Transcriptional coupling between the divergent promoters of a prototypic LysR-type regulatory system, the ilvYC operon of Escherichia coli

Abstract

The twin-domain model [Liu, L. F. & Wang, J. C. (1987) Proc. Natl. Acad. Sci. USA 84, 7024–7027] suggests that closely spaced, divergent, superhelically sensitive promoters can affect the transcriptional activity of one another by transcriptionally induced negative DNA supercoiling generated in the divergent promoter region. This gene arrangement is observed for many LysR-type-regulated operons in bacteria. We have examined the effects of divergent transcription in the prototypic LysR-type system, the ilvYC operon of Escherichia coli. Double-reporter constructs with the lacZ gene under transcriptional control of the ilvC promoter and the galK gene under control of the divergent ilvY promoter were used to demonstrate that a down-promoter mutation in the ilvY promoter severely decreases in vivo transcription from the ilvC promoter. However, a down-promoter mutation in the ilvC promoter only slightly affects transcription from the ilvY promoter. In vitro transcription assays with DNA topoisomers showed that transcription from the ilvC promoter increases over the entire range of physiological superhelical densities, whereas transcription initiation from the ilvY promoter exhibits a broad optimum at a midphysiological superhelical density. Evidence that this promoter coupling is DNA supercoiling-dependent is provided by the observation that a novobiocin-induced decrease in global negative superhelicity results in an increase in ilvY promoter activity and a decrease in ilvC promoter activity predicted by the in vitro data. We suggest that this transcriptional coupling is important for coordinating basal level expression of the ilvYC operon with the nutritional and environmental conditions of cell growth.

LysR proteins comprise the most common type of positive regulatory proteins in bacteria. They are found in prokaryotic families ranging from Enterobacteriaceae to Rhizobiaceae and regulate a diverse array of target genes. LysR proteins are characterized as sequence-related, coinducer-responsive, transcriptional activators that exhibit several common regulatory properties. LysR-type proteins activate transcription of a target gene(s) and autoregulate their own synthesis. In most cases, the target gene is divergently transcribed from the adjacent structural gene encoding the LysR-type protein. Activation of the target gene requires binding of a metabolically important coinducer molecule, whereas autoregulation of the LysR protein gene is coinducer-independent. Binding of coinducer does not significantly alter the DNA-binding affinity of the protein (1). Thus, unlike other regulatory proteins, the binding of a coinducer molecule to a LysR protein activates transcription of a target gene by altering a property of the protein distinct from its DNA-binding activity.

The ilvYC operon of Escherichia coli is a prototypic LysR protein-regulated system. It encodes a structural gene (ilvC) for the second enzyme (acetohydroxy acid isomeroreductase, EC 1.1.1.86) in the common pathway for the biosynthesis of l-isoleucine and l-valine (2). The ilvY gene encodes the LysR-type regulatory protein (IlvY protein). IlvY protein binds in a highly cooperative fashion to two tandem operator sequences in the divergent-overlapping ilvYC promoter region (Fig. 1) and negatively autoregulates transcription from the ilvY promoter (3). Binding of IlvY protein to the same operator sequences, however, is not sufficient to activate transcription from the divergently transcribed ilvC promoter. Activation of ilvC transcription requires the additional binding of a coinducer molecule (either α-acetolactate or α-acetohydoxybutyrate, the substrates for acetohydroxy acid isomeroreductase) to a preformed IlvY protein–DNA complex. This directs a conformational change in the structure of the protein–DNA complex that is correlated with the relief of an IlvY protein-induced DNA bend that effectively remodels the −35 region of the ilvC promoter and increases RNA polymerase binding nearly 100-fold (3).

Figure 1.

Nucleotide sequence of the divergent promoter region of the ilvYC operon in pSAE1. Base pair positions from +6 to −60 relative to the ilvC transcriptional start site are shown (6). The transcription start sites for ilvY and ilvC are identified by arrows. The −10 and −35 hexanucleotide regions of the ilvY (bottom strand) and ilvC (top strand) promoters are identified with horizontal lines. The O₁ (truncated) and O₂ operator sites are denoted by brackets. The mutated bases in the ilvC promoter contained in pSAE2 and in the ilvY promoter contained in pSAE3 are identified by dashed lines and bold characters.

