Nucleotide sequences that define promoters that are used by Bacillus subtilis sigma-29 RNA polymerase

.I. Mol. Bid. (1986) 192, 557-565

Nucleotide Sequences that Define Promoters that are Used by Bacillus subtilis Sigma-29 RNA Polymerase Philip N. Rather I, Regine E. HayI, G. Luann Ray2 William G. Haldenwang2 and Charles P. Moran Jr-l-f ‘Department of Microbiology and Immunology Emory Cnicersity School of &fedicine Atla.nta, Georgia 30322, L’3.A. 2Department of Microbiology University of Texas Health Science Center San Antonio, Texas, G.S.A.

(Received

17 April

1986, and in revised form

9 July

1986’)

There are at least five different forms of RNA polymerase holoenzyme in Bacillus subtilis. These enzymes differ in their sigma subunit. and their specificity for promoter utilization. One form of RNA polymerase (EC?‘) that contains a 29,000 fiIr sigma appears in B. subtilis about two hours after the initiation of endospore formation. The determination of the nucleotide sequences that govern utilizat,ion of promoters by Ea29 has been limited by the small number of cloned promoters that are recognized by Ea29. We have determined the nucleotide sequence of a recently isolated promoter (G4) that, is used exclusively by Ea29 was identified by S1 nuclease both in vitro and in vivo. The start-point of transcription mapping and dinucleotide priming experiments and the probable promoter element was sequenced. We compared the sequence with that’ of six promoters that are used to varying degrees in vitro by Ea29 and found these sequences to be highly conserved at the - 10 and near the -35 regions of these promoters. Single base substitutions were generated at posit’ions - 12, - 15 and - 36 of the G4 promot’er and assayed for their influence on utilization by Ea29 in in-vitro competition experiments. The effects of these mutations in G4 on its use by Ea29 support a model in which Ea29 utilizes its cognate promoters bJ interacting with unique nucleotide sequences at t’he - 10 region and near the -35 region of these promoters.

1. Introduction The sigma subunit of RNA polymerase from eubacteria enables the holoenzyme to recognize promoters (Travers & Burgess, 1969). In Bacillus subtilis and fischurichia coli there are different forms of RNA polymerase that differ in their sigma subunit’ (for a review. see Reznikoff et aE., 1985). These alternative sigma subunits bind to core RNA polymerase and direct the polymerase to different classes of promoters containing unique base sequences. I+?e expect that’ the identification of the nueleot.ide sequences that specifically signal recognition by forms of RNA polymerase that differ only in t’heir sigma subunit will lead to the t Author addrrssrd.

to whom

correspondence should be

ident,ification of the nucleotides within a promoter that interact directly with each novel holoenzpme and perhaps even the alternative sigma subunit itself. The most abundant form of RNA polymerase that is present in growing B. subtilis ( EcF~~)contains a 43.000 M, sigma subunit (Price et al., 1983; formerly c?‘, Shorenstein & Losick. 1973). The promoters recognized by Eo43 have common nucleotide sequences centered 10 and 35 base-pairs upstream from the start-point of transcription; the - 10 region and the -35 region, respectively. These sequences are highly conserved with those that govern recognition of promoters by t’he most abundant form of RNA polymerase in E’. coli, EC” (Moran et al.: 1982). A minor form of RNA polymerase found in growing B. subtilis. t,he sigma37 RNA polymerase (Eo3’), recognizes a different

(‘lass of’ promoters.

