reduced capacity of alternative s to melt promoters ensures stringent promoter recognition

Reduced capacity of alternative ssto melt promoters ensures stringentpromoter recognition

Byoung-Mo Koo,1 Virgil A. Rhodius,1 Gen Nonaka,1,5 Pieter L. deHaseth,2,3 and Carol A. Gross1,4,6

1Department of Microbiology and Immunology, University of California at San Francisco, San Francisco, California 94158, USA;2Center for RNA Molecular Biology, Case Western Reserve University, Cleveland, Ohio 44106, USA; 3Department ofBiochemistry, Case Western Reserve University, Cleveland, Ohio 44106, USA; 4Department of Cell and Tissue Biology,University of California at San Francisco, San Francisco, California 94158, USA

In bacteria, multiple ss direct RNA polymerase to distinct sets of promoters. Housekeeping ss direct transcriptionfrom thousands of promoters, whereas most alternative ss are more selective, recognizing more highly conservedpromoter motifs. For s32 and s28, two Escherichia coli Group 3 ss, altering a few residues in Region 2.3, theportion of s implicated in promoter melting, to those universally conserved in housekeeping ss relaxed theirstringent promoter requirements and significantly enhanced melting of suboptimal promoters. All Group 3 ss andthe more divergent Group 4 ss have nonconserved amino acids at these positions and rarely transcribe >100promoters. We suggest that the balance of ‘‘melting’’ and ‘‘recognition’’ functions of ss is critical to setting thestringency of promoter recognition. Divergent ss may generally use a nonoptimal Region 2.3 to increase promoterstringency, enabling them to mount a focused response to altered conditions.

[Keywords: s factor; Region 2.3; melting proficiency; promoter stringency]

Supplemental material is available at http://www.genesdev.org.

Received July 14, 2009; revised version accepted August 31, 2009.

In bacteria, promoter recognition is accomplished pri-marily by the s subunit of RNA polymerase (RNAP). Asingle housekeeping s directs RNAP to thousands ofpromoters, whereas alternative ss generally orchestratetranscription from substantially fewer promoters, allow-ing focused responses to cellular, environmental, anddevelopmental signals (Gruber and Gross 2003; Pagetand Helmann 2003; Gama-Castro et al. 2008). This un-equal division of labor requires a housekeeping s withbroad, flexible promoter recognition (Hook-Barnard andHinton 2007), and alternative ss with more restrictedrecognition (Amaya et al. 2001; Britton et al. 2002;Eichenberger et al. 2003; Nonaka et al. 2006; van Schaiket al. 2007; Zhao et al. 2007; Asayama and Imamura 2008;Koo et al. 2009a,b). More extensive use of activators andrepressors by the housekeeping s than by alternative sspartially explains this distinction (Browning and Busby2004). Here, we provide evidence for an additional strat-egy intrinsic to the ss themselves, demonstrating thatsome alternative ss are specifically constructed to limittheir ability to transcribe a wide range of promoters.

ss are sequence-specific DNA-binding proteins witha modular architecture, consisting of globular domains(Murakami and Darst 2003) subdivided into conservedregions (Fig. 1A). In free s, DNA-binding determinants aremasked by domain interactions (Dombroski et al. 1993;Sorenson and Darst 2006). However, strong interactionsbetween RNAP and domains 2–4 (s2–4) expose DNA-binding determinants and position domains for interac-tion with promoter motifs (Fig. 1A; Kuznedelov et al.2002; Murakami et al. 2002). s4 recognizes the �35 motif(Gardella et al. 1989; Siegele et al. 1989), s3 recognizes theextended �10 (E-10) motif (Barne et al. 1997; Koo et al.2009a,b), and s2 facilitates strand opening by three se-quential activities: (1) recognition of the �10 and dis-criminator regions (Siegele et al. 1989; Daniels et al. 1990;Waldburger et al. 1990; Tatti et al. 1991; Feklistov et al.2006; Haugen et al. 2006; Koo et al. 2009a,b), (2) participa-tion in melting (Juang and Helmann 1994; Fenton et al.2000; Tomsic et al. 2001; Lee and Gralla 2003), and (3) in-teraction with the �10 region nontemplate strand DNAto stabilize the melted state (Helmann and deHaseth1999; Schroeder et al. 2009). s1 (Region 1.1) is unique tohousekeeping ss and has regulatory roles (Dombroskiet al. 1992, 1993; Hook-Barnard and Hinton 2009).

The s70 superfamily includes three subfamilies, dividedaccording to phylogenetic relatedness to the essential

5Present address: Ajinomoto Co., Inc., 15-1, Kyobashi 1-chome, Chuo-ku, Tokyo 104-8315, Japan6Corresponding author.E-MAIL [email protected]; FAX (415) 514-4080.Article is online at http://www.genesdev.org/cgi/doi/10.1101/gad.1843709.

2426 GENES & DEVELOPMENT 23:2426–2436 � 2009 by Cold Spring Harbor Laboratory Press ISSN 0890-9369/09; www.genesdev.org

housekeeping (Group 1) ss (Lonetto et al. 1992; Gruberand Gross 2003; Paget and Helmann 2003). Group 2 ss,most closely related to the Group 1 ss, have s2–4 but arenonessential; Group 3 ss are less related but also contains2–4; Group 4 ss (ECF ss) are the most minimal ss,containing only s2 and s4. Here, we examine Group 3ss. Most bacteria have multiple ss of this type, andEscherichia coli has two: s28 and s32 (Gruber and Gross2003; Paget and Helmann 2003). s28, the most widelydistributed alternative s, controls flagella-related genesin all motile Gram-negative and Gram-positive bacteriaand development in some nonmotile bacteria (Chilcottand Hughes 2000; Yu and Tan 2003; Serizawa et al. 2004;Shen et al. 2006). s32 controls the heat-shock responseand is present in most proteobacteria (Nakahigashi et al.1995; Guisbert et al. 2008). s32 and s28 each direct RNAPto a limited number of promoters (;50 for s32 and ;25 fors28 in E. coli) (Nonaka et al. 2006; Shen et al. 2006; Wadeet al. 2006; Zhao et al. 2007). Expression from these pro-moters is instituted synchronously in response to achange in the amount and/or activity of the respective s.

We previously found an important distinction betweenthe promoter recognition properties of the E. coli Group 3ss and their housekeeping s, s70. In s70, the �35, �10,and E-10 motifs are partially redundant, and functionalpromoters are constructed from subsets of motifs (Hook-Barnard and Hinton 2007). Most commonly, promotershave �10 and �35 motifs but lack the E-10 (Lisser andMargalit 1993). Conversely, promoters with the E-10motif do not have good matches to both the �10 and�35 motifs (Mitchell et al. 2003). Most E-10 promotersrequire neither the �35 motif nor its s recognition region(4.2) (Ponnambalam et al. 1986; Kumar et al. 1993;Minakhin and Severinov 2003; Young et al. 2004); a fewhave an excellent �35 and a reduced requirement for �10(Thouvenot et al. 2004; Hook-Barnard et al. 2006). Insharp contrast, functional s28 and s32 promoters requiregood matches to the �35, �10, and E-10 motifs (Fig. 1B;Koo et al. 2009a,b). Consistent with this, the informationcontent of the core regions of s32 promoters (18.3 bits)(Nonaka et al. 2006) and s28 promoters (21.3 bits) (de-termined from sequences of all E. coli s28 promoterslisted in BioCyc [http://biocyc.org]; V Rhodius, unpubl.;data not shown) is much higher than that for s70 pro-moters (9.2 bits) (Shultzaberger et al. 2007). Requiringextensive recognition determinants would limit tran-scription to a relatively restricted set of promoters, mak-ing it important to understand how these ss maintaindependence on all three promoter motifs.

