Unusual location and function of the operator in the Escherichia coli galactose operon

Coming full circle: On the origin and evolution of the looping model for enhancer-promoter communication

In mammalian organisms, enhancers can regulate transcription from great genomic distances. How enhancers affect distal gene expression has been a major question in the field of gene regulation. One model to explain how enhancers communicate with their target promoters, the chromatin looping model, posits that enhancers and promoters come in close spatial proximity to mediate communication. Chromatin looping has been broadly accepted as a means for enhancer-promoter communication, driven by accumulating in vitro and in vivo evidence. The genome is now known to be folded into a complex 3D arrangement, created and maintained in part by the interplay of the Cohesin complex and the DNA-binding protein CTCF. In the last few years, however, doubt over the relationship between looping and transcriptional activation has emerged, driven by studies finding that only a modest number of genes are perturbed with acute degradation of looping machinery components. In parallel, newer models describing distal enhancer action have also come to prominence. In this article, we explore the emergence and development of the looping model as a means for enhancer-promoter communication and review the contrasting evidence between historical gene-specific and current global data for the role of chromatin looping in transcriptional regulation. We also discuss evidence for alternative models to chromatin looping and their support in the literature. We suggest that, while there is abundant evidence for chromatin looping as a major mechanism for enhancer function, enhancer-promoter communication is likely mediated by more than one mechanism in an enhancer- and context-dependent manner.

Differential role of base pairs on gal promoters strength

Sequence alignments of promoters in prokaryotes postulated that the frequency of occurrence of a base pair at a given position of promoter elements reflects its contribution to intrinsic promoter strength. We directly assessed the contribution of the four base pairs in each position in the intrinsic promoter strength by keeping the context constant in Escherichia coli cAMP-CRP (cAMP receptor protein) regulated gal promoters by in vitro transcription assays. First, we show that base pair frequency within known consensus elements correlates well with promoter strength. Second, we observe some substitutions upstream of the ex-10 TG motif that are important for promoter function. Although the galP1 and P2 promoters overlap, only three positions where substitutions inactivated both promoters were found. We propose that RNA polymerase binds to the -12T base pair as part of double-stranded DNA while opening base pairs from -11A to +3 to form the single-stranded transcription bubble DNA during isomerization. The cAMP-CRP complex rescued some deleterious substitutions in the promoter region. The base pair roles and their flexibilities reported here for E. coli gal promoters may help construction of synthetic promoters in gene circuitry experiments in which overlapping promoters with differential controls may be warranted.

A common role of CRP in transcription activation: CRP acts transiently to stimulate events leading to open complex formation at a diverse set of promoters

We have shown previously that the cyclic AMP receptor protein (CRP) is not required after the formation of the open complex at the lac promoter (Tagami and Aiba, 1995, Nucleic Acids Res., 19, 6705-6712). In this paper, we investigate the role of CRP in transcription activation at the malT and gal promoters. At the malT promoter, RNA polymerase (RNAP) forms a nonproductive RNAP-promoter binary complex in the absence of CRP and a productive CRP-RNAP-promoter ternary complex in the presence of CRP. CRP can be removed from the malT ternary complex by a moderate concentration of heparin. The resulting binary complex is functionally identical to the ternary complex. At the gal promoter, RNAP predominantly forms a binary complex at the P2 promoter in the absence of CRP and a ternary complex at the P1 promoter in the presence of CRP. A very high concentration of heparin is able to dissociate CRP from the galP1 ternary complex without changing the properties of the complex. These data indicate that CRP is not required for the maintenance of the ternary complex and plays no role in the subsequent steps, irrespective of the promoter. We conclude that the common role of CRP in the activation of transcription is to stimulate events leading to the formation of a productive open complex at a diverse set of CRP-dependent promoters. We suggest that the interaction between CRP and RNAP is needed only transiently for the activation of transcription.

Identification of the nucleotide sequence recognized by the cAMP-CRP dependent CytR repressor protein in the deoP2 promoter in E. coli

In E. coli repression of transcription initiation by the CytR protein relies on CytR-DNA interactions as well as on interactions between CytR and the cAMP-CRP activator complex. To identify the nucleotide sequence recognized by CytR, mutants of the deoP2 promoter with a reduced regulatory response to CytR have been isolated. Five single bp mutation derivatives of deoP2 with a 2-5-fold decrease in CytR regulation have been characterized. In vitro, the only effect of the mutations was a decrease in the binding affinity of CytR, and a clear correlation was observed between the reduction in CytR regulation in vivo and the reduction in CytR binding in vitro. The mutations all reside in a sequence element that contains an imperfect direct as well as an imperfect inverted repeat. As the active form of CytR, most likely, is an oligomer with two-fold rotational symmetry, CytR probably interacts with the inverted repeat. Degenerate versions of the inverted repeat are present in all CytR binding sites characterized so far, however, the distance between the half-sites varies.

