Evolution of Genes and Enzymes of Tryptophan Biosynthesis
Chapter
143
BRIAN P. NICHOLS
The tryptophan biosynthetic (trp) genes rank second to the 16S rRNA genes as a paradigm for microbial molecular evolution. trp genes have been studied for many years, and their organization and regulation have been studied in diverse bacterial and fungal species. Since the advent of rapid DNA sequence determination, the nucleotide sequences of a number of the trp genes have been elucidated, and sequences from diverse organisms have been compared. Several reviews have summarized much of the available information concerning the variation found in amino acid sequence comparisons of trp enzymes, as well as gene organizations, regulatory processes, and enzymatic reaction mechanisms (23, 24, 25, 26, 27, 72, 112, 114, 120, 121).
It is the intent of this review to collect the available sequence information on the trp enzymes and to apply some currently available and popularly used analytical tools to compare the relationships of the amino acid sequences for the seven enzymes of the tryptophan biosynthetic pathway (Fig. 1).
All of the data used in the preparation of this chapter were taken from the established databases, especially the GenBank and Swiss-Prot databases. Analysis of the relationships and cladistics of the amino acid sequences were done by using several available programs. Initial alignments were made with at least five "seed" sequences, using MultAlin 4.0 with the gap penalty set at 8 (22). Additional sequences were added manually, and finally gaps were removed or adjusted manually. Sequence alignments are not presented here in the interest of saving space but will be provided upon request. Cladistic analysis was performed by using PAUP 3.0 (103) and MacClade 3.0 (68). For consistency, the eucaryotic Saccharomyces trp sequences were defined as the outgroup, since all trp sequences are available from this organism.
As of April 1994, at least 12 complete sequences of each of the tryptophan biosynthetic genes had been determined from procaryotic sources. Only complete gene sequences have been used in this analysis. Additional sequences for some of the genes are available from fungal and plant species, but these have not been included in the present analysis. Table 1 shows the organisms, the genes sequenced, their GenBank accession numbers, and the citations in which the sequences appear. The sequences for all seven trp genes are known for eight bacterial species, including the three enteric species Escherichia coli, Salmonella typhimurium (official designation, Salmonella enterica serovar Typhimurium), and Vibrio parahaemolyticus, the low-G+C species Bacillus subtilis and Lactococcus lactis, the high-G+C species Brevibacterium lactofermentum, and the two euryarchaeota Methanobacterium thermoautotrophicum and Haloferax volcanii. Acinetobacter calcoaceticus lacks only the trpA sequence, and the two fluorescent pseudomonads Pseudomonas aeruginosa and P. putida lack only trpF sequences. An additional 14 bacterial species are represented by at least one and in most cases several trp gene sequences. To date, both major subdivisions of the domain Archaea are represented, as are the α, β,and γ subdivisions of the purple bacteria, spirochetes, both the high- and low-G+C gram-positive bacteria, and cyanobacteria.
Table 1Sequenced bacterial tryptophan genes |
A variety of tryptophan gene arrangements have been characterized in bacteria, and these are summarized in Table 2. Adjacent genes are represented by adjacent letters, clusters are separated by spaces, and fused genes are enclosed in parentheses. Genes whose locations are not known to be either clustered or separated are not included in Table 2. Clearly, there is a wide variety of organizations and clusterings of trp genes in bacteria, suggesting that there is no compelling selection for a specific operon structure. Similarly, the regulatory systems that have evolved to control different clusters or solitary genes are quite varied among the bacteria. Evidence demonstrating negative regulation (112), positive regulation (7, 19, 69), attenuation effected by translation (9, 29, 47, 61, 93, 112, 115), attenuation effected by RNA binding proteins (8, 96, 97), and no regulation at all has been gathered for various genes and gene clusters in the various organisms. The evolution of the regulatory systems will not be further considered here.
Table 2Organization of tryptophan genes in bacteria |
The tendency for clustering of the trp genes has led in a few instances to fusions between adjacent genes. While fusions are common among the fungi, only a limited number appear in procaryotes, and the types of fusions are not necessarily specific to the major cladistic groupings. For example, trpG is found fused with trpD in some of the members of the family Enterobacteriaceae, including E. coli and S. typhimurium, but is separate in others like Serratia marcescens and V. parahaemolyticus (29, 74, 79). The only other procaryotic trp(GD) fusion known is found in Thermotoga maritima (61). trpE is found fused with trpG in Rhizobium meliloti, the only known organism in which the genes encoding these two subunits of the same enzyme are fused (10). trpC and trpF are found fused in all of the enteric bacteria so far studied, and the trp(CF) fusion is also found in the high-G+C gram-positive organism Brevibacterium lactofermentum (20, 27, 52, 70).
Fusions between trp genes have not necessarily resulted in a functional cooperativity or channeling of substrates between enzymes activities. Only in the trp(EG) fusion can it be logically expected that a degree of cooperativity between the encoded polypeptide domains might exist, since these activities, when separate, are subunits of the same enzyme. Surprisingly, it is the fusion, rather than an evolution toward, and conservation of, a single polypeptide chain with both activities, that is the rare event. Individual activities of the remaining bifunctional enzymes resulting from gene fusions have been shown to remain independent (21, 62). There is little communication between the domains of the fused trp(GD) products, except for slight inhibition of the phosphoribosyltransferase activity by tryptophan when present in the anthranilate synthase complex (120). Similarly, the two domains of the fused trp(CF) products both fold and function independently of one another. In most cases, only a few nucleotide alterations are required to convert adjacent genes into a fused gene (27, 74, 87). Presumably, if there is not a negative selective pressure on the fusion product, then the fusion may persist without the necessity of evolving cooperative or coordinate activities. As has been concluded earlier, the significance of the appearance of various gene fusions may not be any greater than the significance of the differences in the organization and clustering of the genes (27).