We show that s32 and s28 require all three core pro-moter motifs for function largely because of alteredresidues in Region 2.3, a 17-amino-acid section of s2

implicated in promoter melting. Both ss deviate insequence from critical Region 2.3 residues that areuniversally conserved in the housekeeping ss and areknown to be important in melting. Converting these s32

and s28 residues to those present in the housekeeping ssdecreases the requirement for extensive recognition mo-tifs and increases melting capacity at nonoptimal pro-moters. Our results suggest that extensive recognitionmotifs compensate for nonoptimal Regions 2.3 in s32 ands28. All Group 3 and Group 4 ss have nonconsensusamino acids at the Region 2.3 positions we investigated.Decreasing melting capacity may be a general strategyenabling divergent alternative ss to mount a discrete,focused, and structured response to altered conditions.

Results

Phylogenetic analysis identifies elements requiredto bypass the s32 requirement for an E-10 motif

s32s from different proteobacterial groups differ in theirstringency of promoter recognition and in their recogni-tion determinants. s32s from E. coli and otherg-/b-proteobacteria exhibit high dependence on the E-10motif and recognize it with a universally conservedresidue (K130) in s3 (Koo et al. 2009a). In contrast,a-/d-/e-proteobacteria show little to no dependence onthe E-10 motif, and their comparable s3 residue is alanine,serine, or glutamine (Green and Donohue 2006; McGrath

Figure 1. Promoter utilization by s. (A) Domain organizationand conserved regions of ss and their functional assignmentsbased on those of s70. A schematic of a typical promoter withconserved DNA regions �35, �10, and E-10, and discriminator(DSR) region from the transcription start site is shown belowthe linear representation of s. Interactions between conservedregion of s and promoter regions are indicated by arrows. In s2,Region 2.4 is implicated in both duplex recognition and in-teraction with nontemplate strand, Region 2.3 is implicated induplex recognition and melting, and Region 1.2 recognizes thediscriminator. The crystallized domains of s are indicated above

the linear representation of s. (B) Consensus promoter sequencesof s70, s32, and s28. E-10 motifs of each promoter are boxed ingray, and the �10 motifs of each promoter are underlined. (C)Recognition of the �10 region of promoters by s32 and s28.Amino acid sequences of Regions 2.3, 2.4, and 3.0 of s32 and s28

are shown. The �10 consensus promoter sequences for each s

are indicated above or below the amino acid sequences. Con-tacts between s and promoter are indicated by arrows. Solidarrows indicate direct contacts, and the arrow with the dottedline indicates contact to increase selectivity (Koo et al. 2009a,b).

s promoter melting and recognition

GENES & DEVELOPMENT 2427

et al. 2007; Koo et al. 2009a). s32 of the a-proteobacteriumCaulobacter crescentus has little dependence on the E-10motif even in combination with E. coli core RNAP (Kooet al. 2009a), indicating that the distinction in promoterrecognition stringency resided within s32 itself. Thisraises the question of how C. crescentus s32 bypassesthe requirement for the E-10 motif.

Comparison of E. coli and C. crescentus s32 revealedthat the two differ in important residues in Region 2.3(Fig. 2A), but not in residues implicated in base-specificrecognition of promoters (Fig. 1C) or in Region 2.2, themost important core-binding motif (data not shown). FourRegion 2.3 aromatic amino acids are universally con-served in housekeeping ss (bold in Fig. 2A). Of these, F427

is buried and is likely to play a structural role (Murakamiet al. 2002), W434 is likely to be involved in promoterrecognition (Juang and Helmann 1994), and Y430 andW433 are most directly involved in promoter DNAmelting (Juang and Helmann 1994; Fenton et al. 2000;Tomsic et al. 2001). Three of these residues are present inC. crescentus s32 (and in s32 of all a-proteobacteria), butonly one is present in E. coli s32 (and in s32 of all otherg-proteobacteria) (Kourennaia and deHaseth 2007). Wetested whether transplantation of C. crescentus Region2.3 residues (103LATYAMWW110) into E. coli s32 reduceddependence on the E-10 motif using lacZ reporter assaysdriven from the groE s32 promoter (Fig. 2B) or derivativesof this promoter. Whereas authentic E. coli s32 exhibiteda 20-fold dependence (Fig. 2C, columns 1,2), the hybrids32 showed only a fourfold dependence (Fig. 2C, columns5,6); introducing the K130A substitution to more fullymimic authentic C. crescentus s32 eliminated depen-dence on the E-10 motif (Fig. 2C, columns 7,8) andreduced the deleterious effect of the K130A substitutionalone (Fig. 2C, cf. columns 7,8 and 3,4). Importantly, thiseffect is recapitulated by simply changing the residues inE. coli s32 analogous to Y430 and W433 to their counter-parts in the housekeeping ss and in C. crescentus s32

(creating F104YH107Ws32; YWs32) (Fig. 2C, cf. columns9,10 and 5,6, and columns 11,12, and 7,8). Note that Y430and W433 are the residues most directly implicatedin promoter DNA melting (Juang and Helmann 1994;Fenton et al. 2000; Tomsic et al. 2001). Single aminoacid substitutions showed a partial effect, and a triplemutant (L101F, F104Y, H107W) is virtually indistinguish-able from YWs32 (data not shown). These results suggestthat these YW residues in C. crescentus (and othera-/d-/e-proteobacteria) enable their s32s to bypass therequirement for an E-10 motif.

YWs32 generally suppresses defects of nonoptimalpromoters

We tested whether YWs32 was generally proficient intranscribing nonconsensus promoters (Fig. 3). YWs32

significantly suppressed the two �35 and the three �10mutations in vivo: Whereas most mutations resulted ina fourfold to fivefold decrease in expression with wild-type s32, they showed only a 1.5-fold to twofold decreasein expression with YWs32 (Fig. 3A). Likewise, YWs32

showed reduced dependence on the length of the spacerDNA separating the �35 and�10 elements (Fig. 3B). Thiseffect was reproduced in vitro. All mutant promoterswith a significant transcription defect (#40% of the wild-type promoter) showed significantly higher transcriptionwith YWs32 than with wild-type s32, indicating thateffects were direct (Fig. 3C,D). We conclude that YWs32

generally exhibited higher activity than authentic s32 onweak promoters.