Regulation of open complex formation at the Escherichia coli galactose operon promoters. Simultaneous interaction of RNA polymerase, gal repressor and CAP/cAMP

The regulation of open complex formation at the Escherichia coli galactose operon promoters by galactose repressor and catabolite activator protein/cyclic AMP (CAP/cAMP) was investigated in DNA-binding and kinetic experiments performed in vitro. We found that gal repressor and CAP/cAMP bind to the gal regulatory region independently, resulting in simultaneous occupancy of the two gal operators and the CAP/cAMP binding site. Both CAP/cAMP and gal repressor altered the partitioning of RNA polymerase between the two overlapping gal promoters. Open complexes formed in the absence of added regulatory proteins were partitioned between gal P1 and P2 with occupancies of 25% and 75%, respectively. CAP/cAMP caused open complexes to be formed nearly exclusively at P1 (98% occupancy). gal repressor caused a co-ordinated, but incomplete, switch in promoter partitioning from P1 to P2 in both the absence and presence of CAP/cAMP. We measured the kinetic constants governing open complex formation and decay at the gal promoters in the absence and presence of gal repressor and CAP/cAMP. CAP/cAMP had the largest effect on the kinetics of open complex formation, resulting in a 30-fold increase in the apparent binding constant. We conclude that the regulation of open complex formation at the gal promoters does not result from competition between gal repressor, CAP/cAMP and RNA polymerase for binding at the gal operon regulatory region, but instead results from the interactions of the three proteins during the formation of a nucleoprotein complex on the gal DNA fragment. Finally, we present a kinetic model for the regulation of open complex formation at the gal operon.

Footprinting analysis of mammalian RNA polymerase II along its transcript: an alternative view of transcription elongation

Ternary complexes of RNA polymerase II, bearing the nascent RNA transcript, are intermediates in the synthesis of all eukaryotic mRNAs and are implicated as regulatory targets of factors that control RNA chain elongation and termination. Information as to the structure of such complexes is essential in understanding the catalytic and regulatory properties of the RNA polymerase. We have prepared complexes of purified RNA polymerase II halted at defined positions along a DNA template and used RNase footprinting to map interactions of the polymerase with the nascent RNA. Unexpectedly, the transcript is sensitive to cleavage by RNases A and T1 at positions as close as 3 nucleotides from the 3'-terminal growing point. Ternary complexes in which the transcript has been cleaved to give a short fragment can retain that fragment and remain active and able to continue elongation. Since DNA.RNA hybrid structures are completely resistant to cleavage under our reaction conditions, the results suggest that any DNA.RNA hybrid intermediate can extend for no more than 3 base pairs, in dramatic contrast to recent models for transcription elongation. At lower RNase concentrations, the transcript is protected from cleavage out to about 24 nucleotides from the 3' terminus. We interpret this partial protection as due to the presence of an RNA binding site on the polymerase that binds the nascent transcript during elongation, a model proposed earlier by several workers in preference to the hybrid model. The properties of this RNA binding site are likely to play a central role in the process of transcription elongation and termination and in their regulation.

Autoregulation of Escherichia coli purR requires two control sites downstream of the promoter

The expression of Escherichia coli purR, which encodes the pur regulon repressor protein, is autoregulated. Autoregulation at the level of transcription requires two operator sites, designated purRo1 and purRo2 (O1 and O2). Operator O1 is in the region of DNA between the transcription start site and the site for translation initiation, and O2 is in the protein-coding region. The repressor protein binds noncooperatively to O1 with a sixfold-higher affinity than to O2, and saturation of O1 by the repressor precedes saturation of O2. Both O1 and O2 function in the two- to threefold autoregulation in vivo, as determined by measurement of beta-galactosidase and mRNA from purR-lacZ translational fusions. Of all the genes thus far known to be regulated by the Pur repressor, only purR employs a two-operator mechanism.

Analysis of relative reversion frequencies for IS2 insertion mutations in the regulatory region of the galOPETK operon of Escherichia coli

Two previously characterized mutations in the galOPETK operon of Escherichia coli, galOP-3 and galOPE-490, contain IS2 insertions only 1 bp apart in the gal regulatory region; yet only the former yields Gal+ phenotypic revertants at a detectable frequency. We have shown that the galOPE-490 allele comprises two mutations--an IS2(I) insertion at bp+(2-6) (relative to the gal mRNA start site) plus a C/G to A/T transversion at bp + 59. The latter creates an ochre stop codon and lies within the internal site of the bipartite gal operator; it acts as an operator mutation in an in vivo repressor titration assay. Analysis of a newly isolated allele (galOP-490*) which retains the IS2 of galOPE-490 but is galE+ reveals a reversion frequency approximately 30-fold higher than that of galOP-3. Reversion of galOPE-490 is at least 10,000-fold lower and has not been detectable even under conditions conducive to enhanced double mutations in other systems.