including

thosr for gwes that

iIre transcribed during an earl? stage of endospore fijrmation (Haldenwang & I,oslc*k. 1980). 13:a3’ also rwognixrs its cognate promoters by interacting \vith a sequenw of nucleotides centered at the - 10 rcbgion and a sequence at the - 35 region of t,he I)romotrrs but, these sequences are different from those that are utilized by Ea4” (Tatti & Sloran. l!W). The interaction of distinct RKA polytnerase holoenx~mes with novel base sequences at the - 10 uric1 -35 regions ma,?- not be universal. An E. wli RSA polymerasc h o 1oenzyme containing the T-C p1G5 o recognizes T4 latr promotSers with conserved src~rrrncw only at the - IO region (Elliott & (:riduwht~k. 1984). while another holoenz\-me with t hv 13. coli “heat shock” sigma facator (a”“) is ~wlirvetl to require a unique consensus sequenc*e at and a sequen(~e similar to that ttw - IO region rwogniwd by t&770 at its - 45 region, although this sec~uenc~~ is displawd by eight nueleotides ((‘owing P/ CL/.. 1985). These dat)a have been interpreted to mean that the sigma subunit of the holotwz~rr~r interacts only bvith the -10 region while other (components of RNA pal\-merase make eontact at 1h(‘ - 35 region (C’owing it ol.. 19X6). A form of RNA polymerase ( Eaz9) that contains a 29.000 ~11, siyma appears in 11. ,subtili,s about two hours after thr initiation of endospore formation (Haldenn-ang 4 rrl.. 1981 ), Ihes fto29 havr unique sequenc~e requirements at both the - 10 and -35 regions or will it interact with novel bases only at the - IO wgion md use a sequence at the - 3,5 region that cwuld be “common” to Zj. subtilis RNA polymerase! In order t,o det,ermine the nuclrotide srquewes that 1JV FCJ*". \V(' signal recwgnit ion of a promoter 1 scyuencwl a promoter (G4) from /I. shtilis that is uwd IJot h irr c*iw a,nd in vitro by IGJ*~. but not 1)~ Kc43 or 11:~~‘. 1Ve then wmpared the sequenw of (:J to those of’ six addit)ional promoters that are ul ilixetl by IQ729 and tested the effects of’ base subst,itutions in this promoter on its utilizatiorl I)!, Fo29 i,/ /~ifro. Th(. results Ithad 11~ to ~~011~*l~tl~ that IL*’ utilizes its eognat,e promoters by intrrac+tin~ with the uniqw nuvleotide srquen~es at t\vo rryiows of' these promoters.

2. Materials and Methods The nucleotidr sequence of t,hr G4 promoter region was detrrrnined by thr dideoxy chain terrninat,ion method. The strategy for sryuencing is outlined in Fig. I. The I’S&AhaITI. I’.s~TLS~USA. PstTLHindIII, PstI-CM and UnT--HirLdIII DSA fragments (Fig. 1) were each cloned into M13mp18 and Ml3mp19. The 15.base primer from Kew England Riolabs was used for priming of DSA t~rmplat~rs. svnt,hesis of t,hr phage and plasmid I;ropapation of phagr DKA and sequencing reactions were done as suggrsted by Sew England Riolabs in t’hrir technical bulletin M7J Cloning ZIJA Sequencircy System. The sequence of the mutant promoters that were cloned in pUC?-19 was determined by the dideoxy chain termination method using denatured linear templates.

Sigm:t-2!) KS;\ pol,vrrirr~as~~from I{. scthfili.u $1I\- ~1its purified 1)~ gradient rlution from 1)NA-c~~ll11l~w il.5 described (Tatti 8r Moran. 1985). E:a2” was assa~.c~tlin a run-off transcription using a pMH9 promoter trrrrplirtt~ ah described (Tatti & Moran. 1985). A sample of’ t 1112 l)uriticvl Ea2’ was analyzwl b?- c,l~~c~troI)horrsisin sotliunl clotl~v~~~l sulfate~pol~ac~r~lalnidr gel. .-\ftrr staining this pl \vith thr /I. [Y. I a11(1 29.000 .M, sigma silver reagent only subunit were visible. The l)rotocol for thv prr~binding tranac*ription wac+iotlh has brrn deswibed (Tatti r/ (11.. I!%j). E:a’” wah incubat,ed at WC’ for 10 min with 2 pg of DX.1 tvmplatt~ (unlrss noted in Figure Irgrnd) in a 40.pl rtwtion containing IO”, (v/v) plywrol and no caarrirr I
Plasmid pR,Hl a-as cwnstructed by ligating the 2i4 bpt 1’stTL.~haIIT I)SA fragment that contains the (:-I prornotrr (Fig. 1) between the P&I and SntaT sites of pL’C”19. In order to do high-resolution S, nwlraw mapping of thr (:4 transcript. this plasmid (pKHI) was I I5 bp c+arrd at the EcoRT site that is about downstrranl from thr promoter and labt+d at thtl 5’ kinasr excshange termini with j2P in a pol~naclrotidr reacntion (Maniatis rt trl.. I9ti2). Aftrr calra\agt> with /‘s/I. this labrlrd 289 bp f)K;1 fragment \vas f)uritirtl aftet vlrctrophorrsis into agaroscl grl with a Ion- melting tr~mperatrrrr. Xon-labeled RSX that was genrrat,ed in a run-off transcription reaction (Moran ut al.. 1981 ) by ECr2’ from the (:4 promot)er in plasmid pRHf that had been (aleaved at thr Rgl I site is about 260 by downst~rram from the promot,cr. This RICA was used to protect the rntllabrlrd I>KX from S, nwlease. The hyhridizat)ion and S, c&onditions were those originall? nnclease reaction suggested to us by ($. 1Vl’rr and drwribed by Ray rt ~1. (1985). The 6, digestion products were subjec%ed t,o rlect~rophorrsis into a high-resolution DS.4 srquenving gel next to thr dideoxy seyurncing reaction lwoducts of t,his US=\. Thrsr sequrnving reaction produ& were generated as dewribetl in the previous section except that they were c*lravrtl with EcwRT beforr t~lt,c,trolJhorc~sis in c,rtlrr, to t .%bbwriations dodrcyl sulfattk.

used: bp. base-pairs: SOS. scldium

Promoter

compare t,heir sizes with the S, digestion product was la,belrd at t,hr EcoRI site (Kudo it al.. 1985).