Altering Region 2.3 in s28 also broadens promoterrecognition

s28 is highly divergent from s32 (Gruber and Gross 2003;Paget and Helmann 2003), enabling us to test whether the

Figure 2. Identification of elements that bypass the need forthe E-10 motif in s32. (A) Amino acid sequence alignment ofa segment of Region 2.3 and Region 2.4 of E. coli s70, E. coli s32,and C. crescentus s32. Bold letters are residues universallyconserved in housekeeping and Group 2 ss, and are implicatedin promoter melting in s70; analogous residues that deviate fromconsensus in s32s are in gray. The numbers indicate amino acidsequence position. (B) Sequence of s32-dependent promoter(PgroE) used in this study. Only the wild-type sequence of groE

promoter is shown; promoter variants were constructed fromthis sequence (Koo et al. 2009a). The native sequence of the groEpromoter region is shown in capital letters, vector sequence is inlowercase, and �35 and �10 regions and transcription start siteare shown in bold. (C) Expression from wild-type or �16,�15 AAmutant groE promoters by E. coli RNAP containing s32 variants.b-galactosidase activity (Miller units) of promoterTlacZ fusionsfor each s32 variant on the wild-type promoter (black bar) or the�16,�15 AA mutant promoter lacking the extended �10 region(gray bar) are shown. These assays were performed in CAG57101(DrpoH) with plasmids expressing s32 variants and bearing groE

promoter variantsTlacZ. For the b-galactosidase assays shown inthis and subsequent figures, all values (indicated above each bar)are averages of at least three independent experiments; error barsindicate one standard deviation. Wild type (WT) is E. coli s32;variants have substitutions in E. coli s32 as follows: R2.3CCR,sequence encoding Region 2.3 is replaced with Region 2.3 of C.crescentus s32; R2.3CCR K130A is R2.3CCR with K130A; YW isF104YH107Ws32; and YW K130A is YW with K130A.

Koo et al.

2428 GENES & DEVELOPMENT

relationship between Region 2.3 consensus and increasedtranscription of weak promoters was likely to be gener-alizable to other Group 3 ss. Comparison with s70

indicated that only two of the four aromatic Region 2.3residues universally conserved in the housekeeping ss arepresent in E. coli s28 (Fig. 4A). R74 (W434 in s70) was notconsidered further as it is highly conserved among s28

orthologs and participates in nonspecific promoter bind-ing (Koo et al. 2009b). Q73 (W433 in s70) varies in

different s28s but is never W. We investigated whetherQ73Ws28 has broadened promoter recognition.

s28 is exceptionally sensitive to substitution in the�10region of the promoter, and the Q73W substitutiondramatically increases the ability of s28 to recognizea variety of nonoptimal promoters. We examined pro-moter activity using lacZ reporter assays driven from thetar s28 promoter (Fig. 4B) and its derivates. Eliminatingthe E-10 motif (�14G, �13C) or altering any one of threebases in the �10 motif reduces expression by wild-types28 ;30-fold to 400-fold in vivo (Fig. 4C; Koo et al. 2009b).In sharp contrast, Q73Ws28 exhibits only a threefold to18-fold decrease on this same set of promoters (Fig. 4C).Importantly, Q73A does not increase expression fromthese mutant promoters, indicating that the suppressiveeffect resulted from adding the W residue at position 73,rather than removing the naturally occurring Q residue(data not shown). Likewise, Q73W exhibits 10-fold sup-pression of the transcription defects resulting from re-moving each of the three s28 residues known to partici-pate in base-specific recognition of the s28 promoter (R91[E-10 motif] and D81/R84 [�10 motif]) (Fig. 4D; Koo et al.2009b) and shows enhanced tolerance for variation in thelength of the spacer DNA separating the �35 and �10elements (Fig. 4E). All promoters with significant tran-scriptional defects in vitro (#40% of the wild-type pro-moter) showed significantly increased transcription withQ73Ws28 as compared with wild-type s28 (Fig. 4F). Thus,Q73Ws28 is directly responsible for significantly en-hanced transcription of a broad range of nonoptimalpromoters.

YWs32 broadens promoter recognition of naturalpromoters

Thus far, we examined the effects of YWs32 on a promoterset, each differing from the groE consensus promoter atonly a single position. Natural promoters may have manychanges from the consensus and an UP-element (Zhaoet al. 2005; Nonaka et al. 2006; Wade et al. 2006),recognized by the C-terminal domain of the a-subunit(Ross et al. 1993; Gourse et al. 2000), which mightobscure the effect of YWs32 on natural promoters. Wecompared expression of natural s32 promoters in E. colifused to a green fluorescent protein (GFP) reporter drivenby either wild-type s32 or YWs32 (see the Materials andMethods). The promoter library encompassed sequences�65 to +20 relative to the start site of transcription, andtherefore included any UP-element present. We foundthat promoters very weakly transcribed by s32 RNAPwere more strongly transcribed by YWs32 RNAP; con-versely, promoters strongly transcribed by s32 RNAPhave similar or less expression with YWs32 RNAP. Thisis illustrated by comparing the ratio of promoterstrengths by s32 or YWs32 RNAP against the strength ofthe promoter in the presence of s32 RNAP (Fig. 5A; seealso Supplemental Fig. S2 for the raw data). Therefore,YWs32 preferentially increases expression from weaker,less conserved natural promoters. We consider the impli-cations of this finding in the Discussion.

Figure 3. Suppression of nonoptimal promoter mutations byYWs32 in vivo and in vitro. Expression from wild-type andmutant groE promoters (A) or promoters with nonoptimalspacing between �10 and �35 (B) by wild-type and YW s32 invivo as assayed by b-galactosidase activities (Miller units) foreach s32 variant. Assays were performed as described in Figure2C. Numbers above each bar indicate average values of Millerunits. (C) Effect of YW substitution on transcription from non-optimal groE promoter in vitro. Single-round run-off transcrip-tions were performed as described in the Materials and Methods;representative gels of each transcription reaction are shown.The top band in each lane is from the wild-type groE promoter(internal control); the bottom band originates from either thewild-type groE promoter or the mutant to be tested. (D) Quan-tification of the in vitro transcripts. The bars indicate the rela-tive transcription from mutant promoters as a percentage oftranscript from wild-type promoter for each s32 variant. Eachvalue was calculated as follows: (1) Each short transcript (lowerband) was expressed as a percentage of the long transcript in thesame lane to obtain the normalized short transcript; and (2)normalized short transcripts from each promoter variant (shownin C, lanes 2–11) were divided by normalized short transcriptfrom the wild-type promoter (shown in C, lane 1). All values areaverages of three independent experiments.



Altered Region 2.3 residues exert their effectson a step beyond duplex promoter binding

The s70 residues analogous to those altered in s28 and/ors32 (Y430 and W433) have a minor effect on initialrecognition of duplex DNA and a major effect on opencomplex formation (Fenton et al. 2000; Tomsic et al.2001; Schroeder et al. 2009). The crystal structure ofAquifex aeolicus s28 complexed to its anti-s indicatesthat these Region 2.3 residues in Group 3 ss are roughlyin the same position as in the housekeeping ss with

respect to the most C-terminal helix of s2 (Sorenson et al.

