ABSTRACT We studied the expressionof the cya promoter ( s ) in cya-lac fusion strains ofSalmonella typhimurium and demonstrated cAMP-receptor-protein ( CRP ) - dependent repressiobny CAMP.Expression of cya was reduced about fourfold in cultures grown in acetate minimal-medium as compared to cultures grown in glucose-6-phosphate minimal-medium .
Expression of cya wasalso reduced about fourfold by addition of 5 mM cAMP to cultures grown in glucose minimal-medium .
We constructed in-vitro deletion and insertion mutations alteringa major cya promoter ( P2 ) and a putative CRP binding site overlapping P2 .
These mutationswere recombined into the chromosome by allele replacement with M 13mp : : cyarecombinant phages and the regulationof the mutant promoters was analyzed .
A 4-bp deletion of the CRP binding site and a 4-bp insertion in this site nearly eliminated repression by CAMP .
A mutant with the P2 promoter and the CRP binding site both deleted exhibited an 80 % reduction in cya expression ; the 20 % residual expression was insensitive to cAMP repression .
This mutant retaineda Cya ' phenotype .
Taken together , the results establish thcayta gtheene is transcribed frommultiplepromotersoneof which , P2 , is negativelyregulated by the CAMP-CRPcomplex .
Correctionforthecontributiontotranscription by theCAMP-CRPnonregulated cya promoters indicates that the P2 promoteris repressed at least eightfold by CAMP-CRP .
THE observation that crp mutantsoverproduce adenosine-3 ' ,5 ' - cyclic-monophosphate ( CAMP ) in both Salmonella typhimurium ( RAPHAELaInd SAIER 1976 ) and Escherichia coli ( POTTERC , HALMERS-LAR SON and YAMAZAKI1974 ; WAYNEand ROSEN1974 ) suggested that the synthesis of cAMP is inhibited by CAMP-CRP receptor protein ( CRP ) .
Eithcyear expression or adenylate cyclase activity might be negatively regulated by the CAMP-CRPcomplex .
Although the positive control function of CAMP-CRPin regulating gene expression has been studied in greatest detail , it is important to note that thesynthesis of an approximately equal number of proteins is repressed as is stimulated by cAMP ( MALLICKand HERRLICH 1979 ) .
Studies of the regulation of cya expression in enteric bacteria have led to confusing results and contradictory conclusions .
MAJERFELDet al. ( 1981 ) and BOTSFORD and DREXLER ( 1978 ) reportedapproximately 4-fold and 20-fold CRP-dependent inhibition of cAMP synthesis in Escherichia coli , respectively , but it was not determined whether thseynthesis of adenylate cyclase was directly affected .
BANKAITaInSd BASSFOR ( D1 982 ) isolated chromosomal cya-lac fusions in E. coli and observed atwofold or less repression of cya expression ' Presentaddress : RegeneronPharmaceuticals , Inc. , Suite 10 , 777 Old Saw Mill River Road , Tarrytown , New York 10591 .
` Presentaddress : Division of BiochemistryandMolecularBiology , Department of Molecular and Cell Biology , University of California , Berkeley , California 94720 .
` To whom correspondence should be addressed .
Genetics 125 : 719-727 ( August , 1990 by addition of 5 mM cAMP to the growth medium , which they concluded was not physiologically significant .
KAWAMUKAIet al. ( 1985 ) reported a similar extent of repression of cya expression by cAMP in E. coli using cya-lac fusions carried on multicopy plasmids , and concluded the repression was physiologically significant .
T h e findings of KAWAMUKAeIt al. ( 1985 ) disagreed with those of ROY , HAZIZAand DANCHIN 9 ( 813 ) who also studied the regulation of E. coli cya-lac fusions carried on multicopy plasmids and reported that CAMP-CRPdid not repress cya expression .
In fact , they observed CRP-dependent stimulation of cya expression by CAMP .
AIBA ( 1985 ) and MORI and AIBA ( 1985 ) observed binding of CAMP-CRP to the E. coli cya promoter region in-vitro and demonstratedCRP-dependentinhibition of cya mRNA synthesis by cAMP in-vivo .
In work from this laboratory , strong CRP-dependent repression ( about ninefold ) was observed for S. typhimurium chromosomal cya-lac fusions when 25 mM cAMP was added to the growthmedium ( JOVANOVICH 1985 ) .
The nucleotide sequence of the regulatory region of cya of S. typhimurium was reported in a companion study ( THORNERF , ANDLand ARTZ1990 ) and shows regions of strong homology to the cya sequence reported for E. coli ( AIBAet al. 1984 ) as well as regions of weak homology .
The sequences are most homologous in regionsproposed tobeimportantfor cya expression in E. coli ( AIBA 1985 ) .
