modifications of R299 of S on transcription of osmEp wild-type and mutant derivatives. Supercoiled pJCD01osmEp+, pJCD01osmEp13T, pJCD01osmEp31C and pJCD01osmEp32G were transcribed in vitro with the indicated holoenzyme (45 nM final concentration). Autoradiography of a fixed and dried electrophoresis gel of the products is shown. osmE' mRNA and RNA I were quantitated with a PhosphorImager and the relative ratio osmE' RNA/RNA I (with the ratio obtained with pJCD01osmEp+ set as 1 for each form of holoenzyme) is shown in the histograms under the gels.
Effect of mutations in rpoS on transcription from ficp
The results observed with the osmE promoter suggest that S and 70 interact with this particular target in a similar manner, using domains 2.4 and 4.2 to recognize the –10 and –35 elements, respectively. To investigate whether these conclusions could be extended to other S-dependent promoters, we tested the effect of all the S variants obtained in this work on transcription from ficp, a promoter that exhibits a different organization. Indeed, ficp is completely dependent on S for its transcription (4,34) but, in contrast to osmEp, it has a very poor –35 element and a –10 element much closer to the consensus but without a C in position –12 (note that position –12 in ficp corresponds to position –13 in osmEp, Fig. 2). We first introduced a C in position –12 of ficp and, as shown in Figure 7, this resulted in a 2.5-fold increase in expression of a transcriptional fusion. This expression was abolished in an rpoS::Tn10 background, confirming that it is completely S-dependent (not shown). The variants of S modified at position 299 did not stimulate transcription from ficp. On the contrary, they resulted in an 2-fold reduction in expression of the ficp–lac fusion. Furthermore, the stimulation of expression due to the ficp12C mutation was still observed with all the variants of R299 of S. Therefore, these data indicate that R299 is not playing the same role in the recognition of ficp and of osmEp.
Figure 7. Effect of the rpoS variants on transcription in vivo of the ficp+ and ficp12C promoters. Strains carrying transcriptional ficp+–lac or ficp12C–lac fusions were transformed with the indicated derivatives of pBADrpoS. Overnight cultures of each strain were diluted 200-fold and cultures were grown aerobically for 8 h at 37°C in LB medium + ampicillin before ?-galactosidase activity was measured.
Substitutions of Q152 in S were not observed to stimulate transcription from ficp. However, with the exception of Q152H, they abolish the preference for a C 5' of the –10 element, because the expression from wild-type ficp and ficp12C was equivalent. Similarly, modifications of E155 also abolished the preference for a C in –12 of ficp. Therefore, these data indicate that the 2.4 region of S interacts with the –10 region of ficp and, furthermore, they suggest that the amino acids Q152 and E155 are important for the specific recognition of the CG base pair 5' of the –10 element.
DISCUSSION
In this work, we used a genetic suppressor approach with variants of the osmE promoter altered either at position –13 (osmEp13T) or in the –35 element (osmEp32G), to investigate the mechanisms of promoter recognition by the stress-specific factor S of E.coli. We were able to isolate mutations in rpoS able to suppress both promoter defects and, strikingly, they affected only three codons of rpoS (Table 2). Because changes at each of the three positions appeared several times, it is likely that changes at these three codons are the only possible single modifications that fulfill the criteria of the screen. A BLAST search in the sequenced genomes of Gram-negative bacteria (http://www.ncbi.nlm.nih.gov/blast/Blast.cgi) allowed us to identify 21 clear homologs of E.coli S (not shown). All these homologs carry Q, E and R at the positions corresponding to Q152, E155 and R299 of S, emphasizing the importance of these residues for the functioning of S. We will discuss the role of these amino acids in promoter recognition in light of the available genetic and structural data.
Does S interact with –35 elements at its target promoters?