(d) Oliyonuclrotide

facilitat,e the determination of the nucleotide sequence of this region. The nucleot,ide sequence of the 474 bp that includes the putative promoter region was determined as described in Materials and Methods and is shown in Figure 2.

that

niutapwsi.s

To construcst base subst,itutions on the CS promoter we used an oligonucleotide mutagenesis strategy as outlined by Zoller & Smith (1983). The 274 bp PA-AhaITI DK;A fragment containing the wild-type G4 promoter was cloned between the Z’stT and SmaI restriction sites of M13mpI9. The 3 tnut,agenic oligonucleotides used were 5’L4A4(:AAATBT.~‘C(:ATTTATT(::~.4TTCATTAC: 3’. TTAGGCTTTT(~ 3’. and ~‘TTATTGTAT(‘(IATTAQ(‘TC’A 3’. The caonditions for the annealing reactions. second-strand synt,hesis and transfection were the same as described (Tatti d al., 1985). except that the doublestranded hrtrroduplrxes were treated with 5 units of S, nuclease for PO tnin at 37°C in order to eliminate singlest)randrd mol~ulrs. After transfection into E. coli 71.18. the plaques were screened by hybridization with the radiolabeled tnutagenic oligonueleotide (Zollrr & Smit,h, 1983). The seyuenc*r of the phage D?iA from each positive plaque w-as det,rrmined and the mutated promot,ers rrcloned between the PstT and SacT restriction sites of pl’C”19. The srquc~nc~rsof t,htA Z’s&Sac1 D?U’A fragments

(b) The stwt-poinf

Smnl

restriction

about) 115 addItiona

rlrctrophoresis

fragment

Several DNA fragments from B. subtilis that contain promoters that are utilized by EONS. both in vitro and in rGo. have been cloned into plasmid pBR322 of E. coli (Ray & Haldenwang. 1986). The recombinant plasmid pGLR4, which contained the promoter that was used most efficiently in vitro by F,o~~, was selected for further characterization. The analysis of the size of several run-off transcripbs that were generated in vitro by Eaz9 and S, nuclease mapping of transcripts made in vitro and in viz:0 indicated that the promoter (referred to here as the G4 promoter) was located near the ClaI restriction endonuclease cleavage site shown in Figure 1 (Ra\- 6 Haldenwang, 1986). The region of H. subfil,is Dr;‘A around the CZaI site in pG LR4 was subcloned into M13mp19 and M13mp18 in order to 55

PSf t

Cl0

169

. .

4

1)~

(Fig. 3. lanes a, 1) and c), The DKA

that

resulted

from

S,

digestion

terminated position one

+ I in Figure

triphosphate than is required for subsequent elongation of the transcript. The requirement, for a

high concentration of initiating ribonucleotide triphosphate can be circumvented by, a high concentration of dinucleotides that are complementar? to the region near the start-point of transcription (Minkley & PriImow. 1973). Transcription

from

the

(2

330

was

not

*

* Ah0

100

promoter

+

215

SOU

114

4 .

from dig&ion

Furt,hermore. since any 5’ terminus located bv S, mapping could be a processed end. we determined which dinucleotides were capable of priming transcription frotn the G4 promot’er. Initiation of transcription in vitro requires a higher concentration of the initial ribonucleotide

.w

100

nucleotides

plasmid

at the 5’ described of this from the la,bel and

2.

t t

to form

(somigrated with the dideoxynucleotidr product that corresponded to the nuclrotide upstream from t,hr positim

region

t

of pl!@lS

S, nuclease (Fig. 3, lane e). Size st,andards were generated by the dideoxy sequencing reactions from a primer that was annealed near the promot.er distal side of the EcoKT restriction sitBe and subsequent cleavage of these products with EcoRT before

3. Results of the promoter

sites

pRH1. This DKA was labeled with 32P terminus of the EcoRI restriction site as in Mat,erials and Methods. Hybridization DNA fragment with RNA made ,in z&-o (24 promoter by Eoz9 protected the 5’ end

were determined and found to contain only the desired single base substitutions.