The regulatory region appears to contain a major promoter ( P2 ) , two minor promoters ( P1 , P 1 ' ) located approximately 200 bp upstream of P2 , and a putative CRP binding site that straddles the -10 hexamer of P2 .
In this paper , we report experiments with chromosomal cya-lac fusions thatsupport unequivocally a model of negative regulation of cya by CAMP-CRPin S. typhimurium .
We present genetic evidence for this model based on analysis of site-directed mutations in the putative CRP bindingsite .
We also report genetic evidence confirming the existence and physiological importance of multiple cya promoters in S. typhimu-rium .
MATERIALS AND METHODS Bacterialstrains , bacteriophages and media : S. typhi-murium strains are described in Table 1 .
The E. coli strains JM103 and JM110 have been described ( MESSING1983 ; YANISCH-PERROVNIE , IRA andMESSING 1985 ) .
Ml3mp recombinant phages are derivatives of either M13mp8 or M13mp9 ( MESSING and VIEIRA1982 ) and are described in Table 2 .
Transductions were done using phage P22 HT105/int -201 ( DAVIS , BOTSTEINand ROTH 1980 ) .
The virulent male-specific phage R 17 was from D. PRATT .
Nutrient broth ( NB ) , MacConkey agar , and MacConkey agar base were from Difco Laboratories , Detroit , Michigan .
2X Y T and VBG ( MILLER 1972 ) , M56 ( MONOD , COHENBAZIREand COHN195 l ) , C-minimal-medium ( ALPERand AMES1978 ) , and green indicator plates ( DAVISB , OTSTEIN and ROTH1980 ) have been described .
Carbon sources were added to agar plates at 1 % ( w/v ) and amino-acids at 50pg/ml .
Antibiotics were used at the following concentrations : 50 pg/ml ampicillin , 10 pg/ml tetracycline and 200 pg/ml streptomycin .
CyclicAMPwas added at 1 mM unless otherwise specified .
Constructionof cyu promoter mutations : The construction of mutations that disrupt the putative CRP bindingsite ( Figure 1 ) was facilitated by the presence of a unique BclI recognition site adjacent tothe P2 -10 hexamer .
Phage M13mp9 : : cyaA33R ( Table 2 ) replicative form ( RF ) DNA , isolated from strain JM110 ( lacking the dam methylase ) , was digested with BclI then treated with either nuclease S1 to remove the 4-base single-strand ends , Klenow fragment of DNA polymerase I plus all four deoxyribonucleotides to fill in the single-strand ends , or BAL-31 exonuclease to generate larger deletions ( MANIATISF , RITSCHand SAMBROOK 1982 ) .
Ligated DNAs were transformed into strain JM103 as described ( HANAHA1N983 ) .
The nucleotide sequence of an isolate from each class of mutations ( Figure 1 ) was determined as previously described ( THORNEFRA , NDL anAdRTZ 1990 ) .
Deletion 703 resulted in the loss of 4 bp immediately downstream of the P2 -10 hexamer and within the downstream half of the consensus CRP binding site palindrome .
Insertion 704 resulted in the addition of 4 bp immediately downstream of the P2 -10 hexamer and increases the spacing between the two highly conserved halves of the CRP binding site .
Deletion 703and insertion 704alteredthe putative CRP binding site without changing the adjacent P2 -10 hexamer .
Deletion 702 removed 32 bp including all of the -10 hexamer and all but 1 bp of the -35 hexamer of P2 .
The nucleotide sequence of the entire cya regulatory region of each mutant revealed no additional changes .
Mutations removing the beginning of the cya gene ( A700 and A701 ) were constructedfor usein recombining the promoter mutations described above into the chromosome and for fusing the mutant promoter to lac .
Deletions-700 and 701 were made by combining existing deletions of th promoter region ( THORNER , FANDL and ARTZ 1990 ) ( 2 ) .
The HindIII-BclI fragment of M13mp9 : : cyaA33R was subcloned into the HindIII-BamHI site of either M13mpS : : cyaA121 , resulting in M13mpS : : cyaA700 , or M13mpS : : cyaA23 resulting in M13mpS : : cyaA701 .
The strategy is showninFigure 2 focr onstruction of M13mpS : : cyaA700 .
These deletions extendfromthe P2 promoter to 13 codons ( A700 ) or 35codons ( A701 ) into the cya structuralgene relative to the T T G initiation codon ( THORNER , FANDL A anRdTZ1990 ) .
Allele replacement : The allele replacement theory and procedures have been described in detail ( ARTZet al. 1983 , BLUMet al. 1989 ) .
Briefly , the method is based on theability to select for M13 lysogens and segregantsof lysogens .
The amber mutation in gene I1 of M13mp8and 9 prevents phage replication in nonsuppressing host cells , but the phage can integrate into the chromosome if homology exists between the chromosome and the insert DNA in the recombinant phage .
