A Family of Anti-σ70 Proteins in T4-type Phages and Bacteria that are Similar to AsiA, a...

doi:10.1016/j.jmb.2004.10.003 J. Mol. Biol. (2004) 344, 1183–1197

A Family of Anti-s70 Proteins in T4-type Phages andBacteria that are Similar to AsiA, a TranscriptionInhibitor and Co-activator of Bacteriophage T4

Melissa Pineda1, Brian D. Gregory2, Bridget Szczypinski3

Kimberly R. Baxter1, Ann Hochschild2, Eric S. Miller3 andDeborah M. Hinton1*

1Laboratory of Molecular andCellular Biology, NationalInstitute of Diabetes andDigestive and Kidney DiseasesBuilding 8, Room 2A-13National Institutes of HealthBethesda, MD 20892-0830USA

2Department of Microbiologyand Molecular GeneticsHarvard Medical School, BostonMA 02115, USA

3Department of MicrobiologyNorth Carolina State UniversityRaleigh, NC 27695, USA

0022-2836/$ - see front matter Published

Present address: M. Pineda, EasteSchool, Norfolk, VA, USA.Abbreviations used: PCR, polym

ITPG, isopropyl-beta-D-thiogalactopE-mail address of the correspond

[email protected]

Anti-s70 factors interact with s70 proteins, the specificity subunits ofprokaryotic RNA polymerase. The bacteriophage T4 anti-s70 protein, AsiA,binds tightly to regions 4.1 and 4.2 of the s70 subunit of Escherichia coli RNApolymerase and inhibits transcription from s70 promoters that requirerecognition of the canonical s70 K35 DNA sequence. In the presence of theT4 transcription activator MotA, AsiA also functions as a co-activator oftranscription from T4 middle promoters, which retain the canonical s70

K10 consensus sequence but have a MotA box sequence centered at K30rather than the s70 K35 sequence. The E. coli anti-s70 protein Rsd alsointeracts with region 4.2 of s70 and inhibits transcription from s70

promoters. Our sequence comparisons of T4 AsiA with Rsd, with thepredicted AsiA orthologs of the T4-type phages RB69, 44RR, KVP40, andAeh1, and with AlgQ, a regulator of alginate production in Pseudomonasaeruginosa indicate that these proteins share conserved amino acid residuesat positions known to be important for the binding of T4 AsiA to s70 region4. We show that, like T4 AsiA, Rsd binds to s70 in a native protein gel and,as with T4 AsiA, a L18S substitution in Rsd disrupts this complex. Previouswork has assigned s70 amino acid F563, within region 4.1, as a criticaldeterminant for AsiA binding. This residue is also involved in the bindingof s70 to the b-flap of core, suggesting that AsiA inhibits transcription bydisrupting the interaction between s70 region 4.1 and the b-flap. We findthat as with T4 AsiA, the interaction of KVP40 AsiA, Rsd, or AlgQ with s70

region 4 is diminished by the substitution F563Y. We also demonstrate thatlike T4 AsiA and Rsd, KVP40 AsiA inhibits transcription from s70-dependent promoters. We speculate that the phage AsiA orthologs, Rsd,and AlgQ are members of a related family in T4-type phage and bacteria,which interact similarly with primary s factors. In addition, we show thateven though a clear MotA ortholog has not been identified in the KVP40genome and the phage genome appears to lack typical middle promotersequences, KVP40 AsiA activates transcription from T4 middle promotersin the presence of T4MotA.We speculate that KVP40 encodes a protein thatis dissimilar in sequence, but functionally equivalent, to T4 MotA.

Published by Elsevier Ltd.

Keywords: phage; AsiA; Rsd; AlgQ; transcription
*Corresponding author
by Elsevier Ltd.

rn Virginia Medical

erase chain reaction;yranoside.ing author:

Introduction

s Factors provide RNA polymerase with theability to recognize and initiate transcription fromdiscrete positions within the prokaryotic genome.1,2

Primary s factors, such as s70 of Escherichia coli, arethe predominant recognition factors for the tran-scription of housekeeping genes, while other s

1184 AsiA-like Proteins in Phage and Bacteria

factors are used under certain growth conditions orat times of stress. Anti-s factors interact with sproteins, inhibiting or modulating the activity ofRNA polymerase (reviewed by Hughes &Mathee3).Two such proteins are the AsiA protein of bacterio-phage T44,5 and the Rsd protein of E. coli.6 AsiA actsas both an inhibitor of s70-dependent promotertranscription and as a co-activator for the middleclass of T4 promoters.5,7,8 AsiA binds tightly toregion 4 of s70,9–15 the region which also includesamino acid residues that interact with the K35element of s70 promoters.16–21 The presence of AsiAinhibits the ability of s70 to recognize promotersthat use these canonical K35 elements.10,22,23

However, in the presence of the T4 transcriptionactivator MotA, AsiA co-activates transcriptionfrom T4 middle promoters. These promoters con-tain the s70 K10 recognition element but have a9 bp motif (MotA box) centered at K30 rather thanthe s70K35 DNA element.8,24 Like AsiA, MotA alsointeracts directly with region 4 of s70.25,26 Thus,AsiA and MotA, through their interactions withregion 4 of s70, modify RNA polymerase and enableit to recognize a different DNA binding elementwithin the upstream portion of promoter DNA.

The E. coli anti-s70 Rsd protein was first identifiedas a protein that associates with s70 as cells enterstationary phase.6 Like AsiA, Rsd interacts withregion 4 of s70 and functions as an inhibitor of s70

transcription.6,27–30 However, it has been proposedthat in this case the association of Rsd with s70

prevents s70 from binding to core, thus increasingthe pool of free core that is available for bindingwith stationary phase ss.6,27,31 Recent evidenceindicates that Rsd can also bind to core and toholoenzyme,31 suggesting that the mechanism ofRsd inhibition may be more complex. Rsd alsoshares significant sequence homology with aregulator of alginate production in Pseudomonasaeruginosa, AlgQ,6,32,33 but the mechanism of AlgQaction is not yet clear.34

T4 is a member of the Myoviridae family ofphages, which includes bacteriophages having adistinctive contractile tail morphology (reviewed byAckermann & Krisch35). Despite general commonmorphological features, T4-type phages within thisfamily may share only limited sequence homology.Phylogenetic analyses using phage capsid geneshave shown that the classic T-even coliphages T4and T6 are nearly identical and are similar to theT-even coliphage RB69while other T4-type members(pseudoT-even and schizoT-even), including thoseisolated on non-E. coli hosts, are more distantlyrelated to T4.36–38 One such phage, KVP40, wasisolated from sea water and infects a broad range ofVibrio species.39 Genomic sequencing of this phageidentified a 99 amino acid open reading framewhoseoverall predicted sequence is 27% identical with thatof T4 AsiA.40 Curiously, KVP40 was not observed tocontain a recognizable MotA ortholog or typicalmiddle promoter sequences.40

Here, we have compared the sequence andfunctions of some of the phage AsiA orthologs to

those of Rsd and AlgQ. We provide evidence thatthese phage and bacterial proteins containregions of sequence similarity and share specificbinding epitopes within s70 regions 4.1 and 4.2.We speculate that these proteins belong to afamily of anti-s proteins from T4-type phages andbacteria, which interact similarly with primary sfactors.

Results

Sequence comparisons of T4 AsiA protein withputative AsiA proteins from T4-type phages andwith putative regulators of bacterial primarys factors

Examination of the genomic sequences of acollection of T4-type phages isolated from a rangeof bacterial hosts has revealed the presence ofproteins that are similar to the 90 amino acid AsiAprotein of bacteriophage T4. These putative AsiAproteins were identified in the T-even coliphageRB69, the pseudoT-even Aeromonas salmonicidaphage 44RR, the schizoT-even Aeromonas hydrophilaphage Aeh1, and the schizoT-even Vibrioparahaemolyticus phage KVP40 (Figure 1). Thedegree of sequence identity among these proteinsequences (Table 1) is consistent with previousphylogenetic analyses using the protein sequence ofthe phage capsid gene 23.38 As was seen with gp23,the AsiA protein sequences of the T-even phages T4and RB69 are quite similar to each other, but aremuch less similar to those of schizoT-even orpseudoT-even phages. Thus, T4 AsiA and theKVP40 AsiA ortholog represent very disparatemembers of this phage group.