RNA chain initiation by Escherichia coli RNA polymerase. Structural transitions of the enzyme in early ternary complexes

We have studied the properties and structures of a series of Escherichia coli RNA polymerase ternary complexes formed during the initial steps of RNA chain initiation and elongation. Five different templates were used that contained the bacteriophage T7 A1 promoter or the E. coli Tac or the lac UV5 promoter, as well as variant templates with alterations in the initial transcribed regions. The majority of ternary complexes bearing short transcripts (from two to nine nucleotides) are highly unstable and cannot be easily studied. This includes transcripts from the phage T7 A1 promoter, for which the stability of complexes bearing transcripts as short as four nucleotides has previously been postulated. However, with one Tac promoter template, RNA polymerase forms ternary complexes with transcripts as short as five nucleotides that are stable enough for biochemical study. We describe several approaches to identifying and isolating such stable complexes and show that stringent criteria are needed in carrying out such experiments if the results are to be meaningful. Deoxyribonuclease I (DNase I) footprinting has been used to probe the general structure of the stable ternary complexes formed as the polymerase begins transcription and moves away from the start site. The enzyme undergoes a sequence of structural changes during initiation and transition to an elongating complex. Complexes with five to eight nucleotide transcripts, designated initial transcribing complexes (ITC), have identical footprints; they all retain the sigma factor and have a slightly extended DNase I footprint (-57 to +24) as compared to the open promoter complex (-57 to +20). ITC complexes all show a region of marked DNase I hypersensitivity in the -25 region that may reflect bending or distortion of the DNA template. Complexes with 10 or 11 nucleotide transcripts, designated initial elongating complexes (IEC), have lost the sigma factor and have a slightly reduced and shifted DNase I footprint (-32 to +30). However, these IEC have not yet achieved the much smaller footprint (approximately 30 bp) reported as characteristic of elongating ternary complexes bearing longer RNA chains. During the initial phase of transcription, the RNA polymerase does not move monotonically along the DNA template as RNA chains are extended, but instead, the upstream and downstream contacts remain more or less fixed as the nascent transcript is elongated up to about eight nucleotides in length. Only after incorporation of 10 nucleotides is there significant movement of the enzyme away from the promoter region and a commitment to elongation.

Effect of ethylation of operator-phosphates on Gal repressor binding. DNA contortion by repressor

Gal repressor inhibits transcription by binding to two operators (OE and OI) in the gal operon. By ethylating DNA, we have identified 23 phosphate groups (11 on OE and 12 in OI) in the DNA backbone of gal operators that when ethylated interfere with repressor binding. By inference, either (1) such a phosphate is contacted or closely approached by Gal repressor, or (2) the structure of DNA generated by ethylation of such a phosphate, although not a site of direct contact, is not compatible with repressor binding. Within an operator, these phosphates are arranged with a perfect symmetry aligned with the operator dyad symmetry, indicating that each half-symmetry is contacted by a subunit of repressor dimer. Unlike in many other similar DNA-protein systems, the same phosphates in the gal operator are distributed around a B-form of DNA helix cylinder covering greater than 180 degrees. Models have been proposed to describe the disposition of the Gal repressor-operator complex, which would explain the layout of the participating phosphate groups around the surface of the DNA helix. We have discussed two ways by which Gal repressor can induce structural changes in DNA.

Cyclic-AMP-dependent switch in initiation of transcription from the two promoters of the Escherichia coli gal operon: identification and assay of 5'-triphosphate ends of mRNA by GTP:RNA guanyltransferase

We have studied the initiation of transcription of the gal operon in Escherichia coli (i) by analyzing the 5'-triphosphate ends and (ii) by measuring the level of promoter-proximal gal mRNA made in vivo. The 5' termini were identified and quantified by capping with GTP:mRNA guanyltransferase, and the mRNA levels were determined by hybridization of pulse-labeled [32P]RNA with a specific DNA probe. Our results conclusively demonstrate the in vivo activities of two promoters, P1 and P2, with separate initiation sites (S1 and S2) as suggested before from in vitro and in vivo experiments (S. Adhya and W. Miller, Nature [London] 279:492-494, 1979; R. E. Musso, R. DiLauro, S. Adhya, and B. de Crombrugghe, Cell 12:847-854, 1977). We have also studied the effect of cyclic AMP (cAMP) on in vivo gal transcription and found that whereas total gal transcription remains largely unchanged, the relative proportions of the S1 and S2 mRNAs are influenced by the level of cAMP in the cell. In strains devoid of cAMP (cya), transcription initiates equally at S1 and S2; in cAMP-proficient cells (cya+), the S1 initiation increases twofold with a concomitant decrease in S2 initiation. Addition of a saturating amount of exogenous cAMP to cya mutant cells results in a relatively larger switch from S2 to S1. Our results clearly show that while cAMP is an inhibitor of S2, it is not an absolute requirement for transcription initiation at S1, but only acts to increase low-level transcription from the P1 promoter. Using these approaches, we have also studied gal promoter mutants (P211, P18, and P35) which show altered behavior in transcription initiations and in response to cAMP. On the basis of these results, we have discussed models by which transcription initiates at the two overlapping gal promoters (P1 and P2) and discussed how cAMP level modulates the switch between them.