M 13 lysogens are resistant to killing by malespecific phages ( e.g. , phage R17 ) because F-pili function is repressed .
This provides a positive selection for integrants of the recombinantM 13mp phage.M13 lysogens are sensitive to deoxycholate .
This provides aselection for segregants o f lysogens .
Allele replacement using M 13mp recombinant phages can be accomplished by selecting for either theM13 lysogen phenotype or the phenotype conferred by an inserted DNA fragment .
Phages M 13rnpS : : cyaAi'OO and M13mpS : : cyaA70 1 ( Table 2 ) were recombined into the chromosomeby selecting for the M13 lysogen phenotype ( Figure 3 , left ) .
A fresh and the segregant colonies appearing red ( 11703 ) or pink ( A702 ) and INS704 ) were saved .
Segregants containing A702 , A703 andINS704andan intact cya structural gene were designatedstrainsAZ2620 ( cyap702 ) , AZ2618 ( cyap703 ) and AZ2621 ( cyap704 ) ( Table 1 ) .
In all cases , segregation of M13 prophageswas confirmed by testing segregants for sensitivity to killing by R17 phage .
Cells were patched on NB agar supplemented with 5 mM CaCI2and 2 pl of an R 17 lysate ( 1O9 pfu ) was immediately spotted in the centerof each patch .
After incubation at 37 '' for approximately 5 hours a clearing appeared if the host cells had segregated the prophage .
Immobilization of ( cya-lac ) 696 and duplication of the cya locus : T o facilitate genetic manipulationand growth at highertemperatures ofstrainscontaining the cya-lac fusion , it was necessary to immobilize the fusion ( BLUM , BLAHAand ARTZ 1986 ) in strain AZ4344 .
Heat-resistant derivatives of strain A24344 were obtained by spreading 0.1 mo lf anovernight NB cultureon NB agar plates supplemented with cysteine and CAMP , followed by incubation at 42 '' overnight .
Colonies were pooled in 1 ml NB , diluted 1:10 into NB + cysteine + CAMP , and incubated at 37 '' with shaking for 7 hr .
A phage P22 lysate was made on the pooled cultureandthephage were pelleted ( DAVIS , BOTSTEINand ROTH 1980 ) andresuspended in an equal volume of T 2 buffer ( 1 mM MgS04 , 0.1 mM CaCln , 20 mM NanHP04 , 8.6 mM KH2P04 , 68mM NaCI , 28.7 mM KnS04 , and0 .001 % gelatin , pH 7.0 ) to removeextracellular/3-lactamase .
This lysate was then used totransduce strain LT2Zto ampicillin resistance on MacConkey glycerol + ampicillin agar , and white colonies ( Cya - ) were picked and purified .
A phage P22 lysate was made on each isolate and used to transducestrain LT2Z toampicillin resistance again scoring for Cya - .
This procedurewas repeated until a lysate was obtained that yielded a high frequency ( 99 % ) of Cyatransductants .
Transduction of the immobilized fusion fromstrain AZ2509into a Cya + recipientshould yield 100 % Cyatransductants .
However , we repeatedly observed that about 1 % of the transductants were Cya + , Lac + , and Ap ' .
Transduction of the immobilized fusion into one cya locus of a recipient containing a preexisting duplication of cya would yield this phenotype .
T h e duplications were confirmed as such by the following tests : ( 1 ) T h e duplication phenotype was unstable ( Cya - , Lac + , Ap ' and Cya + , Lac - , Aps segregants were observed ) but was stabilized by introduction of a recA mutation into the strain .
Feoxrample , the duplication in strain AZ2590 was stabilized by transducing the recAl mutationfrom strain AZ403toconstructstrainAZ2599 ( Table 1 ) .
( 2 ) Lysates grown on duplication strain AZ2599 were shown to transduce Cya-recipients to Cya + , Lac - , Ap '' or Cya - , Lac + , Ap ' .
( 3 ) Genesproximal to cya ( e.g. , metE and ilv ) were frequently included in the duplications .
Genetic complementation experiments in S. typhimurium often are done with heterologous E. coli F ' episomes since thereare few S. typhimurium F ' episomes available .
T h e method of isolating and stabilizing duplications with immobilized Mu d 1insertions ( similar to the methoodf ANDERSON and ROTH ( 1 98 1 ) fo isrolation of tandem duplications with TnlO ) should be generally useful for studyingcomplemented gene expression of homologous genes .
Construction of cya promoter mutations fused to lac : To facilitate study of the regulationof mutant cya promoters we placed the cya-lac fusion downstream of the mutations .
A lysate grown onthe Lac-strain AZ2617 ( 9 ( cyaP701-l ~ c ) 6 9 6 , w ~ as ~ u ) sed to transduce thecya promoter mutants AZ2618 , AZ2620 and AZ2621 to ampicillin resistance on MacConkey lactose agar .