Although the basic mechanisms for IlvY protein-mediated regulation in the ilvYC operon have been elucidated, the functional importance of the evolutionarily conserved divergent gene arrangement common to operons regulated by LysR proteins has not been investigated. However, it has been proposed that the activity of a superhelically sensitive promoter can be influenced by the local negative superhelicity generated by transcription from a nearby divergent promoter (4). Therefore, we wished to investigate the possibility that the local domain of negative DNA supercoiling generated in the region between RNA polymerase molecules transcribing the divergently arranged ilvY and ilvC genes might function to transcriptionally couple the activities of these promoters. In this report, we present in vivo and in vitro evidence that the DNA supercoiling-dependent transcriptional activities of the ilvY and ilvC promoters are transcriptionally coupled and that this effect is dependent on the sum of the global and transcriptionally generated local superhelical density of the DNA template in the divergently transcribed region.

Materials and Methods

Chemicals and Reagents.

Restriction endonucleases, T4 ligase, and T4 polynucleotide kinase were purchased from Boehringer Mannheim. Shrimp alkaline phosphatase was obtained from United States Biochemicals. Radiolabeled nucleotides were purchased from DuPont/NEN. DNA oligonucleotides were synthesized by Operon Technologies. Site-directed mutagenesis was performed by using the oligonucleotide-directed in vitro mutagenesis kit (Version 2.0) from Amersham Pharmacia. DNA sequencing was performed by using the T7-Gen kit of United States Biochemicals.

Plasmids and Bacterial Strains.

Plasmid DNA isolation and all recombinant DNA manipulations were carried out by using standard procedures (5). Plasmid pDD3Y was constructed by ligating an end-filled, 220-bp BglII-EcoRI DNA restriction fragment [ilvYC sequence from −220 to +1 relative to the ilvC transcriptional start site (6)] into the BamHI (end-filled) site of pDD3 (7). Plasmid pSAE was constructed by ligation of a 1,340-bp BamHI-BalI, 3′ truncated lacZ′ gene fragment from pRS552Δ (8) into the BamHI site of the galK transcriptional fusion plasmid, pKO9 (7). pSAE1 contains a BamHI PCR-generated fragment encoding the 220-bp E. coli ilvYC promoter-regulatory region ligated into the BamHI site of pSAE. In this plasmid the ilvC promoter is transcriptionally fused to the truncated lacZ′ gene and the ilvY promoter to the galK gene. Plasmids pSAE2 and pSAE3 differ from pSAE1 by the presence of down-promoter mutations in the −10 hexanucleotide region (Fig. 1) of either the ilvC promoter (pSAE2) or the ilvY promoter (pSAE3).

Strains IH5500, IH5550, and IH5580 were constructed by homologous recombination of pSAE1, pSAE2, and pSAE3, respectively, into the chromosome of IH55 lysA, polA, Δgal. ilvY-deleted strains IH5510, IH5560, and IH5590 were constructed by bacteriophage P1 transduction of strains IH5500, IH5550, and IH5580, respectively, by using lysates of IH15 rbs∷Tn5Δ(ilvDAYC-rep)115. Single-copy plasmid integrants into the lacZ locus of the bacterial chromosome were confirmed by Southern analysis and DNA sequencing (7).

Cell Growth Conditions.

Bacterial strains were grown overnight and diluted 1:100 into 1 liter of the same M63 minimal medium containing 0.5 mM l-lysine, 0.4 mM l-leucine, 0.4 mM l-isoleucine; 1 mM l-valine, 5 mg of thiamine, 75 mg of ampicillin, 25 mg of kanamycin, and 0.4% glucose. Cell growth was monitored by measuring the optical density of the culture at 600 nM. Cells were harvested during the exponential growth phase (OD₆₀₀ = 0.2–0.7). To inhibit DNA gyrase activity, novobiocin was added to growing cultures at a final concentration of 60 μg/ml.