The –35 region is not essential in the strictly S-dependent fic promoter (34), and biochemical analysis indicated that S interacts only weakly with the –35 region of several promoters (35). Taken together with the poor sequence conservation in the –35 region of S targets (10,11,36), these observations suggested that a peculiarity of ES might be to recognize its targets without a strong and specific binding to the –35 element. Using synthetic promoters, Gaal et al. (9) demonstrated that a consensus –35 element increases the binding of ES and the efficiency of the promoter, suggesting that a –35 element can play a prominent role in the recognition of at least some promoters by ES. The data presented here demonstrate that an interaction with the –35 element is indeed important in vivo for the recognition of osmEp by ES. Therefore, S seems able to recognize promoters through different mechanisms, which do or do not involve strong interactions with a –35 element. Because most S-dependent promoters have a very poor –35 element (36), the case of osmEp may seem to be an exception. However, other S-dependent promoters having a good –35 element are known , and the identification of new targets of S may reveal additional members in this class of promoters. Interestingly, it is worth noting that this situation is not so different from that of 70 which can also recognize promoters without a –35 element, provided that they carry an extended –10 element (15,38).
How does S interact with the –35 elements at osmEp?
R299 of S is the homolog of R584 of 70 and R409 of A from T.aquaticus (Fig. 1). Genetic evidence suggested that R584 of 70 interacts with the CG base pair (TTGACA) in the –35 element (12,13), and structural data demonstrated that R409 of A donates two hydrogen bonds to the O6 and N5 of the guanine in the major groove (19). Wild-type S also exhibits a preference for a CG bp at position –31 (9), and we show here that R299 is responsible for this preference at the osmE promoter (Fig. 6). Therefore, our data strongly suggest that R299 of S can make direct, base-specific interactions with the guanine of the CG in position –31 of its target promoters. Such contacts are not possible between S and osmEp+ that harbors an AT base pair in position –31 (TTGAAA; Fig. 2). Furthermore, we have shown previously that this deviation is more deleterious for interaction with E70 than with ES and thereby contributes to the selectivity of recognition of osmEp (8). The structural data (structure coordinates 1KU7 in the Protein Data Bank) suggest that the replacement of the consensus G by a T on the template strand would place the methyl group on C5 of thymine in conflict with the arginine residues R584 of 70 or R299 of S. The four suppressor mutations isolated here substitute R299 by short side chain amino acids, and that should remove the conflict with –31T. We believe that the 2-fold stimulation of transcription from the wild-type and from most variants of osmEp (Fig. 5) can be explained by the elimination of this negative interaction. However, why the substitutions of R299 specifically suppress the effect of osmEp32G is not readily explained by the available structural data. Indeed, no base-specific contacts have been identified with the nucleotides in position –32 of the –35 element (19). One possibility is that such contacts do exist, but only transiently during the kinetic pathway of promoter recognition, and that they could not be seen in the crystal structures. Alternatively, the effect of osmEp32G could be due to a modification of the DNA helical structure and/or bending that could reinforce the negative effect of the repulsion between –31T and R299 of S, which would then explain the strong suppressor effect of the substitutions of R299 on osmEp32G. This hypothesis would be in agreement with the ability of suppressing osmEp32G by substituting R299 with four different amino acids that only have in common that they carry short side chains. An important role for the DNA structure would also be consistent with our previous observation that the efficiency of transcription of osmEp is modulated by supercoiling density (22).
