thermal energy requirement for strand separation during

3840-3845 Nucleic Acids Research, 1994, Vol. 22, No. 19 © 1994 Oxford University Press

Thermal energy requirement for strand separation duringtranscription initiation: the effect of supercoiling andextended protein DNA contacts

Helen Burns and Stephen Minchin*School of Biochemistry, The University of Birmingham, PO Box 363, Birmingham B15 2TT, UK

Received July 22, 1994; Revised and Accepted August 23, 1994

ABSTRACT

We have studied the role of extended protein DNAcontacts and DNA topology on the ability of Escherichiacoli RNA polymerase to form open complexes atseveral related promoters. The - 35 region of severalEscherichia coli promoters do not have homology withthe consensus sequence, but still drive activatorindependent transcription initiation. This is due to thepresence of a TG motif upstream from the - 1 0hexamer creating an 'extended - 1 0 ' promoter. Wehave previously shown that two 'extended - 1 0 'promoters, ga/P1 and pBIa, can form open complexesat lower temperatures than the ga/P1 derivative,galPcon6, which has a consensus - 35 hexamer. Herewe report further investigations into the mechanism ofopen complex formation by RNA polymerase, inparticular the thermal energy requirement. A singlebase pair change in ga/Pcon6 creating an 'extended- 1 0 ' sequence, results in a 20°C reduction in thetemperature requirement for open complex formation.The DNA topology has also been shown to effect thethermal energy requirement for strand separation.Promoters carried on supercoiled plasmids form opencomplexes at lower temperatures than when presenton linear DNA templates. We have also shown that invivo, RNA polymerase can form open complexes atlower temperatures than those observed for lineartemplates in vitro, but requires slightly highertemperatures than supercoiled templates in vitro,however the promoter hierachy remains the same.

INTRODUCTION

The bacterium Escherichia coli has one major RNA polymerase(a2)3/3'a70) which recognises promoters containing two hexamersequences 10 and 35 base pairs upstream from the transcriptionalstartpoint (1). The rate of transcription from constitutivepromoters is dependent upon the nucleotide sequence of theindividual —10, —35 hexamers. Work by many scientists hasled to the elucidation of consensus sequences for both hexamers(2,3). The - 1 0 and —35 regions provide contact points for the

a70 subunit of RNA polymerase. Most constitutive promotersmust have close homology to both the consensus - 3 5 and - 1 0sequences. At promoters containing -35 or - 1 0 sequences withno homology to the consensus, RNA polymerase can notgenerally initiate transcription without enlisting the help of othertranscription factors (1). This is due to loss of contacts betweenthe promoter and the a70 subunit, transcription factorscompensating for this loss.

The 'extended - 1 0 ' class of promoters drive transcription bya70 RNA polymerase but generally contain -35 sequences withno homology to consensus (4—6). These promoters are stillfunctional because RNA polymerase makes additional contactsoutside the - 1 0 region, specifically with a 5'-TG-3' motif at— 15/-14 (7). The TG motif is a weakly conserved sequencefound in many E.coli promoters (2). The galPl promoter is anexample of an 'extended - 1 0 ' promoter which has a - 1 0hexamer with close homology to the consensus and the TG motif,but a —35 sequence with no homology to the consensus. ThegalPl promoter also requires additional sequences positionedaround —50 for maximum activity, RNA polymerase binds tothese sequences which results in extended DNAsel footprints (8).A synthetic derivative of galPl has been constructed by cloninga DNA sequence upstream from —12 which contains a consensus-35 sequence, this is named galPcm6 [formerly called galPcon

in (9)]. The galPcon6 promoter drives higher rates oftranscription initiation than galPl both in vivo and in vitro at 37°C(10,11). Studies have shown that RNA polymerase makesdifferent contacts at ga/Pcon6 than at galPl; there are no contactsupstream from —45 or with the —15/—14 region, but importantcontacts are make with the consensus —35 region (7,9).