Figure 1 also shows an alignment of the phageAsiA proteins with the N-terminal 82 amino acidresidues of the E. coli protein Rsd41 and theN-terminal 77 amino acid residues of theP. aeruginosa protein AlgQ.32 Rsd is known to bindto s70 and is thought to aid in switching the activepolymerase from core plus s70 to core plus ss, thestationary phase specific s factor.6 AlgQ interactswith the primary s factor of P. aeruginosa29 and isinvolved in the regulation of alginate production.32,33

Previous analyses of the Rsd and AlgQ proteinsequences revealed that they share significantsequence identity within the first half of theproteins6 (see also Figure 1 and Table 1). Further-more, both Rsd and AlgQ are known to interactwith region 4 of their respective target sfactors.28,29,31 Although previous sequence com-parisons between just T4 AsiA and E. coli Rsd led tothe conclusion that the proteins are not similar,6 thealignment of these seven proteins together suggestsotherwise (Figure 1). We propose that these proteinscomprise an anti-s family present in both T4-typephages and bacteria.

Figure 1. Sequence similarities among the T4 AsiA protein (90 amino acid residues4), the KVP40 AsiA protein (99 amino acid residues40), the N-terminal 1–82 amino acidresidues of the E. coli anti-s70 protein Rsd (158 amino acid residues total),6 the N-terminal 1–77 amino acid residues of AlgQ (160 amino acid residues total),6,41 a putativeregulator of the primary s factor of P. aeruginosa, and the predicted AsiA proteins of the bacteriophages RB69 (90 amino acid residues), Aeh1 (97 amino acid residues), and 44RR(89 amino acid residues) (see Experimental Procedures for sequence accession numbers). Amino acid positions are based on the T4 AsiA sequence. Amino acid residues that areconserved in at least four out of the seven proteins are darker with identical amino acid residues in black. The positions of a-helices in the reported solution structure of T4AsiA15,50 are indicated at the top. The phage consensus is based on a 5 out of 5 conservation match among the phage proteins. If the amino acid is identical, it is shown as acapital letter; if the amino acid is conserved, then the most prevalent amino acid at that site is shown in lower case. The overall consensus is based on a 7 out of 7 conservationmatch, with the letter indicating the most prevalent amino acid. Arrows indicate positions where mutations within T4 AsiA have been found to weaken its interaction with s70

region 4.42,51

Table 1. % Identity among proteins in the AsiA anti-s70 family

RB69 (%) 44RR (%) T4 (%) Aeh1 (%) KVP40 (%) Rsd (%) AlgQ (%)

RB69 (%) 10044RR (%) 37 100T4 (%) 73 34 100Aeh1 (%) 31 40 28 100KVP40 (%) 24 38 27 38 100Rsd (%) 18 15 17 16 17 100AlgQ (%) 13 11 16 13 9 41 100

Protein sequences were compared and identical amino acid residues determined using the alignment shown in Figure 1.


KVP40 AsiA, like T4 AsiA and E. coli Rsd,interacts with region 4 of s70

An indirect measure of the ability of T4 AsiA toinhibit s70-directed transcription is an in vivotoxicity assay, in which cell growth is monitoredafter the induction of T4 AsiA protein synthesis.12,42

We constructed plasmids containing His6-taggedderivatives of T4 AsiA, KVP40 AsiA, Aeh1 AsiA,and the bacterial proteins Rsd and AlgQ andfollowed cell growth after addition of IPTG toinduce the synthesis of the recombinant proteins(Figure 2). As has been seen for T4 AsiA,42 thesynthesis of the KVP40 or the Aeh1 AsiA orthologsresulted in an arrest of cell growth within about onehour. Continued incubation resulted in a decreasein the A600, consistent with cell death. In contrast,the induced synthesis of Rsd only modestlyinhibited cell growth, and the synthesis of AlgQhad practically no effect. This result was notbecause the level of His6Rsd or His6AlgQ was low.SDS-PAGE of cell extracts indicated these proteinswere expressed at least as well as the phage AsiAproteins (data not shown). Previously, Jishage &Ishihama27 observed no inhibition of cell growthafter the induction of Rsd synthesis with arabinosewhen the rsd gene was under the control of the PBAD

promoter.

Figure 2. Effect of the synthesis of various anti-s proteincontaining either the vector, pet28(aC), or the indicated antibefore and after the induction of anti-s protein synthesis by

A L18S substitution in T4 AsiA results in aprotein that is significantly less toxic whenexpressed in E. coli.42 As seen in Figure 2B, thecorresponding substitution in Rsd also reduced thetoxicity of this protein.

The interaction of T4 AsiA with s70 can bedetected using a two-hybrid assay in E. coli.29,42 Inthis assay, contact between a protein (or proteindomain) fused to a DNA-binding protein and apartner domain fused to a component of RNApolymerase activates transcription from a testpromoter bearing a recognition site for the DNA-binding protein in the upstream region of thepromoter (see Figure 3A). Thus, fusion of T4 AsiAto the bacteriophage l cI protein (lcI) and fusion ofthe C-terminal region of s70 (from region 3.2through region 4) to the N-terminal domain of thea subunit of RNA polymerase (the a-NTD) permitsdetection of the protein–protein interaction betweenthe fused AsiA and s70 region 4 moieties.

We used this two-hybrid assay to test whetherKVP40 AsiA, like T4 AsiA, interacts with s70 region4. Figure 3B shows that the lcI-KVP40 AsiA fusionprotein activated transcription from the test pro-moter specifically in the presence of the a-s70

chimera suggesting that, like the T4 AsiA protein,the KVP40 AsiA ortholog interacts with region 4 ofs70. Because region 4 of s70 in V. cholerae is nearly

s on E. coli growth. Growth of E. coli pLysE/BL21(DE3)-s factor plasmids was monitored by measuring the A600

the addition of IPTG (indicated by the arrow).

Figure 3. Bacterial two-hybridassay detects interaction betweenKVP40 AsiA and s70 region 4.A, The diagram depicts test pro-moter plac2lop, which bears aconsensus l operator sequencecentered 93 bp upstream and thel operator OL2 centered 62 bpupstream from the initiation pointof the lac core promoter. In strainF 093C62 this test promoter islinked to lacZ on an F 0 episome.Replacement of the RNA polymer-ase a-CTD by a fragment of s70

that harbors region 4 permits inter-action with the AsiA moiety of alcI-KVP40 AsiA fusion proteinbound to either of the two loperator sites associated with thetest promoter. B, The graph showsthe effect of the lcI-KVP40 AsiAfusion protein on transcriptionin vivo from plac2lop in the pre-sence of the a-s70 D581G chimeraor the a-s70 D581G/F563Y chi-mera. F 093C62 cells harboringcompatible plasmids directing thesynthesis of the indicated proteinswere grown in the presence ofdifferent concentrations of IPTGand assayed for b-galactosidase.Plasmid pACD-35lcI-KVP40 AsiAdirected the synthesis of the lcI-KVP40 AsiA fusion protein andplasmids pBRa-s70 D581G, pBRa-s70 D581G/F563Y, and pBRadirected the synthesis of the a-s70

D581G chimera, the a-s70

D581G/F563Y chimera and full-length a. C, The graph shows theeffect of the lcI-b flap fusion

protein on transcription in vivo from plac2lop in the presence of the a-s70 D581G chimera or the a-s70 D581G/F563Ychimera. Here, replacement of the RNA polymerase a-CTD by a fragment of s70 that harbors region 4 permits interactionwith the b flap moiety of the DNA-bound lcI-b flap fusion protein. F 093C62 cells harboring compatible plasmidsdirecting the synthesis of the indicated proteins were grown in the presence of different concentrations of IPTG andassayed for b-galactosidase. Plasmid pAClcI-b flap directed the synthesis of the lcI-b flap fusion protein.

AsiA-like Proteins in Phage and Bacteria 1187

identical with that of s70 from E. coli,43 it isreasonable to assume that the KVP40 AsiA proteinwould bind tightly to region 4 of its host s70.