Anthranilate synthase catalyzes the first step in the conversion of chorismate to tryptophan, in which chorismate is aminated and then aromatized to anthranilate (o-aminobenzoate), concomitant with the loss of the enol-pyruvyl group as pyruvate. Anthranilate synthase is generally composed of two dissimilar subunits. The larger subunit, termed component I, is encoded by trpE and converts chorismate and ammonia to anthranilate, while the smaller subunit, termed component II, is encoded by trpG and confers upon the enzyme the ability to use glutamine as the ammonia source (120, 121). It has recently been documented that the reaction catalyzed by S. typhimurium anthranilate synthase proceeds in two discrete steps (75). The first step is thought to involve attack by an "ammonia equivalent" at the 2 position of chorismate and the loss of the hydroxyl group at the 4 position to yield 2-amino-4-deoxychorismate (aminodeoxyisochorismate [ADIC]). The second step is the β elimination of the enol-pyruvyl moiety to form the aromatic product anthranilate. The ADIC intermediate remains enzyme bound, except in certain mutant anthranilate synthases in which its accumulation can be detected. Thus, anthranilate synthase can be thought of as a bifunctional enzyme containing ADIC synthase and ADIC lyase activities. The bifunctional nature of anthranilate synthase is particularly relevant because of its paralogous relationships to isochorismate synthases and 4-amino-4-deoxychorismate (ADC) synthase, each of which performs a reaction similar to the ADIC synthase reaction of anthranilate synthesis (41, 42, 46, 80, 84, 118), as shown in Fig. 2.
Phenazine-Specific Anthranilate Synthase.
In addition to the anthranilate synthase involved in tryptophan metabolism, P. aeruginosa expresses a second anthranilate synthase that is involved in secondary metabolism. The alternate anthranilate synthase is involved in phenazine biosynthesis, and the genes encoding it have been designated phnA and phnB (37). Although the phnAB anthranilate synthase is not involved in tryptophan synthesis in P. aeruginosa, it was discovered by the ability of phnA to complement a trpE mutation in E. coli (32). phnA and phnB are adjacent on the P. aeruginosa chromosome and are apparently expressed as an operon. The enzyme is presumed to function in a manner analogous to the anthranilate synthase of tryptophan biosynthesis.
Isochorismate Synthase.
There are two distinct isochorismate synthases in E. coli. One, encoded by entC, is involved in enterochelin synthesis and is regulated by iron availability (84, 107). entC is clustered with additional genes encoding enterochelin synthetic enzymes. The second, encoded by menF, is involved in the anaerobic synthesis of the respiratory cofactor menaquinone. menF lies near additional genes involved in menaquinone biosynthesis (R. Meganathan, personal communication). Isochorismate (2-hydroxy-4-deoxychorismate) synthase can be envisioned to function by using a reaction mechanism similar to that used by the ADIC synthase activity of anthranilate synthase, with the ammonia nucleophile being replaced by an hydroxyl anion.
p
-Aminobenzoate Synthase.
p-Aminobenzoate is synthesized by a pair of enzymes that catalyze two reactions analogous to the two reactions of anthranilate synthase. ADC synthase is encoded by pabA and pabB (42, 60) and is responsible for the amination of chorismate at the 4 position with concomitant loss of the hydroxyl group (46, 80, 118). The stereochemistry of the aminochorismate product is identical to that of chorismate (3). The mechanism of action of ADC synthase must be slightly different from that of anthranilate synthase, since the position of attack by the nucleophile is different. pabB encodes the chorismate amination subunit (component I) and is homologous with trpE (42, 43). pabA encodes the glutamine amidotransferase subunit (component II) of ADC synthase and is homologous to trpG (59, 60). The two glutamine amidotransferase functions appear to be identical, and in some species (A. calcoaceticus, Pseudomonas acidovorans, and B. subtilis), both ADIC synthase and ADC synthase use the same glutamine amidotransferase subunit (16, 57, 94). In E. coli and S. typhimurium, pabA and pabB are unlinked (53).
The second step in p-aminobenzoate synthesis is catalyzed by ADC lyase, encoded by pabC (45, 46, 80). ADC lyase performs a β-elimination/aromatization reaction to form the p-aminobenzoate product. On paper, the reaction appears to be identical to the ADIC lyase reaction of anthranilate synthase. However, ADC lyase shows no significant similarity to anthranilate synthase, and ADC lyase utilizes a pyridoxal cofactor (45). pabC is unlinked to either pabA or pabB.
Interestingly, a third lyase appears to perform a reaction mechanistically similar to the ADC and ADIC lyase reactions. Chorismate lyase (chorismate → p-hydroxybenzoate) catalyzes the first step in the synthesis of the aerobic respiratory cofactor ubiquinone and is encoded by ubiC (41, 67). Although chorismate lyase can complement a deficiency in ADC lyase, no amino acid similarity is seen among ADC, ADIC, and chorismate lyases (78).
The similarity among anthranilate synthase, ADC synthase, and isochorismate synthase is relatively high for the C-terminal halves of the sequences but is quite low toward the N-terminal halves. Nevertheless, the amino acid sequence similarity has been interpreted as reflecting a common ancestry for these enzymes (42, 43, 84). This interpretation is in keeping with the widely held model for the acquisition of new enzyme functions by gene duplication followed by sequence divergence (56). In this model, the duplication of a gene encoding an enzyme with broad substrate specificity would allow for sequence changes that would alter the specificity of one copy without loss of the original function. It can be envisioned, and has been shown in the laboratory, that such a mechanism can lead to the acquisition of novel metabolic functions as well as regulatory responses. While it is easy to imagine how one or a few changes in amino acid sequence near the substrate binding region might alter the specificity for substrate binding, it is more difficult to imagine (perhaps because our understanding is less sophisticated) how a few amino acid sequence changes might alter the chemical mechanism of transformation from substrate to a novel product.
Yet it appears that the ADIC synthase activities of anthranilate synthase, ADC synthase, and isochorismate synthase perform different types of reaction chemistry on the same substrate and retain sequence similarity reflective of common ancestry. On the other hand, the three lyase activities, which would appear to use the same β-elimination/aromatization chemistry on slightly different substrates, are dissimilar enzymes and appear to have dissimilar evolutionary origins. One might argue that substrate binding could be an overriding consideration in conservation of particular amino acid sequences, but if that were the case, then all enzymes using chorismate as a substrate would be expected to show similarity. The substrate binding rationale is not supported by the data. It may be more likely that the three synthases have an additional, as yet uncharacterized portion of their reaction chemistry in common. (Several such mechanisms are possible and have been proposed by Walsh et al. [107].) In the case of the β-elimination/aromatization reactions, which can proceed spontaneously under relatively mild conditions, it appears that the need for pathway compartmentalization may supersede the requirement for a very specifically evolved catalyst.
The reason for the high variability at the N termini of the anthranilate synthase component I sequences is not clear. Within about 70 residues of the N terminus of each sequence is found a highly conserved segment (consensus, -LLES-X10-S-) that is functionally involved in the feedback inhibition of anthranilate synthase activity by tryptophan (17, 44).