.wquenrrs

of tmnscripfiot/

sCl nuclease mapping. dinucle&de priming experiments and precise measurements of t)he size of run-off transcripts were used to determine the exact start-point of transcription from the C:4 promoter. For the S1 nuclease mapping experiments, the 274 bp Pstl-AhaIIT DNA fragment that contains the promoter was cloned between the I’stT rind

that contained the mutant promoters in each plasmid

(a) The nurlrotidr

559

in B. suhtilis

Recognition

116

H/f?

* * * *

4

Figure 1. Restriction map of the G4 promoter region. The restriction endonuclease cleavage sit’es of P&I, C’ZaI. AhaITI. Nnu3A and Hind111 are presented. The numbers indicate the distance in nucleotides between sites. The straight, arrows below the restriction map represents t,he direction and extent of the sequencing reactions used to determine the I)SA sequence shown in Fig. 2. The wavy lines above the map represent the run-off transcript’s that have been generated by Ea29 except for the 55 nurleotide long transcript. which appears to be a prematurel,v terminated transcript.

1

CTGCAGTATC P.91

TGATGAAAAT

TTGGTTTACT

GCATCAGGAA

TTGCTCCTCT

GGAATTTGCG

61

GTAGAAGTAT

TGAAAAAATT

AGATGTATTT

GCACCAGATA

CATTGCGTTC

ATGGTATAAG

121 CTAAATAAGA

CTCAAAAGCC

-35 TGAATATTTC

TTTGAGCTAA

-10 TGAATACAAT

AAATCGk%?

(Ed')

+1 c/c71

181

AAAACACTTT

AGCAGTTTTT

TGTAGT-AACTTATITTT

TTTATATGTG

241

TCTAGGAGCT

AAATAAAGAA

ATGCAGAAAC

TTTTAAAAAG AhoIU

AATATAGATT

GGTAAGCCTT

301

ATTTCGAATA

GAGGTTGATT

TGGGTTAAAA

CTGTTCTTAT

TAAGGCTCAT

TTCATGGTGC

(EONS)

-35 361

TAAAGATTCA

TTTGGAGAAG

ACTGTCTTGA TCAGAAAAGA Sau 3A

421

AAAAGATATC

GGGAGGAACT CTTTTGAAAA

AGATGTTGAT

CCCTTCCAAT

-10 ATATTGAAGT

GTTA

Figure 2. Structure of t,he G4 promoter region. The nucleotide sequence of non-transcribed strand of the G4 promoter region is shown. Transcription occurs from left to right with the start-point of transcription indicated as + I. Restriction endonuclease cleavage sites are indicated below the sequence. The converging arrows above the sequence identify a region of dvad symmetry followed by a series of T residues around nucleotidr number 630 that may cause the rho-independent termination that generates the 55 nucleotide transcript. The - 10 and - 35 regions of the (kl promoter. which is recognized by Ecrz9 is indicated. as is the - 10 and -35 regions of a promoter that is utilized by lCa4” (E,:os5) (data

not shown).

observed in a run-off transcription assay with only 2 PM-ribonucleotide triphosphates present (Fig. 4. lane e). However, when 150 ,UM-ATP was added to the 2 PM-ribonucleotide triphosphates, transcription was initiated (Fig. 4; lane d). Tn the presence of 2 PM-ribonucleotide triphosphates, only the dinucleotides CpG and GpA were capable of priming t,ranscription from the G4 promoter (Fig. 4, lanes f and g). This result is in good agreement with the S, mapping result because the sequence CGA is found at the position of t,he 5’ terminus determined by S, (see Fig. 2). nuclease mapping experiments Moreover, the run-off transcript that was initiated with (:pG migrated more slowly (to a position 1 to 2 nucleotides larger) than did the transcript that was init,iated with ATP in a high-resolution DSA sequencing gel (Fig. 3, lanes f and g). The position of these two transcripts relative t,o the DXA size markers corresponded to a size about’ nine nucleotides larger than the S, protected DNA (Fig.

3. lane e). This

is the

expected

result

since

RSA migrates slight’ly more slowly than its c*omplementary DNA sequence in this system (Moran et nl., 1981). It should also be noted that’ in all of the run-off transcription assays (Figs 3 and 4) we observed a 55 nucleotide long transcript predicted run-off transcript.