Transcription initiation is a complex process involving aplethora of protein—protein and protein—DNA contacts. Severalsteps take place before a competent binary complex is formedcapable of catalysing the synthesis of a nascent RNA chain (1).RNA polymerase first binds DNA non-specifically and migratesalong the DNA searching for promoter sequences, onceencountered RNA polymerase binds specifically to the promoterDNA to form a closed complex. The closed complex is nottranscriptionally competent since the DNA template is still doublestranded and hence no bases are available to pair with incoming

*To whom correspondence should be addressed

Downloaded from https://academic.oup.com/nar/article-abstract/22/19/3840/2400089by gueston 19 February 2018

Nucleic Acids Research, 1994, Vol. 22, No. 19 3841

nucleoside triphosphates. The next step in transcription initiationis the isomerisation of the closed complex to the transcriptionallycompetent open complex; strand separation provides a singlestranded template. The first nucleoside triphosphates may nowbind, synthesis of the nascent RNA chain begins and continueswith the formation of a stable elongation complex. Theseprocesses all require energy, but with the exception of formationof the phosphodiester bonds, no chemical energy is used. Wehave previously looked at the thermal energy requirement fortranscription initiation at galPcon6 and galPl (9). We haveshown that at these two promoters the thermally limiting stepin transcription initiation is the isomerisation of the closedcomplex to the open complex. At galPl and galPcon6 opencomplex formation is blocked at temperatures below 6°C and20°C repectively on linear DNA templates. The surprisingfinding was that RNA polymerase could form an open complexat galPl at much lower temperatures than at galPcon6. It isprobable that at galPl RNA polymerase overcomes the thermalenergy requirement by making additional contacts with thepromoter DNA and that it is not possible to make such contactsat promoters containing only the —10, —35 hexamers.

In this paper, we show that, in the gal context, promoterscontaining the 'extended - 1 0 ' TG motif have a lower thermalenergy requirement for open complex formation. Negativesupercoiling lowers the thermal energy requirement for opencomplex formation. The in vivo temperature profiles of thepromoters studied retain the same hierachy as observed in vitro.

MATERIALS AND METHODSPromotersAll promoters were carried as EcoRl-HindM fragments in thegalK fusion vector pAA121 as described previously (12). Allplasmids were purified via caesium chloride gradients (13).Supercoiled templates were uncut plasmids as isolated from theEscherichia coli host using the mehod of lysis by sodium dodecylsulphate and then purified on a CsCl—ethidium bromide gradientas desribed in (13). Linear templates were prepared by digestionof the supercoiled plasmid DNA with the restriction endonucleasePstl, which cuts all constructs once within the bla gene onpAA121.

Formation of binary complexesThe DNA templates (lOnM) were incubated in transcriptionbuffer (20mM Tris-HCl pH8.0, lOOmM NaCl, 5mM MgCl2,

O.lmM EDTA, lmM DTT, 50/ig/ml BSA, 5% glycerol), in afinal volume of 20/tl at the appropriate temperature for 20 minutesto allow temperature equilibration. RNA polymerase (200nM)(supplied by NBL Gene Sciences) was added and the mixtureincubated for 20 minutes to allow binary complex formation.

In vitro potassium permanganate probingPotassium permanganate was used to probe binary complexesat the appropriate temperature; 1/il of a freshly prepared 200mMsolution of potassium permanganate was added to the reactioncontaining the binary complex. The reaction was stopped after4 minutes by the addition of 50/il KMnO4 stop solution (3Mammonium acetate, O.lmM EDTA, 1.5M /3-mercaptoethanol).The DNA was purified by phenol/chloroform extraction andethanol precipitation. Site of KMnO4 modification was detectedby primer extension.

In vivo potassium permanganate probingAn overnight culture of M182ACRP containing the pAA121promoter derivative was used to inoculate 5ml minimal media(13) and the cells grown at 37°C to an OD650 0.5-0.6. Thecells were then transferred to the appropriate temperature for invivo probing and incubated for 1 hour. If required 20/il of50mgml~' rifampicin was added to the cells 5 minutes prior totreatment with 135/tl 0.37M KMnO4. After two minutespermanganate treatment the cells were harvested by centrifugationand the DNA recovered using the Boiling lysis 'mini-prep'method (13). Site of KMnO4 modification was detected byprimer extension.