Logically, the tryptophan feedback site is not found in the P. aeruginosa phenazine-specific anthranilate synthase, but more surprisingly, it is also not present in the R. meliloti sequence, even though the enzyme has been shown to be inhibited by tryptophan (9). It may be that the fusion of the two subunits of anthranilate synthase has altered the requirements necessary for binding tryptophan to this site or that another site has evolved for feedback inhibition of the Rhizobium enzyme by tryptophan.
While the next 100 residues are difficult to align, except among closely related species, the C-terminal 300 residues align easily, requiring the insertion of few gaps. Of 31 invariant residues in the anthranilate synthase sequences, only one occurs in the N-terminal portion of the molecule. Twenty-two of the invariant residues are shared by ADC synthase sequences, and 14 are shared by isochorismate synthase.
The topology of a cladistic analysis of the complete anthranilate synthase component I sequences is not very congruent with the topology of trees based on 16S rRNA sequences (83) and improves only slightly by using just the C-terminal portions of the sequences. Sequences from closely related genera (e.g., enteric bacteria, pseudomonads, and bacilli) tend to cluster closely but do not branch as might be expected on the basis of their 16S rRNA relationships. The sequences of the archaeal components I do not form a discrete cluster but are topologically mixed among the sequences.
The paralogously related sequence encoded by P. aeruginosa phnA is most closely related to the enteric anthranilate synthase components I. The ADC synthase component I sequences encoded by pabB genes form a separate cluster, and within this cluster, the enteric bacteria of the genera Escherichia, Salmonella, and Klebsiella form a discrete subdivision, as might be expected from their close relationships. Overall, the pabB-encoded sequences are more easily aligned as a group throughout the entire sequence and do not show the extreme variability at their N termini that the trpE-encoded sequences show. Apparently there is greater selection for conservation of the entire enzyme for the ADC synthases than for the anthranilate synthases.
The single complete isochorismate synthase encoded by entC aligns only in the C-terminal portion of the sequence. Even so, several of the most highly conserved segments observed in the trpE-, pabB-, and phnA-encoded components I are not conserved in isochorismate synthase. This divergence may reflect the differences in substrate binding (no ammonia equivalent is used), subunit composition (no additional subunit is necessary), or reaction mechanism for this enzyme. Alternatively, isochorismate synthase may not require as strict conservation in the regions in question, although additional sequences would be needed to address this issue.
The smaller subunit of anthranilate synthase, or component II, is encoded by trpG and confers upon the enzyme the ability to use glutamine as the source of ammonia for the amination of chorismate by component I. Component II, or its domain within a fusion polypeptide, is small, approximately 200 amino acid residues. trpG usually is present as a separate gene but sometimes is fused with trpD, as in E. coli and S. typhimurium (74), and in R. meliloti it is fused with trpE (10). Zalkin has reviewed the glutamine amidotransferases recently, and for a comprehensive consideration of these activities, the reader is referred to reference 121.
trpG is a member of a family of related glutamine amidotransferases containing at least seven members. The glutamine amidotransferases most closely related to the anthranilate synthase TrpG subunits are those of the ADC synthase components II, encoded by pabA. As mentioned previously, three bacterial species have been found to contain an "amphibolic" component II subunit that functions in both the tryptophan and p-aminobenzoate pathways. Two of these sequences (from B. subtilis and A. calcoaceticus) are currently available (58, 100), as is the sequence encoded by P. aeruginosa phnB, which participates in the anthranilate synthase reaction leading to phenazine synthesis (37). The remaining 19 sequences are presumed to be specific for anthranilate synthesis for the tryptophan pathway and for the most part are genetically clustered with other trp genes.
In addition to the close relationship of anthranilate synthase and ADC synthase components II, additional glutamine amidotransferase subunits or glutamine amidotransferase domains within larger enzymes are homologous with trpG-encoded sequences. These include, in rough order of distance from the TrpG sequence, carbamyl phosphate synthase (carA) (86), GMP synthase (guaA) (104), CTP synthetase (pyrG) (108), imidazole glycerol phosphate synthase (hisH) (18), and formylglycinamidine synthase (purL) (95). Each of the glutamine amidotransferases contains three conserved regions, one containing a conserved Pro-Gly structural feature, one containing an essential cysteine residue that covalently reacts with the substrate, and one containing an essential histidine that has been shown to be required for catalysis (2, 85, 91). However, regions between the conserved sequences have diverged to a very high degree. Yet it should be kept in mind that the glutamine amidotransferases of the anthranilate and p-aminobenzoate synthesis are members of one of two different superfamilies of glutamine amidotransferases that have been characterized.
Alignment of the 26 trpG, pabA, and phnB sequences is relatively straightforward, with few gaps needed to achieve high similarity among them. Within the alignment, 24 amino acid residues are invariant. The sequences fall into two general clusters, those that are similar to E. coli trpG and those that are similar to E. coli pabA. The pabA-like cluster includes both the amphibolic sequences from B. subtilis and A. calcoaceticus but also includes the tryptophan pathway-specific sequences from the fluorescent pseudomonads and others, including the archaeal trpG-encoded sequences. However, the separation into pabA- and trpG-like sequences does not correlate with the major divisions or subdivisions of the bacterial taxonomy. For example, two different genera from the α subdivision of the purple bacteria (Azospirillum and Rhizobium) fall into the two different classes, as do members of the low-G+C gram-positive division (B. subtilis and L. lactis). It is also interesting that two different anthranilate synthase components II from P. aeruginosa fall into the two different groups. Clearly, the relationships among the components II are not based on pathway specificity but represent two sequence classes of glutamine amidotransferases that are compatible as either ADC or anthranilate synthase subunits. (It should be noted, however, that the pabA- and trpG-encoded sequences are not completely intermixable, since fusion hybrids expressed from E. coli trpG-pabA fusions function in neither reaction [S. Z. Doktor and B. P. Nichols, unpublished data].)
trpD encodes anthranilate phosphoribosyltransferase, the enzyme that catalyzes the second step in tryptophan biosynthesis. trpD has so far always been found clustered with additional trp genes. In the enteric bacteria and in T. maritima, trpD is fused with trpG, but no other fusions of trpD have been noted. There seems to be no particular selection for a fusion, since trpD-encoded activity remains essentially independent of the fused anthranilate synthase, except for a partial inhibition by tryptophan (120). The steps leading from the adjacent gene organization of trpG and trpD to their fusion have been summarized elsewhere (27, 74).