Lac + transductants ( red or pink colonies ) contained theCya + mutant promotersof the recipientstrainsfused to lac .
Isogenic crp mutants were constructed by transducing the ( cya-lac ) fusion strains to streptomycin resistance with a P22 lysate made on strainJ V l l 4 ( crp404 rpsL20I ) .
& Galactosidase assay : Cultures were grown overnight i M56 minimal-medium supplemented with the indicated carbon source ( 40 mM ) .
Cultures were then diluted 1:100 into 25 ml fresh medium in 250-ml Erlenmeyer flasks and incubated at 37 '' in a New Brunswick rotary shaking water bath with vigorous shaking .
Growth was followed by meas-uring optical density ( OD ) at 650 nm with a Gilford model 250 spectrophotometer .
The cultures were sampled ( 1 .
O ml ) into iced test tubes and assayed for & galactosidase activity ( MILLER1972 ) .
Differential rate determinations were made with five samples taken at ODtiso ` sin the range of 0.2-0.6 and were the averageof two or more experiments .
Chemicals and enzymes : N ' , N ' - Methylene-bis-acrylam-ide , acrylamide , and ammonium persulfate were from BioRad Laboratories , Richmond , California .
Agarose , BAL-31 , DNA polymerase I ( Klenowfragment ) , and all restriction endonucleases were from Bethesda Research Laboratories , Gaithersburg , Maryland .
Nuclease S1 was from P-L Bio ¬ Glucose - & phosphate 62 1 58 52 1,750 Glucose 442 68 4,470 Gluconate 550 65 5,590 Glycerol 526 88 13,770 Glucosamine 0 189 6,180 Acetate 164-312-31070 ,890 ~ Cultures were grown at 37 '' inM56 minimal media with the indicated carbon source at 40 m M as described in MATERIALS AND METHODS .
IPTG was added at 1 m M to AZ105 cultures .
` Specific activitieswere determined from differential rates plots and the units are A4so/min/OD6aoX 1000 .
Values are the average of two experiments .
Isopropyl - @ , D-thiogalac-topyranoside ( IPTG ) , O-nitrophenyl - / 3,D-galactopyranoside ( ONPG ) , CAMP , low gelling temperature agarose ( type VII ) , and sodium deoxycholate were from Sigma Chemical Co. , St. Louis , Missouri .
RESULTS Carbonsource regulation of @ ( cyu-Zuc ) 696 and its correlation withcAMP levels : Table 3 shows the results of experiments in which @ - galactosidaseactivity was assayedas a measuroef cya-lac expression in strain AZ2599 grown on differentcarbon sources .
Strain AZ2599 carries animmobilized cya-lac fusion ( Mu d 1 insertion ) and a cya + gene as the result of a duplication .
The strain , therefore , is Cya + owing to complementation by the homologous S. typhimurium cya + gene .
T h e differential rate of P-galactosidase synthesis in strainAZ2599 was 3.8-fold higher in culturesthat were grown in glucose-6-phosphate minimal-medium than in cultures grown in acetate minimal-medium .
Growth in glucose , gluconate , glycerol and glucosa-mine gave intermediate levelsof cya-lac expression .
The differential rates of the homologously complemented cya-lac fusion were similar to those of a heterologously complemented fusion ( E. coli F ' cya + - data not shown ) .
In addition , the results for the immobilized fusion were similar to those previously observed for the nonimmobilized fusion ( JOVANOVICH 1985 ) .
These dataindicate thatcya expression and regulation were not affected by the immobilization procedure , and that carbon source regulation of cya-lac expression is the same whether the complementing cya + gene is derived from S. typhimurium or E. coli .
The carbon sources shown in Table 3 are arranged in the order of the intracellular level of cAMP that they elicit in E. coli when acting as the sole source of carbon in the medium.Forexample , in theorder reported by EPSTEIN,ROTHMAN-DENEaSnd HESS A22642 cYaP + 706 ( 559 ) b A22638 cyap702 147 A22634 cyap703 681 ( 534 ) AZ2640 cyap704 235 ( 88 ) Cultures weregrown at 37 '' in M56 minimal-medium with 40 nlM glucose as described in MATERIALS AND METHODS .
a Specific activities were determined from differential rateplots and the units are A4n , , / nin/ODti5X0 1000 .
Values are the average of at least two experiments .
Values in parentheses are the specific activities attributed to the P2 promoter and werecalculated by subtracting the specific activity attributedtotheupstreampromoters ( Pl , Pl ' ) in strain AZ2638 .
For example , the activity attributed to P1 , P1 ' in strain AZ2638 ( 147 ) was subtracted from the activity in strain A22642 ( 706 ) to get the activity attributed to P2 in strain A22642 ( 559 ) .