Modulation of transcription by DNA supercoiling: a deletion analysis of the Escherichia coli gyrA and gyrB promoters

Expression of the genes determining the subunits of Escherichia coli DNA gyrase (gyrA and gyrB) is known to be induced by relaxation of the template DNA. In this paper we report a deletion analysis of the gyrA and gyrB promoter regions. We find that a DNA sequence 20 base pairs long that includes the -10 consensus region, the transcription start point, and the first few transcribed bases is responsible for the property of induction by DNA relaxation. We propose a model for relaxation-stimulated transcription in which promoter clearance is the rate-limiting step.

RNA polymerase molecules initiating transcription at tandem promoters can collide and cause premature transcription termination

Using purified E. coli RNA polymerase we have studied the transcription in vitro of a series of DNA fragments carrying two tandemly arranged promoters, where the corresponding transcription start points were separated by 263, 138, 83 and 78 base pairs. In the case where the transcription start points are 83 base pairs apart, there is an interaction between RNA polymerase molecules transcribing from the two promoters. This interaction results in premature termination of the upstream transcript at a precise site. We propose that this is the result of RNA polymerase transcribing from the upstream promoter bumping into polymerase at the downstream promoter. The interaction between the two polymerase molecules is crucially dependent on the distance between the two promoters.

A stressed intermediate in the formation of stably initiated RNA chains at the Escherichia coli lac UV5 promoter

We report experiments designed to elucidate the mechanism by which RNA polymerase advances from the open complex to synthesis of a stably bound RNA chain during transcription initiation. Techniques used include deoxyribonuclease I footprinting, methylation protection, and exonuclease III digestion through upstream domains, each applied to the open, abortive and productive transcription complexes of Escherichia coli RNA polymerase with the lac promoter. The results show a slight loss of upstream open complex contacts during abortive transcription of a 6-mer and 8-mer, but a large loss of these contacts upon escape from abortive cycling into productive transcription at the 11-mer. We propose a model for early initiation in which competition between open complex polymerase-DNA contacts on one hand and initiated complex polymerase-DNA-RNA interactions on the other produces a "stressed intermediate" during formation of a short RNA-DNA duplex. The strain energy is relieved either by ejecting the short RNA, resulting in aborted initiation, or by eliminating the sigma subunit and breaking the open complex contacts, thereby escaping abortive cycling into productive transcription. Further evidence for this model is based on the observation that destabilization of interactions specific for either open complex or initiated complex has the predicted effect on the amount of abortive cycling. The model predicts a complicated relationship between overall promoter strength and DNA sequence changes that alter polymerase-DNA interactions.

RNA polymerase makes important contacts upstream from base pair -49 at the Escherichia coli galactose operon P1 promoter

A G:C to T:A transversion at bp position -19 in the gal operon promoter region relieves the dependence of galP1 promoter activity on the cAMP-CRP complex. Deletion analysis shows that expression from the promoter is decreased on replacement of the sequence between 49 and 54 bp upstream from the P1 start point. Moreover, protection experiments show that RNA polymerase interacts with this region in open complexes at P1. We propose that this contact is necessary for optimal P1 activity; point mutations in the gal promoter region can alter DNA flexibility and hence the strength of this contact; CRP factor activates P1 transcription by favouring formation of this contact; and the gal repressor blocks P1 activity by binding to this zone.

Nucleotide sequence homologies in control regions of prokaryotic genomes

Functional recognition sites for several regulatory factors, including RNA polymerase, cyclic adenosine monophosphate receptor protein and ribosomes, do not always have strong consensus nucleotide sequence homology, yet they are capable of biological activity. Using the computer, other nucleotide sequences can be found that have equal or significantly greater consensus homology, but whose biological function has not been characterized. This analysis shows that no arbitrary 'cutoff score' can successfully distinguish active recognition sites from uncharacterized homologies, due to the great natural diversity in the strength and conservation of functional sites. It also predicts that the strong 'cryptic' homologies presented here are of two types: some might already have a biological function which has so far not been detected, whereas certain single-point mutations might be able to confer activity upon the others by correcting a key structural defect.

Nucleotide sequence of the penicillinase repressor gene penI of Bacillus licheniformis and regulation of penP and penI by the repressor

