Biosynthesis and Regulation of the Branched-Chain Amino Acids
KIRSTY A. SALMON,1 CHIN-RANG YANG,2 AND G. WESLEY HATFIELD1*
[SECTION EDITOR: GEORGES COHEN]
Posted February 28, 2006
Department of Microbiology and Molecular Genetics, College of Medicine, University of California, Irvine, Irvine, CA 92697,1 and Computational Systems Biology and Bioinformatics Laboratory, Simmons Comprehensive Cancer Center, University of Texas Southwestern Medical Center at Dallas, 5323 Harry Hines Blvd., Y4.206, Dallas, Texas 75390-72802
*Corresponding author. Phone: (949) 824-5344, Fax: (949) 824-8598, E-mail:
This e-mail address is being protected from spambots. You need JavaScript enabled to view it
† This chapter is dedicated to H. Edwin Umbarger (1921–1999), my mentor (G.W.H.) and author of previous editions of this chapter. In addition to defining the biosynthetic pathways for the branched-chain amino acids in Escherichia coli, he reported in 1956 that isoleucine inhibits the first enzyme for its biosynthesis from threonine (167). This discovery led to the definition of allosteric enzymes and the realization of the general biological principle that most metabolic pathway end products feedback inhibit the first enzyme specific for their biosynthesis (168). It would be hubristic for me and my coauthors to attempt to improve on his discussions of these fundamental discoveries or his elucidations of the biochemistry and genetics of these pathways that are so well described in previous editions of this chapter and elsewhere (166). Instead, we will only briefly review these topics as necessary to focus on more recent studies concerning the systems biology of branched-chain amino acid biosynthesis, that is, the pathway-specific and global metabolic and genetic regulatory networks that enable the cell to adjust branched-chain amino acid synthesis rates to changing nutritional and environmental conditions.
We begin with an overview of the enzymatic steps and metabolic regulatory mechanisms of the pathways and descriptions of the genetic regulatory mechanisms of the individual operons of the isoleucine-leucine-valine (ilv) regulon. These introductory sections are followed by more-detailed discussions of recent evidence that global control mechanisms that coordinate the expression of the operons of this regulon with one another and the growth conditions of the cell are mediated by changes in DNA supercoiling that occur in response to changes in cellular energy charge levels that, in turn, are modulated by nutrient and environmental signals.
The pathways for the biosynthesis of the branched-chain amino acids are shown in Fig. 1. The first step of isoleucine biosynthesis, the pyridoxal 5'-phosphate-dependent dehydration deamination of threonine to α-ketobutyrate (2-oxobutanoate) and ammonia, is catalyzed by threonine deaminase [threonine dehydratase (biosynthetic); l-threonine hydro-lyase (deaminating); EC 4.3.1.19]. The remaining four steps of the isoleucine pathway are catalyzed by the same enzymes that catalyze the corresponding steps of the valine pathway. The first reaction of these parallel pathways, catalyzed by one of three acetohydroxy acid synthase (acetohydroxybutanoate synthase, EC 4.1.3.18) isozymes (AHAS I, II, and III), involves the decarboxylation of a central metabolite, pyruvate, to generate an acetal group that, depending on each isozyme’s second substrate preference, is transferred to the α-carbonyl carbon of either a second pyruvate for five-carbon valine biosynthesis or α-ketobutyrate for six-carbon isoleucine biosynthesis. The resulting acetohydroxy acids (either α-aceto-α-hydroxybutyrate, for isoleucine biosynthesis, or α-acetolactate, for valine biosynthesis) are branched-chain acids, but the branch is at the α-carbon rather than at the β-carbon of valine and isoleucine. This situation is corrected by an intramolecular rearrangement in which the methyl group of the α-acetolactate valine intermediate or the ethyl group of the α-aceto-α-hydroxybutyrate isoleucine intermediate is transferred from an α-carbon to a β-carbon. The acetohydroxy acid isomeroreductase (EC 1.1.1.86) that catalyzes this rearrangement also catalyzes a NADPH-dependent reduction to yield the α,β-dihydroxy acid intermediates α,β-dihydroxyisovalerate (valine biosynthesis) and α,β-dihydroxy-β-methylvalerate (isoleucine biosynthesis), respectively. Removal of a water molecule from these intermediates by dihydroxyacid dehydratase (EC 4.2.1.9) yields the α-keto acids that undergo transamination with glutamate by the branched-chain amino acid aminotransferase, transaminase B (EC 2.6.1.42), to produce valine or isoleucine. Valine also can undergo transamination with alanine by another aminotransferase, transaminase C (EC 2.6.1.66).
The first step in the lengthening of the five-carbon α-ketoisovalerate branch-point precursor to the six-carbon α-keto acid precursor of leucine is catalyzed by α-isopropylmalate synthase (EC 4.1.3.12). This reaction involves the condensation of the acetyl group of acetyl coenyzme A with α-ketoisovalerate to yield α-isopropylmalate (3-carboxy-3-hydroxyisocaproate). The second step in leucine biosynthesis is catalyzed by α-isopropylmalate isomerase (EC 4.2.1.33). This reaction involves a trans elimination of water from α-isopropylmalate to yield the intermediate dimethylcitraconate, which diffuses freely from the enzyme. To complete the reaction, it is necessary for dimethylcitraconate to rebind to the enzyme in an orientation rotated 180° from the orientation that it had during its formation (43). The addition of a water molecule to this reoriented dimethylcitraconate yields β-isopropylmalate (2-d-threo-hydroxy-3-carboxyisocaproate). The penultimate step of the leucine pathway is the oxidative decarboxylation of β-isopropylmalate to yield α-ketoisocaproate (2-keto-4-methylpentanoate). This reaction is catalyzed by a NAD-dependent β-isopropylmalate dehydrogenase (EC 1.1.1.85). The keto acid precursor of leucine, α-ketoisocaproate, is finally transaminated to leucine by transaminase B (EC 2.6.1.42). The regulatory mechanisms of enzymes for isoleucine, valine, and leucine biosynthesis, and their roles in channeling carbon through these branched-chain amino acid pathways, are discussed below.
Early studies of threonine deaminase by Umbarger and Changeux (26, 27, 28, 29, 30, 31, 32, 33, 64, 169, 171, 172) provided the first experimental support for the idea that active and inactive conformations of regulatory enzymes could be stabilized by ligand binding to allosteric sites (sites distinct from active sites). These studies were influential in the development of the two-state, concerted, Monod-Wyman-Changeux (MWC) model for allosteric regulation first described in 1965 (117). This work showed that threonine deaminase was cooperatively inhibited by isoleucine, the end product of the pathway, and that this inhibition was reversed by valine, the product of the parallel biosynthetic pathway. With isoleucine end product inhibition of threonine deaminase as the example, this early work of Umbarger taught us that carbon flow through most biosynthetic pathways is regulated by feedback inhibition of the first enzyme specific for the pathway by the end product of that pathway (167).
Despite these early studies, the three-dimensional structure of this historical enzyme was only recently elucidated, in 1998, by Eisenstein and his colleagues (70, 71). These workers have published a cohesive explanation of the allosteric properties of the E. coli threonine deaminase based on its structure and detailed thermodynamic and kinetic studies (69, 71). It is of interest to note that, with only minor modifications to account for promiscuous ligand binding, the original two-state concerted MWC model remains adequate for the accurate mathematical modeling of this allosteric regulatory enzyme (81, 84, 119).
As mentioned above, there are three AHAS isoenzymes that function to decarboxylate either pyruvate (valine biosynthesis) or α-ketobutyrate (isoleucine biosynthesis) (14, 20, 50, 79, 122) and to carry out a specific carboligation reaction to form the acetohydroxy acids shown in Fig. 1. Each of the AHAS isoenzymes is a flavoprotein with a requirement for FAD (25, 41, 179) and is composed of two large subunits and two small subunits (166). For example, AHAS I (encoded by the ilvBN genes) is composed of two large catalytic subunits (IlvB) and two small regulatory subunits (IlvN). Full activity requires that both subunits be present (179). The nucleotide and amino acid sequences encoding each of the AHAS isoenzymes of Salmonella enterica serovar Typhimurium are nearly identical to those found in E. coli (41).
Since the parallel pathways for isoleucine and valine biosynthesis are catalyzed by a single set of enzymes, and because the AHAS-catalyzed reaction is the first step specific for valine biosynthesis but the second step of isoleucine biosynthesis, valine inhibition of a single enzyme for this enzymatic step might compromise the cell for isoleucine or result in the accumulation of toxic intermediates (5, 93, 105, 133, 166). This is a common regulatory problem encountered with branched pathways that is often solved by the presence of multiple isozymes that are differentially regulated by multiple end-products. While the three AHAS isoenzymes are capable of performing the same reactions, their substrate specificities differ significantly (5, 58). AHAS I (ilvBN encoded) has substrate preference for the condensation of two pyruvate molecules required for valine biosynthesis (41, 44) to produce the acetohydroxy acid α-acetolactate and is end product inhibited by valine (57, 166). AHAS III (ilvIH encoded) is also end product inhibited by valine (and, to a lesser extent, by isoleucine and leucine [53, 76]), but it has equal preferences for pyruvate and α-ketobutyrate (58). The third isozyme, AHAS II (ilvGM encoded), has a substrate preference for the condensation of pyruvate and α-ketobutyrate to form α-hydroxybutyrate, the precursor leading to isoleucine biosynthesis. This enzyme is not inhibited by any of the branched-chain amino acids (14, 122).
More definitive studies of the three isozymes was done by Chipman and Barak and their colleagues (5, 77), in which analytical procedures were developed for determining α-acetolactate (the valine precursor) and α-aceto-α-hydroxybutyrate (the isoleucine precursor) levels. They defined a specificity ratio, R, which describes the substrate preferences of each of the AHAS isozymes for these substrates as:
R = (Vα-aceto-α-hydroxybutyrate /Vα-acetolactate)/([α-ketobutyrate]/[pyruvate])
For E. coli, the R value for AHAS I is 1; for AHAS II, R is 185; and for AHAS III, R is 60. These quantitative estimates of the relative contribution of each AHAS isozyme for the biosynthesis of isoleucine versus valine and leucine suggest that AHAS I enables cells to survive in the presence of poor carbon sources that lead to low endogenous pyruvate concentrations. AHAS II and III are suited to producing all three of the branched-chain amino acid precursors only during growth on glucose.
The AHAS isozyme reactions are catalyzed by a Ping Pong Bi Bi enzyme mechanism (54, 77, 164, 192). In this mechanism, the substrates bind in an ordered fashion (Fig. 2). That is, a pyruvate molecule (S1) must bind to the free enzyme (En) and transfer an acetaldehyde to the enzyme-bound thiamine pyrophosphate cofactor and release the first product, CO2 (P1), before the second substrate, pyruvate or α-ketobutyrate (S2), can bind to the enzyme intermediate (EnX). Next, the enzyme-bound active acetaldehyde group must be transferred to the second substrate, pyruvate or α-ketobutyrate, to form the second product, α-acetolactate or α-aceto-α-hydoxybutyrate (P2) to release the free enzyme (En). This is a Bi Bi mechanism because the enzyme binds two substrates and releases two products. It is a Ping Pong mechanism because the enzyme shuttles between a free (En) and a substrate-modified (EnX) intermediate state. Since the first substrate, pyruvate, is the same for all three AHAS isozymes, carbon is channeled through the branched-chain amino acids according to the relative concentrations of each AHAS isozyme and their affinities for the second substrate (89). Thus, the AHAS isozymes control the partitioning of carbon flow into the isoleucine and the valine-leucine biosynthetic pathways.
Over a half century ago, Tatum (162) reported that, in the absence of isoleucine, valine inhibits the growth of the E. coli strain K-12. This sensitivity to exogenous valine is due to the feedback inhibition pattern of the AHAS isozymes described above. The genomes of wild-type E. coli and serovar Typhimurium strains contain genes for the three heterodimeric AHAS isozymes IlvBN (AHAS I), IlvGM (AHAS II), and IlvIH (AHAS III). However, because of a frameshift in the middle of the ilvG gene, only the AHAS I and AHAS III activities are expressed in the E. coli K-12 strain (106, 107, 159). (In serovar Typhimurium LT2 only, AHAS I and AHAS II activities are expressed due to a mutation in the ilvI gene of AHAS III [23, 140].) Since in E. coli K-12 both of these isozymes are inhibited by valine (AHAS I and III), it was rationalized that, in the presence of high levels of valine, growth was inhibited because valine inhibition of the remaining AHAS isozyme I and III activities would starve the cell for isoleucine (reviewed in reference 52). However, later studies demonstrated that the intracellular isoleucine level is not suppressed by valine but actually elevated (80, 192). This is because isoleucine biosynthesis is sustained by the AHAS III isozyme that remains 15 to 20% active even at saturating valine concentrations (89, 93) and by the activation of threonine deaminase by valine that shuttles more substrate into the isoleucine pathway (192). Indeed, a mathematical simulation of branched-chain amino acid biosynthesis described later in this chapter shows that, in the presence of extracellular valine, the intracellular isoleucine level increases nearly ninefold (Fig. 3A) (192). At the same time, the pathway precursor of isoleucine, α-ketobutyrate (αKB), increases about fourfold (Fig. 3B). This buildup of α-ketobutyrate is caused by valine activation of threonine deaminase (Fig. 3C), which increases its production, and valine inhibition of the AHAS I and AHAS III isozymes, which reduces its consumption. It is now known that this αKB accumulation is toxic to cells because of its ability to inhibit the glucose phosphotransferase system PTS (47, 48). Thus, valine inhibition of E. coli K-12 growth is not a consequence of isoleucine starvation.
α-Isopropylmalate synthase is the least intensively studied regulatory enzyme of the branched-chain amino acid pathways. It catalyzes the first step in the lengthening of the five-carbon α-keto acid for valine (α-ketoisovalerate) to the six-carbon α-keto acid (α-ketocaproate) for leucine (Fig. 1). This reaction involves the condensation of an acetyl group of acetyl coenzyme A with α-ketoisovalerate to yield α-isopropylmalate. α-Isopropylmalate synthase has been purified from serovar Typhimurium and observed to be a tetramer of approximately 50-kDa subunits (101, 108). Leucine is a noncompetitive inhibitor with respect to both substrates (acetyl-CoA and α-ketoisovalerate) (156) and causes the tetramer to dissociate into dimers and monomers (108). A feedback-resistant α-isopropylmalate synthase mutant was described that remains in the active tetrameric form. This enzyme has a higher substrate specificity for α-ketoisovalerate, and leucine is now a weak competitive inhibitor (156). Kohlhaw and colleagues concluded that end product inhibition of α-isopropylmalate synthase by leucine is facilitated by the relatively loose association of the active tetramer subunits (156).