While the three-dimensional structure of the phosphoribosyltransferase has not been solved, it has been suggested that it has a "moderate probability" of forming a β/α barrel structure, as do several of the other enzymes of the pathway (109). Currently, 15 trpD sequences from microorganisms are available for analysis and can be aligned over 385 characters with 21 invariant residues.
trpF encodes N-phosphoribosylanthranilate isomerase (PRAI), the enzyme that catalyzes the third step in the biosynthetic pathway. The reaction proceeds by an Amadori rearrangement of N-(5'-phosphoribosyl)anthranilate to form 1-(o-carboxyphenylamino)-1-deoxyribulose-5-phosphate. In most cases, PRAI is encoded as a monomeric enzyme, but in all of the enteric bacteria studied to date and in the high-G+C gram-positive organism Brevibacterium lactofermentum, PRAI is fused as an independently folded C-terminal domain of a bifunctional indoleglycerol-3-phosphate synthase (IGPS)-PRAI. Although it might be suspected that there would be some functional rationale for the fusion of the PRAI and IGPS activities, the two domains fold independently and function kinetically independently, with no evidence of substrate or product channeling or of cooperativity (62). The crystal structure of the bifunctional E. coli enzyme has been determined to 2.0-Å (1 Å = 0.1 nm) resolution (111). The PRAI domain folds as an eightfold β/α barrel (except that the fifth α helix is replaced by a long loop), and the active site is formed near the C termini of the central barrel β sheets. Four loops between β sheets and α helices shield the active site, forming a channel to the surface of the enzyme.
Genetically, trpF is usually clustered with other trp genes; in the fluorescent pseudomonads, however, it is isolated from any other trp gene cluster, and as mentioned above, in several species trpF has become fused to trpC. There are only 11 trpF sequences from bacterial species in the current databases, the fewest number for any of the trp genes, and notably absent from this group are sequences from the fluorescent pseudomonads. Alignment of the sequences encompasses 250 character positions, and as has been noted, gaps in the alignment fall primarily in the regions corresponding to the positions between C termini of the α helices and N termini of the β sheets, while the active site and most conserved residues lie on the opposite side of the structure, in the loops between the C termini of the β sheets and the N termini of the α helices (31).
As a group, the PRAI sequences are the most highly divergent of all of the trp enzymes. For example, the PRAI domains of E. coli and S. typhimurium are 20.0% divergent, compared with 3 to 6% divergence for all other enzymes except the tryptophan synthase α subunits, which are 15% divergent. The sequences from E. coli and V. parahaemolyticus are 55.5% divergent, whereas the divergence between other sequences for this pair is 20 to 40%. A difference matrix of all sequences indicates divergences ranging from 19.8% (E. coli-S. typhimurium) to over 80% (Brevibacterium lactofermentum-Lactococcus lactis). Nevertheless, there are 11 invariant residues in the sequences, and nearly all of these have been found to be involved in substrate binding or structural integrity of the active site (111). Some of the conserved residues are oriented toward the active site and have been proposed to be involved in catalysis.
trpC encodes IGPS, the enzyme catalyzing the penultimate step in the tryptophan biosynthetic pathway, the conversion of 1-(o-carboxyphenylamino)-1-deoxyribulose-5-phosphate to indoleglycerol-3-phosphate (IGP). Fifteen sequences from bacterial species can be aligned over 316 characters with 21 invariant residues. As mentioned above, the trpC-encoded IGPS activity is found fused as the N-terminal domain of the bifunctional IGPS-PRAI in enteric bacteria and in Brevibacterium lactofermentum. In most organisms, however, IGPS is encoded as a separate monomeric enzyme.
The structure of the IGPS domain of the bifunctional E. coli enzyme is an eightfold β/α barrel, similar to that of the PRAI domain but also containing an N-terminal α helix that contributes to the formation of the active site. As is the case with PRAI and other β/α-barrel proteins, the active site lies at the C-terminal end of the β barrel (87, 111).
Even though the PRAI and IGPS activities are encoded by the same gene in the Enterobacteriaceae, the evolutionary divergence of the two domains is quite distinct. Comparison of the divergence rates of the two domains between E. coli and S. typhimurium shows that the IGPS portion has diverged to about the same extent as have the trpE-, trpG-, trpD-, and trpB-encoded proteins, while the PRAI domain is more than threefold more divergent. Crawford has suggested that the different divergence rates may reflect that the PRAI activity is often in excess and is less sensitive to partially inactivating sequence alterations and therefore more tolerant to a wider array of amino acid substitutions (27).
The last step in tryptophan biosynthesis is catalyzed by tryptophan synthase, a tetrameric enzyme composed of two copies of two dissimilar subunits. The α and β subunits are encoded by trpA and trpB, respectively. The overall reaction is IGP + serine → tryptophan + glyceraldehyde-3-phosphate. This enzyme has received a great deal of attention and has been reviewed extensively (72, 73, 113, 114). The reaction occurs in two steps, each catalyzed by one of the two types of subunits of the enzyme. The α subunit cleaves IGP to indole and glyceraldehyde-3-phosphate, and the β subunit condenses indole and serine to form tryptophan in a pyridoxal phosphate-dependent reaction. The partial reactions can be catalyzed, albeit relatively inefficiently, by the separated α monomers and β-β dimers, but the reaction proceeds with maximum efficiency only when the subunits are in complex as the α-β-β-α tetramer. Under these conditions, the intermediate indole is not released into the medium. The determination of the three-dimensional structure of tryptophan synthase (54), along with a stop-flow kinetic analysis (4), has provided a basis for understanding both the channeling of indole between active sites that are separated by 25 Å and the coordination of the reactions occurring at two separated active sites. The cleavage of IGP at the active site of the α subunit does not occur readily until serine reacts with pyridoxal phosphate to form an aminoacrylate at the active site of the β subunit. The enzyme then undergoes a conformational change that activates the α subunit (15). The indole cleavage product is channeled rapidly through a "tunnel" in the β subunit that extends from the active site on the α subunit near the α-β interface, through to the β-subunit active site near the middle of the molecule (1). The reaction of the indole with the aminoacrylate is very fast and essentially irreversible.
trpB and trpA are nearly always found adjacent to one another on the chromosomes of bacterial species and are nearly always cotranscribed in the order trpBA. In fungi, however, the two genes are fused in the order trp(AB) (123). An exception to the bacterial organization scheme is found in A. calcoaceticus. While genetic characterization initially suggested a trpFBA cluster (94), nucleotide sequence analysis has demonstrated a trpFB transcriptional unit, but trpA has not been found within 250 bp of trpB (63).