Effect of F563Y substitution in s70 region 4 on itsinteraction with anti-s factors

Two patches of amino acid residues within region4 of s70, amino acid residues 551–563 in region 4.1and amino acid residues 580–598 in region 4.2interact with T4 AsiA.15,44,45 F563 of s70 is a criticalamino acid for the s70/T4 AsiA interaction. Itmakes direct contact with amino acid residueswithin T4 AsiA15 and a F563Y substitution specifi-cally disrupts the interaction between s70 region 4and T4 AsiA in the two-hybrid assay as well as theinteraction between full-length s70 and T4 AsiA asmeasured in vitro.14 To determine whether KVP40

AsiA also recognizes this s70 determinant, we usedthe two-hybrid assay to test the effect of the F563Ysubstitution in s70 region 4 on its interaction withKVP40 AsiA (Figure 3B). We found that thissubstitution disrupted the interaction of the fusedKVP40 AsiA moiety with s70 region 4, analogous toits effect on the T4 AsiA/s70 region 4 interaction.14

The experiment shown in Figure 3C is a controlthat demonstrates that the F563Y substitution in thefused s70 region 4 moiety had no effect on theinteraction of s70 region 4 with the flexible flapdomain of the RNA polymerase b subunit (theb-flap), which normally binds s70 region 4 in thecontext of the RNA polymerase holoenzyme. Wenote that for this set of experiments we used avariant of the a-s70 chimera bearing an amino acidsubstitution (D581G) that apparently stabilizes thefolded structure of the tethered s70 region 4 moiety

Figure 4. Effect of F563Y substi-tution in s70 region 4 on its inter-actions with Rsd and AlgQ, asdetected in the bacterial two-hybrid assay. A, The diagramdepicts test promoter placOR2–62,which bears a single l operator(OR2) centered 62 bp upstreamfrom the initiation point of the laccore promoter. In strain F 0OR2–62this test promoter is linked to lacZon an F 0 episome. B, The graphshows the effect of the F563Ysubstitution in the s70 moiety ofthe a-s70 D581G chimera on lcI-Rsd-stimulated transcription in vivofrom test promoter placOR2–62.F 0OR2–62 cells harboring com-patible plasmids directing the syn-thesis of the indicated proteinswere grown in the presence ofdifferent concentrations of IPTGand assayed for b-galactosidase.Plasmid pAClcI-Rsd directed thesynthesis of the lcI-Rsd fusionprotein and plasmids pBRa-s70

D581G, pBRa-s70 D581G/F563Y,and pBRa directed the synthesisof the a-s70 D581G chimera, the a-s70 D581G/F563Y chimera andfull-length a. C, The graph showsthe effect of the F563Y substitutionin the s70 moiety of the a-s70

D581G chimera on lcI-AlgQ-stimu-lated transcription in vivo from testpromoter placOR2–62. F 0OR2–62cells harboring compatible plas-mids directing the synthesis of theindicated proteins were grown inthe presence of different concen-trations of IPTG and assayed for b-galactosidase. Plasmid pAClcI-AlgQ directed the synthesis of thelcI-AlgQ fusion protein. D, Thegraph shows the effect of the F563Ysubstitution in the s70 moiety of thea-s70 D581G chimera on lcI-b flap-stimulated transcription in vivofrom test promoter placOR2–62.F 0OR2–62 cells harboring compati-ble plasmids directing the syn-thesis of the indicated proteinswere grown in the presence ofdifferent concentrations of IPTGand assayed for b-galactosidase.Plasmid pAClcI-b flap directedthe synthesis of the lcI-b flapfusion protein.


and therefore facilitates the detection of inter-actions, such as the s70 region 4/b-flap interaction,that are near the threshold of detection in the two-hybrid assay.46

Previous work has shown that the interactionbetween s70 region 4 and Rsd or AlgQ can also bedetected using the bacterial two-hybrid assay.29

Accordingly, we used the two-hybrid assay to testthe effect of the F563Y substitution in s70 region 4 onits interaction with either Rsd or AlgQ (Figure 4).Figure 4B and C shows that the F563Y substitutionin the fused s70 region 4 moiety weakenedthe interaction of both Rsd and AlgQ with s70

region 4, the effect being particularly dramatic in

Figure 5. Rsd protein, like T4 AsiA, makes a complex with s70 that is stable to gel electrophoresis. The gels (native 6%polyacrylamide) show s70 alone (lanes 1 and 4) or s70 after incubation with the indicated proteins (lanes 2, 3, 5–10). Thepositions of free s70 and the various protein–protein complexes are indicated.


the case of AlgQ. Once again, a control experimentdemonstrates that the F563Y substitution had noeffect on the interaction of s70 with the b-flap(Figure 4D).

From these experiments, we conclude that resi-due F563 of s70 defines a common determinant inregion 4.1 of primary s factors that is recognized bythe phage proteins, T4 AsiA and KVP40 AsiA, andthe bacterial proteins, Rsd and AlgQ.

Like T4 AsiA, Rsd forms a stable complex withs70 in a native protein gel

To investigate the ability of Rsd or AlgQ to formstable complexes with s70, we isolated His6Rsd, and

105 dpm/pmol) in a final volume of 5 ml containing 11 mMNaCl, 9.4% glycerol, 0.34 mM EDTA, 0.14 mM DTT, 0.001magnesium acetate, 100 mM bovine serum albumin, and 5denaturing gels as described in Experimental Procedures. TMotA-independent Pminor RNA are indicated.

His6AlgQ after NiCC resin affinity chromatography(see Experimental Procedures) and performednative protein gel assays in vitro. Like T4 AsiA(Figure 5, lane 6; see also the work done by Gerber& Hinton25), the Rsd protein formed a discretecomplex with s70 that was stable to gel electro-phoresis (lanes 2 and 5), supporting the idea thatRsd is capable of forming a relatively tight complexwith free s70. AlgQ failed to form a complex withs70 that was stable to electrophoresis (data notshown), consistent with its inability to significantlyinhibit the growth of E. coli in vivo. We could not testthe KVP40 protein in this assay because we wereunable to obtain a sufficiently purified fraction ofthe KVP40 protein (see Experimental Procedures).

Figure 6. KVP40 AsiA and E. coliRsd inhibit transcription from PuvsX

in the absence of T4 MotA andKVP40 AsiA activates transcriptionfrom PuvsX in the presence of T4MotA. s70 (0.4 pmol in 0.8 ml of s70

buffer) and either T4 AsiAHis6,KVP40 AsiAHis6 or His6Rsd(4 pmol in 0.25 pmol of anti-s70

buffer) were incubated at 37 8C forten minutes. Transcription wasinitiated by the addition of a0.2 pmol core, 0.02 pmol of linearpDKT90 DNA (PuvsX template),1.9 pmol MotA (where indicated),200 mM each ATP, CTP, andGTP and 10 mM UTP (7!

Tris–Cl (pH 7.9), 40 mM Tris–acetate (pH 7.9), 39 mM6% Triton X-100, 150 mM potassium glutamate, 4 mMmM immidizole. Samples were prepared and run onhe positions of the MotA-activated PuvsX RNA and the

Table 2. Effect of Rsd, RsdL18S, or AlgQ proteins ontranscription from PuvsX

Relative level of PuvsX RNA

ProteinNone 1.0G0.06Rsd 0.63G0.07RsdL18S 0.72G0.13AlgQ 0.85G0.11

s70 (7 pmol in 4 ml of s70 buffer) and the indicated anti-s protein(48 pmol in 4 pmol of anti-s70 buffer) were first incubated at 37 8Cfor ten minutes and then on ice for five minutes. Transcriptionreactions were assembled by adding 0.5 ml of this solution(0.44 pmol s70, 3.75 pmol anti-s) to 0.6 pmol core and 0.02 pmolof linear pDKT90 DNA (PuvsX template) in a final reactionsolution (5 ml) containing 200 mM each ATP, CTP, and GTP, 10 mMUTP (7!105 dpm/pmol), 8 mM Tris–Cl (pH 7.9), 40 mM Tris–acetate (pH 7.9), 45 mM NaCl, 6.2% glycerol, 0.24 mM EDTA,0.17 mM DTT, 0.0005% Triton X-100, 150 mM potassiumglutamate, 4 mM magnesium acetate, 100 mM bovine serumalbumin, and 5 mM immidizole. Samples were prepared and runon denaturing gels as described in Experimental Procedures. Thevalues are relative to that obtained in the absence of anti-sprotein and were averaged from three experiments.


A L18S substitution in T4 AsiA decreases AsiAinhibition of s70-dependent transcription because itdecreases the ability of T4 AsiA to form a complexwith s70.42 To test the effect of this substitution inRsd, we purified the His6RsdL18S protein andassayed its ability to form a complex with s70 inthe native protein gel (Figure 5, lane 3). As had beenobservedwith T4 AsiA, the L18S substitution in Rsdalso resulted in a protein that was defective informing a complex with s70.

In assays in which both Rsd and T4 AsiA werepresent, only the T4 AsiA/s70 complex wasobserved, even when Rsd was incubated with s70

before the addition of T4 AsiA (data not shown).This result suggests that T4 AsiA can compete withRsd for s70 and is consistent with the idea that theT4 anti-s70 protein binds more tightly to s70 thandoes Rsd. In the presence of T4 MotA, T4 AsiA, ands70, a s70/T4 AsiA/T4 MotA ternary complex isobserved (Figure 5, lane 10; see also the work doneby Gerber & Hinton25). A species that migrates justslightly more slowly than the s70/T4 MotA com-plex was also observed upon addition of both Rsdand T4 MotA to s70 (lane 9). This result suggeststhat a ternary complex of s70/Rsd/T4 MotA is alsopossible.