Role of the 2.4 region of S in the recognition of the –10 element of promoters
Q152 and E155 of S are the homologs of Q437 and T440 of 70 and Q260 and N263 of A from T.aquaticus, respectively (Fig. 1). In the crystal structure of T.aquaticus RNAP holoenzyme complexed with an open promoter-mimicking DNA fragment, both Q260 and N263 are exposed on the same face of an amphipathic -helix and point to nucleotides in the major groove near –12 (20). Works based on suppressor analyses identified changes in these residues and led to the model that Q437 and T440 of 70 could interact directly with the base pair at position –12 (13,14). However, this conclusion has been challenged by both biochemical (39) and structural (20) data, and these residues may contribute to the recognition of nucleotides in –12 without direct contact but, for instance, via the repositioning of a nearby essential residue. In any case, it is striking that genetic screens for 70 suppressors of promoters mutated at position –12 or S suppressors of a promoter mutated at position –13 identified modifications affecting the same two residues. Recently, the analysis of a collection of rpoS mutations identified Q152 but not E155 as essential for the functioning of S (40). The same work concluded that 70 and S have a common global organization but several crucial amino acids differ between the two factors, suggesting that they may differ in the details of the promoter recognition process. Our work demonstrates that E155 is one of those residues important for promoter recognition that differ between 70 and S. Notably, the mutation leading to Q152R, that we identified here as a suppressor of osmEp13T (Table 2, Fig. 3), was found to abolish transcription from several S-dependent promoters (40). The origin of this difference is not known, but once again it may highlight the fact that different promoters could be recognized via different interactions, even within the –10 element, and it also emphasizes the importance of investigating the recognition of a variety of targets in order to fully understand the differential recognition of promoters by ES and E70.
Role of Q152 and E155 in S preference for a C at the –13 position of its target promoters
Previous work, also based on allele-specific suppressors, indicated that the amino acid K173 (corresponding to E458 in 70 and E281 in A of T.aquaticus) has a discriminatory role for the nucleotide at position –13 of S-dependent promoters (10). Our data show that Q152 and E155 are also important for this distinctive property of S, because modifications at both residues abolish the preference for a C at position –13 of osmEp (Fig. 3) and –12 of ficp (Fig. 6). Structural data in the Protein Data Bank] show that in A, Q260 and N263 on the one hand and E281 on the other hand belong to two -helices that form a clip facing the major groove in front of the positions –12/–13 on the template strand. Therefore, these residues are likely to participate in a structure involved in a direct interaction with the nucleotide in position –13 of promoters, in agreement with the genetic data.
ACKNOWLEDGMENTS
We are grateful to F. Norel for the gift of anti-S antibodies, Y.N. Zhou and S. Gottesman for pBADrpoS and A.J. Carpousis for language improvements. Part of this work was supported by grants from the French Ministère de l’Enseignement Supérieur et de la Recherche (Programme de Recherche Fondamentale en Microbiologie, Maladies Infectieuses et Parasitaires) and from the Génop?le of Toulouse to C.G.
REFERENCES
Ishihama,A. (2000) Functional modulation of Escherichia coli RNA polymerase. Annu. Rev. Microbiol., 54, 499–518.
Mulvey,M.R. and Loewen,P.C. (1989) Nucleotide sequence of katF of Escherichia coli suggests KatF protein is a novel sigma transcription factor. Nucleic Acids Res., 17, 9979–9991.
Lange,R. and Hengge-Aronis,R. (1991) Identification of a central regulator of stationary-phase gene expression in Escherichia coli. Mol. Microbiol., 5, 49–59.
Tanaka,K., Takayanagi,Y., Fujita,N., Ishihama,A. and Takahashi,H. (1993) Heterogeneity of the principal factor in Escherichia coli: the rpoS gene product, 38, is a second principal factor of RNA polymerase in stationary-phase Escherichia coli. Proc. Natl Acad. Sci. USA, 90, 3511–3515.
Loewen,P.C. and Hengge-Aronis,R. (1994) The role of the sigma factor S (KatF) in bacterial global regulation. Annu. Rev. Microbiol., 48, 53–80.
Hengge-Aronis,R. (1996) Regulation of gene expression during entry into stationary phase. In Neidhardt,F.C. (ed.), Escherichia coli and Salmonella: Cellular and Molecular Biology. American Society for Microbiology Press, Washington, DC, pp. 1497–1512.