The β subunit of tryptophan synthase is composed of two independently folding domains of approximately equal size (54). The cores of the two domains, composed of three α helices and four parallel β sheets, are structurally similar and nearly superimposable but show no significant amino acid sequence similarity. The structural conservation is suggestive of an ancient gene duplication event in the evolutionary history of the β subunit, yet the β subunit of tryptophan synthase is the most highly conserved across species of all the enzymes of tryptophan biosynthesis. The E. coli and S. typhimurium sequences differ by only 14 residues over a length of 397 amino acids, corresponding to a 3.5% divergence. The highest divergence in the difference matrix of 18 sequences is 55%. The high conservation may be because of selective constraints based on the several functions of the molecule. In addition to maintenance of the active site for substrate and pyridoxal phosphate binding, the indole tunnel between the active sites must be maintained, as well as the extensive regions comprising the β-β and α-β subunit interfaces (1, 15, 101, 102). As a result, 92 of the 445 residues in the alignment are invariant among 18 species.
trpA is second only to trpF in its high divergence among the bacterial species, and as has been noted previously, the divergence is most extreme at the C-terminal portion of the sequences (27). An alignment of the 17 known sequences contains 19 invariant residues.
The α subunit assumes an eightfold β/α-barrel structure as do the PRAI and IGPS domains discussed earlier (54). The α subunit also contains an N-terminal helix that covers the "bottom" of the barrel, that is, the side opposite the active site. One additional deviation from the consensus α/β barrel is a segment of 26 residues that is highly conserved in the bacterial species. This feature lies near the subunit interface and contains an Asp residue that has been implicated in catalysis. A variety of residues lying near or at the active site are conserved or invariant among the sequences.
It has been suggested that metabolic pathways might evolve in retrograde fashion from the enzyme catalyzing the last step in the pathway. For example, after an enzyme that could convert indole to tryptophan had evolved, then following a duplication, a new activity that could convert a new substrate to indole might evolve. Additional duplication and divergence events might lead to the evolution of an entire metabolic pathway. While the data are not compatible with this hypothesis for the entire tryptophan pathway, structural comparisons suggest that it may be a reasonable hypothesis for the evolution of at least a portion of the pathway, in particular the three sequential steps between phophoribosylanthranilate and indole (110).
Comparisons of the three-dimensional structures of PRAI, IGPS, and the α subunit of tryptophan synthase have led to the provocative finding that not only do these enzymes share an overall structural motif of an eightfold β/α barrel, but portions of the active sites of the different enzymes are essentially superimposable (110). For example, the phosphate binding sites of the PRAI and IGPS active sites are equivalent with regard to positioning within the β/α barrels, conformations of the residues involved in phosphate binding, and the hydrogen bonding networks between the phosphate ions and the active-site residues. Similarly, the hydrophobic pockets of the various active sites are composed of similarly disposed residues on the two structures, as is the channel leading from the active site to the enzyme surface. The conservation of the structural aspects, especially in regard to their functional equivalency, has led Wilmanns et al. (110) to propose a common ancestry for the IGPS and PRAI domains of the bifunctional enzyme.
Further, the phosphate ions that have been found in the structures of PRAI and IGPS occupy a position equivalent to the phosphate moiety of the inhibitor indolepropanol phosphate when bound with the α subunit of tryptophan synthase. The phosphate binding pocket present in each of the three enzymes is formed by structurally equivalent residues, and close inspection of the region even indicates a degree of amino acid sequence conservation at these sites. While the common ancestry of many β/α-barrel structures has been considered, it has not been possible to convincingly argue for divergent rather than convergent evolutionary histories (40). However, the presence of a conserved overall structure coupled with conservation of amino acid sequence for the purpose of a common binding function strengthens the idea of divergence of these tryptophan pathway enzymes from a common ancestor.
Additional predictions of three-dimensional structure from amino acid sequences have led to the proposal that other enzymes of the tryptophan biosynthetic pathway may have a β/α-barrel structure (109). Both anthranilate phosphoribosyltransferase and anthranilate synthase component II (glutamine amidotransferase) have been predicted to fold as β/α-barrel structures. While consideration of the anthranilate synthase component I amino acid sequence did not lead to a β/α-barrel prediction, the sequence of isochorismate synthase did yield a "moderately probable" β/α-barrel structure. This prediction leads to a further speculation that the C-terminal portion of anthranilate synthase component I, which is homologous to isochorismate synthase, may also assume a β/α-barrel structure, while the N-terminal portion is an independently folding domain. If subsequent three-dimensional studies confirm these predictions, it would be very interesting to see if the retrograde evolution hypothesis extends to additional enzymes of the pathway.
Cladistic analyses indicate that each gene yields a unique tree topology and that the topology may depend on the sequences available for analysis, the care taken with each alignment, and the overall divergence contained within a data set of genes. Clearly, such an analysis should not be taken as contradictory of established phylogenies for the species under consideration nor as a method for attempting to establish such phylogenies. Rather, we should consider the data in the light of existing phylogenies established by 16S rRNA sequence comparisons and other phylogenetic data (83). In this sense, we can analyze the general trends established by this analysis.
While closely related species nearly always cluster together, notably lacking in such analyses is a tree topology representative of the major subdivisions of the domain Bacteria. Within the purple subdivision of gram-negative organisms, for example, two members of the α subdivision, Caulobacter crescentus and Azospirillum brasilense, generally cluster with the fluorescent pseudomonads, but the fluorescent pseudomonads do not group with the enteric bacteria, even though both are members of the γ subdivision of the purple bacteria. There is also no evidence of a clear demarcation between other divisions of eubacteria, although there is a general, but not very strict, trend for species from the Archaea to lie topologically near one another.
The study of the evolution of the trp pathway genes may be most useful taxonomically when limited to strains or species with very closely related or identical 16S rRNA sequences. Even comparisons between the nucleotide sequences of trp genes from E. coli and S. typhimurium indicate that the divergence of nucleotides in synonymous positions of codons has become saturated. By looking at the evolution across a broad range of species, the data may be more useful in the light of detecting horizontal transfers of trp gene sequences in species as may be the case for Brevibacterium lactofermentum or perhaps the trpFB cluster of A. calcoaceticus. Finally, the comparative sequence and structural studies can lead to broader conclusions concerning the sequence constraints on enzyme function, as well evolutionary processes involved in the evolution of metabolic pathways.