Figure 7.KVP40 AsiA associates with RNA polymeraseholoenzyme. The indicated anti-s proteins were firstincubated with s70 and then with added core beforechromatography through S-200HR spin columns asdescribed in Experimental Procedures. Proteins in thefirst fraction off the column were separated by SDS-PAGEand detected by silver staining. The positions of the RNApolymerase subunits b, b 0, s70, a, and u, and of T4 AsiAand KVP40 AsiA are indicated. The expected position ofRsd, which is not found associated with polymerase afterthis treatment, is also shown.

Both Rsd and KVP40 AsiA inhibit s70-dependenttranscription, but only KVP40 AsiA alsofunctions as a co-activator for T4 MotA-dependent transcription

The T4 middle promoter PuvsX, like other T4middle promoters, has a K10 region that matchesthe s70 K10 consensus sequence, but contains aMotA box motif centered at K30 rather than acanonical s70 K35 recognition sequence. As wehave documented,8 polymerase alone transcribesfrom PuvsX in vitro, presumably because of thepromoter’s perfect match to the s70 K10 recog-nition sequence. This transcription is activated bythe presence of both T4 AsiA and MotA, but it isinhibited by T4 AsiA in the absence of MotA.8 Asshown in Figure 6, KVP40 AsiA behaved similarlyto T4 AsiA in a transcription assay with PuvsX. In theabsence of T4 MotA, addition of KVP40 AsiAinhibited transcription (lane 5) while in thepresence of T4 MotA, addition of KVP40 AsiAactivated transcription (lane 6). As expected, aMotA-independent promoter (Pminor) was not acti-vated by the presence of AsiA and MotA. Toinvestigate whether there were differences betweenKVP40 AsiA and T4 AsiA that were not apparentwith this PuvsX template, we assayed transcriptionusing a minimal PuvsX that contains T4 sequencesfrom K34 to C83 (Pmotbox) and a MotA-dependentpromoter47 that has a MotA box sequence thatdeviates significantly from the MotA box consensussequence (PC39–C53). As has been seen with T4AsiA,47 KVP40 AsiA functioned as a co-activator ateither of these promoters also (data not shown). Weconclude that the KVP40 protein, like T4 AsiA, iscapable of acting both as a transcription inhibitor at

s70-dependent promoters and as a transcription co-activator at T4 MotA-dependent promoters. Thus,the T4 AsiA protein is not unique in being able tofunction as a co-activator for MotA-dependenttranscription.

In contrast to KVP40 AsiA, Rsd only weaklyinhibited PuvsX transcription either in the presenceor absence of MotA (Figure 6, compare lanes 7versus 11 and lanes 8 versus 12, Table 2) and AlgQ


did not significantly inhibit s70-dependent tran-scription from PuvsX in vitro (Table 2). The weaklevel of inhibition by Rsd is consistent withprevious reports, indicating that for most of thepromoters that have been tested in vitro, Rsdinhibits transcription twofold or less.6 In addition,it is consistent with our finding that the growth ofE. coli containing pHis6Rsd was only somewhatslowed, but not prevented, upon the induction ofRsd protein synthesis (Figure 2). RsdL18S still hadsome inhibitory activity, which was also consistentwith its retaining some inhibition of E. coli growthin vivo.

Addition of T4 MotA to the reaction containingRsd slightly increased the level of PuvsX transcrip-tion (Figure 6, lane 12), but this level was still lessthan that observed in the presence of T4MotA alone(lane 8). Transcription reactions with Rsd and theminimal PuvsX promoter also resulted in modestinhibition in the absence of MotA and no activationin the presence of MotA (data not shown). Theseresults suggest that unlike T4 AsiA and KVP40AsiA, Rsd cannot function as a co-activator with T4MotA.

KVP40 AsiA associates with RNA polymeraseholoenzyme

T4 AsiA associates with RNA polymeraseholoenzyme by a two-step process.48 It first bindsto free s70 and then the T4 AsiA/s70 complexassociates with core to form the AsiA-containingholoenzyme (T4 AsiA/s70/core). AsiA associatedwith polymerase can be separated from free AsiA,free s70 and the AsiA/s70 complex by chromato-graphy of the incubated proteins through S-200 HRspin columns48 (Figure 7). Native protein gelsindicated that the amount of AsiA protein used inthis experiment was sufficient to bind about half ofthe s70 (data not shown). After chromatography,SDS-PAGE of the excluded fraction showed that T4AsiA protein had associated with polymerase(Figure 7, lane 2). A similar result was observed inthe experiment using KVP40 AsiA (lane 3). Incontrast to the result with the phage AsiA proteins,Rsd was not found associated with polymerase inthis experiment (lane 4). This result suggests thatunder these conditions, either Rsd does not associ-ate with s70 when it is present in holoenzyme or theRsd/RNA polymerase complex is less stable thanAsiA/polymerase.

Discussion

Identification of a set of anti-s70 proteins inT4-type phages and bacteria

Many phages can be classified as T4-type bymorphology.35 However, genomic sequencing ofseveral of these phages has indicated that there arewide variations in the degree of sequence homologywithin the T4-type genomes.38 Based on sequence

homology, these phages have been classified intothree general groups: T-even, which includes thenearly identical coliphages T4, T2, and T6 as well asthe less homologous coliphage RB69; pseudoT-even, which includes A. salmonicida phage 44RR;and schizoT-even, which includes A. hydrophilaphage AehI and V. parahaemolyticus phage KVP40.38

Sequence analyses suggested that the genomes ofeach of these phages contain an ORF that isorthologous to the anti-s70 protein AsiA of T4. Wecompared the known functions of T4 AsiA withthose of the KVP40 AsiA, a protein whose sequenceis one of the most divergent from the T4 proteinamong the sequenced T4-type phages. Our charac-terization of KVP40 AsiA demonstrates that itinteracts with region 4 of s70, it inhibits transcrip-tion from s70 promoters in the absence of MotA,and it activates transcription from T4 middlepromoters in the presence of T4 MotA. Thus,despite their sequence differences, the KVP40 andT4 AsiA proteins appear to be functionallyequivalent.T4 AsiA exits as a homodimer in solution, and the

solution structure shows an AsiA monomer con-sisting of six a-helices with a helix-turn-helix motiflocated from helix 3 through helix 415,49,50 (Figure 1).The extensive AsiA-AsiA homodimer interfaceinvolves amino acid residues within the N-terminalhalf of the protein, and the heterodimer interface,which is created when the AsiA homodimerbecomes the AsiA/s70 heterodimer, uses many ofthe same amino acid residues.15,45,49 Our sequencecomparisons of the phage AsiA orthologs (Figure 1)reveal that the amino acid residues in T4 AsiA atpositions 10, 13, 16, 17, 18, 20, 21, and 36, which aredirectly involved in contacting s70 region 4,15 areconserved among the phage AsiA orthologs. Pre-vious characterizations of randomly generatedmutations within T4 AsiA have shown that substi-tutions at E10 or L18 weaken the interaction of AsiAwith s70 region 4,42,51 strengthening the idea thatthese amino acid residues are important in theheterodimer. Furthermore, a V42I substitution in T4AsiA generates a protein that behaves like wild-type,42 and this conservative replacement is foundnaturally in RB69 AsiA.E. coli Rsd and P. aeruginosa AlgQ are also

known to interact with their respective primary sfactors.6,29 We find that Rsd behaves similarly to T4AsiA in some assays. It forms a complex with s70

that is stable to electrophoresis, it inhibits transcrip-tion in vitro, and it is inhibitory for E. coli growthwhen expressed in vivo. In addition, our datasuggest that, as with T4 AsiA, L18S of Rsd andF563 of s70 are involved in the anti-s70 protein/s70

interaction. Although AlgQ is known to interactwith region 4.2 of its primary s and we show herethat F563 of s70 region 4.1 is also important for theAlgQ/s interaction, our other assays did not detectthe formation of an AlgQ/s70 complex in a nativegel or significant inhibition of s70 transcriptionin vitro or E. coli growth in vivo. However, these


results reflect the interaction of AlgQ with s70 ofE. coli rather than that of P. aeruginosa.

When the bacterial proteins are included in theAsiA family, we find that the bacterial proteinsshare sequence with the phage proteins throughoutthe length of AsiA (Figure 1). In particular, aminoacid residues corresponding to T4 AsiA positions10, 18, and 36, which are directly involved in the T4AsiA/s70 interaction,15 are conserved.We speculatethat all the phage AsiA proteins and Rsd and AlgQmay interact with region 4 of their respective s70

factors, using an interface composed of residueswithin the N-terminal half of a phage AsiA or forRsd and AlgQ, the first third of each protein.Interestingly the PfrA protein, which is a positiveregulator of siderophore biosynthesis in P. putida,shares a high degree of similarity with Rsd andAlgQ, and can substitute for AlgQ in vivo.52 Thus,PfrA may also be a member of this anti-s family.However, it has not been determined whether PfrAinteracts with region 4 of s70, and the extent towhich PfrA is functionally similar to AsiA or Rsd isnot known.