( 1975 ) , growth onglucose-6-phosphate resulted in the lowest relative intracellular cAMP concentration in E. coli whereas growth on glucosamine elicited the highest concentration .
It is clear that E. coli F ' lac expressionin S. typhimurium strain AZ105 is derepressed during-growth on carbonsources known to elicit high intracellular cAMP levels ( Table 3 ) .
Therefore , F ' lac expression in S. typhimurium parallels lac expression in E. coli during-growth on different carbon sources .
Since lac expression varies directly with intracellular cAMP concentrations in E. coli ( EPSTEIN , ROTHMAN DENESand HESSE1975 ) , we concludethat F ' lac expression in S. typhimurium can be used as a measure of relative cAMP levels .
The intracellular concentration of cAMP resulting from growth on acetate was notdetermined by EPSTEIN,ROTHMAN-DENE anSd HESSE ( 1975 ) .
We found that growth on acetate resulted in the highest level of F ' lac expression in S. typhimurium strainAZ105 ( Table 3 ) indicating that , of thecarbon sources tested , acetate results in the highest intracellular concentrationof CAMP.Since we observed that cya-lac expression and F ' lac expression were inversely related during-growthondifferent carbon sources , we concludethat cya expression is negatively regulated by CAMP .
Effects of cya P2 mutations on promoter activity : Table 4 shows the effects of c y P2 mutations ( Figure 1 ) on transcriptionally fused lac expression .
P-galac-tosidase activities were determined in a constitutively derepressed crp mutant background in order to quantitate promoter activities independent of regulation by CAMP-CRP ( the results were unaffected by addition of CAMP-data not shown ) .
T h e wild-type strain ( AZ2642 ) grown in minimal glucose medium gave a P-galactosidase specificactivity of about 700 .
The 32-bpdeletionmutation ( cyap702 ) in strainAZ2638 , which removes the -10 and -35 hexamer sequences of cya P2 , reduced totalcya expression about fivefold .
A22643 cyap + 573 ( 389 ) 154 ( 50 ) 3.7 ( 7.8 ) AZ2639 cyap702 184 104 1.8 A22635 cyap703 870 ( 686 ) 559 ( 455 ) 1.6 ( 1.5 ) A22641 cyap704 358 ( 174 ) 265 ( 161 ) 1.4 ( 1 .
1 ) Culture conditions and specific activity determinations were as indicated in Table 4 .
a Specific activities attributed to theP2 promoter ( in parentheses ) werecalculatedas indicated in Table 4 using cyap702 deletion mutant ( strain AZ2639 ) as the reference for noP2 activity .
' Repression ratios for the P2 promoter ( in parentheses ) were calculated from specific activities attributed to P2 .
Therefore , cya P2 accounts for approximately80 % of derepressed cya expression ; the remaining 20 % is the consequence of upstream promoter sequences ( presumably the PI and P1 ' promoter ( sT , ) HORNFEARNDL and ARTZ 1990 ) .
In E. coli , the P1 andPI ' promoters were identified by S 1 mapping ( AIBA1985 ) andexist in a region that is nearly identical with the analogous region in S. typhimurium ( THORNER , FANDL AaR ndTZ 1990 ) .
The 4-bp deletion mutation ( cyap703 ) in strain AZ2634 , which removes bp -3 to -6 , but leaves the cya P2 RNA polymerase-binding sequence intact , had essentially no effect on cya expression .
In contrast , the 4-bp insertion mutation ( cyap704 ) between bp -3 and -4 in strain AZ2640 resulted in a threefold decrease in cya expression even though the cya P2 RNA polym-erase binding sequence remained intact .
Subtraction of the contribution of the PIandPI ' promoters indicated that the reduction of P2 promoter activity caused by cyap704 is about sixfold ( Table 4 , values in parentheses ) .
StrainsAZ2620 ( cyap702 ) and AZ2621 ( cyap704 ) have an intactcya structural gene andcarry mutations that eliminate or greatly reduce P2 promoter activity ( Table 4 ) .
These strains grew onthepoorcarbon source citrate indicating that , although it is the major promoter , cya P2 is not required for a Cya + phenotype .
Effects of cya P2 mutations on regulation by CAMP-CRP : Table 5 shows the response of wild-type andmutant cya promoters ( Figure 1 ) to exogenous cAMP in crpc strains grown in minimal glucose me-dium .
Additionof 5 mM cAMP to thegrowth medium repressed wild-type cya expression 3.7-fold in strain AZ2643 .
When the contributionof the upstream promoters ( defined by strain AZ2639 which lacks cya P2 ) was subtracted from the totalactivity in the wild-type strain , we calculated thatP2 expression in strain A22643 was repressed 7.8-fold by the addition of 5 mM CAMP .
Repression by cAMP was eliminated in the isogenic crp mutant ( datanot shown ) .