in addition to the This 55 nucleotide long

1ranscript was found after transcription of templates that were cut either at the BglI site, 260 bp downstream or at the EcoRT site. 115 bp

downstream from the G4 promoter (Fig. 4, lanes a and b. respectively): therefore. this .55 nucleotide transcript appears t’o he t,he product of premature termination of transcript’ion from the G4 promoter. This t,ranscript appears to terminate near a sequence

of dyad

symmetry

followed

by a series of

T residues (Fig. 2). It would be interesting to know if the function of this rho-independent type of t)ermination sequence is modulated in vine. ((2) ICffectx

of mutations

in the G4 promoter

The effects of base substitutions in another promoter (ctc) that is utilized by several forms of H. suhtilis RNA polymerase in vitro, including E~J’~. have been reported (Tatti & Moran. 1985; Tatti rf al.. 1985). (‘omparison of t’he nucleotide sequences of the ctc promoter and the G4 promoter lead us to predict probable single base substitut’ions in the G4 promoter. which should influence it’s utilization in 1itro by Eo 29 (Fig. 5). We thoug ht that an A to T in G4 at the position t,hat is transversion homologous to - 12 in t’he ctc promoter may by EcT~~, since a similar decrease utilization substitution in cfr decreased its utilization by ECUS. Similarly. transit’ions at positions homologous to -36 and - 15 also were expected to decrease utilization by Ea29 if these nucleotides serve homologous roles in the G4 promoter. Oligonucleotide-directed mutagenesis was used t,o construct these single

base substitutions

in the G-I promoter

561

Promoter Recognition in B. subtilis

CAT

-+

G IT I T ‘B

r, T\

A c lC T A

T

-I I I c I c iI ‘T hI IT

c

E I TI T I f

’f t

abc

de

fg

hijk

Figure 3. S, nuclease mapping of the .5’ terminus of the G4 transcript. Eon-labeled RNA generated in oitro by Eaz9 from the G4 promoter was incubated under hybridization conditions with Pst-EcoRI DE.4 fragment that had been labeled with 32P at the 5’ terminus of EcoRI sit,e (see Materials and Methods). These hybrids were digested with S, nuclease. Tn lane e is shown the digestion products of this reaction after electrophoresis int,o a high-resolution DNA sequencing gel and autoradiography. The left-hand arrow points to t’he predominant product’. The end-labeled PstIEcoRT DK’A fragment also was digested without the addition of RP\‘A (lane d). The products of the dideoxy sequencing react,ions produced as described in Materials and Methods were run on the same sequencing gel (lanes a to c and h to k). The lett,ers above each lane indicates which dideoxynucleotide was used to terminate DKA synthesis. Also run on a high-resolution sequencing gel are [cz-~*P]UMP labeled run-off transcripts (lanes f and g), The transcript in lane f was initiated with ATP. while the transcript in lane g was initiated with the dinuclrotide CpG. The right’-hand arrow points to t,he run-off transcript,s. and the asterisks indicate the positions of the prematurely terminated transcripts. as described in Materials and Methods. Dideoxywas used to determine nucleotide sequencing whether the mutant promoters contained only the one desired base substitution. In order to determine the effects of the base substitutions on utilization of this promoter by Ea29 we used a competition assay. In this assay each mutant promoter competed with the wild-type promoter for a limiting amount of EONS. In order to visualize transcripts generated from two different

templates in the same reaction, the two templat’es were cut with different restriction enzymes so that transcription from each promoter generated a different-sized run-off transcript. In each case where equal amounts of mutant and wild-t,ype templates were mixed, most of the transcription by the limiting amount of Ea29 was initiated from the wild-type promoter (Fig. 6: lanes b. d and f), which generated the shorter run-off transcript (transcript e). The mutant promoters. however. were able to

A

-

CG GA AU AC CC CU GC GU UC

t-

abcde

f

g

hij

k

Im

n

Figure 4. I)inuc~l~:ot,ide-Erimed tranacript,ion.

Run-off transcript)s were generated from (:4 by F:(rz9 ilk thca ~~res~n~ of only 2 phi-ribonucleotide triphosphates and 150 ELMof the dinuclrot~idr indicated above the lane (lanes f to n) or in thr absence of added dinucleot,ide (lane e). The tranwriIkion reaction analyzed in lane d was init,iated by t,hc. addition of I.50 ptw-ATI’ and 2 ~$1 of each of the (&her ribonucleotide triphosphatrs. The G-l promoter templa.t,r (pRHI) in path of thr above rrac*tions and also lane b had been cleaved at the EwRT rrhtriction site about I 15 bp downstrram from the start-point of transcription. The arrow marks the position of the I 15 nuclrotidr rurl-off transcript. Thr I)osition of the 5.5 nr&otidr prematurely terminated transcripts is indicated by the t. Sate that, this t t,ranscript also is generated from a trmplatr that was oleavrd at the BylT site that is 260 bp downstream from the promoter (lane a). Plasmid pKRSI:! tLat had been c.ut with Hpcrl I were used for size standards (Iant’ c,).