Primer extension analysis of modified DNAThis a modified version of the method described by Sasse-Dwightand Gralla (14). An oligonucleotide primer (Alta Bioscience, TheUniversity of Birmingham) complementary to the template strandupstream from the £c<?RI site (with respect to the promoter) was5' end-labelled with [7-32P]ATP (Amersham). Labelled primer(20nM), dNTP's (100/iM) (Pharmacia) and Taq DNApolymerase (2 Units) (Boehringer Mannheim) or Vent Exo~ (2Units) (New England Biolabs) were added to the modified DNAin a final volume of 50/tl. A mineral oil overlay (Sigma) wasadded and the extension reaction was started: 1 cycle of 3 minat 94°C, 2 min at 50°C, 1.5 min at 72°C; 15 cycles of 1 minat 94°C, 2 min at 50°C, 1.5 min at 72°C; and 1 cycle of 1 minat 94°C, 2 min at 50°C, 10 min at 72°C. The aqueous layerwas ethanol precipitated and analysed on a 6% sequencing gel(Sequagel, National Diagnostics).

-50 -30 -10 +1 +20

galFl TTGTGTAAACGATTCCACTAATTTATTCCATGTCACACTTTTCGCATCTTTTTTATGCTATGGTTATTTCATACCATAAGCCTAATGGAGC

galPcon6 CGTCTTCAAGAATTCTTGACAGCTGCATGCATCTTTTlTATGGTTATTTCATACCATAAGCCTAATGGAGC

gaIPconTG6 CGTCTTCAAGAATTCTTGACAGCTGCATGCATCTTTglTATGGTTATTTCATACCATAAGCCTAATGGAGC

ga 1 PconTG5 CGTCTTa^GAATTCTAGACAGCTGCATGCATCTTTG'lTATGGTTATTTCATACCATAAGCCTAATGGAGC

Consensus TTGACA TATAAT

Figure 1. The nudeotide sequence of the four promoters used in this work; galPl, galP^, galPmnTG6 and ga&CB^G5. The promoter consensus sequence recognisedby the major E.coli RNA polymerase, Eo70, is shown below. The - 3 5 , - 1 0 and 'extended - 1 0 ' sequence elements are underlined. The * shows the positionof the G to T transversion in galP\ that inactivates the galP2 promoter. The synthetic promoter gatPcon6 was produced by cloning a synthetic sequence containingthe consensus -35 hexamer into galPl upstream of —12. A single transition mutation at - 1 4 introduced the 'TG' motif into galP^^} forming gaffconTG6. Theshaded box indicates the region of sequence identity between all four promoters. The sequences are numbered with the transcription start as + 1 .


3842 Nucleic Acids Research, 1994, Vol. 22, No. 19

RESULTSPromotersThe sequences of the promoters used in this work are shown infigure 1. We have previously shown that ga/Pl can form opencomplexes at lower temperatures than ga/Pcon6 (9). We havecontinued the investigations into the phenomenon of lowtemperature open complex formation. In this paper we haveinvestigated the role of the 5TG3 ' motif at - 1 5 / - 1 4 . Twoderivatives of galPcon6 have been analysed; ga/PconTG6 whichcontains a consensus —35 sequence and the TG motif, andgalPconTG5 which contains a —35 sequence with a 5/6 matchwith consensus and the TG motif. The in vivo activity ofgalPC(m6 and galP-conTG5 is comparable, whereas galPconTG6promotes higher levels of transcription than the other twopromoters (15).

Investigation into the role of the TG motif in the thermalenergy requirement for open complex formation on lineartemplates in vitro

The ability of RNA polymerase to form open complexes has beeninvestigated. Potassium permanganate was used to modify DNAat the site of strand separation (16,17) and the position ofmodification detected using primer extension. We have previouslyshown that the galPcoJS promoter can not form open complexesat temperatures below 25°C, whereas the galPl promoter iscapable of forming open complexes at lower temperatures (9).Analysis of the galPcon6 and galPl promoter sequences revealsthat there are several sequences within galPl that could contributeto low temperature melting. One candidate is the TG motif thatis present upstream from the galPl -10 hexamer, thisdinucleotide is required for galPl activity both in vitro and invivo. Mutagenesis of the TG motif in galPl stops open complexformation at all temperatures (18) and (Burns, unpublished).