In addition to the regulation of enzyme activities to optimize the transformation of nutrients into cellular building blocks such as amino acids, gene expression patterns also must be stringently regulated to optimize cellular energetics. For bacteria, their nutritional and environmental growth conditions can change rapidly and drastically. An enteric organism like E. coli that is growing under stable, nutrient-rich conditions can suddenly find itself relocated to a physically hostile environment that is essentially devoid of nutrients. To survive such an assault, this organism possesses a wide variety of metabolic and genetic regulatory networks that enable it to successfully transition from growth in one environment to another. The genetic regulatory systems and networks that coordinate these growth condition transitions involve at least three hierarchical levels: (i) local control of individual operons/genes, (ii) regional control of multiple operons, and (iii) global control and coordination of cellular gene expression levels.
The most well-defined and best-understood gene regulatory mechanisms are found at the level of individual operons. While many types of operon-specific control mechanisms have been described, each operon is regulated by signals that are closely related to its biological role. For example, the expression of operons encoding genes for biosynthetic pathways are commonly repressed by pathway-specific end products, while the expression of operons encoding genes for catabolic pathways often are activated by pathway-specific substrates. Another common feature of operon-specific control mechanisms is that they are regulated by DNA-binding proteins that are present in small numbers in the cell and bind in a highly site-specific manner to only one or a few genome target sites.
At the next intermediate hierarchical level are regulons, modulons, and stimulons. Regulons are defined by Neidhardt and Savageau in chapter 84 (see the second edition table of contents) as a group of functionally related operons controlled by a common regulator. Modulons are a group of regulons also controlled by a common regulator. The operons within a regulon or modulon usually participate in a common function, such as nitrogen or carbon utilization, and share a common regulator(s), usually a repressor or activator protein that recognizes a DNA target sequence common to all members. The DNA-binding proteins that regulate regulons and modulons are more abundant than operon-specific regulatory proteins in the cell, and they bind to multiple, more degenerate DNA target sites on the chromosome.
Stimulons are defined as an ensemble of genes responding to a defined stimulus without regard to their regulatory organization or mechanism of induction (155). For example, Neidhardt and Savageau (120) point out that while heat shock of growing cells results in the activation of operons of the RpoH regulon, many genes other than those under the control of sigma-32 are induced and that phosphate starvation affects the expression levels of many genes other than those of the Pho regulon. Likewise, a growth transition from aerobiosis to anaerobiosis results in changes in the expression levels of many genes other than those of the ArcA- and FNR-regulated regulons and modulons (111, 146, 147). In fact, recent DNA microarray studies reveal that basal level changes in the expression of nearly one-third of the E. coli genes expressed during growth in minimal medium occur during aerobic versus anaerobic conditions (146).
These levels of regulation target specific operons or networks of operons, adjusting their relative expression levels to optimize their activities to the current conditions of the cell. They do not address the highest level of regulation, the global coordination of basal level expression of all of the genes in the genome to optimize cell growth and survival under a broad range of rapidly changing conditions. This level of gene regulation does not override operon-, regulon-, or stimulon-specific controls but rather tunes global basal level gene expression patterns to the demands of the cell under the prevailing nutritional and environmental growth conditions. For example, a lower basal level of amino acid biosynthesis is needed during stationary-phase than during log-phase growth. Thus the basal expression levels of genes encoding enzymes required for amino acid biosynthesis are lowered in stationary phase, but must maintain the ability to rapidly increase during transition to log phase. Throughout this transition, all operon-, regulon-, and stimulon-specific controls on these genes must remain operative, fine-tuning their expression to specific circumstances.
The following sections discuss the operon-specific regulatory mechanisms of the operons of the ilv regulon followed by a consideration and brief review of global regulatory proteins such as integration host factor (IHF), Lrp, and CAP (CRP) that affect the expression of these operons.
The ilv regulon contains 15 structural genes organized into five operons: ilvGMEDA, ilvBN, ilvIH, ilvYC, and leuABCD (Fig. 4). A variety of operon-specific genetic regulatory mechanisms are involved in regulating these operons. Three operons, ilvGMEDA, ilvBN, and leuABCD, are controlled by transcriptional attenuation mechanisms (reviewed in reference 166). The two remaining operons, ilvIH and ilvYC, are each regulated by operon-specific mechanisms that are uniquely suited to the biosynthetic roles of their gene products.
Attenuation of the ilvGMEDA, ilvBN, and leuABCD Operons.
Attenuation is the regulation of transcription termination at a site preceding the structural gene(s) of an operon (Fig. 5). Attenuation is recognized as a common mechanism for the regulation of gene expression in bacteria (4, 75, 88, 115, 188), and more general forms of attenuation have been described in eukaryotic cells and their viruses (136, 157, 195). In addition to the ilvGMEDA, ilvBN, and leu operons of the branched-chain amino acid biosynthetic regulon, attenuation has been demonstrated to regulate the expression of bacterial operons required for the biosynthesis of the amino acids tryptophan (4, 194, 196, 197), threonine (49, 96, 113, 114, 130), histidine (6, 141, 142, 174, 187, 188, 194), and phenylalanine (18, 40, 63, 72, 73, 74, 78, 126, 127, 158, 174, 191).
The leader-attenuator regions of these operons share certain structural similarities and a common attenuation mechanism. They all contain a relatively long leader region, the DNA sequence between the site of transcription initiation and the first structural gene of the operon (Fig. 5). Each leader region encodes a short polypeptide-coding sequence followed by a transcription termination site, the attenuator, near the 3' end of the leader sequence. The leader polypeptide-coding sequence contains multiple codons for the amino acid(s) synthesized by the gene products of the operon (Fig. 5 and 6). These are the regulatory codons. The role of this short polypeptide-coding sequence is to monitor the cellular supply of aminoacylated tRNA(s) for the regulatory codons. The leader polypeptide itself serves no other apparent functional role in the cell. If the operon-specific amino acids, and hence the aminoacylated tRNA(s) for the regulatory codons, are in ample supply, then the leader polypeptide can be synthesized without interruption and transcription termination at the attenuator will be maximized. Thus, the structural genes of the operon will be expressed at a low, basal, level. If, on the other hand, the cells are starving for the amino acid(s) produced by the gene products of the operon then the aminoacylated tRNA(s) for the regulatory codons will be in short supply and transcription will proceed through the attenuator into the structural genes. Consequently, the catalysts for the biosynthesis of the limiting amino acid(s) will be produced and its supply will be replenished (reviewed in references 75, 88, 115, 195, 196, and 197).
The question becomes, how does impeded translation of the leader polypeptide influence transcription termination at a downstream site? This linkage of translation and transcription functions is brought about by the alternative secondary structures that these leader RNAs can assume. The alternative, mutually exclusive, secondary structures of the ilvGMEDA, ilvBN, and leuABCD leader RNAs are shown in Fig. 6. The 3:4 stem-loop of each structure defines a Rho-independent transcription termination signal (a G + C-rich stem-loop followed by a tract of uridines (Fig. 6A, C, and E). This is the attenuator structure. The 1:2 stem-loop structures result in the formation of a transcriptional pause signal (86). The formation of these structures in the nascent leader RNA causes the transcribing RNA polymerase to pause at the end of stem 2. Stem 2 of stem-loop 1:2 and stem 3 of stem-loop 3:4 can form yet a third alternative structure, the antiterminator, stem-loop 2:3 (Fig. 6B, D, and F). The formation of the antiterminator preempts the formation of the stem-loop 3:4 attenuator structure and effects deattenuation, transcription through the attenuator into the structural genes of the operon.
The formation of the alternative structures (stem-loops 1:2 and 3:4 or stem-loop 2:3) is directed by the temporal positioning of a translating ribosome in the leader polypeptide-coding region in relation to the transcribing RNA polymerase (Fig. 7). The synchronization of ribosome and RNA polymerase movement through the leader-attenuator region is ensured by the presence of the transcriptional pause site ( Fig. 6). The RNA polymerase, paused at the end of stem-loop 1:2, is released to resume transcription when a translating ribosome enters the stem 1 coding region and disrupts the stem-loop 1:2 pause structure. Once the translating ribosome releases the paused RNA polymerase, transcription and translation are synchronized. If the cellular levels of aminoacylated tRNAs are high, then translation through the leader polypeptide-coding region proceeds unimpeded, and translation and transcription remain synchronized until the ribosome reaches the stop codon. The presence of a ribosome at this site inhibits base pairings between stems 1 and 2 as well as 2 and 3 (Fig. 7A). The basal level of transcriptional read-through at the attenuator is determined by the time it takes the ribosome to release the leader RNA template relative to the transcriptional speed of the RNA polymerase. If the ribosome were to dissociate from the leader RNA as soon as it reached the stop codon, before the synthesis of stem 3, then stem-loop 1:2 always would be free to form followed by the synthesis, and the unchallenged formation, of the stem-loop 3:4 attenuator structure. This situation would result in a maximal level of transcription termination (superattenuation) at the attenuator. However, since the average ribosome release time is thought to be about 0.7 s (21) and the average transcription rate is about 63 nucleotides per s (21), the released RNA polymerase will have synthesized stem 3 and be in stem 4 at the time the ribosome releases. In this case, the formation of the antiterminator structure, stem-loop 2:3 will compete with the formation of the alternative stem-loop 1:2 and 3:4 structures when the ribosome releases to set an average basal level of transcriptional read-through at the attenuator (Fig. 7B). Thus, the basal level of attenuation is determined by the location of the transcribing RNA polymerase when the ribosome dissociates from the leader RNA template and by the relative probabilities for the formation of the competing RNA structures. For example, Roesser and Yanofsky (143) have shown that mutations in prfB, encoding Release Factor 2 (RF2; UGA and UAA specific [150]), which delay ribosome release from the RNA template, increase transcription termination at the attenuator of the trp operon approximately twofold. These prfB mutations exert no effect on basal level expression in strains in which the naturally occurring trp leader polypeptide stop codon, UGA, is replaced with UAG. However, transcription termination at the attenuator is increased in these strains when they contain mutations in prfA, which encodes Release Factor 1 (RF1; UAG and UAA specific [150]). These experiments emphasize the importance of the timing of ribosome release from the leader RNA for setting a basal level of transcription through the attenuator.
Unlike the trp operon, the ilvGMEDA and ilvBN operons are end product regulated solely by attenuation; no repressor proteins are present for these operons. On the other hand, evidence that the leuABCD operon is regulated by a repressor, the gene product of the leuO gene, has been offered by Wu and his colleagues (36, 37, 38, 39, 61, 62, 189, 190).
Activation and Repression of the Genes of the ilvIH Operon.
The ilvIH operon of E. coli (Fig. 4A), encoding AHAS III, is directly activated by Lrp (131). In the presence of leucine, decreased Lrp binding in the ilvIH promoter-regulatory region results in a 5- to 10-fold reduction in transcription from the ilvIH promoter (51). The precise mechanism for Lrp-mediated activation of the ilvIH operon remains to be elucidated. However, Wang and Calvo (176) have demonstrated that Lrp binds to six sites within a 200-bp sequence of DNA upstream of the ilvIH promoter (Fig. 4A) and that the binding of Lrp to sites 1 and 2 (upstream region; −240 to −190) is highly cooperative, as is binding to sites 3, 4, 5, and 6 (downstream sites −150 to −40). More recently, Jafri et al. (94) demonstrated that DNA at or upstream of position −160 is absolutely required for ilvIH promoter expression and that Lrp bound at the downstream sites (−150 to −40) is necessary but not sufficient for promoter activation. The addition of DNA (between 2 and 10 bp) between positions −160 and −161 substantially reduced ilvIH promoter expression, suggesting that perhaps additional transcription factors or alterations of DNA topology in the promoter-regulatory region are required. It also is possible that cooperation between Lrp bound at the upstream or downstream sites is needed for promoter activity.
Activation and Repression of the Genes of the ilvYC Operon.
Each of the four operons of the ilv regulon described to this point is regulated at the genetic level by mechanisms that respond to the intracellular levels of their pathway-specific end products, either a free branched-chain amino acid (ilvIH) or the cognate branched-chain aminoacyl-tRNAs (ilvGMEDA, ilvBN, and leu). In contrast, the ilvC gene of the ilvYC operon is regulated by its substrates, not by pathway end products. To understand this unusual situation, consider the biochemical role of the ilvC gene product, α-acetohydroxy acid isomeroreductase (Fig. 1). This enzyme catalyzes the rate-limiting step in the parallel isoleucine and valine biosynthetic pathways and hence must be responsive to changing levels of substrates produced by the three differentially regulated AHAS isozymes. By tuning its gene expression level to the concentration of its substrates, it can keep carbon efficiently flowing through these parallel pathways.
The ilvY and ilvC genes of the ilvYC operon are oppositely oriented and divergently transcribed from overlapping promoter sites (Fig. 4B and 8). Operon-specific regulation is mediated by the IlvY protein, the product of the ilvY gene (13, 15, 177, 181, 182). IlvY is a member of the LysR family of regulatory proteins. It is a helix-turn-helix homodimer that cooperatively binds to adjacent operator sites O1 and O2 in the divergent promoter region (Fig. 8) in a manner independent of substrate-inducer concentration (182). The O1 site covers the region between positions –10 and +1 of the ilvY promoter. By modulating the expression of its own gene, IlvY effectively regulates its own synthesis. The O2 site is located on the opposite face of the helix from O1 and overlaps the –35 region of the ilvC promoter (182). In the absence of substrate inducers, the IlvY protein causes a 60° bend centered in the –35 region of the inactive ilvC promoter (Fig. 9). When substrate inducers bind to this preformed IlvY protein-DNA complex, the bend is relaxed and the affinity for RNA polymerase is increased 100-fold (182). The IlvY protein is autoregulated at a level that keeps these operators nearly saturated at all times. In this way the IlvY protein-DNA nucleoprotein complex acts as a sensor of the intracellular abundance of the α-acetohydroxy acid isomeroreductase substrates, thereby continuously adjusting expression of the ilvC gene to the abundance of its substrates, synthesized by the AHAS isozymes.