2004), suggesting that they might play similar roles in

both s families. We examined whether aromatic amino

acid-substituted s32 and s28 holoenzymes (Es32 and Es28)

were altered in initial duplex DNA recognition or in open

complex formation using assays developed to character-

ize s70 holoenzyme (Es70). Consistent with their effects

in Es70, these residues have little or no effect on duplex

binding by Es32 and Es28 (Fig. 6A,B) and exert their major

effects on strand opening (Fig. 6C,D).

Figure 4. Enhanced promoter utilization by Q73W s28 in vivo and in vitro. (A) Amino acid sequence alignment of Regions 2.3 and 2.4of E. coli s70 and s28. Bold letters are residues implicated in promoter melting in s70, and the residues not conserved in s28 are in gray.The numbers indicate amino acid sequence position. (B) Sequence of s28-dependent promoter (Ptar) used in this study. Only the wild-type sequence of the tar promoter is shown; promoter variants were constructed from this sequence (Koo et al. 2009b). The nativesequence of the tar promoter region is shown in capital letters, vector sequence is in lowercase, and �35 and �10 regions andtranscription start site are shown in bold. (C) Expression from tar promoter mutants by RNAP containing wild-type or Q73W s28 invivo. b-Galactosidase activities (Miller units) of promoterTlacZ fusions are shown. Assays were performed in CAG57115 (DfliA, DflgM)with plasmids expressing s28 variants and bearing tar promoter variantsTlacZ. All values are averages of three independentexperiments. (D) In vivo effects of the Q73W substitution on mutations in s28 that eliminate promoter recognition determinants.b-Galactosidase activities (Miller units) driven by each s28 variant on tested promoters are shown. All values are averages of threeindependent experiments. (Gray bar) Activity of each mutation in wild-type background; (black bar) activity of each mutation in Q73Wbackground. (E) Expression from tar promoter mutants with nonoptimal spacers by RNAP containing wild-type or Q73W s28 in vivo.b-Galactosidase activities (Miller units) are shown. All values are averages of three independent experiments. The spacer length wasvaried from 9 to 13 base pairs (bp); the wild-type tar promoter has an 11-bp spacer. (F) Effects of the Q73W substitution on mutations ins28 that eliminate promoter recognition determinants and of tar promoter variants in vitro. Relative transcription determined fromsingle-round run-off transcription assays (Supplemental Fig. S1) is depicted as bar graphs. Each experiment was repeated a minimum ofthree times, and numbers above each bar indicate average values of relative transcription. (Left graph) Effect of Q73W substitution onmutations eliminating promoter recognition determinants in s28. Percent transcription was calculated as (intensity of transcript forRNAP with wild-type or Q73W s28 carrying the specified additional substitutions in s28)/(intensity of transcript for RNAP with wild-type or Q73W s28). (Middle and right graphs) Expression from tar promoters with base substitutions (middle) and nonoptimal spacerlengths (right) by RNAP containing wild-type or Q73W s28, normalized as described in Figure 3D.

Koo et al.


We assessed initial recognition by determining Es32

and Es28 binding to duplex DNA at 4°C, a temperaturethat prevents strand opening and traps the initial rapidlydissociating complex between RNAP and DNA, as shownfor Es70 (Fenton et al. 2000). The low-temperature com-plex is on the pathway to the open complex, and thereforeprovides an accurate estimate of initial binding (Li andMcClure 1998). These trapped complexes are expected tobe sensitive to inhibition by heparin, which binds freeRNAP irreversibly, thereby inactivating rapidly dissoci-ating RNAP. We assessed binding both to the completepromoter (consensus�35, E-10, and�10 elements) and toa suboptimal promoter (lacking the E-10 motif). YWs32

holoenzyme (E-YWs32) and wild-type s32 holoenzyme(E-wts32) show equivalent extents of heparin-sensitivebinding (Fig. 6A), providing clear evidence that YWs32

does not affect initial binding. Likewise, Q73Ws28 holo-enzyme (E-Q73Ws28) and wild-type s28 holoenzyme(E-wts28) show equivalent extents of binding (Fig. 6B).This binding is partially heparin-resistant, possiblyreflecting formation of an ‘‘intermediate’’ complex fur-ther down the pathway. In any case, E-Q73Ws28 andE-wts28 do not display any distinction in behavior in thisassay, consistent with the idea that initial steps in theprocess are unaffected. Very similar results were obtained

for Es32 and Es28 using shorter templates truncated justdownstream from the �10 regions of each promoter (datanot shown). Additionally, we validated that the observedbinding of Es28 and Es32, although weak, is dependent onspecific promoter sequences, as it is not observed withrandom sequence DNA (Supplemental Fig. S3). Takentogether, these results support the conclusion that theRegion 2.3 alterations in s28 and s32 have a minimaleffect on duplex recognition.

To examine strand opening, we assessed heparin-resistant binding to ‘‘fork junction’’ templates at 4°C. Afork junction template is one in which the templatestrand is truncated just upstream of the position of strandopening (T-11 for PgroE with Es32 [Mecsas et al. 1991] andT-9 for Ptar with Es28 [Givens et al. 2001]), and thenontemplate strand continues as a single-strand over-hang. This assay has been validated both kinetically andstructurally to be an excellent mimic of open complexformation in Es70 (Guo and Gralla 1998; Murakami et al.2002; Tsujikawa et al. 2002). Indeed, the extent to whicha particular combination of fork junction template andholoenzyme is able to form a stable complex (i.e., re-sistant to heparin challenge) reflects the propensity foropen complex formation with the particular set of re-actants used (Guo and Gralla 1998; Fenton et al. 2000;Tsujikawa et al. 2002). We observe clear evidence thatboth YWs32 and Q73Ws28 are more proficient than theirwild-type counterparts at promoting formation of theopen complex, when assayed with the appropriate tem-plates. E-YWs32 shows fivefold enhancement of opencomplex formation when assayed on a suboptimal tem-plate (no E-10) whose fork extends to �9 (8% E-wts32 vs.41% E-YWs32) (Fig. 6C, �16,�15 AA). The other tem-plates are almost completely shifted by E-wts32 andtherefore cannot provide distinction between the twoholoenzymes. The effect is even more dramatic for s28,where E-Q73Ws28 exhibits $20-fold more open complexformation than E-wts28 at both the long and short sub-optimal fork junction templates (Fig. 6D; �14,�13 AA).We also validated that the observed binding of Es28 andEs32 to fork junction templates is dependent on specificpromoter sequences, as no specific binding to fork junc-tion templates bearing their anti-�10 promoter se-quences was observed (Supplemental Fig. S3). Takentogether, these results provide strong support for the ideathat YWs32 and Q73Ws28 significantly promote opencomplex formation at suboptimal promoter templates.