We therefore investigated the ability of the TG motif todecrease the temperature requirement for open complex formationin the context of the galPcon6 promoter. The TG motif wasintroduced into galPcon6 thus creating galPcmTG6, galPconTG6promotes high levels of transcription. The temperaturerequirement for open complex formation at galPcon6 and

galPconTG6 was compared. Binary complexes were allowed toform at 37°C, 25°C, 14°C and 6°C and any open complexesthen probed using potassium permanganate. Figure 2 shows thatthe temperature requirement for open complex formation atgalPcm6 was as observed previously (9), full opening is seen at37°C, minimal opening at 25°C and none at 14°C and 6°C.Introduction of the G into galPcon6 creating ga{PmnTG6 reducedthe temperature requirement by approximately 20°C, opencomplexes being observed at 6°C. We have previously observedthat the temperature requirement for open complex formation isnot a function of the in vivo strength of the promoter at 37°C,galPl opening at a lower temperature than the stronger promotergalPcm6 (9,11). However, galPcon7G6 is a very strongpromoter and therefore promoter strength may be affecting thetemperature requirement. The ga/PconTG5 promoter has asimilar activity to ga/Pcon6 see figure 1 in (11) (Comparepromoters d and e, galPcon6 and galPconTG5 respectively). Wecompared the temperature requirement for open complexformation at g a / P ^ and ga/PconTG5 at 37°C, 25°C, 14°C and6°C (fig. 3). The temperature profile of galPconTG5 wasidentical to that of galPconTG6, confirming that the temperaturerequirement in the 'gal' context is a function of the TG motifand not the in vivo strength of the promoter.

Investigation into the effect of negative supercoiling on thethermal energy requirement for open complex formation invitroPrevious studies of open complex formation at galPl, galPcon6and derivatives have used linear DNA [fig. 2 and 3 this paperand (9)]. It is important to note that in vivo the DNA templatewill be supercoiled. Negative supercoiling promotes strandseparation due to the stress introduced into the DNA which canbe relieved by local unwinding (19,20). It can therefore be arguedthat in the specific case of open complex formation by RNApolymerase, promoters present on supercoiled DNA templatesshould have a lower temperature requirement than those on lineartemplates. The effect of negative supercoiling on the ability of RNA

X«/Pcon6 Linear K«/PconTG5 Linear

Linear .i,'(f/PconTG6 Linear

37°C 25°C 14°C 6°C 37°C 25CC 14°C 6°C

RNAP

+3 -+ 1 "-5 -

"•- +3- +1

-—II

| M 37°C 25°C 14°C 6°C

RNAP '- + " - +"- +" - +'

37°C 25°C 14°C 6°C

Figure 2. The potassium permanganate modifications of galPcon6 andga/PconTG6, cloned in pAA121 and linearised at the unique Pstl site. Theexperiments were done at the temperatures indicated, in the presence and absenceot KNA poiymerase (RNAP), and analysed by PCR primer extension. The arrowsindicate single stranded thy mine residues, numbered relative to the transcriptionstart site + 1 . Lane M is a Maxam and Gilbert sequencing reaction specific forG residues.

Figure 3. The potassium permanganate modifications of galPcon6 andga/PconTG5, cloned in pAA121 and linearised at the unique Psll site. Theexperiments were done at the temperatures indicated, in the presence and absenceof RNA polymerase (RNAP), and analysed by PCR primer extension. The arrowsindicate single stranded thymine residues, numbered relative to the transcriptionstart site +1 . Lane M is a Maxam and Gilbert sequencing reaction specific forG residues.



polymerase to form open complexes at different temperatureswas studied. The supercoiled DNA template was as purified fromthe host E.coli cell (see materials and methods). The ability ofRNA polymerase to form open complexes on supercoiled DNAtemplates at 37°C, 25°C, 14°C and 6°C was determined usingpotassium permanganate. Figure 4 shows that as expectednegative supercoiling reduces the temperature requirement foropen complex formation. The galPcm6 promoter can form'open' complexes at temperatures as low as 6°C when on asupercoiled template, whereas on a linear template no openingis observed at temperatures below 25°C (fig. 2). At ga/PconTG6some opening is observed at 6°C on a linear template, howeversupercoiling leads to increased levels of opening. As mentionedpreviously gatPl is an 'extended —10' promoter, which formsopen complexes at low temperatures, the effect of supercoilingon open complex formation at galPl was determined. Figure 5shows that as with galPcon6 and galPconTG6, supercoilingreduces the temperature requirement for open complex formation.