This work was supported in part by funds from the Public Health Service (GM44199) and the Department of Energy (DEFG0293ER6179). I thank Mary Ashley for helpful discussions and Natasha Austria and Frank Rogaczewski for help with the manuscript.
References
1. Ahmed, S. A., S. B. Ruvinov, A. M. Kayastha, and E. W. Miles. 1991. Mechanism of mutual activation of the tryptophan synthase α and β subunits. Analysis of the reaction specificity and substrate-induced inactivation of active site and tunnel mutants of the β subunit. J. Biol. Chem. 266:21548–21557.
2. Amuro, N., J. L. Paluh, and H. Zalkin. 1985. Replacement by site-directed mutagenesis indicates a role for histidine 170 in the glutamine amide transfer function of anthranilate synthase. J. Biol. Chem. 260:14844–14849.
3. Anderson, K. S., W. M. Kati, Q.-Z. Ye, J. Liu, C. T. Walsh, A. J. Benesi, and K. A. Johnson. 1991. Isolation and structure elucidation of the 4-amino-4-deoxychorismate intermediate in the PABA enzymatic pathway. J. Am. Chem. Soc. 113:3198–3200.
4. Anderson, K. S., E. W. Miles, and K. A. Johnson. 1991. Serine modulates substrate channeling in tryptophan synthase: a novel intersubunit triggering mechanism. J. Biol. Chem. 266:8020–8033.
5. Arhin, F. F., and L. C. Vining. 1993. Cloning, nucleotide sequence and expression in Streptomyces lividans and Escherichia coli of pabB from Lactococcus lactis subsp. lactis NCDO 496. J. Gen. Microbiol. 139:1785–1793.
6. Arhin, F. F., and L. C. Vining. 1993. Organization of the genes encoding p-aminobenzoic acid synthetase from Streptomyces lividans 1326. Gene 126:129–133.
7. Auerbach, S. D., J. Gao, and G. N. Gussin. 1993. Nucleotide sequence of trpI, trpB, and trpA of Pseudomonas syringae: positive control in fluorescent pseudomonads. Gene 123:25–32.
8. Babitzke, P., and C. Yanofsky. 1993. Reconstitution of Bacillus subtilis trp attenuation in vitro with TRAP, the trp RNA-binding attenuation protein. Proc. Natl. Acad. Sci. USA 90:133–137.
9. Bae, Y. M., and I. P. Crawford. 1990. The Rhizobium meliloti trpE(G) gene is regulated by attenuation, and its product, anthranilate synthase, is regulated by feedback inhibition. J. Bacteriol. 172:3318–3327.
10. Bae, Y. M., E. Homgren, and I. P. Crawford. 1989. Rhizobium meliloti anthranilate synthase: cloning, sequence, and expression in Escherichia coli. J. Bacteriol. 171:3471–3478.
11. Band, L., H. Shimotsu, and D. J. Henner. 1984. Nucleotide sequence of the Bacillus subtilis trpE and trpD genes. Gene 27:55–65.
12. Bardowski, J., S. D. Ehrlich, and A. Chopin. 1992. Tryptophan biosynthesis genes in Lactococcus lactis subsp. lactis. J. Bacteriol. 174:6563–6570.
13. Becker-Rudzik, M., D. A. Young, and B. L. Marrs. 1992. Sequence of the indoleglycerol phosphate synthase (trpC) gene from Rhodobacter capsulatus. J. Bacteriol. 174:5482–5484.
14. Brahamsha, B., C. Y. Han, I. P. Crawford, and E. P. Greenberg. 1991. Nucleotide sequence and analysis of a gene encoding anthranilate synthase component I in Spirochaeta aurantia. J. Bacteriol. 173:541–548.
15. Brzovic, P. S., Y. Sawa, C. C. Hyde, E. W. Miles, and M. F. Dunn. 1992. Evidence that mutations in a loop region of the α-subunit inhibit the transition from an open to a closed conformation in the tryptophan synthase bienzyme complex. J. Biol. Chem. 267:13028–13038.
16. Buvinger, W. E., L. C. Stone, and H. E. Heath. 1981. Biochemical genetics of tryptophan synthesis in Pseudomonas acidovorans. J. Bacteriol. 147:62–68.
17. Caligiuri, M. G., and R. Bauerle. 1991. Identification of amino acid residues involved in feedback regulation of the anthranilate synthase complex from Salmonella typhimurium. Evidence for an amino-terminal regulatory site. J. Biol. Chem. 266:8328–8335.
18. Carlomagno, M. S., L. Chiariotti, P. Alifano, A. G. Nappo, and C. B. Bruni. 1988. Structure and function of the Salmonella typhimurium and Escherichia coli K-12 histidine operons. J. Mol. Biol. 203:585–606.
19. Chang, M., A. Hadero, and I. P. Crawford. 1989. Sequence of the Pseudomonas aeruginosa trpI activator gene and relatedness of trpI to other prokaryotic regulatory genes. J. Bacteriol. 171:172–183.
20. Christie, G. E., and T. Platt. 1980. Gene structure in the tryptophan operon of Escherichia coli: nucleotide sequence of trpC and the flanking intercistronic regions. J. Mol. Biol. 142:519–530.
21. Cohn, W., K. Kirschner, and C. Paul. 1979. N-(5'-phosphoribosyl)anthranilate isomerase-indoleglycerol-phosphate synthase. 2. Fast-reaction studies show that a fluorescent substrate analogue binds independently on two different sites. Biochemistry 18:5953–5959.
22. Corpet, F. 1988. Multiple sequence alignment with hierarchical clustering. Nucleic Acids Res. 16:10881–10890.
23. Crawford, I. P. 1975. Gene rearrangements in the evolution of the tryptophan synthetic pathway. Bacteriol. Rev. 39:87–120.
24. Crawford, I. P. 1980. Comparative studies in the regulation of tryptophan synthesis. Crit. Rev. Biochem. 8:175–189.
25. Crawford, I. P. 1980. Gene fusions in the tryptophan pathway: tryptophan synthase and phosphoribosyl-anthranilate isomerase: indoleglycerol-phosphate synthase, p. 151–173. In H. Bisswanger and E. Schmincke-Ott (ed.), Multifunctional Proteins. Wiley, New York.
26. Crawford, I. P. 1986. Regulation of tryptophan synthesis in Pseudomonas, p. 251–263. In J. Sokatch (ed.), The Bacteria, vol. 10. Academic Press, New York.