direct t~ranscription by Ea2’ (t,ransrript p) when no wild-type template was added to the rewtion (Fig. 6, lanes c. e, and g). We concluded that each of these three single-base substitutions reduced the efl’iciency of ut’ilization of this promoter in <%itro 1)~

-35 GAA-AA-T

Fa2”. , ctc

recognition

AA A& TTTTCGA~TTTAAATCCTTATCGTTATGGGT~TTGTTTGTPATA A

4. Discussion (a) Promoter

-10 CATATT-T

hj Ea29

Base substitutions in a promoter (ctc) used 1)~ at least two forms of RNA polymerase have been described that either increase or decrease t,he utilizat’ion of the promoter in vitro by Ea2’ (Tatti Hi Moran. 1985: Tatti et al., 1985). We found that homologous base substitutions in a promoter ((24) used exclusively by ,Ea 29 had qualitatively similar effects on its actlwty. Both promot.ers are used more eficientlg in vitro by Eaz9 when there is an A rather than a T at position - 12, an A rather than a G at. - 15, and G rather than a,n A at position - 36 (Fig. 6). Although the base substitutions in G4 had qualitatively similar effects to the homologous suhstitution in ctc, these substit.utions did not have

G4

T

AAAAGCCTQAATATTTCTTTGAGCTAATGeATFAATAAATCGLj A

Figure

+1

5. Base substitutions

G

T

in the &

and (2 promoters. The nuclrotide sequences of thr ~tr and (X are ShOWI (noIt-transcribed strand). promoters Transcription proceeds from left to right and the startpoints are indicated by the underlined nualeotide. The arrowheads indicate the effects of the base substitutions on utilization of the promoter by Eaz9 (Tatt.i rt n/.. 1985. and this paper). Arrowheads t,hat point upwards indicat,r cahanges that increase the efficiency of promote1 utilization; and conversely those that point downwards indicate changes that decrease utilization by Eaz9. The sequences shown over the - 10 and -35 regions are sryuences that are highly conserved among promoters that, are ut.ilized by Eaz9 (see Fig. 7).

563

Promoter Recognition id H. subtilis

11

2233

555 65 6 GAA-AA-T

566744 6 CATATT-T

-35 TTTTCCAGGTTTAAATCCTTATC

-10 +1 GTTATGCATATTGTTTGTAATA

G4

AAAAGCCTGAATATTTCTTTGAG

CTAATGAATIlCAATAAATCGfi

BV

ATTTCTTCGAATAAATACTATAA

ATGAAAACTATGATGTCAGALJJ

cfc CA

t

ATATTTTTTGAAAAAATAGGATA

TAGTTPJXAATTAGGTCATA~

EB9

TATGAAGTTAAAAGCTATGTGTT

CAATAGCATATTTTGAJlTATGGfi

0.3 ’

ATCTGATTTAACAAAAGATA

CAGTCACATATTATCGTGACGTC

spoIID

APJXAGAGTCATATTAGCTTGTCCCTGCCCATAGACTAGACTAGijG

Figure 7. Promoters that are ut,ilized by RUDE. The nucleotidr seyuences of the non-transcribed strands of 7 promoters that are ut,ilized ire Gtro by Eo29 are shou n. Transcription starts at the underlined nucleotidr in each case and proceeds from left to right. The rtr (‘-4 promoter is the mutant ctc promoter that is used most efficiently 1)~ E:az9 (Tatti it al.. 1985). The nucleotitir sequence and start-point of transc*ription of spoIlI> Mas described 1)) Rang rt ul. (19%); promoters t. pMK9 and 0.3’ 1)~ I’nnasch (1983): and KV by Hay & Moran (unpublished results). The numbers above the sequences at the - 10 and near the -35 regions indicate the number of times that the nuclrot~idr appears in the 7 promoters.

abcdefg Fig. 6. Vompetition of mutant G4 promoters. Plasmid containing t’he wild-type (X promoter was cut with EcoRT. while each mutant promoter template was cut with /‘PUTI. Run-off RR’As generated by initiation at these promoters were 115 bases (e) and dOti bases (p). respectively. The template included in each reaction (I pg each) is indicated above each lane: W. wild-type: mutant I. t~ransversion at position -12; mutant 2. transit,ion at -Xi; and mutant 3. transition at position - 15. An autoradiograph of the transcripts after rlectrof~horesis into a 7 wurea/polyacrylamide gel is shown. The transcript labeled t IS a result of the premature termination event that occurs on all of these t.emplater.