Altuvia,S., Almiron,M., Huisman,G., Kolter,R. and Storz,G. (1994) The dps promoter is activated by OxyR during growth and by IHF and S in stationary phase. Mol. Microbiol., 13, 265–272.
Bordes,P., Repoila,F., Kolb,A. and Gutierrez,C. (2000) Involvement of differential efficiency of transcription by Es and E70 RNA polymerase holoenzymes in growth phase regulation of the Escherichia coli osmE promoter. Mol. Microbiol., 35, 845–853.
Gaal,T., Ross,W., Estrem,S.T., Nguyen,L.H., Burgess,R.R. and Gourse,R.L. (2001) Promoter recognition and discrimination by ES RNA polymerase. Mol. Microbiol., 42, 939–954.
Becker,G. and Hengge-Aronis,R. (2001) What makes an Escherichia coli promoter S dependent? Role of the –13/–14 nucleotide promoter positions and region 2.5 of S. Mol. Microbiol., 39, 1153–1165.
Espinosa-Urgel,M., Chamizo,C. and Tormo,A. (1996) A consensus structure for S-dependent promoters. Mol. Microbiol., 21, 657–659.
Gardella,T., Moyle,H. and Susskind,M.M. (1989) A mutant Escherichia coli 70 subunit of RNA polymerase with altered promoter specificity. J. Mol. Biol., 206, 579–590.
Siegele,D.A., Hu,J.C., Walter,W.A. and Gross,C.A. (1989) Altered promoter recognition by mutant forms of the 70 subunit of Escherichia coli RNA polymerase. J. Mol. Biol., 206, 591–603.
Waldburger,C., Gardella,T., Wong,R. and Susskind,M.M. (1990) Changes in conserved region 2 of Escherichia coli 70 affecting promoter recognition. J. Mol. Biol., 215, 267–276.
Barne,K.A., Bown,J.A., Busby,S.J. and Minchin,S.D. (1997) Region 2.5 of the Escherichia coli RNA polymerase 70 subunit is responsible for the recognition of the ‘extended-10’ motif at promoters. EMBO J., 16, 4034–4040.
Lonetto,M., Gribskov,M. and Gross,C.A. (1992) The 70 family: sequence conservation and evolutionary relationships. J. Bacteriol., 174, 3843–3849.
Malhotra,A., Severinova,E. and Darst,S.A. (1996) Crystal structure of a 70 subunit fragment from E.coli RNA polymerase. Cell, 87, 127–136.
Vassylyev,D.G., Sekine,S., Laptenko,O., Lee,J., Vassylyeva,M.N., Borukhov,S. and Yokoyama,S. (2002) Crystal structure of a bacterial RNA polymerase holoenzyme at 2.6 ? resolution. Nature, 417, 712–719.
Campbell,E.A., Muzzin,O., Chlenov,M., Sun,J.L., Olson,C.A., Weinman,O., Trester-Zedlitz,M.L. and Darst,S.A. (2002) Structure of the bacterial RNA polymerase promoter specificity subunit. Mol. Cell, 9, 527–539.
Murakami,K.S., Masuda,S., Campbell,E.A., Muzzin,O. and Darst,S.A. (2002) Structural basis of transcription initiation: an RNA polymerase holoenzyme–DNA complex. Science, 296, 1285–1290.
Conter,A., Menchon,C. and Gutierrez,C. (1997) Role of DNA supercoiling and RpoS sigma factor in the osmotic and growth phase-dependent induction of the gene osmE of Escherichia coli K12. J. Mol. Biol., 273, 75–83.
Bordes,P., Conter,A., Morales,V., Bouvier,J., Kolb,A. and Gutierrez,C. (2003) DNA supercoiling contributes to disconnect S accumulation from S-dependent transcription in Escherichia coli. Mol. Microbiol., 48, 561–571.
Guzman,L.M., Belin,D., Carson,M.J. and Beckwith,J. (1995) Tight regulation, modulation and high-level expression by vectors containing the arabinose PBAD promoter. J. Bacteriol., 177, 4121–4130.