Previous work has indicated that only amino acidresidues within the N-terminal half of T4 AsiA(amino acid residues 1–46) are absolutely requiredfor its interaction with s70 or for it to function as aninhibitor or co-activator.42,44 However, sequencecomparisons of the phage AsiA proteins revealhighly conserved regions throughout the length ofthe protein. Within the C-terminal half of thevarious phage proteins, positions 46–57, position61, positions 66–73, and position 84 are particularlyconserved (Figure 1). This observation suggests thatportions of the C-terminal half are just as critical forthe function of a phage AsiA in vivo as the aminoacid residues located along the dimer interface.However, exactly how these amino acid residuesexert their influence on AsiA function is not known.The far C-terminal region (amino acid residues74–90) is the least conserved region, which isconsistent with the fact that a T4 AsiA lacking thisregion behaves like the wild-type protein in anumber of different assays in vivo and in vitro.42,51

s70 determinants for interacting with AsiA familyproteins

Within s70 region 4, the binding surface for T4AsiA has been mapped to two patches of aminoacid residues, one in region 4.1 (551–563) andanother in region 4.2 (580–598).15,44,45 Within region4.1, F563 is evidently a critical determinant for AsiAbinding, since the relatively conservative F563Ysubstitution has been shown to weaken the inter-action of s70 with T4 AsiA without affecting otherinteractions of s70 region 4.14 This substitution hasthe same effect on the interaction of s70 region 4with KVP40 AsiA, Rsd, or AlgQ. Previous work haslocated another determinant for Rsd and AlgQwithin the region 4.2 binding patch for T4 AsiA, atamino acid R596 (or for AlgQ, the equivalent R600in the P. aeruginosa primary s factor). Although

R596 is within the binding patch for T4 AsiA, theR596H substitution only very slightly weakens theinteraction of s70 with T4 AsiA in the two-hybridassay.26,29 However, single amino acid substitutionsat s70 residues known to interact with T4 AsiAfrequently either do not affect or only modestlyweaken the AsiA/s70 interaction in the two-hybridassay (Pande et al., unpublished experiments). Thismay stem from the fact that T4 AsiA binds so tightlyto s70 5 that most single amino acid changes in s70

do not significantly disrupt the interface. Takentogether, our results argue that all the members ofthe AsiA anti-s70 family inhibit transcriptionthrough their interactions with similar or at leastoverlapping patches of amino acid residues in s70

regions 4.1 and 4.2.Recent evidence indicates that region 4 of s70 also

binds to the b-flap region of core in the context ofthe RNA polymerase holoenzyme,20,21,46,53,54 andthat T4 AsiA, which exerts its effects on transcrip-tion as a subunit of the holoenzyme, inhibitstranscription from s70-dependent promoters bydisrupting this interaction between s70 region 4and the b-flap14 and creating a completely differentstructural conformation of s70 region 4.15 Thus,AsiA and the b-flap interact with overlappingdeterminants of s70 region 4, and hence competefor access to s70 region 4. The s70 region 4/b-flapinteraction is required to correctly position region 4for contact with the K35 element of promoters.46

Thus, in the presence of AsiA, this positioningcannot occur and region 4 is not available forbinding to the K35 region of promoter DNA. Thefact that AsiA inhibits transcription fromK10/K35promoters, but not from a consensus extended K10promoter that lacks aK35 element,22,23 is consistentwith this model. Our finding that T4 AsiA, KVP40AsiA, Rsd, and AlgQ all interact with commonepitopes on s70 region 4 suggests that KVP40 AsiA,Rsd, and AlgQ can also prevent s70 region 4 frombinding the b-flap. In fact, previous genetic analysisimplicates residue F563 of s70 in both the s70 region4/b-flap and the s70 region 4/AsiA interactions,with the F563Y substitution specifically affecting thes70 region 4/AsiA interaction and an F563Lsubstitution specifically affecting the s70 region4/b-flap interaction14 (Nickels & A.H., unpublishedexperiments). Thus, it seems reasonable to assumethat KVP40 AsiA inhibits transcription in theabsence of MotA because it, like T4 AsiA, disruptsthe s70 region 4/b-flap interaction in the context ofthe RNA polymerase holoenzyme. The action ofRsd is unclear, because under some circumstances(here and in the work done by Jishage & Ishihama6)it fails to remain associated with s70 in the contextof holoenzyme. However, recent evidence indicatesthat under other conditions it associates withholoenzyme or just with core.31 Thus, whetherRsd also exerts some or all of its inhibition bycompeting with the b-flap is yet to be determined.In addition, whether the binding of Rsd, like T4AsiA,15 can induce a novel conformation of s70

region 4 is not yet known.


The T4 and KVP40 AsiA proteins also differ fromRsd by their ability to activate transcription from T4middle promoters in the presence of T4 MotA. Thefact that the KVP40 AsiA can function as a co-activator with the heterologous T4 MotA proteinindicates that it retains all of the functions neededfor co-activation. However, this result wasunexpected since sequence analyses have failed tofind aMotA ortholog or typical T4 middle promotersequences within the KVP40 genome40 or within thegenome of Aeh1, which also has an AsiA protein.We speculate that these schizoT-even phages havefunctional equivalents for MotA and middlepromoters, which share at best only minimalsequence similarity with those of the T-even phages.Given the finding that Rsd and AlgQ are membersof the AsiA family, it may be that they too can workas co-activators using as yet unidentified activators.

Experimental Procedures

Phage and phage DNA

KVP40 phage was grown and the genomic DNAisolated as described.40

Plasmids

The plasmids pDKT9055 and pMotbox47 contain thebacteriophage T4 sequences fromK94 toC83 orK34 toC83, respectively, surrounding the start of the middlepromoter, PuvsX. pDKT90 also contains a MotA-indepen-dent promoter designated Pminor.

56 pP(C39/C53) is aderivative of pMotbox in which the MotA box sequenceof PuvsX (5 0TTTGCTTAA3 0) has been replaced with theMotA box sequence 5 0AAAGATTAA3 0.47 Previous workhas demonstrated that in vitro this MotA box is nearly aseffective for binding MotA and activating transcription asthe PuvsX MotA box.47

Linear templates used for in vitro transcription reac-tions were obtained after digestion of plasmids withBsaAI and purification by phenol-extraction and ethanol-precipitation. Synthetic oligonucleotides were purchasedfrom Gene Probe Technology. Oligonucleotide sequencesare available upon request.Plasmids containing C-terminal His6-tagged proteins

were constructed so that a sequence coding for sixhistidine residues was added to the C-terminal end ofthe gene. pT4AsiAHis6 was derived from pAsiA8 by site-directed mutagenesis using the method described byKunkel et al.57 modified by adding the T4 polymeraseaccessory proteins. C-terminal His6-tagged KVP40 asiADNA was obtained as a PCR product of Pfu Turbopolymerase (Stratagene), KVP40 genomic DNA, anupstream primer containing a BspHI site at the ATGstart codon of KVP40 asiA, and a downstream primerwhich coded for the end of the KVP40 asiA gene, the His6-tag, and a BamHI site. After digestion with BspHI andBamHI, this DNA was ligated into pet28a(C) (Novagen,Inc.) that had been digested with NcoI and BamHI,resulting in the plasmid pKVP40AsiAHis6. In each case,the C-terminally His6-tagged gene is located downstreamof the bacteriophage T7 F10 promoter.The plasmids pHis6Rsd, pHis6AlgQ, and pHis6Aeh1

were constructed so that a sequence encoding the His6-tagMGSSHHHHHHSSGLVPRGSHwas fused in frame to

the first amino acid of respective gene, which waspositioned downstream of the T7F10 promoter. To obtainthe rsd or algQ genes, PCR products were generated usingpAClcI-Rsd or pAClcI-AlgQ, respectively,29 Pfu Turbopolymerase, an upstream primer that contained a NdeIsite followed by the start of rsd or algQ, and a downstreamprimer that contained the end of rsd or algQ and a BamHIsite. The Aeh1 asiA gene was generated by annealing andligating overlapping oligomers to create the gene andthen this template was used as a PCR substrate asdescribed above using primers that annealed to the aeh1gene sequence. After digestion with NdeI and BamHI,PCR products were ligated with pet28a(C) that had beendigested with BamHI and NdeI to generate the plasmidspHis6Rsd, pHis6AlgQ, or pHis6Aeh1AsiA. pHis6RsdL18Swas derived from pHis6Rsd using the QuikChangeprocedure (Stratagene) and a primer that contained acodon for serine (AGC) rather than the wild-type leucine(CTG) at amino acid position 18.Plasmids pBRa–s70 D581G58 and pBRa–s70 D581G/