identical quant’itative effects. This may reflect the fact, that. the efficiency of interaction of RNA polymerase with a promoter is the out’come of the sum of it’s useful interactions with numerous nucleotides. therefore. the relative importance of arig one nuclcotide contact in determining a promoter’s activitv is likely to differ het’ween promoters. For this reason, we have not made quantitative estimates of the relative strength of each mutant promoter but it is clear even in the run-off transcription assay that the transversion at position - 12 had a more severe effect on utilization of the G4 promoter by Eoz9 than did the other two base substitutions. This is the same substit,ution

that had the most severe effect, on ut,ilization of the ctc promoter bg Ea 29 (Tatti et al.. 1985). In Figure 7 we compare the nucleotide sequences of seven promoters that are utilized irk citro by IGJ~~. To align this sequences we considered the effects of the base substitutions in the ctc and (:4 promoters and the start-point of transcript’ion of each promoter. Ry permitting minor adjustments in the spacing bet,ween the - 10 and - 35 regions of these promot,ers we found that these sequences are highly conserved at the - 10 region and near the - 35 region. This consensus sequence differs slightly from the previously published sequence (Tatti et nl.. 1985) because t*hat sequence was derived by comparison of only four promoters and without knowledge

of the effects of base subst.itntions

in G4.

The consensus sequence shown in Figure 7 may not, represent the one that is utilized most efficiently by E(T29. since none of the known promoters has sequences that are exactly that, of t,he proposed consensus sequence. In run-off transcription assays it appears that the wild-type G4 promoter functions more efIicientlr than at least five of the other promoters in F’igure 7 and clearly t)he (X promoter does not match the consensus sequence at several positions (only the s;uoIID promoter has not been tested: Rather, Hay 8r Moran. unpublished results). Base substitut.ions in b0t.h rtc and (2 promoters that make the sequences in these promoters less like these conserved sequences caused the promoters to function less efficiently (Tatti et nl.. 1985. and this paper) and changes that make the ctc promoter more like the conserved sequence cause the promoter to function more efficiently (Tatt,i et nl..

I’. S. Kathrr

564

1985). Furthermore. transit,ions at each posit iolj oc,cupied by G or (‘ in t’he seyuence shown for f?c in Figure 7 outside these regions had no effect on utilization of that promoter by Err2’ (Tatti & Loran, 1985). Since t)he sequences at the - 10 region and near the -35 region are highly cbonserved among promoters that are used by l&~‘~. and base substitutions in these regions affect this use while base substitut’ions outside the - IO and - 35 regions do not. we conclude that H:a29 nt,ilizes its cognate promoters by interacting wit)h the nucleotides near t)he - 10 and -35 regions of these promoters. Moreover, the sequence of nucaleotides that are conserved at both regions differs significantly from those sequences that are thought to signal recognitjion of promoters by other forms of K.NA polymerase.

et al

presumabI\- in unison to recognize ;I I~r~~tnotet‘ sequence is entirely unknoMn. In summary. we have determined the nuc~leot~itlc sequenw and start,-point of transcription of’ ii promoter that is used efficiently by Ho2 The effects of base sub&itutions in this and another promoter (rtc) t,hat is utilized by Eaz9 support H model in which Eo29 utilizes it’s c*ognat)e promoters by interacting with sequences near the -- 10 and -35 regions. The sequences at) t’hese regions werc~ found to be conserved among promoters that are utilized by Eg29 and are different from the srquenczes at these regions that signal recognition of’ promot’t>rs by Ea43 or Ea3’. This work was supported by l’ubli~ Health Servi(.tl grant ATN519 fi-om the Xational Institute of Allrrgy and Tnkctious IXseases t,o (‘.%I.

(h) Implications

Five forms of B. subtilis RNA polymerase that differ only by their sigma subunits. EcT~~ (Moran et al., 1982), Ea3’ (Tatti dt Moran, 1984), Eo29 (Tatti & Moran, 1985, and this paper), Eo gp33,34 and E:o gp28 (Talkington & Pero, 1979) appear t’o use t’heir cognate promoters by interacting with distinct nucleotide sequences at’ two regions of these promoters near the - 10 region and t’he - 35 region. A straightforward interpretation of this finding is that the sigma subunit acts by direct)lT contacting the nucleotides of these regions (LosIck 8s Pero. 1981). Tf this is true, it is intriguing that the conserved sequences that t)he sigma proteins are recognizing arp in each instancae at) approximately the - 10 and -35 regions, regardless of the overall size of the sigma factor. If the sigma factors directl) contact) these two regions it implies that the promoter recognition regions of the sigma protein have a common topography on the surface of the sigma protein. Presumahl~ the amino acid sequence of the promotter-recogmtmn site would be unique for each sigma factor due to t)he need to interact with a novel nucleotide sequence: however. the scaffolding for presenting this amino acid sequence ma\- be similar among the sigma proteins due t’o sikilar primary sequences elsewhere in the protein or common elements of tertiary structure. The sequences of amino acids at two regions in t)hr (barboxy-terminal portion of several sigmas conform to the cc-helix-P-turn-x-helix structural motif of t,he DXA binding domains of procaryotic repressors and act,ivators (Stragier et nl.. 1985; Reznikoff et al.. 1985). This may represent bhe recognition element scaffold. Models of the mechanism of sigmamediated promot)er recognition that are based on the analysis of promoter sequence requirements for recognition by some RNA polymerases are likely to be complicated by the fact that at least in some systems more than one sigma-like protein may be bound to RSA polymerase at the same time (Pero pt nl., 1975: Malik et al., 1985). At present, the mode proteins acting of action of multiple “sigma-like”