Regulation of the ilvBN Operon by CAP and IHF.
In addition to the ilv-specific regulation by valine- and leucine-mediated attenuation, the ilvBN operon (Fig. 4C) is subject to catabolite repression control by CAP (65, 66, 161, 178, 185). Stimulation of ilvBN expression by binding of CAP at −44 to −82 bp upstream of the transcription start site is thought to prevent binding of the RNA polymerase at a second, nonproductive polymerase-binding site and therefore to increase RNA polymerase binding to the promoter site where transcription of the operon is initiated (66).
In addition to transcriptional activation by CAP, the ilvBN operon is stimulated by IHF (67, 68, 165) (Fig. 4C). Tsui and Freundlich showed that IHF binds to a site immediately upstream of the ilvBN promoter and strongly decreases transcriptional pausing and transcriptional termination in the downstream ilvBN leader-attenuator region (165). They suggested that these upstream IHF-induced DNA conformational changes might explain IHF-mediated decreased transcriptional pausing in the leader-attenuator region.
Lrp Regulation of the ilvGMEDA and leuABCD Operons.
In addition to activating the ilvIH operon, Lrp protein binds to a high-affinity DNA-binding site 266 bp downstream of the transcription initiation site of the ilvG gene and represses transcription into the structural genes of the ilvGMEDA operon (138). Binding to this primary site facilitates cooperative binding of Lrp to adjacent upstream secondary and tertiary sites (Fig. 4D). Rhee et al. (138) suggested that this nucleoprotein complex immediately downstream of the attenuator site enhances transcription termination at the attenuator and thereby decreases transcription into the genes of the operon.
Ricca et al. (139) have shown that leucine decreases the affinity of Lrp for its consensus-like DNA-binding site (site 2) in the ilvIH operon (Fig. 4A). However, even at high concentrations, l-leucine did not abolish the ability of Lrp to bind to this site. Leucine also inhibits but does not abolish Lrp binding to its primary and secondary sites in the leader region of the ilvGMEDA operon (Fig. 4D). Given this observation and the results of the in vitro transcription and in vivo reporter constructs, Rhee et al. (138) suggested that high concentrations of leucine would relieve Lrp-mediated repression of ilvGMEDA operon expression. They further suggested that Lrp-mediated repression of the ilvGMEDA operon allows E. coli cells to coordinate isoleucine and valine biosynthesis rates with leucine availability. They point out that, during leucine starvation, the synthesis of all three of the AHAS isozymes is increased (166). However, the activities of the two isozymes with substrate preferences for leucine (and valine) biosynthesis, AHAS I and AHAS III (5), are feedback inhibited by l-valine, while the activity of AHAS II, with a substrate preference for l-isoleucine biosynthesis, remains unchecked (106). They suggest that this situation compromises the abilities of AHAS I and III to compete with AHAS II for pyruvate required for leucine biosynthesis. If this were the case, then an increase in the rate of pyruvate biosynthesis and a lowering of the expression of the ilvGM genes for AHAS II during leucine starvation might be advantageous (under these conditions, the expression of the distal genes of the ilvGMEDA operon would be sustained by the activity of the internal ilvPE promoter (24, 180). Thus, Lrp might function to repress the production of AHAS II during conditions of leucine starvation.
In addition to documented Lrp-mediated regulation of the ilvIH and ilvGMEDA operons, insertional mutagenesis studies performed by Tchetina and Newman (163) have been used to suggest that the leuABCD operon also is regulated by Lrp (Fig. 4E). However, direct regulation of this operon by Lrp has not been demonstrated.
To enable an organism to be both metabolically efficient and rapidly adaptive, mechanisms must exist to coordinate its global patterns of gene expression to its growth and environmental conditions. It is well known that basal levels of gene expression are coupled to the growth and nutritional states and environmental conditions of the cell through the regulation of transcriptional initiation by mechanisms that are sensitive to DNA superhelicity (83). For example, basal level expression of operons encoding structural genes for the biosynthesis of intermediary metabolites should be coordinated above the operon-specific level, so their basal expression levels are low when the chromosomal superhelical density is low during the stationary phase of growth and high when the chromosomal superhelical density is high during the logarithmic phase of growth. This is the case for the biosynthetic operons of the ilv regulon.
Genes required for the biosynthesis of intermediary metabolites such as amino acids must be continuously expressed at levels tuned to the amounts of their pathway end products. For example, operons regulated by attenuation, such as the ilvGMEDA, ilvBN, and leu operons of the ilv regulon, continuously transcribe a leader RNA whose translation into a leader polypeptide continuously monitors the intracellular levels of their pathway end products (reviewed in references 3a, 4, 75, 88, 194, 195, 197). The ilvY gene of this regulon also must be continuously expressed to maintain an IlvY protein-DNA complex that continuously monitors cellular levels of the α-acetohydroxy acid isomeroreductase substrate and adjusts the expression of the ilvC gene accordingly (13, 15, 177, 181, 182). Because these monitoring activities, which are typical of biosynthetic systems, present a high energy cost to the cell, one expects global mechanisms to exist that coordinate them both with each other and with cellular energy demand. These global mechanisms would be expected to respond to the energy charge of the cell in a manner independent of operon-specific controls.
In this section we describe the ilv regulon, a well-understood system involving hierarchical levels of global and operon-specific regulation that together coordinate the biosynthesis of branched-chain amino acids with the metabolic demands of the cell and its nutritional and environmental growth conditions. This material is prefaced with brief descriptions of the concept of cellular energy charge and how cellular energy charge influences DNA supercoiling, followed by the ways that DNA supercoiling influence the transcription initiation reaction.
Effects of DNA Supercoiling and Cellular Energy Charge on Gene Expression.
Early studies by Atkinson and his coworkers showed that the energy charge of the intracellular adenylate pool, defined as the ratio ([ATP] + ½ [ADP]) / ([ATP] + [ADP] + [AMP]), is the parameter that correctly describes the amount of metabolically available energy for the cell (2, 3, 34, 35). They also demonstrated that, during states of metabolic adjustment when the energy charge transiently decreases, such as during the transition from aerobic to anaerobic growth, enzymes involved in ATP-regenerating reactions are activated and enzymes involved in ATP-utilizing reactions are inhibited. Thus, decreases in the energy charge induce increases in the rates of enzymes that produce ATP and decreases in the rates of enzymes that consume ATP, while increases in the energy charge have the opposite effect. Together these changes maintain the energy charge in homeostatic balance. These findings explain why, although the absolute levels of adenylate pools vary under different growth conditions, the energy charge of the cell remains constant at a value of ~0.85 during balanced growth under all conditions. In stationary phase, however, the energy charge is maintained at a lower level (95, 99). To facilitate the following discussion energy charge sometimes will be referred to as the cellular [ATP/ADP] ratio, realizing that this is a simplification of Atkinson’s definition.
Many studies have shown that the level of global negative supercoiling is controlled by the cellular energy charge. Because the enzymatic activity of gyrase is controlled by the intracellular [ATP]/[ADP] ratio, not by the free ATP concentration (90, 91), high negative superhelical densities occur when cellular energy charge is high, and low negative superhelical densities occur when it is low (55, 56, 91, 183, 184). It is well known that energy charge and DNA supercoiling play coordinated roles in cellular adaptation and survival, both under suboptimal growth conditions and during growth-state transitions. In nongrowing E. coli cells in stationary phase where the [ATP/ADP] ratio is low, the superhelical density of a reporter plasmid is σ ≈ −0.03. As cells recover and enter into log phase the [ATP/ADP] ratio increases and the global negative superhelical density moves into the midphysiological level of σ ≈ −0.05, a value typical during balanced growth (103). Physical stresses alter both cellular energy charge and DNA-supercoiling levels. For example, osmotic stress (salt shock) causes the cellular [ATP/ADP] ratio to transiently increase fourfold and the negative superhelical density of the bacterial chromosome to increase to a value as high as σ = –0.09 (90). During transitions from aerobic to anaerobic growth the cellular [ATP/ADP] ratio decreases, and the global negative superhelical density transiently falls from σ = −0.05 to σ = −0.38 (42).
There are many ways in which DNA template topology (i.e., an imposed linking difference ▵Lk) influences gene expression (for a review, see reference 46). Drlica and coworkers showed that the level of supercoiling which gives optimal expression for promoters correlates with the length of the spacer region between the −35 and the −10 regions (160). Promoters whose maximum expression occurs at low superhelical densities tend to have spacer regions that are shorter than 17 bp. Promoters whose maximal activity occurs at high levels of supercoiling commonly have spacers that are longer than 17 bp. Promoters with 17-bp spacers were preferentially optimized for normal physiological levels of supercoiling and showed less sensitivity to changes in supercoiling than did the others. This is because the relative orientation between the −35 and the −10 promoter regions can strongly affect the ability of the σ70 subunit of RNA polymerase (RNAP) to locate and bind to a promoter (175). As negative supercoiling untwists DNA (i.e., ▵Lk < 0, so ▵Tw < 0), a promoter with a long-spacer region having a larger intrinsic twist than would provide optimal alignment will become more active at higher negative superhelical densities because this deformation decreases twist. Conversely, a short spacer whose twist is less than optimal will have its activity decreased by negative supercoiling and hence would be more active at smaller superhelical densities. Examples of such supercoiling-induced realignment mechanisms have been demonstrated in promoters involved in processes as diverse as the cold shock and osmotic shock responses, amino acid biosynthesis, and carbon utilization (19, 97, 98, 103, 173, 175).
The imposition of a negative linking difference on a topological domain of DNA also may alter the tertiary structure through its effect on writhe, Wr. In short regions, this can cause looping, while in longer regions it can induce plectonemic interwinding. Looped structures can form microdomains, which act as small independent topological domains. This process has been proposed to be involved in prokaryotic transcriptional activation (118). The formation of an interwound structure brings regions that are remote along the sequence into close physical proximity. If the DNA reptates through such a structure, eventually any site will find itself close in space to any other site. In this way plectonemic interwinding can greatly enhance the opportunities for sites remote along the duplex, or molecules bound thereto, to interact. This effect is the basis for the activity of the NtrC-dependent enhancer in E. coli, which can act in a supercoil-dependent manner over large distances, on the order of 2,000 bp (112).
DNA supercoiling also is known to drive transitions to a wide variety of alternative secondary structures, including local denaturation (102), transitions to Z-form (12) and to H-form (92), and cruciform extrusion (109). The formation of alternative DNA structures can serve regulatory functions, either by forming or modifying a regulatory binding site, or by altering the level of unconstrained supercoiling in the balance of the domain (8, 9, 10, 11). Dai and Rothman-Denes have shown that the bacteriophage N4 virion RNA polymerase (vRNAP) promoters contain short inverted repeat sequences centered at position −12 (45). These sites extrude cruciforms that are required for vRNAP recognition at physiological levels of supercoiling.
Finally, negative superhelicity also is known to destabilize the DNA duplex at specific locations (102). This effect is known as stress-induced duplex destabilization (SIDD). Local denaturation is the most extreme form of duplex destabilization. By altering the superhelicity of the region involved, local denaturation diminishes the level of unconstrained superhelicity experienced by the rest of the topological domain. This process can affect any regulatory activity that is attuned to supercoiling levels. The converse process, whereby the binding of a DNA duplex-binding protein to a destabilized site can force it back to B-form, also can be important. For example, protein-binding-induced reassociation can transmit the destabilization from the original SIDD site to the next most susceptible location, which can be a substantial distance away. If this distant site includes the −10 region of a promoter, this protein-mediated translocation of superhelical energy can lower the energy of activation for open complex formation and increase the rate of the closed-to-open complex step of the transcription initiation reaction.
Three DNA-supercoiling-dependent transcriptional regulatory mechanisms related to branched-chain amino acid biosynthesis and utilization have been documented (124, 125, 152, 153). These mechanisms act independently of the operon-specific controls described above to coordinate the expression levels of the operons of the ilv regulon with one another and with the nutritional needs and growth state of the cell in its physical environment. Two of these mechanisms (those governing the ilvGMEDA and leuV operons) modulate basal level transcription into their structural genes by a protein-mediated (IHF or Fis) transmission of local superhelical energy from an upstream SIDD site to a downstream promoter site. This influences the rate of RNAP-promoter open complex formation and/or the rate of RNAP binding (152, 153). In the third case, the basal level expression of the ilvC gene is enhanced by additional local superhelical energy contributed to the ilvC promoter region by divergent transcription of the ilvY gene (124, 125, 137). In each case, local superhelical energy is provided to the promoter regions to amplify promoter activity over the entire range of global physiological superhelical densities in a manner that coordinates the basal level expression of these operons within multiple regulons, both with one another and with the energy charge of the cell. Each of these mechanisms is described separately below.
IHF Regulation of the ilvGMEDA Operon by Protein-Mediated Translocation of Superhelical Energy.
Nested 5' deletions extending into the ilvPG promoter of the ilvGMEDA operon identified an upstream activating region (UAS). This UAS contains a high-affinity IHF target-binding site located 92 bp upstream from the transcriptional start site (Fig. 4D). Biochemical and genetic experiments showed that IHF binding to this site, both in vivo and in vitro on a supercoiled DNA template, but not on a relaxed template, causes a fivefold activation of transcription from the downstream ilvPG promoter (186). Several experimental approaches have established that: this activation occurs in the absence of specific protein interactions between IHF and RNA polymerase; it is not the consequence of a DNA-looping (reptation) mechanism, and it requires a superhelical DNA template (128, 129, 152, 153, 154). The possible presence of a SIDD site in the UAS was first suggested by the observation that the base pair composition of the ilvPG promoter regulatory region is exceptionally A + T rich, the 80-bp segment from base pair positions −67 to −153 being approximately 88% A + T. In order to determine whether this region does indeed contain a SIDD site, SIDD profiles were calculated for the pBR322-based plasmid pDHΔwt (152). This plasmid contains the ilvPG promoter region from positions −248 to +6, together with transcriptional terminators located downstream from the ilvPG start site. The results of these calculations are presented as the destabilization profiles in Fig. 10. Subsequent chemical probing experiments confirmed that, in the absence of IHF, the UAS region was indeed destabilized at the superhelix densities where activation occurs. Moreover, IHF binding in the UAS region of a superhelical DNA template resulted in the transmission of this duplex destabilization to the −10 region of the downstream ilvPG promoter site (Fig. 11). Abortive transcription assays showed that this DNA structural change at the downstream promoter site is correlated with both an increase in the rate of open complex formation and a concomitant increase in the rate of transcriptional initiation (152).