β-Galactosidase and Galactokinase Activities.

Cells were harvested by centrifugation, washed with 20 mM Tris⋅HCl (pH 7.0), and, for β-galactosidase assays, resuspended in 1 ml of 10 mM Tris⋅HCl (pH 8.0) containing 0.1 mM EDTA or, for galactokinase assays, in 1 ml of 20 mM Tris⋅HCl (pH 7.2) containing 8 mM MgCl₂, and 2 mM DTT. The resuspended cells were disrupted by sonication. Cell debris was removed by centrifugation, and samples of the supernatant fluid were assayed for β-galactosidase (9) or galactokinase (10) activity. Assays were performed under conditions in which the activity was linear with respect to time and cell extract concentration. β-Galactosidase-specific activity is reported as nanomoles of o-nitrophenol formed per min per mg of protein. Galactokinase-specific activity is reported as nanomoles of galactose-1-phosphate formed per min per mg of protein. Protein was measured by the method of Bradford (11).

Generation of DNA Plasmid Topoisomers.

Ten-microgram samples of pDD3Y were treated at room temperature for 4 hr with 20 units of Drosophila melanogaster topoisomerase II in a 60-μl reaction mixture containing 10 mM Tris⋅HCl (pH 7.9), 50 mM NaCl, 5 mM MgCl₂, 1 mM EDTA, 15 μg/ml BSA, 1 mM ATP, and 0–50 μM ethidium bromide. Each plasmid DNA sample was extracted three times with phenol to remove the ethidium bromide, precipitated with 2 vol of isopropanol, and resuspended in 20 μl of water. The average negative superhelical density (−σ) for each DNA topoisomer sample was determined by electrophoresis on four separate 1.4% agarose gels in 0.04 M Tris-acetate/1 mM EDTA containing 0.006, 0.02, 0.04, or 0.08 μg/ml ethidium bromide. The average linking-number difference (ΔLk) for each plasmid topoisomer sample was determined by the band-counting method of Keller (12). The linking-number difference range of each topoisomer sample was equal to or less than ±3. The average superhelical density was calculated by using the equation σ = −10.5ΔLk/N (N = total number of base pairs in plasmid). The average superhelical density of reporter plasmids obtained from log phase E. coli cells grown in glucose M63 minimal salts medium is −0.05.

In Vitro Transcription Reactions.

In vitro transcription reactions were performed according to the methods of Wek and Hatfield (7) with pDD3Y plasmid templates of defined superhelical density (3). RNA polymerase–DNA complexes were formed by preincubating 0.5 units (1.2 pmol) of RNA polymerase and 250 ng of plasmid DNA (0.1 pmol) in a 45-μl reaction mixture containing 0.04 M Tris⋅HCl (pH 8.0), 0.1 M KCl, 0.01 M MgCl₂, 1.0 mM DTT, 0.1 mM EDTA, 100 μg/ml BSA, 40 units of RNasin, 200 μM CTP, 20 μM UTP, and 10 μCi (3,000 Ci/mmol) [α-³²P]UTP. Transcription reactions were initiated with the addition of 5 μl of a 2-mM ATP/2-mM GTP solution. Reactions were terminated after 3 and 6 min by transferring 15 μl of reaction mixture into 15 μl of stop solution (95% formamide/0.025% bromphenol blue/0.025% xylene cyanol). The reaction products were separated by electrophoresis on an 8% denaturing polyacrylamide gel (7.6% acrylamide/0.4% N,N′-methylene bisacrylamide) containing 8 M urea in a 0.09-M Tris-borate/2 mM EDTA solution and quantitated by PhosphorImager analysis.

Results

In Vivo Promoter Coupling in the ilvYC Operon.

In 1987, Liu and Wang (4) proposed that, because of its inability to freely rotate around the DNA helix during transcription, negative DNA supercoiling is generated behind and positive DNA supercoiling is generated in front of a transcribing RNA polymerase. This model suggested that closely spaced, divergent, superhelically sensitive promoters can affect the transcriptional activity of one another by the transcriptionally induced negative DNA supercoiling generated in the divergent promoter region. At this time, much information in support of this model has been reported (for reviews see refs. 13–17) Thus, given the divergent spatial organization of many LysR-type regulatory systems, we wished to examine the possibility of transcriptional coupling between the divergently transcribed promoters of the ilvYC operon.