Deletio mutation cyap703 in strain AZ2635 , which removed 4 bp of the putative CRP binding site , nearly eliminated cya repression by CAMP-CRP ; cya P2 was repressed 1.5-fold by addition of 5 mM CAMP.Similarly , repression by CAMP-CRPwas greatly reduced by the 4-bp insertion mutationcyap704 in strain AZ2641and by the P2 deletion mutationcyap702 in strain AZ2639 .
These results indicate that cya P2 is the primary target of CAMP-CRPrepression .
DISCUSSION In this paper we demonstrated genetically that the CAMP-CRP complex represses cya expression in S. typhimurium by interacting at a site overlapping the major cya promoterP t , 2 .
Mutations alteringhe CAMP-CRPbinding site were constructed and recombined into the chromosome allowing us to study the effects of the mutations under physiological conditions , in single copy and at the appropriate chromosomal locus .
The mutant analysis supports the exist-ence of multiple cya promoters consistent with the comparison of the cya DNA sequences of S. typhimu-rium and E. coli ( THORNEFRA , NDL andARTZ 1990 ) and nuclease S1-mapping experimentsin E. coli ( AIBA 1985 ) .
The nucleotide sequenceof theputative CRP binding site at cya P2 is 5 ' AG-TGTTA -- '' - TCACG-TT 3 ' 3 ' TC-ACAAT ------ AGTGC-AA 5 ' ( THORNEFARN , DL and ARTZ 1990w ) hich closely resembles the consensus CRP binding site 5 ' AA-TGTGA ------ TCACA-TT 3 ' 3 ' TT-ACACT ------ AGTGT-AA 5 ' ( EBRIGHTet al. 1987 ; BERG and VON HIPPEL 1988 ) .
Important contacts between CRP and the lac operon CRP binding site were identified by chemical protectionstudies ( REZNIKOFFand ABELSON1980 ) , X-ray crystallography ( MCKAYand STEITZ1981 ) , and binding site mutations ( DICKSON et al. 1977 ; EBRIGHTet al. 1987 ; JANSEN , GRONENBORNandCLORE1987 ) .
These studies indicated that the two TGTGA penta-mers , which occur in the symmetric left-and right-half binding sites are most important in specifying CRP-DNA interactions .
T h e cya CRPbinding site deviates from the TGTGAconsensus at two positions ( shown in boldface ) : position 4 of the left-half binding site ( T instead of G ) and position 1 of the right-half binding site ( C instead of T ) .
T h e G at position 4 is critical for specific binding while the T at position 1 is relatively unimportant ( EBRIGHTet al. 1987 ; JAN-SEN , GRONENBORNanCdLORE1987 ) .
Interestingly , JANSEN , GRONENBORaNnd CLOR ( E198a7n ) d EBRIGHTet al. ( 1987 ) foundthat , of the possible Mutations substitutions at position 4 , T had the least detrimental effect on specific CRP binding .
The A703 mutation destroyed the right-half binding site forCRPat cya P2 ( Figure 1 ) .
When this deletion was recombinedinto the chromosome we observed nearly complete loss of repression of cya expression by CAMP-CRP ( Table 5 ) .
In crp mutants , the activity of the A703 promoter was similar to that of the wild-type promoter ( Table 4 ) , indicating that A703 did not affect binding of RNA polymerase .
In contrast , the 4-bp insertion mutation ( INS704 ) which increased the spacing between the highly conserved CRPbinding site pentamersreducedboth cya P2 activity and CAMP-CRP regulation .
Decreased promoter activity in this mutant was surprising because INS704 did not change the -10 hexamer of P2 .
We speculatethat the presence of tandemGATC sequences ( d a m methylation sites ) neartheP2 -10 hexamer and transcription startpoint in this mutant may interfere with the interaction of RNA polymer-ase .
The presence of dam methylation sites adjacent to certain promoters has been shown to reduce their activity ( HOOPESand MCCLURE1987 ) .
Our results and otherstudies ( BOTSFORaDnd DREX-LER 1978 ) indicate that under conditions where glucose is available and cAMP isless required , the cell derepresses synthesis of ( and presumably accumulates ) adenylate cyclase , while inhibiting the activity of the enzyme .
When glucose in the medium is exhausted , the cell can rapidly react tothe new conditions by activating its store of adenylate cyclaseand accumulate CAMP.This allows activation of transcription of those genesrequiredforthe use of available alternative carbon sources .
When the cell must use acarbon source with a lower energy yield than glucose it conserves energy by repressing the synthesis of adenylate cyclase as well as the synthesis of many other proteins ( MALLICKand HERRLICH 1979 ) thatpresumably are less necessary under these growth-conditions .