According to this mechanism, the primary determinant for IHF-mediated activation is predicted to be superhelically induced DNA destabilization. Neither specific DNA sequences nor specific IHF-RNAP interactions are required. To directly test this prediction, a (CG)13AATT(CG)22 sequence susceptible to a superhelically induced transition to Z-form DNA (135) was inserted into a plasmid approximately 500 bp upstream from the UAS-ilvPG promoter region (153). Because the B-Z DNA transition at this remote site competes effectively with the SIDD site in the UAS- ilvPG promoter region, this insertion can alter the destabilization properties of the UAS without changing the DNA sequence in any part of the ilvPG regulatory promoter region. This transition was shown to absorb 13 negative superhelical turns, thereby relaxing the global superhelical density of the remainder of the supercoiled DNA template by a corresponding amount. Since this B-Z transition occurs at a lower threshold superhelical density (σ = −0.025) than the destabilization of the SIDD site in the ilvPG UAS (σ = −0.038), it inhibits destabilization in the ilvPG UAS region until approximately 13 additional negative superhelical turns have been added to the DNA template. Therefore, if the energy required for IHF-mediated transcriptional activation is indeed solely derived by IHF-binding-induced transfer of superhelical energy (negative twist) from the upstream SIDD region to the downstream ilvPG promoter site, then the global superhelicity required for IHF activation in the plasmid containing the Z-DNA insert should be offset by 13 turns. The results of transcription assays on DNA templates of defined superhelix densities showed this to be the case (Fig. 12). The superhelicities required both for half-maximal basal level and for IHF-activated transcription were indeed offset by 13 helical turns (152). This experiment clearly demonstrated that IHF-mediated transcriptional activation of the ilvPG promoter is solely DNA supercoiling dependent.
Fis Regulation of the leuV Operon by Protein-Mediated Translocation of Superhelical Energy.
The leuV operon of E. coli encodes three of the five genes for tRNALeu isoacceptors. Like other stable RNA-encoding genes, it has a strong promoter, with near-consensus RNAP recognition sequences and a G + C-rich discriminator region located between base pair positions –8 and +1 (Fig. 13 and 14). Transcription from the leuV promoter is enhanced by a third RNAP recognition element located between base pairs −39 and –47. This A + T-rich UP sequence makes contacts with the α-subunits of RNAP, stabilizes closed complex formation, and activates leuV expression more than 10-fold (59, 132, 144). The UAS of this promoter contains a Fis protein-binding site centered at base pair position –71. Fis binding to this site enhances leuV expression an additional threefold (145).
A large body of evidence demonstrates that Fis is a class-I activator, enhancing transcription by increasing RNAP-binding affinity through direct contacts with the C-terminal domain of its α-subunits (16, 17, 22, 121, 198). However, other mechanisms also are involved in Fis activation of stable RNA promoters. For example, Fis binding has been shown to increase the rate of promoter clearance at the rrnD promoter (148). Muskhelishvili and Travers have shown that Fis activates transcription from the tyrT promoter by enhancing the rate of open complex formation and promoter clearance, as well as RNA polymerase-binding affinity (118). Opel et al. have obtained evidence that basal level expression of the leuV promoter is also activated by a Fis-mediated translocation of superhelical energy mechanism similar to the IHF-mediated, DNA-supercoiling-dependent mechanism of the ilvPG operon (123).
A SIDD profile of the promoter regulatory region of the leuV operon is shown in Fig. 14. At a mid-physiological superhelical density of σ = –0.05, this region is predicted to be destabilized from base pair positions +43 to –94. The Fis protein-binding site is located in the upstream region of this SIDD site centered at base pair position –72. An interesting feature of the SIDD profile is the sharp peak of duplex stability at the G + C-rich discriminator region between positions +1 and –8. The presence of this region of high-duplex stability at a site where strand separation must occur predicts a high energy of activation for open complex formation. Indeed, leuV transcription is exceptionally sensitive to negative DNA supercoiling, increasing over 100-fold from its lowest level on a relaxed DNA template to its highest level on a more supercoiled DNA template. The SIDD profile further predicts that the upstream region of the SIDD site should be stabilized by Fis binding to its target site and that this binding should destabilize the downstream portion of this SIDD region containing the leuV promoter sequences. This prediction was confirmed with in vitro transcription assays and KMnO4 structural probing experiments performed with supercoiled DNA template topoisomers in the presence or absence of Fis and/or RNAP. These experiments showed that Fis binding enhances the rate of open complex formation in a DNA-supercoiling-dependent manner. At subsaturating concentrations of RNAP, Fis activation is facilitated both by protein-protein interactions between Fis and RNAP and by DNA-supercoiling-dependent enhancement of open complex formation. At saturating RNAP concentrations only the enhancement of open complex is seen. Mutant Fis proteins that do not form contacts with the α-subunits of RNAP but bind to the target site with wild-type affinities were used to demonstrate that this activation does not require Fis-RNAP interactions (123). These mutant proteins maintain their ability to facilitate DNA-supercoiling-dependent enhancement of open complex formation. Thus, Fis activation of basal level transcription from the leuV promoter involves at least two mechanisms: stabilization of the closed complex by protein interactions with the α-subunit of RNAP and increasing open complex formation by translocation of superhelical energy from the upstream portion of the SIDD region containing the Fis target site to the downstream portion of this region containing the leuV promoter sequences.
Unlike the ilvPG promoter that reaches its peak transcriptional activity at a high physiological superhelical density near σ = −0.10, transcriptional activity of the leuV promoter peaks at a superhelical density near σ = −0.07 and decreases thereafter to the level observed on a relaxed DNA template. This decrease in transcriptional activity at high superhelix densities is accompanied by the in vitro formation of a cruciform structure located between base pair positions +8 to +26 relative to the transcriptional start site (M. Opel, K Aeling, and G. W. Hatfield, unpublished results). Structural probing experiments with DNA topoisomers showed that the threshold superhelical density required for extrusion of this cruciform structure is σ = −0.069, the same superhelical density beyond which transcription from the leuV promoter decreases. This suggests that, at superhelical densities beyond this threshold, the global and Fis-transferred local superhelical energy in the promoter region is absorbed by the cruciform. This transition has two effects on leuV gene expression: it usurps the local superhelical energy that would otherwise have been transferred to the promoter region by Fis for open complex formation and it physically blocks RNAP binding.
A schematic diagram illustrating how Fis binding and DNA-supercoiling-induced structural transitions might regulate transcription from the leuV promoter is presented in Fig. 15. At the low physiological superhelical densities typical of stationary-phase growth, transcription from the leuV promoter is very low. This is due to the energy barrier for open complex formation caused by the G + C-rich discriminator near the transcription start site. Under these conditions Fis can activate transcription about threefold by increasing RNAP binding through interactions with its α-subunits. Since this low level of global superhelicity is insufficient for SIDD site formation, no additional activation by Fis-mediated translocation of superhelical energy to the promoter site is possible. As the global superhelicity of the DNA template is increased to the midphysiological range, the energy barrier for open complex formation caused by the discriminator is decreased and transcription increases up to 100-fold. As the SIDD site now is present, Fis binding can activate transcription both by enhancing RNAP binding and by protein-mediated translocation of superhelical energy to the promoter region. As the global superhelical density passes beyond the midphysiological range, the superhelical energy at the SIDD site is usurped by formation of the cruciform structure near the transcription start site and RNAP binding is inhibited. This might serve as a mechanism to inhibit leuV tRNA expression during stress conditions that cause transient increases in DNA supercoiling. However, the in vivo presence of the cruciform has not been demonstrated.
Regulation of the ilvYC Operon by Transcriptional Coupling.
The ilvYC operon of E. coli K-12 is a prototypic LysR-type regulated system (87, 149). LysR-type regulated operons are the largest class of positively regulated operons and are found in many prokaryotic species (87, 149). The prevalence of a divergent gene arrangement among the LysR-type regulated operons suggests an evolutionary conservation of potential regulatory significance. According to the twin-domain model (110), a local domain having a high level of DNA supercoiling can be generated between, and influence the activities of, divergently transcribed promoters. Mojica and Higgins have used in vivo psoralen cross-linking techniques to demonstrate that localized domains of increased negative DNA supercoiling are indeed generated upstream from an actively transcribed promoter (116). They demonstrated DNA-supercoiling-mediated transcriptional coupling between the divergently oriented tetA and mutant leu-500 promoters (114a, 133a).
Rhee et al. used double-reporter gene constructs to provide the first in vivo evidence for transcriptional coupling in a naturally occurring system, the ilvYC operon of E. coli (137). They showed that each of these promoters is intrinsically sensitive to global DNA supercoiling and that a 13-fold decrease in transcriptional activity from the ilvY promoter results in an 11-fold decrease in transcription from the divergent ilvC promoter. This transcriptional coupling was shown to be the consequence of transcription-induced negative DNA supercoiling. In this situation, a highly stressed local topological domain is created in the promoter region by divergent transcription, in which the total supercoiling is the sum of the basal, global superhelicity plus the supercoiling arising from divergent transcription. This suggested a strategy to document and characterize transcriptional coupling in a purified in vitro transcription system. When a set of DNA topoisomer templates containing the wild-type, divergently oriented ilvY and ilvC promoters was transcribed in a purified system, optimal transcriptional activity was observed to occur at superhelical density σ = −0.065 for the ilvY promoter, and σ = −0.11 for the ilvC promoter (137). If the level of negative DNA supercoiling in the divergently transcribed promoter region were in fact proportional to the sum of transcription-induced (local) DNA supercoiling and the global superhelical density of the DNA template, then a decrease in transcription (hence, in transcription-induced supercoiling) from either promoter should require a compensating increase in global DNA supercoiling to maintain maximal transcription from its divergently transcribed partner. Conversely, an increase in transcription from either promoter should require a corresponding decrease in global supercoiling to maintain maximal transcription from the divergently oriented other promoter. Further, the twin-domain model of Liu and Wang (110) also predicts that the levels of transcription-induced negative DNA supercoiling in the divergent promoter domain region should be proportional to the lengths of the transcripts. Opel and Hatfield performed in vitro experiments using a purified transcription system and DNA topoisomer templates containing the genes of the ilvYC operon that confirm both of these predictions (125).
An ultimate goal of systems biology is to develop dynamic mathematical models of interacting biological networks capable of simulating a living system. Since E. coli is the model organism on which much of our basic knowledge of molecular biology is based, and because of our vast knowledge of the biology of this single organism evidenced by the first and second editions of Escherichia coli and Salmonella: Cellular and Molecular Biology, it is likely that this daunting task might be accomplished in the not too distant future. As a beginning step toward the elucidation of the systems biology of E. coli, Yang et al. (192) have published a mathematical model of the well-studied end product-regulated pathways for the biosynthesis of the branched-chain amino acids l-isoleucine, l-valine, and l-leucine described in this chapter. To accomplish this task, they developed a computer language, kMech, which describes a suite of enzyme mechanisms. kMech is built in a modular manner such that complex enzyme mechanisms may be constructed from simpler ones. Each enzyme mechanism is parsed by kMech into a set of fundamental association-dissociation reactions according to the interactions of molecular components and translated by another software component, Cellerator, into ordinary differential equations (ODEs) based on the law of mass action that are numerically solved by MathematicaTM (193 ). Also, Yang et al. (193) have described methods that use commonly available kinetic measurements to estimate rate constants and genomic and proteomic data to estimate enzyme concentrations required for solving these differential equations.
Traditional modeling approaches use the Michaelis-Menten kinetic equation for one substrate/one product reactions and the King-Altman method to derive equations for multiple-reactant reactions (60). These methods focus on conversion between metabolites (metabolic flux) rather than enzyme mechanisms. While metabolic flux estimates can provide valuable information about biomass conversions, they cannot simulate pathway-specific regulation patterns. Flux analyses cannot capture the control of carbon flow channeling through the three AHAS isozymes shared by the parallel isoleucine and valine pathways, the allosteric control by pathway end products, or the multiple transamination reactions performed by the transaminase B enzyme described in previous sections.
In the same way that flux analyses do not model enzyme mechanisms and regulatory circuits, these details also are not apparent in traditional metabolite conversion pathways. Therefore, Yang et al. (192) developed a method to convert the traditional metabolite conversion pathway of Fig. 1 into an enzyme-centric pathway shown in Fig. 16. This figure depicts pathway-specific regulation patterns, channeling of metabolic flow, isozymes, cross-talk between pathways (bi-functional enzymes), positive/negative regulation, reversible enzyme reactions, and modifications of enzyme states (Fig. 16) (192). This enzyme-centric pathway diagram also includes the threonine biosynthetic pathway that immediately precedes the isoleucine biosynthetic pathway. The aspartate kinase I (AKI) and homoserine dehydrogenase I (HDHI) activities of the threonine pathway are both catalyzed by a single bifunctional enzyme. The AKI active site is in the N-terminal domain of this protein molecule, and the HDHI active site is in the C-terminal domain. Both enzyme activities are allosterically regulated by threonine. Najdi et al. (119) have developed a generalized Monod-Wyman-Changeaux allosteric enzyme model capable of accurately simulating the regulatory properties of this and other complex allosteric enzymes.
Fig. 16Enzyme-centric metabolic pathways for the biosynthesis of threonine and the branched-chain amino acids l-isoleucine, l-valine, and l-leucine. The abbreviations of enzymes used are as follows: AKI, aspartate kinase I; AKIII, aspartate kinase III; HDHI, homoserine dehydrogenase I; ASD, semialdehyde dehydrogenase; HSK, homoserine kinase; TS, threonine synthase; TDA, l-threonine deaminase; AHAS, acetohydroxy acid synthase; IR, acetohydroxy acid isomeroreductase; DAD, dihydroxy acid dehydrase; TB, transaminase B; TC, transaminase C; IPMS, α-isopropylmalate synthase; IPMI, α-isopropylmalate isomerase; IPMDH, α-isopropylmalate dehydrogenase. The abbreviations of metabolites used are as follows: Asp, aspartate; AspP, aspartyl phosphate; ASA, aspartate semialdehyde; Hse, homoserine; HseP, homoserine phosphate; Thr, l-threonine; Ile, l-isoleucine; Val, l-valine; Leu, l-leucine; Glu, l-glutamate; Ala, alanine; Pyr, pyruvate; αKB, α-ketobutyrate; αAL, α-acetolactate; αAHB, α-aceto-α,β-hydroxybutyrate; αDHIV, α,β-dihydroxyisovalerate; αDMV, α,β-dihydroxy-β-methylvalerate; αKIV, α-ketoisovalerate; αKMV, α-keto-β-methylvalerate; αKG, α-ketoglutarate; αIPM, α-isopropylmalate; βIPM, β-isopropylmalate; αKIC, α-ketoisocaproate. Ovals represent enzyme molecules. White ovals indicate free enzyme states, and shaded ovals indicate intermediate enzyme states with a function group attached to enzymes. Enzyme reactions are indicated by lines with arrows. Reversible reactions are indicated by gray lines with arrows. Switching between free and intermediate enzyme states as well as switching between AKI-HDHI activities is indicated by double-arrowed dashed lines. Presented with permission from reference 192.