If transcription from one promoter influences transcription from the other by a transcriptional coupling mechanism, then the transcriptional activity of each promoter should be altered if transcription from its divergent promoter were inhibited. To test this prediction, down-promoter point mutations were introduced into the −10 hexanucleotide region of either the ilvY or the ilvC promoter sites contained in the dual-reporter plasmid construct pSAE (Fig. 1). The point mutations were selected such that the promoter elements of the divergent promoter were unaffected. These plasmids were integrated, in single copy, into the lacZ locus of the bacterial chromosome of ilvY⁺ and ilvY⁻ strains by homologous recombination (7). Promoter activities were monitored in growing cells by measuring the steady-state levels of the β-galactosidase (ilvC) and galactokinase (ilvY) reporter gene activities.

The results presented in Table 1 show that mutations in the ilvY promoter that decrease the transcriptional activity of this promoter approximately 13-fold result in a concomitant 11-fold decrease in the transcriptional activity of the divergent ilvC promoter (compare strains IH5500 and IH5580). Furthermore, deletion of the ilvY structural gene from these strains, which results in a 3- to 5-fold increase in ilvY gene expression (because of the relief of IlvY protein-mediated autoregulation), effects a 3- to 5-fold increase in ilvC gene expression (compare strains IH5500–IH5510 and IH5580–IH5590). Thus, increased and decreased transcription from the ilvY promoter correlates with increased and decreased transcription from the ilvC promoter. These observations are consistent with the transcriptional coupling mechanism discussed above.

Table 1.

ilvPy∷galK and ilvP_C∷lacZ expression in wild-type and mutant strains

Strain	Relevant genotype	Galactokinase-specific activity (ilvY)	β-Galactosidase-specific activity (ilvC)
IH5500	Wild type	0.65  ±  0.08	9.4  ±  1.4
IH5510	ΔilvY	3.21  ±  0.29	27.4  ±  1.4
HI5550	ilvPCdown	0.49  ±  0.08	ND
IH5560	ilvPCdown,ΔilvY	2.45  ±  0.15	ND
IH5580	ilvPYdown	0.05  ±  0.01	0.9  ±  0.1
IH5590	ilvPYdown,ΔilvY	0.14  ±  0.01	4.0  ±  0.2

Galactokinase- and β-galactosidase-specific activities are expressed as the mean ± SD of six independent experiments. ND, not detectable. 

In contrast to the large effect of ilvY promoter activity on transcription from the ilvC promoter, little effect of ilvC promoter activity on transcription from the ilvY promoter was observed. The ilvC promoter mutations that decrease ilvC promoter activity in an ilvY⁺ strain to nondetectable levels only slightly affect transcription from the ilvY promoter (compare strains IH5500 and IH5550). Although ilvY promoter activity is increased in an ilvY⁻ strain (compare strains IH5500 and IH5510), it is still only slightly affected in the ilvC down-promoter strain (compare strains IH5510 and IH5560). These observations might be explained by the fact that, in vitro, the basal level transcriptional activity of the ilvC promoter is about 40-fold weaker than the basal level activity of the ilvY promoter (3). Thus, the small change in local superhelical density contributed by this low basal-level ilvC promoter activity might not be expected to elicit a significant change in ilvY promoter activity. It also remained possible that transcription from the ilvY promoter was relatively insensitive to changes in DNA supercoiling.

In Vitro Effects of DNA Supercoiling on Transcription from the ilvC and ilvY Promoters.