Discussion

Early ‘‘primordial’’ s factors have diverged into majorsubgroups: the housekeeping ss such as s70, and thealternative ss. This divergence spawned evolution of animportant gene expression strategy, allowing differen-tially regulated ss to recognize discrete classes of pro-moters. s factor specialization has another distinguishingfeature: Housekeeping ss recognize a large number ofdiverse promoters (>1000), whereas most alternative ssare much more restrictive in promoter selection, witha tighter requirement for the sequence and spacing of

Figure 5. YWs32 preferentially increases expression of weaks32 promoters in vivo. Promoter activities driven by either wildtype (wt) or YWs32 were determined by the expression of GFPfrom 50 s32 promoters (Nonaka et al. 2006) as described in theMaterials and Methods. (A) Data are displayed as a scatter plotshowing the log2 value of the ratio of promoter activities drivenby YWs32 and wild-type s32 (Y-axis) versus the strength of thepromoter when driven by wild-type s32 (X-axis). Promoteractivity was calculated from the slope of the differential plotof OD600 versus GFP fluorescence (RFU); see the Materials andMethods. (B) Information content of promoters more activewith either wild type (30 promoters) or YW s32 (19 promoters).Values were calculated as described in the Materials andMethods. Each motif covers the following sequences: UPelement, �60 to �45; �35 motif, �37 to �31; spacer between�35 and �10 motif; extended �10 motif, �16 and �15; and �10motif, �14 to �9.



their promoter motifs. In this study, we asked whatfeature(s) of ss is responsible for this important differencein promoter recognition strategy. Our results suggest thehypothesis that the balance of melting and recognitionfunctions of ss is critical to setting the stringency ofpromoter recognition.

We investigated the features of ss required for stringentpromoter recognition by s32 and s28, two highly divergentmembers of the Group 3 s subfamily present in E. coli.Sequence variation in Region 2.3 is largely responsible forthis requirement: Converting one or two residues inRegion 2.3 to their counterpart(s) in the housekeepingss largely alleviated the stringent requirements. Ourresults indicated that the consensus variants were five-fold to $20-fold more proficient than their wild-type s

counterparts at promoting open complex formation onsuboptimal promoters (Fig. 6C,D). Thus, Es32 and Es28

require highly specified promoters because of their re-duced capacity for promoter melting.

ss perform two sequential activities: recognizing pro-moter motifs (to position RNAP), and facilitating sub-sequent steps required for melting the �10 region tocreate the ‘‘open complex’’ poised for transcription initi-ation. These activities are tightly coupled: The Region 2.3residues facilitating melting and general duplex recogni-tion are located in the same a-helix as the Region 2.4

residues that recognize specific bases in the �10 regionand then stabilize the melted state by interaction withthe nontemplate strand. In Es70 promoters, strand sepa-ration is nucleated within the �10 recognition motif,most likely by flipping out the �11A base (Lim et al.2001; Schroeder et al. 2009). W433 may ‘‘push’’ the �11Aout of the helix (Tomsic et al. 2001). Y430 is believed tointeract with and stabilize the conformation of theflipped out �11A (Schroeder et al. 2009). Importantly,it is altering s32 and s28 residues at positions analo-gous to Y430 and W433 to those in housekeeping ss(F104YH107Ws32; Q73Ws28) that increases melting andrelaxes promoter recognition.

How might a completely consensus promoter decreasethe requirements for s melting functions performed byY430 and W433? The kinetics of open complex formationby E. coli s70 holoenzyme at a consensus promoter(consensus �35, E-10, �10) provide a way of thinkingabout this linkage (Schroeder et al. 2009). It is suggestedthat the completely consensus promoter greatly stabi-lizes the transition state of the normally rate-limitingstep in open complex formation so that it becomesa kinetically significant intermediate, possibly throughcontacts between the E-10 region and RNAP. The impli-cation is that base flipping and strand separation, pre-viously coupled kinetically, become two separable steps

Figure 6. The altered Region 2.3 residues in s32 and s28 affect open complex formation but not template binding. Electrophoreticmobility shift assay was performed as described in the Materials and Methods (also contains sequences of the DNA probes). End-labeledprobe (10 nM) and holoenzyme (25 nM) (both as shown in the figure) were incubated for 10 min at 4°C. In heparin challengeexperiments, 100 mg/mL heparin was added and incubation was continued for 5 min. All values shown are averages of at least threeindependent experiments. Standard deviations are #25% of average values. Representative gels are shown. (A,B) Binding to duplex DNAby E-wts32 or E-YW s32 at the groE promoter (A) or by E-wts28 or E-Q73W s28 at the tar promoter (B). Holoenzyme–DNA complexesformed without (top) or with (bottom) heparin challenge are shown. The percent complex formation in the absence of heparin is consideredto reflect initial duplex binding. (C,D) Binding of RNAP to fork junction templates by E-wts32 or E-YW s32 at the groE promoter (C) or byE-wts28 or E- Q73W s28 to the tar promoter (D). Percent binding is the extent of formation of heparin-resistant complexes.

Koo et al.


at the consensus promoter, with the latter being rate-limiting. Thus, the defect in base flipping caused bya suboptimal amino acid sequence in s Region 2.3 wouldhave no or little effect at consensus promoters but wouldlead to much reduced expression in weaker promoters.Consistent with this interpretation, Y430 and W433substitutions do not affect open complex formation atthe consensus s70 promoter, although they have big ef-fects on melting at standard promoters (Tomsic et al.2001; Schroeder et al. 2009). Thus, the conformationalchanges driven by the consensus promoter elementsobviate the requirement for these Region 2.3 residues.This scenario precisely explains our findings for s32 ands28: Their consensus promoters drive melting eventhough substitutions in residues analogous to Y430 and/or W433 reduce the melting proficiency of s32 and s28.

The suboptimal melting capacity of s32 and s28 hasbiological correlates. The necessity of optimal placementand sequence of promoter motifs to create a functionalpromoter means that deviation from consensus has pro-found negative effects on promoter activity. This focusestranscription by these ss on their authentic regulons anddecreases the possibility that their responses will bediminished because they engage in adventitious tran-scription of near-match promoters. Likewise, the exqui-site sensitivity of transcriptional capacity to promotersequence also allows promoter strength to be regulatedover a broad range so that regulon members are producedin optimal amounts relative to each other. Finally, sub-optimal melting allows these ss to maintain the integrityof their recognition across many organisms. s28 directssynthesis of flagellar components in both Gram-negativeand Gram-positive bacteria, separated in evolution bybillions of years. Yet the consensus sequence of thepromoters recognized by s28 remains unchanged.

Our results graphically illustrate the extent to whichthe relative promoter strength is deregulated when the s

factor has consensus melting determinants. Whereas thenatural promoters in the s32 regulon members display an;100-fold range in activity when driven by wild-type s32,only a 20-fold range is seen when they are transcribed bythe melting-proficient YWs32, with the weakest pro-moters having enhanced activity and the strongest pro-moters showing decreased transcription by YWs32 (pos-sibly because of difficulty in promoter clearance) (Fig. 5A;Supplemental Fig. S2). Interestingly, motif comparisonindicates that the predominant difference between pro-moters preferentially transcribed by YWs32 and by wild-type s32 is that the former has a less conserved �35 motif(Fig. 5B). This distinction raises the intriguing possibilitythat wild-type s32, but not YWs32, uses the �35 region asa ‘‘gatekeeper’’ both to determine functional promotersand to set promoter strength. A consensus �35 regionmay be necessary to slow dissociation sufficiently topermit strand opening.