Yang et al. (192) and Najdi et al. (119) have used kMech and Cellerator to mathematically model the regulated flow of metabolites through the threonine and branched-chain amino acid biosynthetic pathways. The mathematical model of these four interacting metabolic pathways (Fig. 12) consists of 17 enzymes described by 131 ordinary differential equations (ODEs) and 189 rate constants. The enzyme mechanisms of these pathways include simple catalytic, Bi Bi, Ping Pong Bi Bi, and Ter Bi systems that are regulated by allosteric, competitive, and noncompetitive inhibition and activation mechanisms. The mathematically simulated attainment of steady-state production levels for the 20 metabolites of these pathways is shown in Fig. 17. Steady-state enzyme activity levels were optimized to properly channel the steady-state flow of metabolic intermediates through these pathways at levels that closely match their reported in vivo levels.
Systems biology may be broadly defined as the integration of diverse data into useful mathematical models that allow scientists to easily observe complex cellular behavior and to predict the outcomes of metabolic and genetic perturbations. Yang et al. (192) have shown that the enzyme-centric model accurately simulates experimentally observed effects of genetic and biochemical perturbations.
Valine-Mediated Inhibition of E. coli K-12 Growth.
It is well known that addition of l-valine at a final concentration of 1 mM to a culture of E. coli K-12 cells growing in a glucose minimal salts medium inhibits their growth and that this valine inhibition can be reversed by the addition of isoleucine to the growth medium (170). Since the AHAS I and AHAS III isozymes of E. coli K-12 strains are inhibited by valine and since the ilvG gene for AHAS II in E. coli K-12 strains contains a frameshift mutation that destroys AHAS II activity (107), it was assumed that l-valine inhibition of AHAS I and AHAS III might inhibit growth by inhibiting isoleucine biosynthesis. However, later studies demonstrated that the intracellular isoleucine level is not suppressed by valine because its biosynthesis is sustained, by AHAS III that remains 15 to 20% active , even at saturating l-valine concentrations (89, 93), and by valine activation of threonine deaminase that shuttles more substrate into the isoleucine pathway (Fig. 3). Indeed, the simulation in Fig. 3A shows that, in the presence of extracellular valine, the intracellular isoleucine level, in fact, accumulates nearly ninefold in the presence of excess valine. At the same time, the pathway precursor of l-isoleucine, αKB, increases about fourfold (Fig. 3B). This buildup of αKB is the consequence of valine activation of threonine deaminase (Fig. 3C), which increases its production, and l-valine inhibition of the AHAS I and AHAS III isozymes, which reduces its consumption. It is now known that this αKB accumulation is toxic to cells because of its ability to inhibit the glucose PTS transport system (47, 48). Thus, as reproduced by the simulations, l-valine growth inhibition of E. coli K-12 is not a consequence of l-isoleucine starvation.
The simulation results in Fig. 3B further show that the growth-inhibiting effects of l-valine-induced αKB accumulation can be reversed by isoleucine through its ability to inhibit threonine deaminase activity. This simulation shows that, in the presence of 1 mM l-valine, the level of αKB increases around fourfold and that, in the presence of 500 μM l-isoleucine, αKB levels are reduced to the control level observed in the absence of l-valine. The simulation results in Fig. 3C show that, concomitant with the rise in αKB observed in the presence of 1 mM l-valine, nearly 18% of the cellular threonine deaminase is converted to the active R state. However, concomitant with the decrease in αKB observed in the presence of 500 μM l-isoleucine, the cellular threonine deaminase in the active R state is reversed to the control level observed in the absence of l-valine. These simulations are consistent with experimental results accumulated from multiple laboratories over three decades (47, 48, 170).
Engineering an Isoleucine-Overproducing E. coli Strain.
An obvious goal of modeling biological systems is to facilitate metabolic engineering for the commercial production of specialty chemicals such as amino acids. In the past, this was largely accomplished by genetic manipulation and selection methods. For example, a common strategy to overproduce an amino acid was to isolate a strain with a feedback-resistant mutation in the gene for the first enzyme for the biosynthesis of that amino acid. Yang et al. (192) used the enzyme-centric model to determine the effects of a feedback-resistant threonine deaminase for the overproduction of l-isoleucine in the E. coli K-12 strain. They simulated a feedback-resistant threonine deaminase strain (TDAR) by simply increasing the Ki for l-isoleucine to a large number (e.g., 100,000 μM). The simulation in Fig. 18A shows that, in the absence of l-isoleucine inhibition, the activator and substrate ligands drive nearly 100% of cellular feedback-resistant TDAR enzyme to the active R state compared with the wild-type enzyme that is only 6% present in the active R state. However, despite this increased amount of enzyme in the active state, the data in Fig. 18B show that AHAS III is able to support only a five- to sixfold increase in l-isoleucine production in a feedback-resistant E. coli K-12 strain compared with a wild-type strain. At the same time, the steady-state level of the AHAS III substrate, αKB, is increased about 40-fold (Fig. 18C). This is because E. coli K-12 does not have an active AHAS II isozyme that favors the condensation of pyruvate and αKB for l-isoleucine production; thus, αKB accumulates to toxic levels. Therefore, these simulation results suggest that, to overproduce isoleucine, αKB accumulation must be reduced. The results in Fig. 18D show that restoring a wild-type AHAS II isozyme and simulating an ilvGMEDA attenuator mutation that elevates the levels of all of the enzymes of the l-isoleucine and l-valine parallel pathways 11-fold (1) both avoids buildup of αKB and subsequent pathway intermediates (Fig. 18C) and results in a 40-fold increase in isoleucine production. These simulated results that predict that high-level overproduction of l- isoleucine in E. coli requires a functional AHAS II isozyme and a deattenuated genetic background (ilvGMEDA-att−) agree with experiments performed by Hashiguchi et al. of the Ajinomoto Co., Tokyo, Japan (80).
As described in the previous sections, the kMech/Cellerator package generates differential equations according to the law of mass action. That is, it is assumed that reactants are spatially homogenously distributed and that concentrations of reactants vary continuously and deterministically. Both assumptions are questionable in the case of gene regulation. In this case, small numbers of molecules involved in the genetic regulatory mechanism compromise the continuity assumption. For example, in the extreme only a single chromosomal site is available for many operon-specific regulatory proteins. Such a paucity of molecular sites makes deterministic change of concentrations problematic because of fluctuations in the timing of cellular events such as RNA polymerase and regulatory protein binding. In other words, motivation for the stochastic approach comes from the realization that many crucial events in living cells depend on the interaction of small numbers of molecules and are sensitive to the underlying stochasticity of reaction processes such as the chance meeting of a repressor molecule with a single genomic operator site.
Shapiro and colleagues are currently developing stochCellar, a stochastic simulation software for biological reactions. stochCellar is a sister package of Cellerator (151). It uses similar input formats for chemical reactions, initial conditions, and rate constants. The major difference is that the values of variables representing individual molecules are numbers of molecules in stochCellar instead of concentrations of molecules in Cellerator. Also, values of variables representing rate constants are reaction probabilities in stochCellar instead of kinetic constants in Cellerator. In collaboration with Eric Mjolsness and Bruce Shapiro at Caltech, we currently are developing gMech, a stochCellar language extension that describes a suite of genetic regulatory mechanisms, such as attenuation, suitable for the stochastic modeling of genetic regulatory circuits in E. coli. In the near future, an integrative environment will be developed for the simultaneous simulation of both deterministic and stochastic models.
References
1. Adams, C. W., M. Rosenberg, and G. W. Hatfield. 1985. Analysis of in vivo RNA transcription products of the ilvGEDA attenuator region of Escherichia coli K12. J. Biol. Chem. 260:8538–8544.[PubMed]
2. Atkinson, D. E. 1965. Biological feedback control at the molecular level. Science 150:851–857.[PubMed] [CrossRef]
3. Atkinson, D. E. 1969. Regulation of enzyme function. Annu. Rev. Microbiol. 23:47–68.[PubMed] [CrossRef]
3a. Babitzke, P. 1997. Regulation of tryptophan biosynthesis: Trp-ing the TRAP or how Bacillus subtilis reinvented the wheel. Mol. Microbiol. 26:1–9.[PubMed] [CrossRef]
4. Babitzke, P. 2004. Regulation of transcription attenuation and translation initiation by allosteric control of an RNA-binding protein: the Bacillus subtilis TRAP protein. Curr. Opin. Microbiol. 7:132–139.[PubMed] [CrossRef]
5. Barak, Z., D. M. Chipman, and N. Gollop. 1987. Physiological implications of the specificity of acetohydroxy acid synthase isozymes of enteric bacteria. J. Bacteriol. 169:3750–3756.[PubMed]
6. Barnes, W. M., and E. Tuley. 1983. DNA sequence changes of mutations in the histidine operon control region that decrease attenuation. J. Mol. Biol. 165:443–459.[PubMed] [CrossRef]
7. Bartkus, J. M., B. Tyler, and J. M. Calvo. 1991. Transcription attenuation-mediated control of leu operon expression: influence of the number of Leu control codons. J. Bacteriol. 173:1634–1641.[PubMed]
8. Benham, C. J. 1996. Duplex destabilization in superhelical DNA is predicted to occur at specific transcriptional regulatory regions. J. Mol. Biol. 255:425–434.[PubMed] [CrossRef]
9. Benham, C. J. 1993. Sites of predicted stress-induced DNA duplex destabilization occur preferentially at regulatory loci. Proc. Natl. Acad. Sci. USA 90:2999–3003.[PubMed] [CrossRef]
10. Benham, C. J. 1980. Theoretical analysis of transitions between B- and Z-conformations in torsionally stressed DNA. Nature 286:637–638.[PubMed] [CrossRef]
11. Benham, C. J. 1979. Torsional stress and local denaturation in supercoiled DNA. Proc. Natl. Acad. Sci. USA 76:3870–3874.[PubMed] [CrossRef]
12. Benham, C. J. 1990. Transitions in superhelical DNA molecules of specified sequence. J. Chem. Phys. 92:6294–6305. [CrossRef]
13. Biel, A. J., and H. E. Umbarger. 1981. Mutations in the ilvY gene of Escherichia coli K-12 that cause constitutive expression of ilvC. J. Bacteriol. 146:718–724.[PubMed]
14. Blatt, J. M., W. J. Pledger, and H. E. Umbarger. 1972. Isoleucine and valine metabolism in Escherichia coli. XX. Multiple forms of acetohydroxy acid synthetase. Biochem. Biophys. Res. Commun. 48:444–450.[PubMed] [CrossRef]
15. Blazey, D. L., and R. O. Burns. 1984. Regulation of Salmonella typhimurium ilvYC genes. J. Bacteriol. 159:951–957.[PubMed]
16. Bokal, A. J., W. Ross, T. Gaal, R. C. Johnson, and R. L. Gourse. 1997. Molecular anatomy of a transcription activation patch: FIS-RNA polymerase interactions at the Escherichia coli rrnB P1 promoter. EMBO J. 16:154–162.[PubMed] [CrossRef]
17. Bokal, A. J. T., W. Ross, and R. L. Gourse. 1995. The transcriptional activator protein FIS: DNA interactions and cooperative interactions with RNA polymerase at the Escherichia coli rrnB P1 promoter. J. Mol. Biol. 245:197–207.[PubMed] [CrossRef]
18. Borg-Olivier, S. A., D. Tarlinton, and K. D. Brown. 1987. Defective regulation of the phenylalanine biosynthetic operon in mutants of the phenylalanyl-tRNA synthetase operon. J. Bacteriol. 169:1949–1953.[PubMed]
19. Borowiec, J. A., and J. D. Gralla. 1985. Supercoiling response of the lac ps promoter in vitro. J. Mol. Biol. 184:587–598.[PubMed] [CrossRef]
20. Bowen, T. L., J. Union, D. L. Tumbula, and W. B. Whitman. 1997. Cloning and phylogenetic analysis of the genes encoding acetohydroxyacid synthase from the archaeon Methanococcus aeolicus. Gene 188: 77-84. [CrossRef]
21. Bremer, H., and P. P. Dennis. 1996. Modulation of chemical composition and other parameters of the cell by growth rate, p. 1553–1569. In F. C. Neidhardt, R. Curtiss III, J. L. Ingraham, E. C. C. Lin, K. B. Low, B. Magasanik, W. S. Reznikoff, M. Riley, M. Schaechter, and H. E. Umbarger (ed.), Escherichia coli and Salmonella: Cellular and Molecular Biology, 2nd ed., vol. 2. ASM Press, Washington, D.C.