F563Y,14 derivatives of plasmid pBRa–s70,59 contain acolE1 replication origin, confer carbenicillin resistance,and direct transcription of an a–s70 chimera gene (codonsfor a amino acid residues 1–248 fused in frame to thecodons for s70 amino acid residues 528–613) under thecontrol of tandem promoters Plpp and PlacUV5. PlasmidpBRa encodes wild-type a and has been described.59

To construct cI-KVP40 AsiA fusion plasmids, weobtained a fragment containing KVP40 asiA as theproduct of a polymerase chain reaction (PCR) usingPfuTurbo DNA polymerase (Stratagene), KVP40 genomicDNA, and appropriate primers that contained thenecessary NotI or BglII sites to allow ligation withpAClcI3260 that had been cleaved with NotI and BglII.The resulting plasmid was designated pcI-KVP40AsiA.Plasmid pACD–35lcI-KVP40 AsiA was constructed byreplacing the T4 asiA gene on pACD–35lcI-AsiA29 withthe KVP40 asiA gene from pcI-KVP40AsiA after gener-ation of an asiA fragment of pcI-KVP40AsiA by PCR.pACD–35lcI-KVP40 AsiA, contains a p15A replicationorigin and confers chloramphenicol resistance; thisplasmid directs the expression of relatively low levels ofthe lcI-KVP40 AsiA fusion gene under the control of alacUV5 promoter variant in which theK35 element of thepromoter has been deleted.Plasmid pAClcI-b-flap46 encodes lcI (residues

1–236) fused to the b-flap moiety of the b subunit ofRNA polymerase (residues 858–946) via three alanineresidues. Plasmid pAClcI-Rsd29 encodes lcI (residues1–236) fused to Rsd (residues 1–158) via three alanineresidues. Plasmid pAClcI-AlgQ29 encodes lcI (residues1–236) fused to AlgQ (residues 1–160) via three alanineresidues. Expression of the lcI-b-flap, lcI-Rsd, and lcI-AlgQ fusion genes is under the control of the intactlacUV5 promoter.Plasmid constructions were confirmed by DNA

sequence analyses.61 In some cases, this sequencing wasperformed by the Center for Agricultural Biotechnology,University of Maryland.

Proteins

MotA protein and wild-type T4 AsiA protein werepurified as described.8 s70 was purified as described.25

E. coli RNA polymerase core and holoenzyme werepurchased from Epicentre Technologies.The various His-tagged anti-s70 proteins were isolated

from the strain pLysE/BL21(DE3)62 containing the appro-priate anti-s70 plasmid following one of the procedures

† http://phage.bioc.tulane.edu/


described by Pal et al.42 For cells containing pHis6Rsd,pT4AsiAHis6, pHis6RsdL18S, or pHis6AlgQ, the pro-cedure was followed in which cultures were grown at37 8C, synthesis of the His-tagged protein was induced atmid-log phase, and protein fractions were obtained aftercell disruption by sonication followed by centrifugation at106,000g (Rsd or T4 AsiA) or 8740g (RsdL18S or AlgQ).When using this procedure with cells containing thepKVP40AsiAHis6 plasmid, most of KVP40 AsiA proteinwas found in inclusion bodies. Therefore, the alternativeprocedure42 was used in which cultures were grown at30 8C, synthesis of KVP40 AsiAHis6 was induced once theculture was at the start of stationary phase, and a proteinfraction was obtained after cell disruption by sonicationfollowed by centrifugation at 8740g for ten minutes.Fractions containing the various His6-tagged proteinswere then obtained after adsorption to and elution fromHis-bind resin (Novagen) as described.42 His6Rsd, T4AsiAHis6, His6RsdL18S, and His6AlgQ were highlypurified while the KVP40 AsiAHis6 protein fraction stillcontained several contaminating proteins. Our attempt topurify Aeh1 AsiA was unsuccessful. In some cases,proteins were concentrated using an Ultrafreemembrane concentrator (molecular mass cut-off of5000 Da) from Millipore. Proteins were stored atK80 8C. The concentration of the anti-s70 protein wasestimated by comparing the level of the protein to aknown amount of wild-type T4 AsiA as seen after SDS-PAGE on Tris–tricine, 10–20% (w/v) polyacrylamide gels(Invitrogen).We have not observed any significant difference in

activity between wild-type T4 AsiA, T4 AsiAHis6, and T4His6AsiA, which has been characterized42 and has thesame N-terminal fusion as does His6Rsd, His6RsdL18S,His6AlgQ.

Protein and transcription buffers

Anti-s70 buffer contained 20 mM Tris–Cl (pH 7.9),500 mM NaCl, 5% (v/v) glycerol, and 100 mM or200 mM immidazole. s70 buffer contained 50 mM Tris–Cl (pH 8.0), 50 mM NaCl, 50% glycerol, 1 mM EDTA,0.1 mM DTT, and 0.01% (w/v) Triton X-100. MotA buffercontained 20 mM Tris–Cl (pH 7.9), 270 mM NaCl, 10%glycerol, 1 mM EDTA, 1 mM 2-mercaptoethanol.

E. coli growth curves and b-galactosidase assays

E. coli pLysE/BL21(DE3) containing the indicatedplasmids was grown with aeration in LB broth sup-plemented with 30 mg/ml kanamycin and 25 mg/mlchloramphenicol at 37 8C until a cell density ofapproximately 6!108 cells/ml was attained. Isopropyl-beta-D-thiogalactopyranoside (IPTG) was added to aconcentration of 1.7 mM to induce synthesis of the anti-s70 protein, and the incubation was continued. Cellgrowth was monitored by following the A600 of thecultures. b-Galactosidase assays were performed asdescribed.29 Reporter strain F 093C62 (used for theexperiment of Figure 3) has been described.29 Reporterstrain F 0OR2-62 (used for the experiment of Figure 4),which was constructed in an analogous manner, differsfrom F 093C62 only in that the test promoter bears a singlelambda OR2 site centered at position K62 relative to thetranscription start site. In either case, cells were grown inLB supplemented with the appropriate antibiotics at thefollowing concentrations: carbenicillin (100 mg/ml),chloramphenicol (25 mg/ml), and kanamycin (50 mg/ml). IPTG was provided at the indicated concentrations.

Assays were performed at least three times in duplicateon separate occasions, with similar results. Values are theaverages from one experiment; duplicate measurementsdiffered by !5%.

In vitro transcriptions

Transcription reactions were assembled as indicated inthe legend to Figure 6 or in Table 2. After the addition ofNTPs, the reactions were placed at 37 8C for 20 secondsbefore the addition of 0.5 ml of rifampicin at 300 mg/ml.Reactions were then incubated at 37 8C for 7.5 minutes,collected on dry ice, and mixed with a fivefold excess ofgel load solution (1!TBE, 7 M urea, 0.1% (w/v) bromo-phenol blue, 0.1% (w/v) xylene cyanol FF). Solutionswere heated at 95 8C for two minutes before electro-phoresis on 4% polyacrylamide, 7 M urea denaturing gelsrun in 1/2!TBE. After autoradiography, films werescanned using a Powerlook 2100XL densitometer andquantification was performed using Quantity One soft-ware from Bio-Rad, Inc.

Native protein gel assay

Protein–protein complexes were separated from freeproteins by gel electrophoresis on native 6% polyacryl-amide gels as described.25 Reactions containing 15 pmolof T4 AsiA or 25 pmol of the other anti-s70 proteins in 2 mlanti-s70 buffer, 7 pmol (Figure 5, lanes 4–10) or 3.5 pmol(Figure 5, lanes 1–3) s70 (in 2 ml s70 buffer), and/or15 pmol of MotA (in 1 ml MotA buffer) were incubated at37 8C for ten minutes before electrophoresis. Afterelectrophoresis, proteins were detected using Silver-Xpressw (Invitrogen).

S-200 HR chromatography

Solutions containing 10 pmol wild-type T4 AsiA, His6-Rsd or KVP40 AsiAHis6 and 10 pmol s70 were firstincubated at 37 8C for ten minutes and then mixed with10 pmol of RNA polymerase core in a final volume of25 ml of a buffer containing 20 mM Tris–Cl (pH 7.9),40 mM Tris–acetate (pH 7.9), 65 mM NaCl, 10% glycerol,0.4 mM EDTA, 0.15 mM DTT, 0.002% Triton X-100,150 mM potassium glutamate, 4 mM magnesium acetate,5 mM immidizole. After incubation for ten minutes at37 8C, mixtures were collected on ice and passed throughS-200HR spin columns (Amersham) as described.48 Theproteins present in the first fraction (void volume)obtained after centrifugation at 735g for two minuteswere separated by SDS-PAGE and detected by silverstaining using SilverXpressw (Invitrogen).