It is interesting to consider why the s32s of a-proto-bacteria might have broadened promoter specificity. In-terestingly, s32s in the a-protobacteria often control pro-cesses in addition to heat shock, such as development inMyxococcus xanthus (Ueki and Inouye 2001) or other

stress responses (e.g., heavy metal stress in C. crescentus)(McGrath et al. 2007). Additionally, groEL, the mostimportant member of the s32 regulon (Kusukawa andYura 1988), is regulated by an alternative mechanism ina-proteobacteria (Yura and Nakahigashi 1999). Thus,a-proteobacterial s32s may transcribe more genes with-out the necessity of finely controlling the extent of theirexpression as compared with g-proteobacterial s32s.

Ever since Helmann’s seminal work (Juang and Helmann1994) demonstrated the involvement of s Region 2.3residues in promoter DNA melting, much effort has beendevoted to defining its mechanism (Juang and Helmann1994; Fenton et al. 2000; Schroeder et al. 2009). However,there has been little consideration of whether s meltingproficiency differs among s subfamilies and whether thisproperty is used to set the promoter recognition pro-miscuity of that family. Given that Group 3 ss rarely use>100 promoters in any bacterial species (in markedcontrast to the housekeeping s70) and that all of themshare nonconsensus residues in Region 2.3 (SupplementalFig. S4), we propose the hypothesis that amino acidsequence differences within Region 2.3 are importantfor the differences in breadth of promoter choice. Impor-tantly, divergent Group 4 ss are also discrepant from thehousekeeping ss in some of these important Region 2.3residues. Melting deficiencies in the more divergentalternative ss may be a universal mechanism to ensuretheir promoter recognition stringency.

Materials and methods

Details of materials and methods are presented in the Supple-mental Material.

Strains, plasmids, and growth conditions

Strains and plasmids used in this study are listed in Supplemen-tal Table S1. Cells were grown at 30°C in Luria-Bertani (LB)media supplemented with appropriate antibiotics such as ampi-cillin (100 mg/mL), chloramphenicol (30 mg/mL), kanamycin (20mg/mL), and spectinomycin (50 mg/mL). For the strains lackingrpoH (CAG57101), cells were grown with 0.1% L-(+)-arabinose toinduce expression of GroESL (Koo et al. 2009a).

b-galactosidase assay, purification of ss, and in vitro

transcription

b-galactosidase assays (used to measure in vivo promoterTlacZ

activities), overproduction and purification of ss, and in vitrosingle-round transcription assays were performed essentially asdescribed (Koo et al. 2009a,b). Details are in the SupplementalMaterial.

Promoter activity determined by expression of GFP

The s32 promoter library was constructed as described pre-viously (Rhodius et al. 2006). Fifty s32 promoters validated inour previous work (Nonaka et al. 2006) were cloned as XhoI–BamHI fragments into the GFP reporter plasmid, pUA66. Re-porter strains were constructed by transforming derivatives ofpSAKT32 and promoter library plasmids into CAG57101 se-quentially using electroporation.



Promoter assays were performed by direct inoculation of LBbroth supplemented with appropriate antibiotics from freshtransformants. Fluorescence and OD600 were measured in aVarioskan spectrofluorometer (Thermo Electron Corporation).s32-dependent promoter activity was determined as described inthe Supplemental Material.

Calculation of information content of promoter motifs

The information content (Iseq) of aligned promoter motifs wascalculated using

Iseq = +i

+b

fb;i log2

fb;i

pb

;

where i is the position within the site, b refers to each of thepossible bases, fb,i is the observed frequency of each base at thatposition, and pb is the frequency of base b in the entire genome(in E. coli, taken to be 0.25 for A/G/C/T) (Schneider et al. 1986).

Electrophoretic mobility shift assay

PAGE-purified synthetic oligonucleotides were used for prepar-ing double-strand and fork junction probes. 32P-labeled 10 nMannealed DNA probe and 25 nM holoenzyme were mixed in 10mL of binding buffer and incubated for 10 min at 4°C. For heparinchallenge, 2 mL of 600 mg/mL heparin were added, and theincubation was continued for an additional 5 min. Electropho-resis was performed in a prechilled 5% acrylamide/TBE gel at4°C. See details in the Supplemental Material.

Acknowledgments

We thank Tania Baker for her critical input in conceptualizingthis work, Vivek Mutalik for technical assistance, and membersof the Gross laboratory for useful comments. This work wassupported by National Institutes of Health Grants GM057755 (toC.A.G.) and GM31808 (to P.L.dH.).

References

Amaya E, Khvorova A, Piggot PJ. 2001. Analysis of promoterrecognition in vivo directed by sF of Bacillus subtilis byusing random-sequence oligonucleotides. J Bacteriol 183:3623–3630.

Asayama M, Imamura S. 2008. Stringent promoter recognitionand autoregulation by the group 3 s-factor SigF in thecyanobacterium Synechocystis sp. strain PCC 6803. Nucleic

Acids Res 36: 5297–5305.Barne KA, Bown JA, Busby SJ, Minchin SD. 1997. Region 2.5 of

the Escherichia coli RNA polymerase s70 subunit is re-sponsible for the recognition of the ‘extended-10’ motif atpromoters. EMBO J 16: 4034–4040.

Britton RA, Eichenberger P, Gonzalez-Pastor JE, Fawcett P,Monson R, Losick R, Grossman AD. 2002. Genome-wideanalysis of the stationary-phase s factor (sH) regulon ofBacillus subtilis. J Bacteriol 184: 4881–4890.

Browning DF, Busby SJ. 2004. The regulation of bacterialtranscription initiation. Nat Rev Microbiol 2: 57–65.

Chilcott GS, Hughes KT. 2000. Coupling of flagellar geneexpression to flagellar assembly in Salmonella entericaserovar typhimurium and Escherichia coli. Microbiol Mol

Biol Rev 64: 694–708.Daniels D, Zuber P, Losick R. 1990. Two amino acids in an RNA

polymerase s factor involved in the recognition of adjacentbase pairs in the �10 region of a cognate promoter. Proc Natl

Acad Sci 87: 8075–8079.

Dombroski AJ, Walter WA, Record MT Jr, Siegele DA, GrossCA. 1992. Polypeptides containing highly conserved regionsof transcription initiation factor s70 exhibit specificity ofbinding to promoter DNA. Cell 70: 501–512.

Dombroski AJ, Walter WA, Gross CA. 1993. Amino-terminalamino acids modulate s-factor DNA-binding activity. Genes& Dev 7: 2446–2455.