22. Breslauer, K. J., R. Frank, H. Blocker, and L. A. Marky. 1986. Predicting DNA duplex stability from the base sequence. Proc. Natl. Acad. Sci. USA 83:3746–3750.[PubMed] [CrossRef]
23. Burns, D. M., M. J. Burger, and I. R. Beacham. 1995. Silent genes in bacteria: the previously designated 'cryptic' ilvHI locus of 'Salmonella typhimurium LT2' is active in natural isolates. FEMS Microbiol. Lett. 131:167–172.[PubMed] [CrossRef]
24. Calhoun, D. H., J. S. Kuska, and G. W. Hatfield. 1975. Threonine deaminase from Escherichia coli. II. Maturation and physical properties of the enzyme from a mutant altered in its regulation of gene expression. J. Biol. Chem. 250:127–131.[PubMed]
25. Chang, Y. Y., and J. E. Cronan, Jr. 1988. Common ancestry of Escherichia coli pyruvate oxidase and the acetohydroxy acid synthases of the branched-chain amino acid biosynthetic pathway. J. Bacteriol. 170:3937–3945.[PubMed]
26. Changeux, J. P. 1962. Effect of L-threonine and L-isoleucine analogs on L-threonine desaminase. J. Mol. Biol. 4:220–225.[PubMed] [CrossRef]
27. Changeux, J. P. 1961. The feedback control mechanisms of biosynthetic L-threonine deaminase by L-isoleucine. Cold Spring Harbor Symp. Quant. Biol. 26:313–318.[PubMed]
28. Changeux, J. P. 1965. On the allosteric properties of biosynthesized L-threonine deaminase. IV. The desensitization phenomenon. Bull. Soc. Chim. Biol. (Paris) 47:113–139.[PubMed]
29. Changeux, J. P. 1965. On the allosteric properties of biosynthesized L-threonine deaminase. V. The allosteric transition. Bull. Soc. Chim. Biol. (Paris) 47:267–280.[PubMed]
30. Changeux, J. P. 1965. On the allosteric properties of biosynthesized L-threonine deaminase. VI. General discussion. Bull. Soc. Chim. Biol. (Paris) 47:281–300.[PubMed]
31. Changeux, J. P. 1964. On the allosteric properties of biosynthetic L-threonine deaminase. 3. Interpretation of the inhibitory effect of L-isoleucine: steric hindrance or allosteric effect. Bull. Soc. Chim. Biol. (Paris) 46:1151–1173.[PubMed]
32. Changeux, J. P. 1964. On the allosteric properties of biosynthetic L-threonine deaminase. II. Kinetics of action of biosynthetic L-threonine deaminase with respect to the natural substrate and inhibitor. Bull. Soc. Chim. Biol. (Paris) 46:947–961.[PubMed]
33. Changeux, J. P. 1964. On the allosteric properties of L-threonine deaminase. I. Methods of studying biosynthetic L-threonine deaminase. Bull. Soc. Chim. Biol. (Paris) 46:927–946.[PubMed]
34. Chapman, A. G., and D. E. Atkinson. 1977. Adenine nucleotide concentrations and turnover rates. Their correlation with biological activity in bacteria and yeast. Adv. Microb. Physiol. 15:253–306.[PubMed] [CrossRef]
35. Chapman, A. G., L. Fall, and D. E. Atkinson. 1971. Adenylate energy charge in Escherichia coli during growth and starvation. J. Bacteriol. 108:1072–1086.[PubMed]
36. Chen, C. C., M. Y. Chou, C. H. Huang, A. Majumder, and H. Y. Wu. 2005. A cis-spreading nucleoprotein filament is responsible for the gene silencing activity found in the promoter relay mechanism. J. Biol. Chem. 280:5101–5112.[PubMed] [CrossRef]
37. Chen, C. C., M. Ghole, A. Majumder, Z. Wang, S. Chandana, and H. Y. Wu. 2003. LeuO-mediated transcriptional derepression. J. Biol. Chem. 278:38094–38103.[PubMed] [CrossRef]
38. Chen, C. C., and H. Y. Wu. 2005. LeuO protein delimits the transcriptionally active and repressive domains on the bacterial chromosome. J. Biol. Chem. 280:15111–15121.[PubMed] [CrossRef]
39. Chen, C. C., and H. Y. Wu. 2003. Transcription-driven DNA supercoiling and gene expression control. Front. Biosci. 8:d430–d439.[PubMed] [CrossRef]
40. Chen, J. W., E. Harms, and H. E. Umbarger. 1991. Mutations replacing the leucine codons or altering the length of the amino acid-coding portion of the ilvGMEDA leader region of Escherichia coli. J. Bacteriol. 173:2341–2353.[PubMed]
41. Chipman, D., Z. Barak, and J. V. Schloss. 1998. Biosynthesis of 2-aceto-2-hydroxy acids: acetolactate synthases and acetohydroxyacid synthases. Biochim. Biophys. Acta 1385:401–419.[PubMed]
42. Cortassa, S., and M. A. Aon. 1993. Altered topoisomerase activities may be involved in the regulation of DNA supercoiling in aerobic-anaerobic transitions in Escherichia coli. Mol. Cell. Biochem. 126:115–124.[PubMed] [CrossRef]
43. Crout, D. H. G. 1990. The chemistry of branched chain amino acid biosynthesis: stereochemical and mechanistic aspects, p. 199–242. In Z. Barak, D. M. Chipman, and J. V. Schloss (ed.), Biosynthesis of Branched Chain Amino Acids. VCH Publishers, New York, N.Y.
44. Crout, D. H. G., E. R. Lee, and D. L. Rathbone. 1990. Absolute configuration of the product of the acetolactate synthase reaction by a novel method of analysis using acetolactate decarboxylase. J. Chem. Soc. Perkin Trans. 1 5:544–549.
45. Dai, X., M. B. Greizerstein, K. Nadas-Chinni, and L. B. Rothman-Denes. 1997. Supercoil-induced extrusion of a regulatory DNA hairpin. Proc. Natl. Acad. Sci. USA 94:2174–2179.[PubMed] [CrossRef]
46. Dai, X., and L. B. Rothman-Denes. 1999. DNA structure and transcription. Curr. Opin. Microbiol. 2:126–130.[PubMed] [CrossRef]
47. Danchin, A., L. Dondon, and J. Daniel. 1984. Metabolic alterations mediated by 2-ketobutyrate in Escherichia coli K12. Mol. Gen. Genet. 193:473–478.[PubMed] [CrossRef]
48. Daniel, J., L. Dondon, and A. Danchin. 1983. 2-Ketobutyrate: a putative alarmone of Escherichia coli. Mol. Gen. Genet. 190:452–458.[PubMed] [CrossRef]
49. Daniel, J., and I. Saint-Girons. 1982. Attenuation in the threonine operon: effects of amino acids present in the presumed leader peptide in addition to threonine and isoleucine. Mol. Gen. Genet. 188:225–227.[PubMed] [CrossRef]
50. De Felice, M., J. Guardiola, B. Esposito, and M. Iaccarino. 1974. Structural genes for a newly recognized acetolactate synthase in Escherichia coli K-12. J. Bacteriol. 120:1068–1077.[PubMed]
51. De Felice, M., and M. Levinthal. 1977. The acetohydroxy acid synthase III isoenzyme of Escherichia coli K-12: regulation of synthesis by leucine. Biochem. Biophys. Res. Commun. 79:82-87. [CrossRef]
52. De Felice, M., M. Levinthal, M. Iaccarino, and J. Guardiola. 1979. Growth inhibition as a consequence of antagonism between related amino acids: effect of valine in Escherichia coli K-12. Microbiol. Rev. 43:42–58.[PubMed]
53. De Felice, M., C. Squires, and M. Levinthal. 1978. A comparative study of the acetohydroxy acid synthase isoenzymes of Escherichia coli K-12. Biochim. Biophys. Acta 541:9-17.
54. Dewolf, W., Jr., and I. H. Segel. 2000. Simplified velocity equations for characterizing the partial inhibition or nonessential activation of bireactant enzymes. J. Enzyme Inhib. 15:311–333.[PubMed]
55. Drlica, K. 1992. Control of bacterial DNA supercoiling. Mol. Microbiol. 6:425–433.[PubMed] [CrossRef]
56. Drlica, K., and X. Zhao. 1997. DNA gyrase, topoisomerase IV, and the 4-quinolones. Microbiol. Mol. Biol. Rev. 61:377–392.[PubMed]
57. Eoyang, L., and P. M. Silverman. 1986. Role of small subunit (IlvN polypeptide) of acetohydroxyacid synthase I from Escherichia coli K-12 in sensitivity of the enzyme to valine inhibition. J. Bacteriol. 166:901–904.[PubMed]
58. Epelbaum, S., R. A. LaRossa, T. K. VanDyk, T. Elkayam, D. M. Chipman, and Z. Barak. 1998. Branched-chain amino acid biosynthesis in Salmonella typhimurium: a quantitative analysis. J. Bacteriol. 180:4056–4067.[PubMed]
59. Estrem, S. T., T. Gaal, W. Ross, and R. L. Gourse. 1998. Identification of an UP element consensus sequence for bacterial promoters. Proc. Natl. Acad. Sci. USA 95:9761–9766.[PubMed] [CrossRef]
60. Falk, S., A. Guay, C. Chenu, S. D. Patil, and A. Berteloot. 1998. Reduction of an eight-state mechanism of cotransport to a six-state model using a new computer program. Biophys. J. 74:816–830.[PubMed] [CrossRef]
61. Fang, M., and H. Y. Wu. 1998. A promoter relay mechanism for sequential gene activation. J. Bacteriol. 180:626–633.[PubMed]
62. Fang, M., and H. Y. Wu. 1998. Suppression of leu-500 mutation in topA+ Salmonella typhimurium strains. The promoter relay at work. J. Biol. Chem. 273:29929–29934.[PubMed] [CrossRef]
63. Fayat, G., J. F. Mayaux, C. Sacerdot, M. Fromant, M. Springer, M. Grunberg-Manago, and S. Blanquet. 1983. Escherichia coli phenylalanyl-tRNA synthetase operon region. Evidence for an attenuation mechanism. Identification of the gene for the ribosomal protein L20. J. Mol. Biol. 171:239–261.[PubMed] [CrossRef]
64. Freundlich, M. 1963. Multivalent repression in the biosynthesis of threonine in Salmonella typhimurium and Escherichia coli. Biochem. Biophys. Res. Commun. 10:277–282.[PubMed] [CrossRef]
65. Friden, P., T. Newman, and M. Freundlich. 1982. Nucleotide sequence of the ilvB promoter-regulatory region: a biosynthetic operon controlled by attenuation and cyclic AMP. Proc. Natl. Acad. Sci. USA 79:6156–6160.[PubMed] [CrossRef]
66. Friden, P., P. Tsui, K. Okamoto, and M. Freundlich. 1984. Interaction of cyclic AMP receptor protein with the ilvB biosynthetic operon in E. coli. Nucleic Acids Res. 12:8145–8160.[PubMed] [CrossRef]
67. Friden, P., K. Voelkel, R. Sternglanz, and M. Freundlich. 1984. Reduced expression of the isoleucine and valine enzymes in integration host factor mutants of Escherichia coli. J. Mol. Biol. 172:573–579.[PubMed] [CrossRef]
68. Friedman, D. I., E. J. Olson, D. Carver, and M. Gellert. 1984. Synergistic effect of himA and gyrB mutations: evidence that him functions control expression of ilv and xyl genes. J. Bacteriol. 157:484–489.[PubMed]
69. Gallagher, D. T., D. Chinchilla, H. Lau, and E. Eisenstein. 2004. Local and global control mechanisms in allosteric threonine deaminase. Methods Enzymol. 380:85–106.[PubMed] [CrossRef]
70. Gallagher, D. T., E. Eisenstein, K. E. Fisher, J. Zondlo, D. Chinchilla, H. D. Yu, J. Dill, E. Winborne, K. Ducote, G. Xiao, and G. L. Gilliland. 1998. Polymorphous crystallization and diffraction of threonine deaminase from Escherichia coli. Acta Crystallogr. D Biol. Crystallogr. 54:467–469.[PubMed] [CrossRef]
71. Gallagher, D. T., G. L. Gilliland, G. Xiao, J. Zondlo, K. E. Fisher, D. Chinchilla, and E. Eisenstein. 1998. Structure and control of pyridoxal phosphate dependent allosteric threonine deaminase. Structure 6:465–475.[PubMed] [CrossRef]
72. Gavini, N., and B. E. Davidson. 1991. Regulation of pheA expression by the pheR product in Escherichia coli is mediated through attenuation of transcription. J. Biol. Chem. 266:7750–7753.[PubMed]
73. Gavini, N., and L. Pulakat. 1991. Role of ribosome release in the basal level of expression of the Escherichia coli gene pheA. J. Gen. Microbiol. 137:679–684.[PubMed]
74. Gavini, N., and L. Pulakat. 1991. Role of translation of the pheA leader peptide coding region in attenuation regulation of the Escherichia coli pheA gene. J. Bacteriol. 173:4904–4907.[PubMed]
75. Gollnick, P., and P. Babitzke. 2002. Transcription attenuation. Biochim. Biophys. Acta 1577:240–250.[PubMed]
76. Gollop, N., D. M. Chipman, and Z. Barak. 1983. Inhibition of acetohydroxy acid synthase by leucine. Biochim. Biophys. Acta 748:34–39.[PubMed]
77. Gollop, N., B. Damri, Z. Barak, and D. M. Chipman. 1989. Kinetics and mechanism of acetohydroxy acid synthase isozyme III from Escherichia coli. Biochemistry 28:6310–6317.[PubMed] [CrossRef]
78. Gowrishankar, J., and J. Pittard. 1982. Regulation of phenylalanine biosynthesis in Escherichia coli K-12: control of transcription of the pheA operon. J. Bacteriol. 150:1130–1137.[PubMed]
79. Green, J. B. 1989. Pyruvate decarboxylase is like acetolactate synthase (ILV2) and not like the pyruvate dehydrogenase E1 subunit. FEBS Lett. 246:1–5.[PubMed] [CrossRef]
80. Hashiguchi, K., H. Takesada, E. Suzuki, and H. Matsui. 1999. Construction of an L-isoleucine overproducing strain of Escherichia coli K-12. Biosci. Biotechnol. Biochem. 63:672–679.[PubMed] [CrossRef]
81. Hatfield, G. W. 1968. The regulation of L-threonine deaminase in Bacillus subtilis by repression and endproduct inhibition. Ph.D. dissertation. Purdue University, Lafayette, Ind.
82. Hatfield, G. W. 1993. A two-ribosome model for attenuation, p. 1–22. In J. Ilan (ed.), Translational Regulation of Gene Expression 2. Plenum Press, New York, N.Y.