Althoughrepressionof cya expression by CAMP-CRP is now clearly established , this regulation can not readily explain the 10-to 100-fold overproduction of cAMP by crp mutantsgrowing in minimal glucose medium ( POTTERC d , HALMERS-LARSaOnN YAMAZAKI 1974 ; WAYNE anRdOSEN1974 ; RAPHAELaInd SAIER 1976 ) .
This is apparent from the observation that a crp mutant ( strain AZ2642 ; Table 4 ) expresses cya less than twofold higher than crp + strains ( AZ2599 and AZ2643 ; Tables 3 and 5 , respectively ) during-growth in minimal glucose medium .
Therefore , overproduction of cAMP by crp mutants must be the direct or indirect consequence of a defect in CRP-mediated regulation of adenylate cyclase activity rather thancya expression , a conclusion realized by others as well ( RAPHAELaInd SAIER1976 ; BOTSFORD and DREXLE 1978 ; MAJERFELDet al. 1981 ; JOSEPH et al. 1982 ; DOBROCOSeZ t al. 1983 ) .
In addition to carbon source regulation involving CAMP-CRP , the cell may utilize other mechanisms to increase its store of adenylate cyclase under rapid growth-conditions .
The location of cya is nearthe origin of replication and therefore will be present in multicopy during rapid growth due to multiple replication forks .
In addition , ANDERSOaNnd ROTH ( 1981 ) observed thatrapidgrowthrates increase thefrequency of rm-mediated duplications .
The finding that the cya locus experiences a high frequency of duplication ( 1 % ) see is consistent MATERIALS AND METHODS ) with its location between rmA and rmC an thdis may provide another method to maximize adenylate cyclase synthesis during rapidgrowth .
Finally , the GATC sequenceproximal to thecya P2 -10 hexamer is a dam methylase recognition site and dam methylation near promoters is known to inhibit transcription except transiently during periods of hemimethylation ( HOOPESand MCCLURE1987 ) .
DNA is hemimethyl-ated immediately after DNA replication .
Therefore , during rapid growth , the GATC adjacent to the P2 promoter would be hemimethylated more often than during slow-growth .
This work was supported by U.S. Public Health Service grant GM27307 from the National Institutes of Health .
We thank LE-WANNA ARCHERfor typing the manuscript .
LITERATURECITED AIBA , H. ,1985 Transcription of the Escherichia coli adenylate cyclase gene is negatively regulated by CAMP-CAMP receptor protein .
MORI , M. TANAKTA001 , A. ROYand A. DANCHIN , 1984Thecompletenucleotidesequence of theadenylate cyclase gene of Escherichia coli .
ALPER , M. D. , and B. N. AMES ,1978 Transport of antibiotics and metabolic analogs by systems under cyclic AMP control : positive selection of Salmonella typhimurium cya and crp mutants .
ROTH , 1981Spontaneoustandemgenetic duplications in Salmonella typhimurium arise by unequal recombination between rRNA ( rrn ) cistrons .
ARTZ , S. , D. HOLZSCHU , P. BLUMand R. SHAND ,1983 Use of Ml3mp phages to study gene regulation , structure and function : cloning and recombinational analysis of genes of the Salmonella typhimurium histidine operon .
A , , and P. J. BASSFORD1 ,982 Regulationof aden-ylate cyclase synthesis in Escherichia coli : studies with cya-lac operonandprotein fusion strains .
VON HIPPEL , 1988 Selection of DNA binding sites by regulatory proteins .
The binding specificity of cyclic AMPreceptorproteinto recognition sites .
, L. BLAHAand S. ARTZ , 1986 Reversion and immobili-zation of phageMud1 cts ( Ap lac ) insertion mutations in Salmonella typhimurium .
BI.UM , P. , D. HOLZSCHUH , .
- S. KWAN , D. RIGGSand S. ARTZ , 9 enreeplacemenatnd retrieval with recombinant M13mp bacteriophages.J .
L. , and M. DREXLER , 1978 The cyclic 3 ' ,5 ' - adeno-sine monophosphate receptor protein and regulation of cyclic 3 ' ,5 ' - adenosinemonophosphate synthesis in Escherichia coli .
DAVIS , R. W. , D. BOTSTEINand J. R. ROTH , 1980 Advanced Bacterial Genetics.Cold Spring Harbor LaboratoryC , old Spring Harbor , N.Y. DICKSONR , .
JOHNSON , W. S. REZNIKOFF anWd .
M. BARNES1 , 977 Nucleotide sequence changes produced by mutations in the lac promoter of Escherichia coli .
, G. W. HALL , D.K. SHERBAD .
SILVA , J.G. HARMENand T. MELTON , 198R3 egulatoriynteractions among the cya , crp , and pts gene products in Salmonella typhi-murium .