Bacillus licheniformis penicillinase genes, penP and penI, are coded on a 4.2-kilobase EcoRI fragment of pTTE21 (T. Imanaka, T. Tanaka, H. Tsunekawa, and S. Aiba, J. Bacteriol. 147:776-186, 1981). The EcoRI fragment was subcloned in a low-copy-number plasmid pTB522 in Bacillus subtilis. B. subtilis carrying the recombinant plasmid pPTB60 (Tcr penP+ penI+) was chemically mutagenized. Of about 150,000 colonies, two penI(Ts) mutant plasmids, pPTB60D13 and pPTB60E24, were screened by the plate assay at 30 and 48 degrees C for penicillinase. By constructing recombinant plasmids between wild-type and mutant plasmids, the mutation points were shown to be located in a 1.7-kilobase EcoRI-PstI fragment. The EcoRI-PstI fragments of the wild-type plasmid and two mutant plasmids were sequenced. A large open reading frame, composed of 384 bases and 128 amino acid residues (molecular weight, 14,983), was found. Since the mutation points were located at different positions in the protein coding region (Ala to Val for pPTB60D13 and Pro to Leu for pPTB60E24), the coding region was concluded to be the penI gene. A Shine-Dalgarno sequence was found 7 bases upstream from the translation start site (ATG). A probable promoter sequence which is very similar to the consensus sequence was also found upstream of the penP promoter, but in the opposite direction. A consensus twofold symmetric sequence (AAAGTATTA CATATGTAAGNTTT) which might have been used as a repressor binding region was found downstream and in the midst of the penP promoter and also downstream of the penI promoter. The regulation of penP and penI by the repressor is discussed.

Tn10 tet operator mutations affecting Tet repressor recognition

The effect of single base pair alterations of the Tn10 encoded tet operator on recognition of Tet repressor was studied in vivo using a repressor titration system and in vitro by dissociation rate determinations of the respective complexes. Both methods reveal that the two operators, O1 and O2, which are in a tandem arrangement in the wild type, are recognized with a two-fold different affinity when separated. Studies on synthetic operator sequences indicate that the Tet repressor binds with higher affinity to the non-palindromic O2 wildtype than to the respective palindromic sequences. The in vivo repressor titration system links the expression of lacZ to the affinity of tet operator to Tet repressor. It was used to isolate tet operator mutations with reduced affinity to the repressor. The in vivo and in vitro obtained results with these mutants agree quantitatively and indicate, that the GC base pairs at positions 2, 6, and 8 are involved in interaction with the Tet repressor. Their importance for recognition decreases in that order. Transitions at position 7 of the tet operator show smaller effects on recognition than transversions.

Overlapping promoters and their control in Escherichia coli: the gal case

Two overlapping promoters compete for RNA polymerase in the region that controls the expression of the galactose operon in Escherichia coli. Kinetics of open complex formation at P1 and P2 can be followed through the rate of formation of two specific abortive transcripts. The corresponding forward kinetic constants appear to be identical over a wide range of enzyme concentrations and temperatures, indicating that the two processes are strongly coupled. We propose a scheme accounting for our observations. In a first step, the competition between the two sites is a simple kinetic process, involving the "on" rate constants. In a second step, a slow reequilibration occurs, implicating the "off" rate constants and the conversion of one open complex to the other through a set of closed complexes. The first step is clearly affected when the complex between cyclic AMP and its receptor is bound at the activator site. An estimate of the various rate constants describing open complex formation at P1 and P2 is provided, as well as a qualitative description of the effect of the activator complex on these two pathways.

Nucleotide sequence of the CytR regulatory gene of E. coli K-12

We have determined the nucleotide sequence of the cytR gene, which codes for the Cyt repressor (CytR). The coding region consists of 1023 or 1029 bp. The subunits of CytR are thus predicted to consist of 341 or 343 residues. It is shown that the N-terminal segment of the polypeptide is structurally similar to the DNA-binding region of known DNA-binding proteins. In addition, there exists an exceptionally high amino acid sequence homology between CytR and the Gal repressor, indicating a common origin of evolution.

AsnC: an autogenously regulated activator of asparagine synthetase A transcription in Escherichia coli

The regulation of the asparagine synthetase A gene of Escherichia coli was studied in vitro with a coupled transcription-translation system. It was shown that the 17-kilodalton gene, which is transcribed divergently from the adjacent asnA gene, codes for an activator of asnA transcription. The synthesis of the 17-kilodalton protein, which we now call AsnC, is autogenously regulated. The stimulating effect of AsnC on asnA transcription is abolished by asparagine, while the autoregulation of asnC is not affected by asparagine. The N-terminal part of the asnC protein, inferred from the DNA sequence, is homologous to the DNA-binding domain of regulatory proteins like catabolite gene activator, cro, and cI. This homology and direct repeats found in the region of the two asn promoters suggest that the asnC protein regulates transcription by binding to DNA. The asn promoters were defined by mapping of the mRNA start sites of in vitro-generated transcripts.

Interaction of RNA polymerase with lacUV5 promoter DNA during mRNA initiation and elongation. Footprinting, methylation, and rifampicin-sensitivity changes accompanying transcription initiation

We have used enzymatic and chemical probes to follow the movement of Escherichia coli RNA polymerase along lacUV5 promoter DNA during transcription initiation. The RNA polymerase does not escape from the promoter but remains tightly bound during the synthesis of the initial bases of the transcript. This initial phase of RNA synthesis involves the reiterative synthesis and release of RNA chains up to ten bases long via the RNA polymerase cycling reaction and the enzyme remains sensitive to rifampicin inhibition. When longer chains are made, promoter-specific binding is disrupted and the enzyme forms a rifampicin-resistant elongation complex with downstream DNA sequences. This elongation complex covers less than half as much DNA and lacks the DNase I-hypersensitive sites and the base-specific contacts that characterize promoter-bound RNA polymerase. These results lead us to suggest that lacUV5 mRNA synthesis is primed by a promoter-bound enzyme complex that synthesizes the initial nine or ten bases in the mRNA chain. Subsequently, when a chain of ten bases, or slightly longer, is made, contacts with promoter DNA are irreversibly disrupted, sigma subunit is lost, and a "true" elongation complex is formed.