83. Hatfield, G. W., and C. J. Benham. 2002. DNA topology-mediated control of global gene expression in Escherichia coli. Annu. Rev. Genet. 36:175–203.[PubMed] [CrossRef]
84. Hatfield, G. W., and H. E. Umbarger. 1970. Threonine deaminase from Bacillus subtilis. II. The steady state kinetic properties. J. Biol. Chem. 245:1742–1747.[PubMed]
85. Hauser, C. A., and G. W. Hatfield. 1983. Nucleotide sequence of the ilvB multivalent attenuator region of Escherichia coli K12. Nucleic Acids Res. 11:127–139.[PubMed] [CrossRef]
86. Hauser, C. A., J. A. Sharp, L. K. Hatfield, and G. W. Hatfield. 1985. Pausing of RNA polymerase during in vitro transcription through the ilvB and ilvGEDA attenuator regions of Escherichia coli K12. J. Biol. Chem. 260:1765–1770.[PubMed]
87. Henikoff, S., G. W. Haughn, J. M. Calvo, and J. C. Wallace. 1988. A large family of bacterial activator proteins. Proc. Natl. Acad. Sci. USA 85:6602–6606.[PubMed] [CrossRef]
88. Henkin, T. M., and C. Yanofsky. 2002. Regulation by transcription attenuation in bacteria: how RNA provides instructions for transcription termination/antitermination decisions. Bioessays 24:700–707.[PubMed] [CrossRef]
89. Herring, P. A., B. L. McKnight, and J. H. Jackson. 1995. Channeling behavior and activity models for Escherichia coli K-12 acetohydroxy acid synthases at physiological substrate levels. Biochem. Biophys. Res. Commun. 207:48–54.[PubMed] [CrossRef]
90. Higgins, C. F., C. J. Dorman, D. A. Stirling, L. Waddell, I. R. Booth, G. May, and E. Bremer. 1988. A physiological role for DNA supercoiling in the osmotic regulation of gene expression in S. typhimurium and E. coli. Cell 52:569–584.[PubMed] [CrossRef]
91. Hsieh, L. S., J. Rouviere-Yaniv, and K. Drlica. 1991. Bacterial DNA supercoiling and [ATP]/[ADP] ratio: changes associated with salt shock. J. Bacteriol. 173:3914–3917.[PubMed]
92. Htun, H., and J. E. Dahlberg. 1989. Topology and formation of triple-stranded H-DNA. Science 243:1571–1576.[PubMed] [CrossRef]
93. Jackson, J. H., P. A. Herring, E. B. Patterson, and J. M. Blatt. 1993. A mechanism for valine-resistant growth of Escherichia coli K-12 supported by the valine-sensitive acetohydroxy acid synthase IV activity from ilvJ662. Biochimie 75:759–765.[PubMed] [CrossRef]
94. Jafri, S., S. Chen, and J. M. Calvo. 2002. ilvIH operon expression in Escherichia coli requires Lrp binding to two distinct regions of DNA. J. Bacteriol. 184:5293–5300.[PubMed] [CrossRef]
95. Jensen, P. R., L. Loman, B. Petra, C. van der Weijden, and H. V. Westerhoff. 1995. Energy buffering of DNA structure fails when Escherichia coli runs out of substrate. J. Bacteriol. 177:3420–3426.[PubMed]
96. Johnson, D. I., and R. L. Somerville. 1983. Evidence that repression mechanisms can exert control over the thr, leu, and ilv operons of Escherichia coli K-12. J. Bacteriol. 155:49–55.[PubMed]
97. Jones, P. G., and M. Inouye. 1994. The cold-shock response—a hot topic. Mol. Microbiol. 11:811–818.[PubMed] [CrossRef]
98. Jones, P. G., R. A. VanBogelen, and F. C. Neidhardt. 1987. Induction of proteins in response to low temperature in Escherichia coli. J. Bacteriol. 169:2092–2095.[PubMed]
99. Kahru, A., and R. Vilu. 1983. On characterization of the growth of Escherichia coli in batch culture. Arch. Microbiol. 135:12–15.[PubMed] [CrossRef]
100. Keseler, I. M., J. Collado-Vides, S. Gama-Castro, J. Ingraham, S. Paley, I. T. Paulsen, M. Peralta-Gil, and P. D. Karp. 2005. EcoCyc: a comprehensive database resource for Escherichia coli. Nucleic Acids Res. 33:D334–D337.[PubMed] [CrossRef]
101. Kohlhaw, G., T. R. Leary, and H. E. Umbarger. 1969. Alpha-isopropylmalate synthase from Salmonella typhimurium. Purification and properties. J. Biol. Chem. 244:2218–2225.[PubMed]
102. Kowalski, D., D. A. Natale, and M. J. Eddy. 1988. Stable DNA unwinding, not "breathing," accounts for single-strand-specific nuclease hypersensitivity of specific A+T-rich sequences. Proc. Natl. Acad. Sci. USA 85:9464–9468.[PubMed] [CrossRef]
103. Kusano, S., Q. Ding, N. Fujita, and A. Ishihama. 1996. Promoter selectivity of Escherichia coli RNA polymerase E sigma 70 and E sigma 38 holoenzymes. Effect of DNA supercoiling. J. Biol. Chem. 271:1998–2004.[PubMed] [CrossRef]
104. Landick, R., C. L. Turnbough, Jr., and C. Yanofsky. 1996. Transcription attenuation, p. 1263–1286. In F. C. Neidhardt, R. Curtiss III, J. Ingraham, E. C. C. Lin, K. B. Low, B. Magasanik, W. S. Reznikoff, M. Riley, M. Schaechter, and H. E. Umbarger (ed.), Escherichia coli and Salmonella: Cellular and Molecular Biology, 2nd ed., vol. 1. ASM Press, Washington, D.C.
105. LaRossa, R. A., T. K. Van Dyk, and D. R. Smulski. 1990. A need for metabolic insulation: lessons from sufonylurea genetics, p. 109–121. In Z. Barak, D. M. Chipman, and J. V. Schloss (ed.), Biosynthesis of Branched Chain Amino Acids. VCH Balaban Publishers, New York, N.Y.
106. Lawther, R. P., D. H. Calhoun, C. W. Adams, C. A. Hauser, J. Gray, and G. W. Hatfield. 1981. Molecular basis of valine resistance in Escherichia coli K-12. Proc. Natl. Acad. Sci. USA 78:922–925.[PubMed] [CrossRef]
107. Lawther, R. P., D. H. Calhoun, J. Gray, C. W. Adams, C. A. Hauser, and G. W. Hatfield. 1982. DNA sequence fine-structure analysis of ilvG (IlvG+) mutations of Escherichia coli K-12. J. Bacteriol. 149:294–298.[PubMed]
108. Leary, T. R., and G. Kohlhaw. 1970. Dissociation of alpha-isopropylmalate synthase from Salmonella typhimurium by its feedback inhibitor leucine. Biochem. Biophys. Res. Commun. 39:494–501.[PubMed] [CrossRef]
109. Lilley, D. M. 1980. The inverted repeat as a recognizable structural feature in supercoiled DNA molecules. Proc. Natl. Acad. Sci. USA 77:6468–6472.[PubMed] [CrossRef]
110. Liu, L. F., and J. C. Wang. 1987. Supercoiling of the DNA template during transcription. Proc. Natl. Acad. Sci. USA 84:7024–7027.[PubMed] [CrossRef]
111. Liu, X., and P. De Wulf. 2004. Probing the ArcA-P modulon of Escherichia coli by whole genome transcriptional analysis and sequence recognition profiling. J. Biol. Chem. 279:12588–12597.[PubMed] [CrossRef]
112. Liu, Y., V. Bondarenko, A. Ninfa, and V. M. Studitsky. 2001. DNA supercoiling allows enhancer action over a large distance. Proc. Natl. Acad. Sci. USA 98:14883–14888.[PubMed] [CrossRef]
113. Lynn, S. P., W. S. Burton, T. J. Donohue, R. M. Gould, R. I. Gumport, and J. F. Gardner. 1987. Specificity of the attenuation response of the threonine operon of Escherichia coli is determined by the threonine and isoleucine codons in the leader transcript. J. Mol. Biol. 194:59–69.[PubMed] [CrossRef]
114. Lynn, S. P., J. F. Gardner, and W. S. Reznikoff. 1982. Attenuation regulation in the thr operon of Escherichia coli K-12: molecular cloning and transcription of the controlling region. J. Bacteriol. 152:363–371.[PubMed]
114a. Margolin, P., L. Zumstein, R. Sternglanz, and J. C. Wang. 1985. The Escherichia coli supX locus is topA, the structural gene for DNA topoisomerase I. Proc. Natl. Acad. Sci. USA 82:5437–5441.[PubMed] [CrossRef]
115. Merino, E., and C. Yanofsky. 2005. Transcription attenuation: a highly conserved regulatory strategy used by bacteria. Trends Genet. 21:260–264.[PubMed] [CrossRef]
116. Mojica, F. J., and C. F. Higgins. 1996. Localized domains of DNA supercoiling: topological coupling between promoters. Mol. Microbiol. 22:919–928.[PubMed] [CrossRef]
117. Monod, J., J. Wyman, and J. P. Changeux. 1965. On the nature of allosteric transitions: a plausible model. J. Mol. Biol. 12:88–118.[PubMed] [CrossRef]
118. Muskhelishvili, G., M. Buckle, H. Heumann, R. Kahmann, and A. A. Travers. 1997. FIS activates sequential steps during transcription initiation at a stable RNA promoter. EMBO J. 16:3655–3665.[PubMed] [CrossRef]
119. Najdi, T. S., C. R. Yang, B. E. Shapiro, E. D. Mjolsness, and G. W. Hatfield. Application of a generalized MWC model for the mathematical simulation of metabolic pathways regulated by allosteric enzymes. J. Bioinform. Comput. Biol., in press.
120. Neidhardt, F. C., and M. A. Savageau. 1996. Regulation beyond the operon, p. 1310–1324. In F. C. Neidhardt, R. Curtiss III, J. L. Ingraham, E. C. C. Lin, K. B. Low, B. Magasanik, W. S. Reznikoff, M. Riley, M. Schaechter, and H. E. Umbarger (ed.), Escherichia coli and Salmonella: Cellular and Molecular Biology, 2nd ed., vol. 1. ASM Press, Washington, D.C.
121. Newlands, J. T., C. A. Josaitis, W. Ross, and R. L. Gourse. 1992. Both fis-dependent and factor-independent upstream activation of the rrnB P1 promoter are face of the helix dependent. Nucleic Acids Res. 20:719–726.[PubMed] [CrossRef]
122. O'Neill, J. P., and M. Freundlich. 1972. Two forms of biosynthetic acetohydroxy acid synthetase in Salmonella typhimurium. Biochem. Biophys. Res. Commun. 48:437–443.[PubMed] [CrossRef]
123. Opel, M. L., K. A. Aeling, W. M. Holmes, R. C. Johnson, C. J. Benham, and G. W. Hatfield. 2004. Activation of transcription initiation from a stable RNA promoter by a Fis protein-mediated DNA structural transmission mechanism. Mol. Microbiol. 53:665–674.[PubMed] [CrossRef]
124. Opel, M. L., S. M. Arfin, and G. W. Hatfield. 2001. The effects of DNA supercoiling on the expression of operons of the ilv regulon of Escherichia coli suggest a physiological rationale for divergently transcribed operons. Mol. Microbiol. 39:1109–1115.[PubMed] [CrossRef]
125. Opel, M. L., and G. W. Hatfield. 2001. DNA supercoiling-dependent transcriptional coupling between the divergently transcribed promoters of the ilvYC operon of Escherichia coli is proportional to promoter strengths and transcript lengths. Mol. Microbiol. 39:191–198.[PubMed] [CrossRef]
126. Pages, D., and R. H. Buckingham. 1990. Mutants of pheV in Escherichia coli affecting control by attenuation of the pheS, T and pheA operons. Two distinct mechanisms for de-attenuation. J. Mol. Biol. 216:17–24.[PubMed] [CrossRef]
127. Panina, E. M., A. G. Vitreschak, A. A. Mironov, and M. S. Gelfand. 2001. Regulation of aromatic amino acid biosynthesis in gamma-proteobacteria. J. Mol. Microbiol. Biotechnol. 3:529–543.[PubMed]
128. Parekh, B. S., and G. W. Hatfield. 1996. Transcriptional activation by protein-induced DNA bending: evidence for a DNA structural transmission model. Proc. Natl. Acad. Sci. USA 93:1173–1177.[PubMed] [CrossRef]
129. Parekh, B. S., S. D. Sheridan, and G. W. Hatfield. 1996. Effects of integration host factor and DNA supercoiling on transcription from the ilvPG promoter of Escherichia coli. J. Biol. Chem. 271:20258–20264.[PubMed] [CrossRef]
130. Parsot, C., I. Saint-Girons, and P. Cossart. 1982. DNA sequence change of a deletion mutation abolishing attenuation control of the threonine operon of E. coli K12. Mol. Gen. Genet. 188:455–458.[PubMed] [CrossRef]
131. Platko, J. V., D. A. Willins, and J. M. Calvo. 1990. The ilvIH operon of Escherichia coli is positively regulated. J. Bacteriol. 172:4563–4570.[PubMed]
132. Pokholok, D. K., M. Redlak, C. L. Turnbough, Jr., S. Dylla, and W. M. Holmes. 1999. Multiple mechanisms are used for growth rate and stringent control of leuV transcriptional initiation in Escherichia coli. J. Bacteriol. 181:5771–5782.[PubMed]
133. Primerano, D. A., and R. O. Burns. 1982. Metabolic basis for the isoleucine, pantothenate or methionine requirement of ilvG strains of Salmonella typhimurium. J. Bacteriol. 150:1202–1211.[PubMed]
133a. Pruss, G. J., and K. Drlica. 1985. DNA supercoiling and suppression of the leu-500 promoter mutation. J. Bacteriol. 164:947–949.[PubMed]
134. Quay, S. C., T. E. Dick, and D. L. Oxender. 1977. Role of transport systems in amino acid metabolism: leucine toxicity and the branched-chain amino acid transport systems. J. Bacteriol. 129:1257–1265.[PubMed]
135. Rahmouni, A. R., and R. D. Wells. 1989. Stabilization of Z DNA in vivo by localized supercoiling. Science 246:358–363.[PubMed] [CrossRef]
136. Remington, J. A. 1979. Attenuation as a general mechanism for the regulation of differential gene transcription in eukaryoties. FEBS Lett. 100:225–229.[PubMed] [CrossRef]