Sequence analyses

Genebank accession numbers for sequences are asfollows: RB69 AsiA, NP_861947; 44RR AsiA, NP_932581;T4 AsiA, NP_049866; Aeh1 AsiA, NP_944219; KVP40AsiA, NP_899542; Rsd, P31690; AlgQ, P15275. (Phagesequences are also available†) Phage and bacterial proteinsequences were initially aligned separately usingCLUSTAL W.63 These alignments were then refined bycomparing the phage protein set to the bacterial proteinset. Conserved amino acid residues were grouped asfollows: STA, NEQK, NHQK, NDEQ, QHRK, MILV,MILF, HY, FYW, SAG, FYM.

http://phage.bioc.tulane.edu/


Acknowledgements

We are grateful to Jeff Gerber for the purificationof the AsiAHis6 protein and to Nancy Nossal, IndiaHook-Barnard, Jennifer Lee, and Lauren Gallinotfor helpful discussions.

References

1. Gruber, T. M. & Gross, C. A. (2003). Multiple sigmasubunits and the partitioning of bacterial transcrip-tion space. Annu. Rev. Microbiol. 57, 441–466.

2. Paget, M. S. & Helmann, J. D. (2003). The sigma70family of sigma factors. Genome Biol. 4, 203.

3. Hughes, K. T. & Mathee, K. (1998). The anti-sigmafactors. Annu. Rev. Microbiol. 52, 231–286.

4. Orsini, G., Ouhammouch, M., Le Caer, J. P. & Brody,E. N. (1993). The asiA gene of bacteriophage T4 codesfor the anti-sigma 70 protein. J. Bacteriol. 175, 85–93.

5. Stevens, A. & Rhoton, J. C. (1975). Characterization ofan inhibitor causing potassium chloride sensitivity ofan RNA polymerase from T4 phage-infectedEscherichia coli. Biochemistry, 14, 5074–5079.

6. Jishage, M. & Ishihama, A. (1998). A stationary phaseprotein in Escherichia coli with binding activity to themajor sigma subunit of RNA polymerase. Proc. NatlAcad. Sci. USA, 95, 4953–4958.

7. Ouhammouch, M., Orsini, G. & Brody, E. N. (1994).The asiA gene product of bacteriophage T4 isrequired for middle mode RNA synthesis.J. Bacteriol. 176, 3956–3965.

8. Hinton, D. M., March-Amegadzie, R., Gerber, J. S. &Sharma, M. (1996). Bacteriophage T4 middle tran-scription system: T4-modified RNA polymerase;AsiA, a sigma 70 binding protein; and transcriptionalactivator MotA. Methods Enzymol. 274, 43–57.

9. Severinova, E., Severinov, K., Fenyo, D., Marr, M.,Brody, E. N., Roberts, J. W. et al. (1996). Domainorganization of the Escherichia coli RNA polymerasesigma 70 subunit. J. Mol. Biol. 263, 637–647.

10. Colland, F., Orsini, G., Brody, E. N., Buc, H. & Kolb, A.(1998). The bacteriophage T4 AsiA protein: a molecu-lar switch for sigma 70-dependent promoters. Mol.Microbiol. 27, 819–829.

11. Severinov, K. & Muir, T. W. (1998). Expressed proteinligation, a novel method for studying protein–proteininteractions in transcription. J. Biol. Chem. 273,16205–16209.

12. Sharma, U. K., Ravishankar, S., Shandil, R. K.,Praveen, P. V. & Balganesh, T. S. (1999). Study of theinteraction between bacteriophage T4 asiA andEscherichia coli sigma(70), using the yeast two-hybridsystem: neutralization of asiA toxicity to E. coli cellsby coexpression of a truncated sigma(70) fragment.J. Bacteriol. 181, 5855–5859.

13. Urbauer, J. L., Adelman, K., Urbauer, R. J., Simeonov,M. F., Gilmore, J. M., Zolkiewski, M. & Brody, E. N.(2001). Conserved regions 4.1 and 4.2 of sigma(70)constitute the recognition sites for the anti-sigmafactor AsiA, and AsiA is a dimer free in solution.J. Biol. Chem. 276, 41128–41132.

14. Gregory, B. D., Nickels, B. E., Garrity, S. J., Severinova,E., Minakhin, L., Bieber Urbauer, R. J. et al. (2004). Aregulator that inhibits transcription by targeting anintersubunit interaction of the RNA polymeraseholoenzyme. Proc. Natl Acad. Sci. USA, 101, 4554–4559.

15. Lambert, L. J., Wei, Y., Schirf, V., Demeler, B. &

Werner, M. H. (2004). T4 AsiA blocks DNA recog-nition by remodeling sigma(70) region 4. EMBO J. 23,2952–2962.

16. Keener, J. & Nomura, M. (1993). Dominant lethalphenotype of a mutation in the K35 recognitionregion of Escherichia coli sigma 70. Proc. Natl Acad. Sci.USA, 90, 1751–1755.

17. Gardella, T., Moyle, H. & Susskind, M. M. (1989). Amutant Escherichia coli sigma 70 subunit of RNApolymerase with altered promoter specificity. J. Mol.Biol. 206, 579–590.

18. Siegele, D. A., Hu, J. C., Walter, W. A. & Gross, C. A.(1989). Altered promoter recognition by mutant formsof the sigma 70 subunit of Escherichia coli RNApolymerase. J. Mol. Biol. 206, 591–603.

19. Campbell, E. A., Muzzin, O., Chlenov, M., Sun, J. L.,Olson, C. A., Weinman, O. et al. (2002). Structure of thebacterial RNA polymerase promoter specificity sigmasubunit. Mol. Cell, 9, 527–539.

20. Murakami, K. S., Masuda, S., Campbell, E. A.,Muzzin, O. & Darst, S. A. (2002). Structural basis oftranscription initiation: an RNA polymerase holo-enzyme–DNA complex. Science, 296, 1285–1290.

21. Vassylyev, D. G., Sekine, S., Laptenko, O., Lee, J.,Vassylyeva, M. N., Borukhov, S. & Yokoyama, S.(2002). Crystal structure of a bacterial RNA polymer-ase holoenzyme at 2.6 A resolution. Nature, 417,712–719.

22. Pahari, S. & Chatterji, D. (1997). Interaction ofbacteriophage T4 AsiA protein with Escherichia colisigma70 and its variant. FEBS Letters, 411, 60–62.

23. Severinova, E., Severinov, K. & Darst, S. A. (1998).Inhibition of Escherichia coli RNA polymerase bybacteriophage T4 AsiA. J. Mol. Biol. 279, 9–18.

24. Ouhammouch, M., Adelman, K., Harvey, S. R., Orsini,G. & Brody, E. N. (1995). Bacteriophage T4 MotA andAsiA proteins suffice to direct Escherichia coli RNApolymerase to initiate transcription at T4 middlepromoters. Proc. Natl Acad. Sci. USA, 92, 1451–1455.

25. Gerber, J. S. & Hinton, D. M. (1996). An N-terminalmutation in the bacteriophage T4 motA gene yields aprotein that binds DNA but is defective for activationof transcription. J. Bacteriol. 178, 6133–6139.

26. Pande, S., Makela, A., Dove, S. L., Nickels, B. E.,Hochschild, A. & Hinton, D. M. (2002). The bacterio-phage T4 transcription activator MotA interacts withthe far-C-terminal region of the sigma70 subunit ofEscherichia coli RNA polymerase. J. Bacteriol. 184,3957–3964.

27. Jishage, M. & Ishihama, A. (1999). Transcriptionalorganization and in vivo role of the Escherichia coli rsdgene, encoding the regulator of RNA polymerasesigma D. J. Bacteriol. 181, 3768–3776.

28. Jishage, M., Dasgupta, D. & Ishihama, A. (2001).Mapping of the Rsd contact site on the sigma 70subunit of Escherichia coli RNA polymerase.J. Bacteriol. 183, 2952–2956.

29. Dove, S. L. & Hochschild, A. (2001). Bacterial two-hybrid analysis of interactions between region 4 of thesigma(70) subunit of RNA polymerase and thetranscriptional regulators Rsd from Escherichia coliand AlgQ from Pseudomonas aeruginosa. J. Bacteriol.183, 6413–6421.

30. Westblade, L. F., Ilag, L. L., Powell, A. K., Kolb, A.,Robinson, C. V. & Busby, S. J. (2004). Studies of theEscherichia coli Rsd-sigma70 complex. J. Mol. Biol. 335,685–692.

31. Ilag, L. L., Westblade, L. F., Deshayes, C., Kolb, A.,Busby, S. J. & Robinson, C. V. (2004). Mass


spectrometry of Escherichia coli RNA polymerase:interactions of the core enzyme with sigma70 andRsd protein. Structure (Camb), 12, 269–275.