Demonstration of two operator elements in gal: in vitro repressor binding studies

Genetic and DNA base sequence analyses of cis-dominant mutations that derepress the gal operon of Escherichia coli suggested the existence of two operator loci needed for gal repression. One (OE) is located immediately upstream to the two overlapping gal promoters and the other (OI) is inside the first structural gene. We have investigated the ability of wild-type and mutant OE and OI DNA sequences to bind to gal repressor. The repressor has been purified from cells containing a multicopy plasmid in which the repressor gene is brought under the control of phage lambda PL promoter. The DNA-repressor interactions are detected by the change in electrophoretic mobility of labeled DNA that accompanies its complex formation with repressor protein. The purified repressor shows concentration-dependent binding to both O+E and O+I but not to OEc and OIc sequences. These results authenticate the proposed operator role of the two homologous gal DNA control elements and thereby establish that the negative control of the gal operon requires repressor binding at both OE and OI, which are separated by greater than 90 base pairs.

Kinetics of RNA polymerase-promoter complex formation: effects of nonspecific DNA-protein interactions

The rates of formation of RNA polymerase-promoter open complexes at the galactose P2 and lactose UV5 promoters of E. coli were studied using polyacrylamide gels to separate the heparin-resistant complexes from unbound DNA. Both the apparent rate and extent of reaction at these promoters are inhibited at excess RNA polymerase. This inhibition, which can be relieved by the addition of non-promoter DNA, is interpreted to be the result of occlusion of the promoter site by nonspecifically bound polymerase. Additionally, biphasic kinetics are observed at both gal P2 and lac UV5, but not at the PR promoter of phage lambda. This behavior disappears when the concentration of RNA polymerase in the binding reaction is less than that of the promoter fragment. It is proposed that at excess enzyme nonspecifically bound polymerase molecules sliding along the DNA may "bump" closed complexes from the promoter site thereby reducing the rate of open complex formation. Kinetics mechanisms quantifying both the occlusion and bumping phenomena are presented.

DNA sequence of the carA gene and the control region of carAB: tandem promoters, respectively controlled by arginine and the pyrimidines, regulate the synthesis of carbamoyl-phosphate synthetase in Escherichia coli K-12

The carAB operon of Escherichia coli K-12, which encodes the two subunits of carbamoyl-phosphate synthetase (glutamine hydrolyzing) [carbon-dioxide: L-glutamine amido-ligase (ADP-forming, carbamate-phosphorylating); EC 6.3.5.5], is cumulatively repressed by arginine and the pyrimidines. We describe the structure of the control region of carAB and the sequence of the carA gene. Nuclease S1 mapping experiments show that two adjacent tandem promoters within the carAB control region serve as initiation sites. The upstream promoter P1 is controlled by pyrimidines; the downstream promoter P2 is regulated by arginine. Attenuation control does not appear to be involved in the expression of carAB. A possible mechanism by which control at these promoters concurs to produce a cumulative pattern of repression is discussed. The translational start of carA is atypical; it consists of a UUG or AUU codon.

Sequence of the Saccharomyces GAL region and its transcription in vivo

In Saccharomyces, the enzymes used to convert galactose to glucose are specified by three coordinately expressed, tightly linked genes, GAL7, GAL10, and GAL1. These genes are induced by galactose and are controlled by the positive regulator gene gal4 and the negative regulator gene gal80. GAL81 mutations, which are known to alter the gal4 protein, produce a constitutive phenotype. We have cloned fragments of Saccharomyces carlsbergensis DNA that span 26.3 kilobases surrounding the three clustered GAL genes. About 5 kilobases of the sequence was determined, which includes the entire GAL1 gene, the two intercistronic regions, and portions of the coding sequences of GAL10 and GAL7. Some amino acid homology between the GAL1 gene product, galactokinase, and the Escherichia coli galactokinase was detected. By using various Saccharomyces DNA fragments, the accumulation of GAL1 and GAL10 RNA in yeast cells after induction with galactose was studied. Our results, using wild-type, gal4-, gal80-, and GAL81-1- yeast cells, support the hypothesis that control is exerted at the transcriptional level.