Investigation into the thermal energy requirement for opencomplex formation in vivo

All previous experiments have concentrated on the thermal energyrequirement for open complex formation in vitro, it was thereforeneccessary to determine if the effects seen in vitro were mirroredin vivo.

Using potassium permanganate as a probe it is possible to detectopen complex formation in vivo (16,18). Escherichia coli hostcells carrying the galPcon6, galPcmTG6 and galPl promoterscloned into pAA121 were grown to mid-log phase, the cells werethen incubated at the appropriate temperature for 1 hour, beforetreatment with rifampicin followed by potassium permanganateto trap and probe open complexes. Plasmid DNA was purifiedfrom the treated cells and the position of modification determinedusing primer extension. The results showed that in vivo lowertemperatures were required for open complex formation thanobserved in vitro on linear templates, but higher temperatureswere required compared to supercoiled templates in vitro (Fig.6). It is important to note that the same promoter hierachy wasobserved in vivo as observed in vitro.

K(//Pcon6 Supercoiled ,t,'<//PconTG6 Supercoiled

RNAP

+ 3 -+ 1 -- 5 -

-II -

Figure 4. The potassium permanganate modifications of galPcm6 andgatPconTG6, cloned in the supercoiled plasmid pAA121. The experiments weredone at the temperatures indicated, in the presence and absence of RNA polymerase(RNAP), and analysed by PCR primer extension. The arrows indicate singlestranded thymine residues, numbered relative to the transcription start site + 1 .Lane M is a Maxam and Gilbert sequencing reaction specific for G residues.

galPl linear %alP\ Supercoiled

G T37°C 25°C I4°C 6°C 37°C 25°C 14°C 6°C

- 5 - a - •Ill+3

-+1- -5--11

DISCUSSION

We had previously shown that galPl, an 'extended - 1 0 'promoter, can form open complexes at low temperatures. In thispaper we have analysed the contribution of the TG motif, presentin 'extended —10' promoters, towards low temperature melting

galPl in vivo

RifampicinT A

37°C 25°C 14°C 6°C

-5-

-11- !

galPcon6 in vivo ga/PconTG6 in vivo

A 37°C 25°C I4°C 6"C | A 37°C 25°C 14°C 6°C |Rifampicin' T ^ r — J { — ^ F T " ? l l I- +11. +11. +11. +1 l

+3-+1"-5-

• • I I S :+3+1-5

Figure 5. The potassium permanganate modifications of galP\, cloned inpAA121,in a linearised and supercoiled form. The plasmid was linearised at the uniquePstl site when necessary. The experiments were done at the temperatures indicated,in the presence and absence of RNA polymerase (RNAP), and analysed by PCRprimer extension. The arrows indicate single stranded thymine residues, numberedrelative to the transcription start site + 1 . Di-deoxy sequencing specific for theindicated base was used as a calibration.

Figure 6. The in vivo potassium permanganate modifications of galP\, ga/Pcon6and ga/PconTG6, cloned in pAA121, in the E.coli strain M\$2bcrp. Culturesin mid log phase (O.D.650 = 0.5) were shifted to the appropriate temperatureand incubated for 1 hour. Rifampicin was added as indicated 5 minutes priorto probing with potassium permanganate. The modified plasmid DNA wasextracted by the boiling lysis miniprep method and analysed by PCR primerextension. The arrows indicate single stranded thymine residues, numbered relativeto the transcription start site +1 . Di-deoxy sequencing specific for the indicatedbase was used as a calibration.