Eichenberger P, Jensen ST, Conlon EM, van Ooij C, Silvaggi J,Gonzalez-Pastor JE, Fujita M, Ben-Yehuda S, Stragier P, LiuJS, et al. 2003. The sE regulon and the identification ofadditional sporulation genes in Bacillus subtilis. J Mol Biol

327: 945–972.Feklistov A, Barinova N, Sevostyanova A, Heyduk E, Bass I,

Vvedenskaya I, Kuznedelov K, Merkiene E, Stavrovskaya E,Klimasauskas S, et al. 2006. A basal promoter elementrecognized by free RNA polymerase s subunit determinespromoter recognition by RNA polymerase holoenzyme. MolCell 23: 97–107.

Fenton MS, Lee SJ, Gralla JD. 2000. Escherichia coli promoteropening and �10 recognition: Mutational analysis of s70.EMBO J 19: 1130–1137.

Gama-Castro S, Jimenez-Jacinto V, Peralta-Gil M, Santos-ZavaletaA, Penaloza-Spinola MI, Contreras-Moreira B, Segura-Salazar J,Muniz-Rascado L, Martinez-Flores I, Salgado H, et al. 2008.RegulonDB (version 6.0): Gene regulation model of Escher-ichia coli K-12 beyond transcription, active (experimental)annotated promoters and Textpresso navigation. Nucleic

Acids Res 36: D120–D124. doi: 10.1093/nar/gkm994.Gardella T, Moyle H, Susskind MM. 1989. A mutant Escher-

ichia coli s70 subunit of RNA polymerase with alteredpromoter specificity. J Mol Biol 206: 579–590.

Givens JR, McGovern CL, Dombroski AJ. 2001. Formation ofintermediate transcription initiation complexes at pfliD andpflgM by s28 RNA polymerase. J Bacteriol 183: 6244–6252.

Gourse RL, Ross W, Gaal T. 2000. UPs and downs in bacterialtranscription initiation: The role of the a subunit of RNApolymerase in promoter recognition. Mol Microbiol 37: 687–695.

Green HA, Donohue TJ. 2006. Activity of Rhodobacter sphaer-

oides RpoHII, a second member of the heat shock s factorfamily. J Bacteriol 188: 5712–5721.

Gruber TM, Gross CA. 2003. Multiple s subunits and thepartitioning of bacterial transcription space. Annu Rev

Microbiol 57: 441–466.Guisbert E, Yura T, Rhodius VA, Gross CA. 2008. Convergence

of molecular, modeling, and systems approaches for anunderstanding of the Escherichia coli heat shock response.Microbiol Mol Biol Rev 72: 545–554.

Guo Y, Gralla JD. 1998. Promoter opening via a DNA forkjunction binding activity. Proc Natl Acad Sci 95: 11655–11660.

Haugen SP, Berkmen MB, Ross W, Gaal T, Ward C, Gourse RL.2006. rRNA promoter regulation by nonoptimal binding of s

region 1.2: An additional recognition element for RNApolymerase. Cell 125: 1069–1082.

Helmann JD, deHaseth PL. 1999. Protein–nucleic acid interac-tions during open complex formation investigated by sys-tematic alteration of the protein and DNA binding partners.Biochemistry 38: 5959–5967.

Hook-Barnard IG, Hinton DM. 2007. Transcription initiation bymix and match elements: Flexibility for polymerase bindingto bacterial promoters. Gene Regul Syst Bio 1: 275–293.

Hook-Barnard IG, Hinton DM. 2009. The promoter spacerinfluences transcription initiation via s70 region 1.1 ofEscherichia coli RNA polymerase. Proc Natl Acad Sci 106:737–742.

Koo et al.


Hook-Barnard I, Johnson XB, Hinton DM. 2006. Escherichia coli

RNA polymerase recognition of a s70-dependent promoterrequiring a �35 DNA element and an extended �10 TGnmotif. J Bacteriol 188: 8352–8359.

Juang YL, Helmann JD. 1994. A promoter melting region in theprimary s factor of Bacillus subtilis. Identification of func-tionally important aromatic amino acids. J Mol Biol 235:1470–1488.

Koo BM, Rhodius VA, Campbell EA, Gross CA. 2009a. Dissec-tion of recognition determinants of Escherichia coli s32

suggests a composite �10 region with an ‘extended �10’motif and a core �10 element. Mol Microbiol 72: 815–829.

Koo BM, Rhodius VA, Campbell EA, Gross CA. 2009b. Muta-tional analysis of Escherichia coli s28 and its target pro-moters reveals recognition of a composite �10 region,comprised of an ‘extended �10 motif’ and a core �10element. Mol Microbiol 72: 830–843.

Kourennaia OV, deHaseth PL. 2007. Substitution of a highlyconserved histidine in the Escherichia coli heat shocktranscription factor, s32, affects promoter utilization in vitroand leads to overexpression of the biofilm-associated fluprotein in vivo. J Bacteriol 189: 8430–8436.

Kumar A, Malloch RA, Fujita N, Smillie DA, Ishihama A,Hayward RS. 1993. The �35-recognition region of Escher-

ichia coli s70 is inessential for initiation of transcription atan ‘extended �10’ promoter. J Mol Biol 232: 406–418.

Kusukawa N, Yura T. 1988. Heat shock protein GroE ofEscherichia coli: Key protective roles against thermal stress.Genes & Dev 2: 874–882.

Kuznedelov K, Minakhin L, Niedziela-Majka A, Dove SL,Rogulja D, Nickels BE, Hochschild A, Heyduk T, SeverinovK. 2002. A role for interaction of the RNA polymerase flapdomain with the s subunit in promoter recognition. Science295: 855–857.

Lee SJ, Gralla JD. 2003. Open complex formation in vitro by s38

(rpoS) RNA polymerase: Roles for region 2 amino acids. J Mol

Biol 329: 941–948.Li XY, McClure WR. 1998. Characterization of the closed

complex intermediate formed during transcription initiationby Escherichia coli RNA polymerase. J Biol Chem 273:23549–23557.

Lim HM, Lee HJ, Roy S, Adhya S. 2001. A ‘master’ in baseunpairing during isomerization of a promoter upon RNApolymerase binding. Proc Natl Acad Sci 98: 14849–14852.

Lisser S, Margalit H. 1993. Compilation of E. coli mRNApromoter sequences. Nucleic Acids Res 21: 1507–1516.

Lonetto M, Gribskov M, Gross CA. 1992. The s70 family:Sequence conservation and evolutionary relationships. J

Bacteriol 174: 3843–3849.McGrath PT, Lee H, Zhang L, Iniesta AA, Hottes AK, Tan MH,

Hillson NJ, Hu P, Shapiro L, McAdams HH. 2007. High-throughput identification of transcription start sites, con-served promoter motifs and predicted regulons. Nat Biotech-nol 25: 584–592.

Mecsas J, Cowing DW, Gross CA. 1991. Development of RNApolymerase–promoter contacts during open complex forma-tion. J Mol Biol 220: 585–597.