On the different binding affinities of CRP at the lac, gal and malT promoter regions

We have determined the stoichiometry of CRP binding to various DNA fragments carrying the lac, malT or gal promoters in the presence of cAMP, using a gel electrophoresis method. In each case, one dimer of CRP binds to the functional CRP site upstream of the transcription start. At the lac promoter, a second CRP dimer can bind to the operator region. Direct binding analysis and competition experiments performed at 200 microM cAMP allow us to measure the affinity of CRP for these different sites and to correlate them with variations in the consensus sequences, already proposed. The order is lac greater than malT greater than gal greater than lac operator greater than lac L8 much greater than non specific sites. No strong coupling exists between the two lac sites when on the same fragment. Conversely, we have studied, at constant CRP concentrations, the cAMP levels required to obtain half maximal binding to a particular DNA site : the required cAMP level increases inversely as the affinity for CRP. These variations may account for the differential activation of various cAMP sensitive operons in vivo. Anomalies in the migrations of the 1:1 complexes between CRP and DNA have been analysed and related to the size and to the position of the CRP site in the fragment. The electrophoretic mobility of the complexes depends not only on the size of the fragment but on the position of the CRP site : the mobility is lower when CRP binds near the center of the fragment. This effect is due to a clear change in the persistence length of the DNA induced by CRP binding. We suggest that, upon binding, the protein introduces a local bend (or a kink) in the DNA structure.

Physical mapping of the exuT and uxaC operators by use of exu plasmids and generation of deletion mutants in vitro

Operons uxaCA and exuT of the hexuronate system are very closely linked on the Escherichia coli genetic map. Using plasmid vectors constructed by Casadaban et al. (J. Bacteriol. 143:971-980, 1980), we formed exuT-lacZ and uxaA-lacZ fusions in vitro. The phenotypic properties of the new plasmids allowed us to confirm that the exuT and uxaCA operons are divergently transcribed. An analysis of these fusion plasmids and derivatives in the presence of multiple copies of the exuR regulatory gene demonstrated that the two operons possess separate control regions. The precise location of the operator site relative to endonuclease restriction sites was determined. In addition, deletions of different lengths were generated on exu plasmids by restriction enzymes and were recombined into the chromosome. The expression of the exu regulon genes in the resulting deletion mutants is in agreement with the postulated location of the exuT and uxaCA operators in the fusion plasmids.

Interactions of RNA polymerase and the cyclic AMP receptor protein on DNA of the E. coli galactose operon

We have examined the interaction site on gal DNA for the cyclic AMP receptor protein and RNA polymerase when both are present together to form a stable initiation complex at the P1 gal promoter. Substitution of the bases to the left of -60 by unrelated DNA sequences does not change the cyclic AMP concentration dependency for in vitro transcription at P1 and inhibition of P2. Although the presence of some DNA to the left of -60 appears to be needed for efficient in vitro transcription at P1, the gal sequence to the left of -60 does not provide any specific interactions for transcription initiation at P1. Similarly, efficient in vitro transcription from P2 also requires non-specific DNA sequences to the left of -60. We have also examined which bases were protected by RNA polymerase and CRP together from the action of DNAase and dimethylsulfate. Some of the interactions that take place when cAMP-CRP alone interacts with gal DNA appear to be preserved in the cAMP-CRP-RNA polymerase-gal DNA complex, suggesting that CRP occupies the same site in the DNA when it is alone or together with RNA polymerase. Our results suggest that the formation of an open complex at different promoters can result from different interaction patterns between RNA polymerase and promoter DNA.

Molecular basis for modulated regulation of gene expression in the arginine regulon of Escherichia coli K-12

We compare the nucleotide sequences of the regulatory regions of five genes or groups of genes of the arginine regulon of Escherichia coli K-12: argF, argI, argR, the bipolar argECBH operon and the carAB operon. All these regions harbour one or two copies of a conserved 18 bp sequence which appears to constitute the basic arginine operator sequence (ARG box). We discuss the influence of ARG box copy number, degree of dyad symmetry, base composition, and position relative to the cognate promoter site on the derepression-repression ratios of the genes of the regulon. A novel hypothesis, based on structural considerations, is also put forward to account for the absence ot attenuation control.

Cyclic AMP-dependent constitutive expression of gal operon: use of repressor titration to isolate operator mutations

When the gal operator region is present in a multicopy plasmid it binds to all ("titrates") the gal repressor and "induces" the chromosomal gal operon. To make operator mutations (Oa) with reduced affinity toward the repressor, plasmid DNA was irradiated with UV light and mutant derivatives were isolated that were unable to release the chromosomal gal genes from repression. Then with such an Oa plasmid operator revertants were isolated that had reacquired the ability to release repression. Both sets of mutations have been localized by DNA sequence analysis. When the Oa mutations were transferred from the plasmid to the chromosome by recombination these mutant operators were found to make gal expression constitutive (independent of repressor) but still dependent on cAMP, whereas the previously reported gal operator mutants (Oc) are constitutive both in the presence and in the absence of cAMP. The titration method of isolating mutants enables the isolation of strains with operator mutations that also affect normal promoter activity, and it provides an easy way to isolate revertants of operator mutations.