To determine the intrinsic effects of DNA supercoiling on transcription from the ilvY and ilvC promoters, in vitro transcription reactions were performed with a set of DNA topoisomers of pDD3Y with superhelical densities ranging from σ = 0.00 to −0.14 (ΔLk from 0 to −48). These reactions were performed with a minimally saturating concentration of RNA polymerase under conditions in which the rate of transcription was directly proportional to the DNA template concentration. The results in Fig. 2 show that transcription from the ilvC promoter increases at least 10-fold as the superhelical density of the DNA template is increased from σ = 0.00 to −0.11. Thus, transcription from the ilvC promoter is intrinsically sensitive to the superhelical state of the DNA template and increases over the entire range of physiological superhelical densities (σ = −0.03 to −0.09; ref. 18). This superhelical-dependence pattern is consistent with the decrease in ilvC promoter activity observed in vivo in the absence of transcription from the ilvY promoter. That is, if the global plus the transcriptionally generated superhelical density in the divergent ilvYC promoter region maintains ilvC at a near-maximal expression level, then loss of transcriptionally generated local DNA supercoiling from the ilvY promoter would be expected to decrease transcription initiation from the ilvC promoter.

Figure 2.

Effects of DNA topology on the transcriptional activity of the ilvC promoter. Relative transcription rates were determined from the phosphorimage intensity of ³²P-labeled transcripts from in vitro transcription reactions by using DNA templates of defined superhelical densities as described in Materials and Methods. Error bars represent the SDs calculated from the results of at least three independent experiments.

The effect of DNA supercoiling on transcription from the ilvY promoter was determined with the same set of DNA topoisomers in the same manner described above. The results in Fig. 3 show that ilvY promoter-specific transcription increases about 3-fold over a physiological range of superhelical densities from σ = −0.03 to −0.06. Above this optimum, transcription is decreased only about 20–30% at the high physiological density of σ = −0.09. Thus, in contrast to the ilvC promoter, transcription from the ilvY promoter is intrinsically less sensitive to the effects of DNA supercoiling and peaks at a midphysiological range. Thus, as observed in vivo, little effect on ilvY expression would be expected when the negative superhelical density in the ilvY promoter region is decreased by the loss of transcription from the relatively weak ilvC promoter.

Figure 3.

Effects of DNA topology on the transcriptional activity of the ilvY promoter. Relative transcription rates were determined from the phosphorimage intensity of ³²P-labeled transcripts from in vitro transcription reactions by using DNA templates of defined superhelical densities as described in Materials and Methods. Error bars represent the SDs calculated from the results of at least three independent experiments.

The Effect of Global Superhelical Density on Promoter Coupling in the Divergently Transcribed ilvYC Operon.

Although the in vivo results presented above demonstrate intracellular promoter coupling and are consistent with the tenets of a transcriptional coupling mechanism, they do not show that this coupling is supercoiling-dependent. However, if these promoters are transcriptionally coupled, then the negative superhelical density in the divergent ilvYC promoter region should be greater than the global superhelical density of the chromosome, and a decrease in global negative superhelical density should decrease the superhelical density in the divergent promoter region and bring it closer to a normal physiological density. Therefore, based on the in vitro effects of DNA supercoiling on the transcriptional activities of these promoters (Figs. 2 and 3), it is predicted that a decrease in global negative superhelical density would increase the transcriptional activity of the ilvY promoter and decrease the transcriptional activity of the ilvC promoter. To test this prediction, in vivo DNA supercoiling was inhibited by addition of a sublethal amount of the gyrase-inhibiting drug novobiocin (19, 20) to a growing culture of strain IH5500. Consistent with the in vitro results, the data depicted in Fig. 4 show that the addition of the drug results in an increase in the differential rate of transcription from the ilvY promoter and a decrease in the differential rate of transcription from the ilvC promoter.

Figure 4.

Effects of novobiocin on the differential rates of synthesis of β-galactosidase (ilvC) and galactokinase (ilvY) in growing cells. Addition of novobiocin (final concentration, 60 μg/ml) is indicated by arrow: ●, plus novobiocin; ○, no novobiocin. (A) β-Galactosidase activity. (B) Galactokinase activity.

Discussion

In Vivo Promoter Coupling in the ilvYC Operon.