H. , A. KOLB , H. Buc , T. H. KUNKEL , JS .
KRAKOW and J. BECKWITH1 ,987 Role of glutamic-acid-181 in DNA-sequence recognition by the catabolic gene activator protein ( CAP ) of Escherichia coli : altered DNA-sequence-recognition properties of [ Val '' '' ] CAP and [ Leu '' '' ] CAP .
EPSTEIN , W. , L. B. ROTHMAN-DENEaSnJd .
HESSE , 1975 Adenosine 3 ' : 5 ' - cyclic monophosphate as mediator of cdtabolite repression in Escherichia coli .
, 1983 Studies on transformation of Escherichia coli with plasmids .
HOOPES , B. C. , and W. R. M C C L U R E. ~ S ~ trr ~ att ~ egies in regulation of transcriptioninitiation,pp.1223-1234 in Escherichia coli and Salmonellatyphimurium : cellular and molecularbiology , edited by F. C. NEIDHARDTA.merican Society for Microbiology , Washington , D.C. JANSEN , C. , A. M. GRONENBORNandG .
M. CLORE ,1987 The binding of cyclic-AMP-receptor-protein tosynthetic DNA sites containing permutations in the consensus sequence TGTGA .
JOSEPH , E. , C. BERNSLEYN , , .
Gu ~ soand A. ULLMANN1982 Multiple regulation of the activity of adenylate cyclase in Esch-erichia coli .
JOVANOVICH , S. B. , 1985 Regulationof a cya-lac fusion by cyclic AMP in Salmonella typhimurium .
KONAMO andH.AIBA ,1985 Negative regulation of adenylate cyclase gene ( cya ) expression by cyclic AMP-cyclic-AMP-receptor-protein in Escherichia coli : studies with cya-lac protein and operon fusion plasmids .
H. , D. MILLER , E. SPITZand H. V. RICKENBERG , 1981 Regulation of the synthesis of adenylate cyclase in Esch-erichia coli by the CAMP-CAMP receptor protein compleMx.ol .
, and P. HERRLICH , 1979 Regulationof synthesis of a major outer membrane protein : cyclic AMP represses Esche-richia coli protein 111synthesis .
E. , E. F. FRITSCHand J. SAMBROOK , 1982Molecular Cloning : A Laboratory ManuaC l.old Spring Harbor Laboratory , Cold Spring Harbor , N.Y. MCKAY , D. B. , and T. STEITZ ,1981 Structureofthe catabolite gene activator protein at 2.9 A resolution suggests binding to left-handed B-DNA .
,1983 New M13 vectors forcloning.Methods Enzymol .
, and J. VIEIRA , 1982 A new pair of M13 vectors for selecting either DNA strand of double-digest restriction fragments .
, 1972 Experiments in Molecular Genetics.Cold Sprin MONOD , J. , G. COHEN-BAZIRanEd M. COHN ,1951 Sur la bio-synthese de la B-galactosidase ( lactase ) chez Escherichia coli : la specificitede I ' induction .
MORI , K. , and H. AIBA ,1985 Evidence for negative control of cya transcription by cAMP and cAMP-receptor-protein in intact Escherichia coli cells.J .
T. , P. H. BLUMand S. W. ARTZ , 1983 Effects of the hisT mutation of Salmonellatyphimurium on translation elongation rate .
POTTER , K. , G. CHALMERS-LARSaOnNd H. YAMAZAKI1 ,974 Abnormally high rate of cyclic AMP excretion from an Esche-richia coli mutant deficient incyclicAMP-receptor-protein .
S. , and M. H. SAIER1976 Effects of crp mutations on adenosine 3 ' ,5 ' - monophosphate metabolism in Salmonella typhimurium.J .
S. , and J. N. ABELSON1 ,980 The lac promoter , pp. 221-243 in The Operon , edited by J. H. MILLERand W. S. REZNIKOFCFo .
ld Spring Harbor Laboratory , Cold Spring Harbor , N.Y. ROY , A. , C. HAZIZA andA .
DANCHIN1 ,983 Regulation of aden-ylate cyclase synthesis in Escherichia coli : nucleotide sequence of the control region.EMBOJ .
K. , J. P. FANDLand S. W. ARTZ , 1990 Analysisof sequence elements important for expression and regulation of the adenylate cyclase gene ( c y ) of Salmonellatyphimurium .
M. ROSENC , yclic 3 ' : 5 ' - adenosine monophosphate in Escherichia coli during transient and catabolite-repression .
YAMAMOTO , K. R. , B. M. ALBERTS , R. BENZINGELR. , LAWHORN and G. TREIBER1 ,970 Rapid bacteriophage sedimentation in the presence of polyethylene-glycol and its application to large-scale virus purification .
, , J. VIEIRAand J. MESSING1 ,985 Improved M I 3 phage cloning vectors and host strains : nucleotide sequences of the M13mp18 and pUCI9vectors .