Miller,J.H. (1992) A Short Course in Molecular Genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
Vidal-Ingigliardi,D. and Raibaud,O. (1985) A convenient technique to compare the efficiency of promoters in Escherichia coli. Nucleic Acids Res., 13, 5919–5926.
Debarbouillé,M., Shuman,H.A., Silhavy,T.J. and Schwartz,M. (1978) Dominant constitutive mutations in malT, the positive regulator gene of the maltose regulon in Escherichia coli. J. Mol. Biol., 124, 359–371.
Gutierrez,C. and Devedjian,J.C. (1991) Osmotic induction of gene osmC expression in Escherichia coli K12. J. Mol. Biol., 220, 959–973.
Fran?ois,V., Louarn,J., Patte,J. and Louarn,J.-M. (1987) A system for in vivo selection of genomic rearrangements with predetermined endpoints in Escherichia coli using modified Tn10 transposons. Gene, 56, 99–108.
Sambrook,J., Fritsch,E.F. and Maniatis,T. (1989) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
Silhavy,T.J., Berman,M. and Enquist,L. (1984) Experiments with Gene Fusions. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
Diaz,J.J., Rhoads,D.D. and Roufa,D.J. (1991) PCR-mediated chemical mutagenesis of cloned duplex DNAs. Biotechniques, 11, 204–211.
Bordes,P., Bouvier,J., Conter,A., Kolb,A. and Gutierrez,C. (2002) Transient repressor effect of Fis on the growth phase-regulated osmE promoter of Escherichia coli K12. Mol. Gen. Genet., 268, 206–213.
Becker,G., Klauck,E. and Hengge-Aronis,R. (1999) Regulation of RpoS proteolysis in Escherichia coli: the response regulator RssB is a recognition factor that interacts with the turnover element in RpoS. Proc. Natl Acad. Sci. USA, 96, 6439–6444.
Hiratsu,K., Shinagawa,H. and Makino,K. (1995) Mode of promoter recognition by the Escherichia coli RNA polymerase holoenzyme containing the S subunit: identification of the recognition sequence of the fic promoter. Mol. Microbiol., 18, 841–850.
Colland,F., Fujita,N., Kotlarz,D., Bown,J.A., Meares,C.F., Ishihama,A. and Kolb,A. (1999) Positioning of S, the stationary phase factor, in Escherichia coli RNA polymerase–promoter open complexes. EMBO J., 18, 4049–4059.
Lee,S.J. and Gralla,J.D. (2001) Sigma38 (rpoS) RNA polymerase promoter engagement via –10 region nucleotides. J. Biol. Chem., 276, 30064–30071.
Lacour,S., Kolb,A., Boris Zehnder,A.J. and Landini,P. (2002) Mechanism of specific recognition of the aidB promoter by S-RNA polymerase. Biochem. Biophys. Res. Commun., 292, 922–930.
Kumar,A., Malloch,R.A., Fujita,N., Smillie,D.A., Ishihama,A. and Hayward,R.S. (1993) The minus 35-recognition region of Escherichia coli 70 is inessential for initiation of transcription at an ‘extended minus 10’ promoter. J. Mol. Biol., 232, 406–418.
Fenton,M.S., Lee,S.J. and Gralla,J.D. (2000) Escherichia coli promoter opening and –10 recognition: mutational analysis of 70. EMBO J., 19, 1130–1137.
Lee,S.J. and Gralla,J.D. (2002) Promoter use by 38 (rpoS) RNA polymerase. Amino acid clusters for DNA binding and isomerization. J. Biol. Chem., 277, 47420–47427.
Casadaban,M.J. (1976) Transposition and fusion of the lac genes to selected promoters in Escherichia coli using bacteriophage lambda and Mu. J. Mol. Biol., 104, 541–555.(Claire Checroun, Patricia Bordes, Olivie)