References (‘oning.

II.

iv,.

f
,J. C’. X..

(‘raig.

E. A..

\Voolford, L.. Hendrix. R. 8’. & Gross. C. A. ( 198.i). Pror. *Vat. ilcad. A%.. I’.S.A 82, 2679--P683. Elliott. T. 8r Geidusrhek. tC. I’. (1984). Cell. 36. %I f-119.

Haldenwang. W. G. & Losick. R. (1980). E’ro~. ;VCI~. ilcad. Sri., f’.AS.rl. 77, 7oo(t7004. Haldenwang. FI’, (i., Lang. K. 8: Losicak. R. (1981). (‘cl/. 23. 6f5--6&f. Hay. Ii. E:. 8r Moran. (‘. I’., .Jr ( 198.5). III Mol. Hiol. 0s BfifmJbio/. Difffwrhation. pp. 170-l 7.5. A.S.hl. Il’ashington. D.C’. Kudo. T.. \-oshitakr, J.. tiato. (“. I’sami. R. & Horikoshi. K. (19%). .J. kKtWiCJ/. 161. IS-~165. Losick, R. 8: I’ero. ,I. (1981). (1e11,25, 58%584. Jlalik. S.. Dimitrov. M. & Goldfarb. A. (1985). ,J. Noi. Biol. 185. 83&91. Maniatis. T.. Fritsch. E. F. & Sambrook. ,J. (1982). Molecular C’lonirq, Cold Spring Harbor Laboratory T’rrss. (‘old Spring Harbor. iY.Y. Minklry, F:. G. & Pribnow. I). (1973). .I. Mol. Biol. 77. 25.5 ~“77. Moran. (‘. I’., *Jr. Lang. 8.. Banner. (‘. I). H.. Haldrnwnng. W. G. & Losick, R. (1981). (Jell, 25. 78:)-x I Moran. (‘. I’.. *Jr. Lang. X.. LeGrice. S. F. .J.. Lee. (i.. Strf)hans. M.. Sonenshein. 4. C’.. Pero, .I. KT Losick. R. (1981). Nol. C&x. Urnet. 186, 339-346. l’ero. .I.. Tjian. R., il;elson. ,J. 8 Losick. It. (197.5). ,\hturc (I,ondon) 257, dlti- “5 I Price. IA. \V.. (:itt. M. A. hi Doi. R. H. (1983). I’roc. AVat. .4cnd. Sri.. f7dS.A. 80. 4074-4078. Ray, C’.. Hay, It. F:.. Carter, H. L. & Moran. (‘. I’., *II (1985). J. Ha&e&l. 163, 6 lo--61 4. Ray, G. L. & Haldenwang. \V. (Z. (1986). .1. Hncteriol. 166. 476-478. Rrznikoff. \V’. S.. Siegefe. II. A.. (‘owing, 11. W’. 8r (iross. (‘. A. (1985). dnnu. Rw. Genet. 19. 35%38i. Rang, S., Rosenkrantz. M. S. 8 Sonenshein. A. T,. (1986). .I. J3actrriol. 165, 771-779. Shorrnstein. R. (:. 81 Losick. R. (1973). ,I. Hiol. C’hrnc. 248. 6170~-6li3. Stragier. I’.. I’arsot. (‘. 8: Bouvirr. .J. (198.5). PEBS Lrttrrs, 187. 1l-15. Talkington, C. & Pero, J. (1979). Proc. LV~t. Acad. hi., f7.S.A. 76, .5465--5469.

Promoter

Recognition

TatA. K. M. & Moran. (‘. P..
in B. suhtilis

565

Travers. A. A. & Burgess. R. R. (1969). TV&w (London), 222. X37-540. I’nnasch. S. L. (1983). Ph.D. thesis. Harvard University. (‘ambridge. MA. Zollrr. &I. J. & Smith. M. (1983). AII&otls En:!ynwl. 100. 46% 600.