3844 Nucleic Acids Research, 1994, Vol. 22, No. 19

of promoter DNA. Determining the role of the TG motif in lowtemperature open complex formation at the galPl promoter isnot possible, since the motif is essential for galPl activity at alltemperatures. We have therefore looked at the effect ofintroducing the TG motif into two 'typical' promoters galPCOD6and galPcon5, these promoters already have 6/6 and 5/6homology to the consensus —35 sequence respectively and bothhave a 4/6 match to the consensus - 1 0 sequence. The ga/Pcon6promoter drives high levels of transcription initiation, however,introduction of the TG motif into this promoter, creatinggalPconTG6, results in increased activity (15). This increasedactivity is emphasised by the observation that when galPconTG6is cloned into the galK promoter probe vector pAA121 therecombinant plasmid is toxic to the cell. A stable derivative ofthe ga/PconTG6/pAA121 construct was isolated which containeda deletion in the galK gene. In addition to increasing the in vivostrength of the galPcon6 promoter introduction of the TG motifconfered the ability to form open complexes at lower temperaturesboth in vitro on linear and supercoiled templates and in vivo. Itcould be possible that the decreased temperature requirement foropen complex formation is a result of increased promoterstrength. However as observed previously the galPl promoterforms open complexes at lower temperatures than galPcon6,while driving lower rates of transcription initiation both in vitroand in vivo at 37°C (9,11). To rule out the direct effect ofpromoter strength on the thermal energy requirement for opencomplex formation, the temperature requirement for galP^rJGSwas measured. The galPCOD5 promoter drives low levels oftranscription (11), however, introduction of the TG motif intogalPcon5 creating galPconTG5 results in a promoter with similarstrength to galPcon6 but still weaker than ga/PconTG6 (11). Thetemperature requirement for open complex formation atgalPamTGS and ga/PconTG6 appeared the same, thus confirmingthat promoter strength per se does not determine the temperaturerequirement for open complex formation.

The most striking finding of this work is the observation thatintroduction of a single base change into a promoter sequence(galPcon6 to galPconTG6) results in approximately a 20°Creduction in the temperature requirement for open complexformation. This raises the question of the role of the TG motifin promoter function. There are at least two possible explanationsfor the role of the TG motif, i) RNA polymerase may make asequence specific contact with the TG motif, i.e. extendingcontacts made by a with the —10 sequence or ii) the TG motifresults in an altered DNA conformation which could eitherdirectly facilitate strand separation or allow additionalDNA—protein contacts which would then promote open complexformation. Neither explanation is mutually exclusive. Theevidence for direct contact between RNA polymerase and theTG motif is that at galPl, the G is protected by RNA polymerasefrom methylation by dimethly sulphate when in the open complex(11). More importantly, methylation interference studies showthat methylation of the G inhibits open complex formation (7).Evidence for the role of the TG motif in driving a DNAconformation comes from the occurance of this motif at positionswhere the DNA is bent in protein—DNA complexes such as theCRP(CAP)-DNA complexes (21). Formation of atranscriptionally competent complex involves several stages, firstRNA polymerase binds the DNA to form the closed complex,this step appears to be temperature independent (9,22), there isthen a conformational change in RNA polymerase in order tonucleate DNA melting and this is followed by a temperature

dependent melting of the DNA (22). It is therefore probable thatthe TG motif is involved in DNA melting rather than nucleation.Further work will be required to determine the role of the'extended —10' motif and to assess the role of other sequenceelements in low temperature opening. There are several other'extended —10' promoters, one example is the cysG promoter.The cysG promoter does not form open complexes at lowtemperatures which therefore suggests that the TG motif isessential but not sufficient for low temperature open complexformation (Belyaeva et al., in preparation). The galPl promotermakes additional contacts upstream from the -35 region, thesecontacts are analogous to the upstream contacts made by the asubunit of RNA polymerase at the rrnB promoter (23). It istherefore probable that the a subunit of RNA polymerase makescontacts with the upstream region of galPl, such contacts andtheir role in the temperature dependence of open complexformation are currently under investigation.