27. Crawford, I. P. 1989. Evolution of a biosynthetic pathway: the tryptophan paradigm. Annu. Rev. Microbiol. 43:567–600.
28. Crawford, I. P., and L. Eberly. 1986. Structure and regulation of the anthranilate synthase genes in Pseudomonas aeruginosa. I. Sequence of trpG encoding the glutamine amidotransferase subunit. Mol. Biol. Evol. 3:436–448.
29. Crawford, I. P., C.-Y. Han, and M. Silverman. 1991. Sequence and features of the tryptophan operon of Vibrio parahemolyticus. DNA Sequences 1:189–196.
30. Crawford, I. P., B. P. Nichols, and C. Yanofsky. 1980. Nucleotide sequence of the trpB gene in Escherichia coli and Salmonella typhimurium. J. Mol. Biol. 142:489–502.
31. Crawford, I. P., T. Niermann, and K. Kirschner. 1987. Prediction of secondary structure by evolutionary comparison: application to the α subunit of tryptophan synthase. Proteins Struct. Funct. Genet. 2:118–129.
32. Crawford, I. P., A. Wilde, E. M. Yelverton, D. Figurski, and R. W. Hedges. 1986. Structure and regulation of the anthranilate synthase genes in Pseudomonas aeruginosa. II. Cloning and expression in Escherichia coli. Mol. Biol. Evol. 3:449–458.
33. da Costa e Silva, O., and T. Kosuge. 1991. Molecular characterization and expression analysis of the anthranilate synthase gene of Pseudomonas syringae subsp. savastanoi. J. Bacteriol. 173:463–471.
34. Eberly, L., and I. P. Crawford. 1989. DNA sequence of the tryptophan synthase genes of Pseudomonas putida. Biochimie 71:521–531.
35. Edman, J. C., A. L. Goldstein, and J. G. Erbe. 1993. Para-aminobenzoate synthase gene of Saccharomyces cerevisiae encodes a bifunctional enzyme. Yeast 9:669–675.
36. Essar, D. W., L. Eberly, and I. P. Crawford. 1990. Evolutionary differences in chromosomal locations of four early genes of the tryptophan pathway in fluorescent pseudomonads: DNA sequences and characterization of Pseudomonas putida trpE and trpGDC. J. Bacteriol. 172:867–883.
37. Essar, D. W., L. Eberly, A. Hadero, and I. P. Crawford. 1990. Identification and characterization of genes for a second anthranilate synthase in Pseudomonas aeruginosa: interchangeability of the two anthranilate synthases and evolutionary implications. J. Bacteriol. 172:884–900.
38. Essar, D. W., L. Eberly, C.-Y. Han, and I. P. Crawford. 1990. DNA sequences and characterization of four early genes of the tryptophan pathway in Pseudomonas aeruginosa. J. Bacteriol. 172:853–866.
39. Farber, G. K., and G. A. Petsko. 1990. The evolution of β/α-barrel enzymes. Trends Biochem. Sci. 15:228–234.
40. Furter, R., G. Paravicini, M. Aebi, G. Braus, F. Prantl, P. Niederberger, and R. Huetter. 1986. The TRP4 gene of Saccharomyces cerevisiae: isolation and structural analysis. Nucleic Acids Res. 14:6357–6373.
41. Gibson, M. I., and F. Gibson. 1964. Preliminary studies on the isolation and metabolism of an intermediate in aromatic biosynthesis: chorismic acid. Biochem. J. 90:248–256.
42. Goncharoff, P., and B. P. Nichols. 1984. Nucleotide sequence of Escherichia coli pabB indicates a common evolutionary origin of p-aminobenzoate synthetase and anthranilate synthetase. J. Bacteriol. 159:57–62.
43. Goncharoff, P., and B. P. Nichols. 1988. Evolution of aminobenzoate synthases: nucleotide sequences of Salmonella typhimurium and Klebsiella aerogenes pabB. Mol. Biol. Evol. 5:531–548.
44. Graf, R., B. Mehmann, and G. H. Braus. 1993. Analysis of feedback-resistant anthranilate synthases from Saccharomyces cerevisiae. J. Bacteriol. 175:1061–1068.
45. Green, J. M., W. K. Merkel, and B. P. Nichols. 1992. Characterization and sequence of Escherichia coli pabC, the gene encoding aminodeoxychorismate lyase, a pyridoxal phosphate-containing enzyme. J. Bacteriol. 174:5317–5323.
46. Green, J. M., and B. P. Nichols. 1991. p-Aminobenzoate biosynthesis in Escherichia coli: purification of aminodeoxychorismate lyase and cloning of pabC. J. Biol. Chem. 266:12971–12975.
47. Guerrero, C., L. M. Mateos, L. Malumbres, and J. F. Martin. 1994. Directed mutagenesis of a regulatory palindromic sequence upstream from the Brevibacterium lactofermentum tryptophan operon. Gene 138:35–41.
48. Hadero, A., and I. P. Crawford. 1986. Nucleotide sequence of the genes for tryptophan synthase in Pseudomonas aeruginosa. Mol. Biol. Evol. 3:191–204.
49. Haspel, G., M. Hunger, R. Schmucker, and W. Hillen. 1990. Identification and nucleotide sequence of the Acinetobacter calcoaceticus encoded trpE gene. Mol. Gen. Genet. 220:475–477.
50. Henner, D. J., L. Band, and H. Shimotsu. 1985. Nucleotide sequence of the Bacillus subtilis tryptophan operon. Gene 34:169–177.
51. Horowitz, H., G. E. Christie, and T. Platt. 1982. Nucleotide sequence of the trpD gene, encoding anthranilate synthetase component II of Escherichia coli. J. Mol. Biol. 156:245–256.
52. Horowitz, H., J. Van Arsdell, and T. Platt. 1983. Nucleotide sequence of the trpD and trpC genes of Salmonella typhimurium. J. Mol. Biol. 169:775–797.
53. Huang, M., and J. Pittard. 1967. Genetic analysis of mutant strains of Escherichia coli requiring p-aminobenzoic acid for growth. J. Bacteriol. 93:1938–1942.
54. Hyde, C. C., S. A. Ahmed, E. A. Padlan, E. W. Miles, and D. R. Davies. 1988. Three dimensional structure of the tryptophan synthase α2β2 multienzyme complex from Salmonella typhimurium. J. Biol. Chem. 263:17857–17871.