Minakhin L, Severinov K. 2003. On the role of the Escherichia

coli RNA polymerase s70 region 4.2 and a-subunit C-terminaldomains in promoter complex formation on the extended�10galP1 promoter. J Biol Chem 278: 29710–29718.

Mitchell JE, Zheng D, Busby SJ, Minchin SD. 2003. Identifica-tion and analysis of ‘extended �10’ promoters in Escherichia

coli. Nucleic Acids Res 31: 4689–4695.Murakami KS, Darst SA. 2003. Bacterial RNA polymerases: The

wholo story. Curr Opin Struct Biol 13: 31–39.

Murakami KS, Masuda S, Campbell EA, Muzzin O, Darst SA.2002. Structural basis of transcription initiation: An RNApolymerase holoenzyme-DNA complex. Science 296: 1285–1290.

Nakahigashi K, Yanagi H, Yura T. 1995. Isolation and sequenceanalysis of rpoH genes encoding s32 homologs from gramnegative bacteria: Conserved mRNA and protein segmentsfor heat shock regulation. Nucleic Acids Res 23: 4383–4390.

Nonaka G, Blankschien M, Herman C, Gross CA, Rhodius VA.2006. Regulon and promoter analysis of the E. coli heat-shock factor, s32, reveals a multifaceted cellular response toheat stress. Genes & Dev 20: 1776–1789.

Paget MS, Helmann JD. 2003. The s70 family of s factors.Genome Biol 4: 203. doi: 10.1186/gb-2003-4-1-203.

Ponnambalam S, Webster C, Bingham A, Busby S. 1986.Transcription initiation at the Escherichia coli galactoseoperon promoters in the absence of the normal �35 regionsequences. J Biol Chem 261: 16043–16048.

Rhodius VA, Suh WC, Nonaka G, West J, Gross CA. 2006.Conserved and variable functions of the sE stress response inrelated genomes. PLoS Biol 4: e2. doi: 10.1371/journal.pbio.0040002.

Ross W, Gosink KK, Salomon J, Igarashi K, Zou C, Ishihama A,Severinov K, Gourse RL. 1993. A third recognition elementin bacterial promoters: DNA binding by the a subunit ofRNA polymerase. Science 262: 1407–1413.

Schneider TD, Stormo GD, Gold L, Ehrenfeucht A. 1986.Information content of binding sites on nucleotide se-quences. J Mol Biol 188: 415–431.

Schroeder LA, Gries TJ, Saecker RM, Record MT Jr, Harris ME,deHaseth PL. 2009. Evidence for a tyrosine-adenine stackinginteraction and for a short-lived open intermediate sub-sequent to initial binding of Escherichia coli RNA poly-merase to promoter DNA. J Mol Biol 385: 339–349.

Serizawa M, Yamamoto H, Yamaguchi H, Fujita Y, Kobayashi K,Ogasawara N, Sekiguchi J. 2004. Systematic analysis of SigD-regulated genes in Bacillus subtilis by DNA microarray andNorthern blotting analyses. Gene 329: 125–136.

Shen L, Feng X, Yuan Y, Luo X, Hatch TP, Hughes KT, Liu JS,Zhang YX. 2006. Selective promoter recognition by chla-mydial s28 holoenzyme. J Bacteriol 188: 7364–7377.

Shultzaberger RK, Chen Z, Lewis KA, Schneider TD. 2007.Anatomy of Escherichia coli s70 promoters. Nucleic Acids

Res 35: 771–788.Siegele DA, Hu JC, Walter WA, Gross CA. 1989. Altered pro-

moter recognition by mutant forms of the s70 subunit ofEscherichia coli RNA polymerase. J Mol Biol 206: 591–603.

Sorenson MK, Darst SA. 2006. Disulfide cross-linking indicatesthat FlgM-bound and free s28 adopt similar conformations.Proc Natl Acad Sci 103: 16722–16727.

Sorenson MK, Ray SS, Darst SA. 2004. Crystal structure of theflagellar s/anti-s complex s28/FlgM reveals an intact s factorin an inactive conformation. Mol Cell 14: 127–138.

Tatti KM, Jones CH, Moran CP Jr. 1991. Genetic evidence forinteraction of sE with the spoIIID promoter in Bacillus

subtilis. J Bacteriol 173: 7828–7833.Thouvenot B, Charpentier B, Branlant C. 2004. The strong

efficiency of the Escherichia coli gapA P1 promoter dependson a complex combination of functional determinants.Biochem J 383: 371–382.

Tomsic M, Tsujikawa L, Panaghie G, Wang Y, Azok J, deHasethPL. 2001. Different roles for basic and aromatic amino acidsin conserved region 2 of Escherichia coli s70 in the nucle-ation and maintenance of the single-stranded DNA bubble inopen RNA polymerase-promoter complexes. J Biol Chem

276: 31891–31896.



Tsujikawa L, Tsodikov OV, deHaseth PL. 2002. Interaction ofRNA polymerase with forked DNA: Evidence for twokinetically significant intermediates on the pathway to thefinal complex. Proc Natl Acad Sci 99: 3493–3498.

Ueki T, Inouye S. 2001. SigB, SigC, and SigE from Myxococcusxanthus homologous to s32 are not required for heat shockresponse but for multicellular differentiation. J Mol Micro-

biol Biotechnol 3: 287–293.van Schaik W, van der Voort M, Molenaar D, Moezelaar R, de

Vos WM, Abee T. 2007. Identification of the sB regulon ofBacillus cereus and conservation of sB-regulated genes inlow-GC-content gram-positive bacteria. J Bacteriol 189:4384–4390.

Wade JT, Roa DC, Grainger DC, Hurd D, Busby SJ, Struhl K,Nudler E. 2006. Extensive functional overlap between s

factors in Escherichia coli. Nat Struct Mol Biol 13: 806–814.Waldburger C, Gardella T, Wong R, Susskind MM. 1990.

Changes in conserved region 2 of Escherichia coli s70

affecting promoter recognition. J Mol Biol 215: 267–276.Young BA, Gruber TM, Gross CA. 2004. Minimal machinery of

RNA polymerase holoenzyme sufficient for promoter melt-ing. Science 303: 1382–1384.

Yu HH, Tan M. 2003. s28 RNA polymerase regulates hctB, a latedevelopmental gene in Chlamydia. Mol Microbiol 50: 577–584.

Yura T, Nakahigashi K. 1999. Regulation of the heat-shockresponse. Curr Opin Microbiol 2: 153–158.

Zhao K, Liu M, Burgess RR. 2005. The global transcriptionalresponse of Escherichia coli to induced s32 protein involvess32 regulon activation followed by inactivation and degrada-tion of s32 in vivo. J Biol Chem 280: 17758–17768.

Zhao K, Liu M, Burgess RR. 2007. Adaptation in bacterialflagellar and motility systems: From regulon members to‘foraging’-like behavior in E. coli. Nucleic Acids Res 35:4441–4452.

Koo et al.


reduced capacity of alternative s to melt promoters ensures stringent promoter recognition

Documents