In this work we have also shown that the topology of the DNAalso effects the temperature requirement for open complexformation. We have presented evidence that when promoters arecarried on negatively supercoiled plasmid DNA, RNApolymerase can form open complexes at lower temperatures.Supercoiling lowered the temperature requirement for opencomplex formation at all promoters studied. This is consistentwith previous observations at the XPR promoter (24). Thisobservation would be predicted since strand separation will beenergetically favoured by negative supercoiling. In addition tothe effect of supercoiling at all temperatures in vitro, as thetemperature is lowered the level of negative supercoiling increasesdue to changes in the helical pitch this will further help RNApolymerase form open complexes at low temperatures (25). Wehave also shown that in vivo RNA polymerase can form opencomplexes at lower temperatures than that observed for lineartemplates in vitro, but requires slightly higher temperatures thansupercoiled templates in vitro. This is consistent with theobservation that the effective level of supercoiling in vivo is lowerthan that observed in vitro due to the constraints put on the DNAby protein binding (26). Another aspect is the role oftopoisomerases in the cell which will affect the DNA topologyin vivo. In vivo as the temperature drops the increase insupercoiling due to helical pitch is offset by topoisomerasesincreasing the linking number of the DNA (27). The importantobservation is that the promoter heirachy is the same in vivo asthat in vitro on both linear and supercoiled templates. ThereforegalPl and other 'extended —10' promoters can form opencomplexes in vivo at lower temperatures than the 'consensus' likepromoters used in this study.

REFERENCES

1. McClure,W.R. (1985) Ann. Rev. Biochem. 54, 171-204.2. Hawley.D.K. and McClure.W.R. (1983) Nucl. Acids Res. 11, 2237-2255.3. Harley.C.B. and Reynolds,R.P. (1987) Nucl. Acids Res. 15, 2343-2361.4. Ponnambalam.S., Webster.C, Bingham,A. and Busby,S. (1986) J. Biol.

Chem. 261, 16043-16048.5. Keilty.S. and Rosenberg.M. (1987) J. Biol. Chem. 262, 6389-6395.6. BelyaevaX, Griffiths,L., Minchin,S., ColeJ. and Busby.S. (1993) Biochem.

J. 296, 851-857.7. Minchin.S.D. and Busby.S.J.W. (1993) Biochem. J. 289, 771-775.8. Busby.S., Spassky.A. and Chan.B. (1987) Gene. 53, 145-152.9. Grimes.E., Busby,S. and Minchin.S. (1991) Nucl. Acids Res. 19,

6113-6118.10. Char.,B. ar.d Busby.S. (1989) Ger.t, 84, 227-236.11. Chan,B., Spassky.A. and Busby.S. (1990) Biochem. J. 270, 141-148.



12. Kelsall,A., Evans.C. and Busby,S. (1985) FEBS Letters, 180, 155-159.13. Sambrook,J., Fritsch,E.F. and Maniatis.T. (1989) Molecular Cloning, A

laboratory manual. Cold Spring Harbor Laboratory Press, New York.14. Sasse-Dwight,S. and Gralla,J.D. (1991) Methods Enzymol, 208, 146-168.15. Chan,B.P.L. (1989) Thesis, The University of Birmingham.16. Sasse-Dwight,S. and GrallaJ.D. (1989) J. Biol. Chem. 264, 8074-8081.17. Gralla,J.D., Hsieh.M. and Wong,C. (1993) In Revzin,A. (ed.), Footprinting

of nucleic acid—protein complexes. Academic Press, Inc. San Diego,California, pp. 107-128.

18. Chan.B., Minchin,S. and Busby,S. (1990) FEBS Letters, 267, 46-50.19. Borowiec,J.A. and Gralla.J.D. (1987) J. Mol. Biol. 195, 89-97.20. Wood.D.C. and LebowitzJ. (1984) J. Biol. Chem. 259, 11184-11187.21. Schultz,S.C, Shields.G.C. and Steitz,T.A. (1991) Science, 253, 1001-1007.22. Roe,J.-H., Burgess,R.R. and Record,M.T. (1985) J. Mol. Biol. 184,

441-453.23. Ross,W., Gosink,K.K., Salomon,J., Igarashi,K., Zou,C, Ishihama.A.,

Severinov.K. and Gourse.R.L. (1993) Science, 262, 1407-1413.24. Suh,W.-C, Ross.W. and Record,M.T. (1993) Science, 259, 358-361.25. WangJ.C. (1969) J. Mol. Biol. 43, 25-39.26. Bliska,J.B. and Cozzarelli,N.R. (1987) J. Mol. Biol. 194, 205-218.27. Goldstein,E. and Drlica,K. (1984) Proc. Natl. Acad. Sci. USA, 81,

4046-4050.


thermal energy requirement for strand separation during

Documents