The twin-domain model of Liu and Wang (4) predicts that the local superhelical density in a divergent promoter region between transcribing RNA polymerase molecules should be increased to a level greater than the global superhelical density of the chromosome and that this incremental increase in local superhelical density should be correlated with the frequency of transcription initiation from the divergent promoters and the amount of negative DNA supercoiling generated by this transcription. Therefore, if transcription initiation from either or both of the promoters is sensitive to the superhelical state of the DNA template, promoter activity will be determined by the sum of the global plus transcriptionally generated superhelicity in the divergent promoter region. This type of a DNA supercoiling-dependent interaction between divergently transcribed promoters is referred to as transcriptional coupling.

In this report, we demonstrate that the transcriptional activity of the ilvC promoter of the ilvYC operon of E. coli is influenced by the transcriptional activity of the divergently transcribed ilvY promoter. The results presented in Table 1 show that mutations in the ilvY promoter, which decrease its in vivo transcriptional activity, effect a similar fold decrease in the in vivo transcriptional activity of the divergent ilvC promoter, and deletion of the trans-acting ilvY structural gene from these strains, which causes an increase in ilvY promoter activity (because of the relief of IlvY protein-mediated autoregulation), effects a similar fold increase in ilvC promoter activity. Thus, increased and decreased transcription from the ilvY promoter correlates with increased and decreased transcription from the ilvC promoter. This result clearly demonstrates promoter coupling between these divergently transcribed promoters. However, no evidence for promoter coupling was obtained when transcription from the ilvY promoter was examined in either an ilvY⁺ or ilvY⁻ strain with ilvC promoter mutations that decrease ilvC promoter activity (Table 1). In these cases, only slight effects on transcription from the ilvY promoter were observed.

The finding that ilvC promoter activity does not affect ilvY promoter activity is explained by two observations. First, because the basal level transcriptional activity of the ilvC promoter is about 40-fold weaker than the basal level activity of the ilvY promoter (3), the small change in transcriptionally generated superhelicity contributed by this low level ilvC promoter activity is not expected to significantly alter the local (global plus transcriptionally generated) superhelical density in the divergent promoter region. Second, the results of the experiments shown in Fig. 3 demonstrate that ilvY promoter activity is relatively insensitive to small changes effected by ilvC transcription in the mid- to high physiological superhelical density ranges encountered in the in vivo experiments reported here.

Evidence for Transcriptional Coupling Between the Divergent Promoters of the ilvYC Operon.

The observation that the activity of the ilvC promoter increases and the activity of the ilvY promoter decreases at superhelical densities above the physiological density of normally growing cells offered an opportunity to test a principal tenet of the transcriptional coupling model, that is, to determine whether the promoter coupling between the divergent promoters of the ilvYC operon is DNA supercoiling-dependent. Because the twin-domain model predicts that the local superhelical density in the divergently transcribed ilvYC promoter region must be greater than the normal global superhelical density of the cell (σ ≈ −0.05) and because global DNA supercoiling is a major contributor to this higher than physiological local superhelicity, any decrease in global superhelical density should bring the local superhelical density in the divergent promoter region closer to its normal physiological level. Thus, if the promoters were transcriptionally coupled, then such a decrease in the global superhelical density of the cell should effect a decrease in ilvC promoter activity and an increase in ilvY promoter activity. On the other hand, if the two promoters were not transcriptionally coupled, and the local superhelical density in the promoter regions were equal to the global superhelical density of the cell, then a decrease in global superhelical density would cause a decrease in the transcriptional activities of both promoters (see Figs. 2 and 3). To distinguish these possibilities, the global DNA superhelical density of growing cells was decreased by the addition of novobiocin to a growing culture and the activities of the ilvY and ilvC reporter gene products were measured. After drug addition, the differential rate of transcription from the ilvC promoter decreased and transcription from the ilvY promoter increased (Fig. 4). These results confirm DNA supercoiling-dependent transcriptional coupling between the divergently transcribed promoters of the ilvYC operon of E. coli.

A Rationale for Transcriptional Coupling Between the Divergent Promoters of the ilvYC Operon.