32. Kato, J., Chu, L., Kitano, K., DeVault, J. D., Kimbara,K., Chakrabarty, A. M. & Misra, T. K. (1989).Nucleotide sequence of a regulatory region control-ling alginate synthesis in Pseudomonas aeruginosa:characterization of the algR2 gene. Gene, 84, 31–38.

33. Deretic, V. & Konyecsni, W. M. (1989). Control ofmucoidy in Pseudomonas aeruginosa: transcriptionalregulation of algR and identification of the secondregulatory gene, algQ. J. Bacteriol. 171, 3680–3688.

34. Deretic, V., Schurr, M. J., Boucher, J. C. &Martin, D. W.(1994). Conversion of Pseudomonas aeruginosa tomucoidy in cystic fibrosis: environmental stress andregulation of bacterial virulence by alternative sigmafactors. J. Bacteriol. 176, 2773–2780.

35. Ackermann, H.W. & Krisch, H.M. (1997). A catalogueof T4-type bacteriophages. Arch. Virol. 142, 2329–2345.

36. Matsuzaki, S., Inoue, T. & Tanaka, S. (1998). Avibriophage, KVP40, with major capsid proteinhomologous to gp23* of coliphage T4. Virology, 242,314–318.

37. Matsuzaki, S., Kuroda, M., Kimura, S. & Tanaka, S.(1999). Major capsid proteins of certain Vibrio andAeromonas phages are homologous to the equivalentprotein, gp23(*), of coliphage T4. Arch. Virol. 144,1647–1651.

38. Tetart, F., Desplats, C., Kutateladze, M., Monod, C.,Ackermann, H. W. & Krisch, H. M. (2001). Phylogenyof the major head and tail genes of the wide-rangingT4-type bacteriophages. J. Bacteriol. 183, 358–366.

39. Matsuzaki, S., Tanaka, S., Koga, T. & Kawata, T.(1992). A broad-host-range vibriophage, KVP40, iso-lated from sea water. Microbiol. Immunol. 36, 93–97.

40. Miller, E. S., Heidelberg, J. F., Eisen, J. A., Nelson,W. C., Durkin, A. S., Ciecko, A. et al. (2003). Completegenome sequence of the broad-host-range vibrio-phage KVP40: comparative genomics of a T4-relatedbacteriophage. J. Bacteriol. 185, 5220–5233.

41. Blattner, F. R., Plunkett, G., III, Bloch, C. A., Perna,N. T., Burland, V., Riley, M. et al. (1997). The completegenome sequence of Escherichia coli K-12. Science, 277,1453–1474.

42. Pal, D., Vuthoori, M., Pande, S., Wheeler, D. & Hinton,D. M. (2003). Analysis of regions within the bacterio-phage T4 AsiA protein involved in its binding to thesigma70 subunit of E. coli RNA polymerase and itsrole as a transcriptional inhibitor and co-activator.J. Mol. Biol. 325, 827–841.

43. Heidelberg, J. F., Eisen, J. A., Nelson, W. C., Clayton,R. A., Gwinn, M. L., Dodson, R. J. et al. (2000). DNAsequence of both chromosomes of the cholerapathogen Vibrio cholerae. Nature, 406, 477–483.

44. Minakhin, L., Camarero, J. A., Holford, M., Parker, C.,Muir, T. W. & Severinov, K. (2001). Mapping themolecular interface between the sigma(70) subunit ofE. coli RNA polymerase and T4 AsiA. J. Mol. Biol. 306,631–642.

45. Simeonov, M. F., Bieber Urbauer, R. J., Gilmore, J. M.,Adelman, K., Brody, E. N., Niedziela-Majka, A. et al.(2003). Characterization of the interactions betweenthe bacteriophage T4 AsiA protein and RNA poly-merase. Biochemistry, 42, 7717–7726.

46. Kuznedelov, K., Minakhin, L., Niedziela-Majka, A.,Dove, S. L., Rogulja, D., Nickels, B. E. et al. (2002). Arole for interaction of the RNA polymerase flapdomain with the sigma subunit in promoter recog-nition. Science, 295, 855–857.

47. Marshall, P., Sharma, M. & Hinton, D. M. (1999). Thebacteriophage T4 transcriptional activator MotAaccepts various base-pair changes within its bindingsequence. J. Mol. Biol. 285, 931–944.

48. Hinton, D. M. & Vuthoori, S. (2000). Efficientinhibition of Escherichia coli RNA polymerase by thebacteriophage T4 AsiA protein requires that AsiAbinds first to free sigma70. J. Mol. Biol. 304, 731–739.

49. Lambert, L. J., Schirf, V., Demeler, B., Cadene, M. &Werner, M. H. (2001). Flipping a genetic switch bysubunit exchange. EMBO J. 20, 7149–7159.

50. Urbauer, J. L., Simeonov, M. F., Urbauer, R. J.,Adelman, K., Gilmore, J. M. & Brody, E. N. (2002).Solution structure and stability of the anti-sigmafactor AsiA: implications for novel functions. Proc.Natl Acad. Sci. USA, 99, 1831–1835.

51. Sharma, U. K., Praveen, P. V. & Balganesh, T. S. (2002).Mutational analysis of bacteriophage T4 AsiA:involvement of N- and C-terminal regions in bindingto sigma(70) of Escherichia coli in vivo. Gene, 295,125–134.

52. Venturi, V., Ottevanger, C., Leong, J. & Weisbeek, P. J.(1993). Identification and characterization of a side-rophore regulatory gene (pfrA) of Pseudomonas putidaWCS358: homology to the alginate regulatory genealgQ of Pseudomonas aeruginosa. Mol. Microbiol. 10,63–73.

53. Murakami, K. S., Masuda, S. & Darst, S. A. (2002).Structural basis of transcription initiation: RNApolymerase holoenzyme at 4 A resolution. Science,296, 1280–1284.

54. Mekler, V., Kortkhonjia, E., Mukhopadhyay, J.,Knight, J., Revyakin, A., Kapanidis, A. N. et al.(2002). Structural organization of bacterial RNApolymerase holoenzyme and the RNA polymerase-promoter open complex. Cell, 108, 599–614.

55. March-Amegadzie, R. & Hinton, D. M. (1995). Thebacteriophage T4 middle promoter PuvsX: analysis ofregions important for binding of the T4 transcrip-tional activator MotA and for activation oftranscription. Mol. Microbiol. 15, 649–660.

56. Vuthoori, S., Bowers, C. W., McCracken, A.,Dombroski, A. J. & Hinton, D. M. (2001). Domain1.1 of the sigma(70) subunit of Escherichia coli RNApolymerase modulates the formation of stablepolymerase/promoter complexes. J. Mol. Biol. 309,561–572.

57. Kunkel, T. A., Roberts, J. D. & Zakour, R. A. (1987).Rapid and efficient site-specific mutagenesis withoutphenotypic selection. Methods Enzymol. 154, 367–382.

58. Nickels, B. E., Dove, S. L., Murakami, K. S., Darst, S. A.&Hochschild, A. (2002). Protein–protein and protein–DNA interactions of sigma70 region 4 involved intranscription activation by lambdacI. J. Mol. Biol. 324,17–34.

59. Dove, S. L., Huang, F. W. & Hochschild, A. (2000).Mechanism for a transcriptional activator that worksat the isomerization step. Proc. Natl Acad. Sci. USA, 97,13215–13220.

60. Hu, J. C., Kornacker, M. G. & Hochschild, A. (2000).Escherichia coli one- and two-hybrid systems for theanalysis and identification of protein–protein inter-actions. Methods, 20, 80–94.

61. Sanger, F., Nicklen, S. & Coulson, A. R. (1977). DNAsequencing with chain-terminating inhibitors. Proc.Natl Acad. Sci. USA, 74, 5463–5467.

62. Studier, F. W., Rosenberg, A. H., Dunn, J. J. &


Dubendorff, J. W. (1990). Use of T7 RNA polymeraseto direct expression of cloned genes.Methods Enzymol.185, 60–89.

63. Thompson, J. D., Higgins, D. G. & Gibson, T. J. (1994).

CLUSTALW: improving the sensitivity of progressivemultiple sequence alignment through sequenceweighting, position-specific gap penalties and weightmatrix choice. Nucl. Acids Res. 22, 4673–4680.

Edited by M. Gottesman

(Received 4 June 2004; received in revised form 30 September 2004; accepted 5 October 2004)

A Family of Anti-σ70 Proteins in T4-type Phages and Bacteria that are Similar to AsiA, a...

Documents

Transcript of A Family of Anti-σ70 Proteins in T4-type Phages and Bacteria that are Similar to AsiA, a...