137. Rhee, K. Y., M. Opel, E. Ito, S. Hung, S. M. Arfin, and G. W. Hatfield. 1999. Transcriptional coupling between the divergent promoters of a prototypic LysR-type regulatory system, the ilvYC operon of Escherichia coli. Proc. Natl. Acad. Sci. USA 96:14294–14299.[PubMed] [CrossRef]
138. Rhee, K. Y., B. S. Parekh, and G. W. Hatfield. 1996. Leucine-responsive regulatory protein-DNA interactions in the leader region of the ilvGMEDA operon of Escherichia coli. J. Biol. Chem. 271:26499–26507.[PubMed] [CrossRef]
139. Ricca, E., D. A. Aker, and J. M. Calvo. 1989. A protein that binds to the regulatory region of the Escherichia coli ilvIH operon. J. Bacteriol. 171:1658–1664.[PubMed]
140. Ricca, E., C. T. Lago, M. Sacco, and M. De Felice. 1991. Absence of acetohydroxy acid synthase III in Salmonella typhimurium is due to early termination of translation within the ilvl gene. Mol. Microbiol. 5:1741–1743.[PubMed] [CrossRef]
141. Riccio, A., C. B. Bruni, M. Rosenberg, M. Gottesman, K. McKenney, and F. Blasi. 1985. Regulation of single and multicopy his operons in Escherichia coli. J. Bacteriol. 163:1172–1179.[PubMed]
142. Rodriguez, R. L., and R. W. West, Jr. 1984. Histidine operon control region of Klebsiella pneumoniae: analysis with an Escherichia coli promoter-probe plasmid vector. J. Bacteriol. 157:764–771.[PubMed]
143. Roesser, J. R., Y. Nakamura, and C. Yanofsky. 1989. Regulation of basal level expression of the tryptophan operon of Escherichia coli. J. Biol. Chem. 264:12284–12288.[PubMed]
144. Ross, W., K. K. Gosink, J. Salomon, K. Igarashi, C. Zou, A. Ishihama, K. Severinov, and R. L. Gourse. 1993. A third recognition element in bacterial promoters: DNA binding by the alpha subunit of RNA polymerase. Science 262:1407–1413.[PubMed] [CrossRef]
145. Ross, W., J. Salomon, W. M. Holmes, and R. L. Gourse. 1999. Activation of Escherichia coli leuV transcription by FIS. J. Bacteriol. 181:3864–3868.[PubMed]
146. Salmon, K., S. P. Hung, K. Mekjian, P. Baldi, G. W. Hatfield, and R. P. Gunsalus. 2003. Global gene expression profiling in Escherichia coli K12. The effects of oxygen availability and FNR. J. Biol. Chem. 278:29837–29855.[PubMed] [CrossRef]
147. Salmon, K. A., S. P. Hung, N. R. Steffen, R. Krupp, P. Baldi, G. W. Hatfield, and R. P. Gunsalus. 2005. Global gene expression profiling in Escherichia coli K12: effects of oxygen availability and ArcA. J. Biol. Chem. 280:15084–15096.[PubMed] [CrossRef]
148. Sander, P., W. Langert, and K. Mueller. 1993. Mechanisms of upstream activation of the rrnD promoter P1 of Escherichia coli. J. Biol. Chem. 268:16907–16916.[PubMed]
149. Schell, M. A. 1993. Molecular biology of the LysR family of transcriptional regulators. Annu. Rev. Microbiol. 47:597–626.[PubMed] [CrossRef]
150. Scolnick, E., R. Tompkins, T. Caskey, and M. Nirenberg. 1968. Release factors differing in specificity for terminator codons. Proc. Natl. Acad. Sci. USA 61:768–774.[PubMed] [CrossRef]
151. Shapiro, B. E., A. Levchenko, E. M. Meyerowitz, B. J. Wold, and E. D. Mjolsness. 2003. Cellerator: extending a computer algebra system to include biochemical arrows for signal transduction simulations. Bioinformatics 19:677–678.[PubMed] [CrossRef]
152. Sheridan, S. D., C. J. Benham, and G. W. Hatfield. 1998. Activation of gene expression by a novel DNA structural transmission mechanism that requires supercoiling-induced DNA duplex destabilization in an upstream activating sequence. J. Biol. Chem. 273:21298–21308.[PubMed] [CrossRef]
153. Sheridan, S. D., C. J. Benham, and G. W. Hatfield. 1999. Inhibition of DNA supercoiling-dependent transcriptional activation by a distant B-DNA to Z-DNA transition. J. Biol. Chem. 274:8169–8174.[PubMed] [CrossRef]
154. Sheridan, S. D., M. L. Opel, and G. W. Hatfield. 2001. Activation and repression of transcription initiation by a distant DNA structural transition. Mol. Microbiol. 40:684–690.[PubMed] [CrossRef]
155. Smith, M. W., and F. C. Neidhardt. 1983. Proteins induced by aerobiosis in Escherichia coli. J. Bacteriol. 154:344–350.[PubMed]
156. Soper, T. S., J. Doellgast, and G. B. Kohlhaw. 1976. Mechanism of feedback inhibition by leucine. Purification and properties of a feedback-resistant alpha-isopropylmalate synthase. Arch. Biochem. Biophys. 173:362–374.[PubMed] [CrossRef]
157. Spencer, C. A., and M. Groudine. 1990. Transcription elongation and eukaryotic gene regulation. Oncogene 5:777–785.[PubMed]
158. Springer, M., M. Trudel, M. Graffe, J. Plumbridge, G. Fayat, J. F. Mayaux, C. Sacerdot, S. Blanquet, and M. Grunberg-Manago. 1983. Escherichia coli phenylalanyl-tRNA synthetase operon is controlled by attenuation in vivo. J. Mol. Biol. 171:263–279.[PubMed] [CrossRef]
159. Squires, C. H., M. De Felice, C. T. Lago, and J. M. Calvo. 1983. IlvHI locus of Salmonella typhimurium. J. Bacteriol. 154:1054–1063.[PubMed]
160. Steck, T. R., R. J. Franco, J. Y. Wang, and K. Drlica. 1993. Topoisomerase mutations affect the relative abundance of many Escherichia coli proteins. Mol. Microbiol. 10:473–481.[PubMed] [CrossRef]
161. Sutton, A., and M. Freundlich. 1980. Regulation of cyclic AMP of the ilvB-encoded biosynthetic acetohydroxy acid synthase in Escherichia coli K-12. Mol. Gen. Genet. 178:179–183.[PubMed] [CrossRef]
162. Tatum, E. L. 1946. Biochemical mutations in bacteria. Cold Spring Harbor Symp. Quant. Biol. 11:278–283.
163. Tchetina, E., and E. B. Newman. 1995. Identification of Lrp-regulated genes by inverse PCR and sequencing: regulation of two mal operons of Escherichia coli by leucine-responsive regulatory protein. J. Bacteriol. 177:2679–2683.[PubMed]
164. Tse, M. T., and J. V. Schloss. 1993. The oxygenase reaction of acetolactate synthase. Biochemistry 32:10398–10403.[PubMed] [CrossRef]
165. Tsui, P., and M. Freundlich. 1990. Integration host factor bends the DNA in the Escherichia coli ilvBN promoter region. Mol. Gen. Genet. 223:349–352.[PubMed] [CrossRef]
166. Umbarger, H. E. 1996. Biosynthesis of the branched-chain amino acids, p. 442–457. In F. C. Neidhardt, R. Curtiss III, J. L. Ingraham, E. C. C. Lin, K. B. Low, B. Magasanik, W. S. Reznikoff, M. Riley, M. Schaechter, and H. E. Umbarger (ed.), Escherichia coli and Salmonella: Cellular and Molecular Biology, 2nd ed., vol. 1. ASM Press, Washington, D.C.
167. Umbarger, H. E. 1956. Evidence for a negative-feedback mechanism in the biosynthesis of isoleucine. Science 123:848. [CrossRef]
168. Umbarger, H. E. 1961. Feedback control by endproduct inhibition. Cold Spring Harbor Symp. Quant. Biol. 26:301–312.[PubMed]
169. Umbarger, H. E. 1973. Threonine deaminases. Adv. Enzymol. Relat. Areas Mol. Biol. 37:349–395.[PubMed] [CrossRef]
170. Umbarger, H. E., and B. Brown. 1955. Isoleucine and valine metabolism in Escherichia coli. V. Antagonism between isoleucine and valine. J. Bacteriol. 70:241–248.[PubMed]
171. Umbarger, H. E., and B. Brown. 1956. Threonine deamination in Escherichia coli. I. D- and L-threonine deaminase activities of cell-free extracts. J. Bacteriol. 71:443–449.[PubMed]
172. Umbarger, H. E., and B. Brown. 1957. Threonine deamination in Escherichia coli. II. Evidence for two L-threonine deaminases. J. Bacteriol. 73:105–112.[PubMed]
173. Urios, A., G. Herrera, V. Aleixandre, and M. Blanco. 1990. Expression of the recA gene is reduced in Escherichia coli topoisomerase I mutants. Mutat. Res. 243:267–272.[PubMed] [CrossRef]
174. Vitreschak, A. G., E. V. Lyubetskaya, M. A. Shirshin, M. S. Gelfand, and V. A. Lyubetsky. 2004. Attenuation regulation of amino acid biosynthetic operons in proteobacteria: comparative genomics analysis. FEMS Microbiol. Lett. 234:357–370.[PubMed] [CrossRef]
175. Wang, J. Y., and M. Syvanen. 1992. DNA twist as a transcriptional sensor for environmental changes. Mol. Microbiol. 6:1861–1866.[PubMed] [CrossRef]
176. Wang, Q., and J. M. Calvo. 1993. Lrp, a global regulatory protein of Escherichia coli, binds co-operatively to multiple sites and activates transcription of ilvIH. J. Mol. Biol. 229:306–318.[PubMed] [CrossRef]
177. Watson, M. D., J. Wild, and H. E. Umbarger. 1979. Positive control of ilvC expression in Escherichia coli K-12; identification and mapping of regulatory gene ilvY. J. Bacteriol. 139:1014–1020.[PubMed]
178. Weinberg, R. A., and R. O. Burns. 1984. Regulation of expression of the ilvB operon in Salmonella typhimurium. J. Bacteriol. 160:833–841.[PubMed]
179. Weinstock, O., C. Sella, D. M. Chipman, and Z. Barak. 1992. Properties of subcloned subunits of bacterial acetohydroxy acid synthases. J. Bacteriol. 174:5560–5566.[PubMed]
180. Wek, R. C., and G. W. Hatfield. 1986. Examination of the internal promoter, PE, in the ilvGMEDA operon of E. coli K-12. Nucleic Acids Res. 14:2763–2777.[PubMed] [CrossRef]
181. Wek, R. C., and G. W. Hatfield. 1986. Nucleotide sequence and in vivo expression of the ilvY and ilvC genes in Escherichia coli K12. Transcription from divergent overlapping promoters. J. Biol. Chem. 261:2441–2450.[PubMed]
182. Wek, R. C., and G. W. Hatfield. 1988. Transcriptional activation at adjacent operators in the divergent-overlapping ilvY and ilvC promoters of Escherichia coli. J. Mol. Biol. 203:643–663.[PubMed] [CrossRef]
183. Westerhoff, H. V., M. H. O'Dea, A. Maxwell, and M. Gellert. 1988. DNA supercoiling by DNA gyrase. A static head analysis. Cell Biophys. 12:157–181.[PubMed]
184. Westerhoff, H. V., and M. van Workum. 1990. Control of DNA structure and gene expression. Biomed. Biochim. Acta 49:839–853.[PubMed]
185. Williams, A. L. 1983. Regulation of acetohydroxy acid synthase activities in adenyl cyclase-deficient strains of Escherichia coli K-12. Mol. Gen. Genet. 191:353–357.[PubMed] [CrossRef]
186. Winkelman, J. W., and G. W. Hatfield. 1990. Characterization of the integration host factor binding site in the ilvPG1 promoter region of the ilvGMEDA operon of Escherichia coli. J. Biol. Chem. 265:10055–10060.[PubMed]
187. Winkler, M. E. 1978. Expression of the histidine operon in rho mutants of Escherichia coli. J. Bacteriol. 135:721–725.[PubMed]
188. Winkler, M. E., D. J. Roth, and P. E. Hartman. 1978. Promoter- and attenuator-related metabolic regulation of the Salmonella typhimurium histidine operon. J. Bacteriol. 133:830–843.[PubMed]
189. Wu, H. Y., and M. Fang. 2003. DNA supercoiling and transcription control: a model from the study of suppression of the leu-500 mutation in Salmonella typhimurium topA- strains. Prog. Nucleic Acids Res. Mol. Biol. 73:43–68.[PubMed] [CrossRef]
190. Wu, H. Y., J. Tan, and M. Fang. 1995. Long-range interaction between two promoters: activation of the leu-500 promoter by a distant upstream promoter. Cell 82:445–451.[PubMed] [CrossRef]
191. Wu, T. H., D. L. Wood, P. L. Stein, and M. M. Comer. 1984. Transcription of a gene cluster coding for two aminoacyl-tRNA synthetases and an initiation factor in Escherichia coli. J. Mol. Biol. 173:177–209.[PubMed] [CrossRef]
192. Yang, C. R., B. E. Shapiro, S. P. Hung, E. D. Mjolsness, and G. W. Hatfield. 2005. A mathematical model for the branched chain amino acid biosynthetic pathways of Escherichia coli K12. J. Biol. Chem. 280:11224–11232. [CrossRef]
193. Yang, C. R., B. E. Shapiro, E. D. Mjolsness, and G. W. Hatfield. 2005. An enzyme mechanism language for the mathematical modeling of metabolic pathways. Bioinformatics 21:774–780.[PubMed] [CrossRef]
194. Yanofsky, C. 1981. Attenuation in the control of expression of bacterial operons. Nature 289:751–758.[PubMed] [CrossRef]
195. Yanofsky, C. 1988. Transcription attenuation. J. Biol. Chem. 263:609–612.[PubMed]
196. Yanofsky, C. 2000. Transcription attenuation: once viewed as a novel regulatory strategy. J. Bacteriol. 182:1–8.[PubMed] [CrossRef]
197. Yanofsky, C., K. V. Konan, and J. P. Sarsero. 1996. Some novel transcription attenuation mechanisms used by bacteria. Biochimie 78:1017–1024.[PubMed] [CrossRef]
198. Zacharias, M., G. Theissen, C. Bradaczek, and R. Wagner. 1991. Analysis of sequence elements important for the synthesis and control of ribosomal RNA in E coli. Biochimie 73:699–712.[PubMed] [CrossRef]
Module3.6.1.5