In vivo transcription of the E. coli uvrB gene: both promoters are inducible by UV

The transcriptional activity of the tandem promoters of the Escherichia coli uvrB gene was measured in vivo. Both promoters are shown to be inducible by UV irradiation. P1, the most proximal promoter, is responsible for the main part of transcription both in uninduced and induced cells. Plasmids have been constructed carrying small deletions in the lexA binding site that overlaps with P2, the distal promoter. These deletions result in constitutive transcription from P1. This indicates that the DNA region which contains P2 functions mainly as a target site for regulation of P1 transcription in vivo.

A control element within a structural gene: the gal operon of Escherichia coli

The gal operon of Escherichia coli is transcribed from two overlapping promoters, PG1 and PG2. Cyclic AMP and its receptor protein (CRP) modulate the two promoters in opposite directions by binding to a single cat locus. Both the promoters are negatively regulated by a single repressor, the product of the galR gene. An operator site, defined by several mutations, has previously been located upstream from the cat locus. We have isolated and characterized a new set of cis-dominant constitutive mutations of the gal operon and determined their locations by DNA sequencing. From these studies, we propose the existence of a second functional gal operator element at an extraordinary site--within galE, the first structural gene. Both the operators, OE (exterior) and OI (interior), are involved in the repression of PG1 and PG2. This would be the first example of the presence of a functional operator element within a structural protein-coding region.

Two catabolite activator protein molecules bind to the galactose promoter region of Escherichia coli in the presence of RNA polymerase

The catabolite activator protein (CAP) of Escherichia coli, complexed with cAMP, is required for efficient initiation of transcription from the galactose P1 promoter (start site at +1) but not from the overlapping P2 promoter (start site at -5) [Musso, R. E., DiLauro, R., Adhya, S. & deCrombrugghe, B. (1977) Cell 12, 847-854]. We investigated the interactions between CAP/cAMP and the gal promoter region in the presence of RNA polymerase. DNase I protection experiments of gal promoter restriction fragments revealed that CAP/cAMP protects the DNA from digestion between positions -50 and -25 and that RNA polymerase protects it from -35 to +10; however, gal DNA in the presence of both CAP/cAMP and RNA polymerase is protected from DNase I digestion between positions -68 and +15. Results of exonuclease III protection experiments show that RNA polymerase alone protects the gal DNA from -30 to +15; when both CAP/cAMP and RNA polymerase are present in the reaction, protection is afforded from -65 to +20. We directly quantified the amount of cAMP and CAP bound to gal promoter DNA in the presence of RNA polymerase by selectively pelleting the ternary complexes (CAP/cAMP-RNA polymerase-gal promoter DNA) in a Beckman Airfuge. We found two CAP molecules specifically bound to the gal promoter, although only one cAMP molecule was found in the complex at low cAMP concentrations (but sufficient to support P1 transcription). Thus, both the DNA protection experiments and the centrifugation results indicate that RNA polymerase induces the binding of a second CAP molecule to the gal promoter in forming stable initiation complexes. It appears that the second CAP molecule is needed to stimulate initiation from the P1 promoter; this may be involved in regulating the relative rates at which transcription begins from the two gal start sites.

Recognition of base sequences by regulatory proteins in procaryotes and eucaryotes

A model is described whereby (i) regulatory proteins recognize one face of the DNA double helix on non-adjacent DNA regions brought close together in space through folding around nucleosomes, (ii) regulatory sequences may occur inside gene structures, and (iii) the recognition of regulatory sequences might be modulated by short (4-6 base-pairs) insertions or deletions in introns or upstream from the transcription start site. This model might apply not only to eucaryotes but also to procaryotic organisms whose DNA is organized through interactions with histone-like proteins. Consequences of the model regarding the binding of regulatory proteins in procaryotes are suggested.

Rate-limiting steps in RNA chain initiation

Promoter-specific lags in the approach to the steady-state rate of abortive initiation were observed when Escherichia coli RNA polymerase was added to initiate the reaction. The lag times were related to the time required for free enzyme and free promoter to combine and isomerize into a functionally active complex. The lag times measured for several bacteriophage and bacterial promoters differed widely (10 sec to several minutes) and in most cases corresponded to the rate-limiting step in the initiation process. The unique advantage in using the abortive initiation reaction to measure the lags was that the binding and isomerization steps in a simple two-state model could be quantitated separately. The separation of the contributions of both steps was effected by deriving an equation to describe the rate of formation of the active binary complex. Results from experiments based on the theory showed a linear relationship between the observed lag times and the reciprocal enzyme concentration. The slope and intercept of the equation yielded quantitative estimates of the binding and isomerization steps in initiation. The analysis was applied to the bacteriophage T7 A2 and D promoters to show the bases for the differences in in vitro initiation frequency that have been observed for these promoters.

Modulation of the two promoters of the galactose operon of Escherichia coli

The gal operon of Escherichia coli is controlled by two independent promotors--one is activated and the other inhibited by cyclic AMP and cyclic AMP receptor protein. The two promotors are modulated, however, by the same operator locus and receptor protein.