The ilvC gene product, α-acetohydroxy acid isomeroreductase, catalyzes the second step of the parallel pathways for the biosynthesis of l-isoleucine and l-valine and is positively regulated by the intracellular levels of its substrates (2, 3). The remaining steps of this pathway are catalyzed by gene products encoded by the ilvGMEDA operon, which is regulated by the intracellular levels of the pathway end products (2). Therefore, because the gene products of these two operons participate in the same biosynthetic pathway, it might be expected that some global mechanism (independent of operon specific controls) should exist for coordinating basal level expression of these two operons. In this regard, we have demonstrated previously that basal level expression of the ilvGMEDA operon is regulated by a negative DNA supercoiling-dependent mechanism (9, 21–23). Like the ilvC gene, basal level expression of the genes of the ilvGMEDA operon also increases 10-fold over the physiological range of superhelical densities (23). Therefore, we suggest that the independent effects of DNA supercoiling on the expression of these two operons might be responsible for coordinating their basal expression levels with one another. We further suggest that DNA supercoiling also might function to coordinate basal level expression of these biosynthetic operons with environmental and nutritional growth states that affect the energy charge (intracellular ATP/ADP ratios) of the cell. This speculation is based on observations that the in vivo negative superhelicity varies with the energy charge of the cell (24–27). High negative superhelical densities are observed in cells with high ATP/ADP ratios, and low negative superhelical densities are observed in cells with low ratios. In the stationary growth phase, where the energy charge is low, the superhelical density of reporter plasmids drops to a value of σ ≈ −0.03. As cells recover and enter into the log growth phase, the energy charge becomes higher and global negative superhelical density increases to a value of about σ = −0.05 (28). Under certain growth-stress conditions such as osmotic shock, the superhelical density can become even more negative (29). Thus, given the parallel in vivo responses of the ilvYC and ilvGMEDA operons to changes in global DNA supercoiling levels, it is possible that changes in global superhelical densities associated with changes in growth conditions might also serve to coordinate the basal level capacity for l-isoleucine and l-valine biosynthesis with the demands of the cell under different growth conditions.

A Rationale for the Evolutionarily Conserved Divergent Gene Arrangement of LysR-Type Operons.

In addition to suggesting DNA supercoiling-dependent mechanisms for the coordination of expression of the ilvYC and ilvGMEDA operons of the ilv regulon, the results described here also suggest a general rationale for the evolutionarily conserved divergent gene arrangement common to the ilvYC operon and other operons regulated by LysR-type proteins. In each case, the LysR-type regulatory protein binds to operator sites in the divergent promoter region in an inducer-independent manner to form a protein–DNA complex. Although the formation of this complex autoregulates the expression of the gene for the LysR-type regulatory protein, it does not activate transcription from the divergent target gene. This requires the additional binding of a small-molecule coinducer to the preformed LysR-type protein–DNA complex (1, 3). Therefore, for efficient induction of the target gene, the operator sites must be nearly saturated at all times. Because it is likely that the stoichiometry of the autoregulated LysR protein and its operator sites does not change with growth conditions, this demands that the synthesis of the LysR-type protein should be constant at all nutritional and environmental growth conditions that can affect the superhelical density of the bacterial chromosome. On the other hand, it is reasonable to assume that like the ilvC gene, many LysR-type system target genes encode proteins whose functions require that their levels must be adjusted over the entire range of physiological superhelical densities effected by the nutritional and environmental growth conditions of the cell. This demands that the promoter for the target gene must exhibit maximal activity at a high superhelical density. In general, promoters with high superhelical density optima are weak promoters because of the high energy of activation required for open complex formation. We suggest that this is accommodated by supplementing the global superhelical density environment of these promoters with locally generated DNA supercoiling from the divergent, constant-activity, LysR gene promoter. In other words, if the target gene were separated from the LysR regulatory gene, its promoter might not be strong enough to support its basal-level metabolic function in the cell and be able to respond to a broad range of physiological superhelical densities.

Finally, it should be emphasized that the global-regulatory mechanisms suggested here are postulated to affect operon expression in a manner independent of the specific controls for these operons that respond to the in vivo levels of small-molecule effectors. Instead, these global control mechanisms serve to adjust the basal level or capacity for operon expression with changes in DNA supercoiling that reflect the growth conditions of the cell.