55. Ishiwata, K., S. Yoshino, S. Iwamori, T. Suzuki, and N. Makiguchi. 1989. Cloning and sequencing of Bacillus stearothermophilus tryptophan synthase genes. Agric. Biol. Chem. 53:2941–2948.
56. Jensen, R. A. 1976. Enzyme recruitment in evolution of new function. Annu. Rev. Microbiol. 30:409–425.
57. Kane, J. F., W. M. Holmes, and R. A. Jensen. 1972. Metabolic interlock: the dual function of a folate pathway gene as an extra-chromosomal gene of tryptophan biosynthesis. J. Biol. Chem. 247:1587–1596.
58. Kaplan, J. B., P. Goncharoff, A. M. Seibold, and B. P. Nichols. 1984. Nucleotide sequence of the Acinetobacter calcoaceticus trpGDC gene cluster. Mol. Biol. Evol. 1:456–472.
59. Kaplan, J. B., W. K. Merkel, and B. P. Nichols. 1985. Evolution of glutamine amidotransferase genes: nucleotide sequence of the pabA genes from Salmonella typhimurium, Klebsiella aerogenes, and Serratia marcescens. J. Mol. Biol. 183:327–340.
60. Kaplan, J. B., and B. P. Nichols. 1983. Nucleotide sequence of Escherichia coli pabA and its evolutionary relationship to trp(G)D. J. Mol. Biol. 168:451–468.
61. Kim, C. W., P. Markiewicz, J. J. Lee, C. F. Schierle, and J. H. Miller. 1993. Studies of the hyperthermophile Thermotoga maritima by random sequencing of cDNA and genomic libraries. Identification and sequencing of the trpEG(D) operon. J. Mol. Biol. 231:960–981.
62. Kirschner, K., H. Szadkowski, A. Henschen, and F. Lottspeich. 1980. Limited propteolysis of N-(5'-phosphoribosyl)anthranilate isomerase: indoleglycerol phosphate synthase from Escherichia coli yields two different enzymatically active, functional domains. J. Mol. Biol. 143:395–409.
63. Kishan, V., and W. Hillen. 1990. Molecular cloning, nucleotide sequence, and promoter structure of the Acinetobacter calcoaceticus trpFB operon. J. Bacteriol. 172:6151–6155.
64. Koyama, Y., and K. Furukawa. Cloning and sequence analysis of tryptophan synthetase genes of an extreme thermophile, Thermus thermophilus HB27: plasmid transfer from replica-plated Escherichia coli recombinant colonies to competent T. thermophilus cells. J. Bacteriol. 172:3490–3495.
65. Lam, W. L., A. Cohen, D. Tsouluhas, and W. F. Doolittle. 1990. Genes for tryptophan biosynthesis in the archaebacterium Haloferax volcanii. Proc. Natl. Acad. Sci. USA 87:6614–6618.
66. Lam, W. L., S. M. Logan, and W. F. Doolittle. 1992. Genes for tryptophan biosynthesis in the halophilic archaebacterium Haloferax volcanii: the trpDFEG cluster. J. Bacteriol. 174:1694–1697.
67. Lawrence, J., G. B. Cox, and F. Gibson. 1974. Biosynthesis of ubiquinone in Escherichia coli K-12: biochemical and genetic characterization of a mutant unable to convert chorismate into 4-hydroxybenzoate. J. Bacteriol. 118:41–45.
68. Maddison, W. P., and D. R. Maddison. 1992. MacClade: Analysis of phylogeny and Character Evolution. Sinauer Associates, Sunderland, Mass.
69. Manch, J. N., and I. P. Crawford. 1982. Genetic evidence for a positive regulatory factor mediating induction in the tryptophan pathway of Pseudomonas aeruginosa. J. Mol. Biol. 156:67–77.
70. Matsui, K., K. Sano, and E. Ohtsubo. 1986. Complete nucleotide and deduced amino acid sequences of the Brevibacterium lactofermentum tryptophan operon. Nucleic Acids Res. 14:10113–10114.
71. Meile, L., R. Stettler, R. Banholzer, M. Kotik, and T. Leisinger. 1991. Tryptophan gene cluster of Methanobacterium thermoautotrophicum Marburg: molecular cloning and nucleotide sequence of a putative trpEGCFBAD operon. J. Bacteriol. 173:5017–5023.
72. Miles, E. W. 1979. Tryptophan synthase: structure, function, and subunit interaction. Adv. Enzymol. Relat. Areas Mol. Biol. 49:127–186.
73. Miles, E. W., R. Bauerle, and S. A. Ahmed. 1987. Tryptophan synthase. Methods Enzymol. 142:398–414.
74. Miozzari, G. F., and C. Yanofsky. 1979. Gene fusion during the evolution of the tryptophan operon in Enterobacteriaceae. Nature (London) 277:486–489.
75. Morollo, A. A., and R. Bauerle. 1993. Characterization of composite aminodeoxyisochorismate synthase and aminodeoxyisochorismate lyase activities of anthranilate synthase. Proc. Natl. Acad. Sci. USA 90:9983–9987.
76. Natori, Y., Y. Kano, and F. Imamoto. 1990. Nucleotide sequence and genomic constitution of five tryptophan genes of Lactobacillus casei. J. Biochem. 107:248–255.
77. Nichols, B. P., M. B. Blumenberg, and C. Yanofsky. 1981. Comparison of the nucleotide sequence of trpA and sequences immediately beyond the trp operon of Klebsiella aerogenes, Salmonella typhimurium, and Escherichia coli. Nucleic Acids Res. 9:1743–1755.
78. Nichols, B. P., and J. M. Green. 1992. Cloning and sequencing of Escherichia coli ubiC and purification of chorismate lyase. J. Bacteriol. 174:5309–5316.
79. Nichols, B. P., G. F. Miozzari, M. van Cleemput, G. N. Bennett, and C. Yanofsky. 1980. Nucleotide sequences of the trpG regions of Escherichia coli, Shigella dysenteriae, Salmonella typhimurium and Serratia marcescens. J. Mol. Biol. 142:503–517.
80. Nichols, B. P., A. Seibold, and S. Z. Doktor. 1989. p-Aminobenzoate synthesis from chorismate occurs in two steps. J. Biol. Chem. 264:8597–8601.
81. Nichols, B. P., M. van Cleemput, and C. Yanofsky. 1981. Nucleotide sequence of Escherichia coli trpE: anthranilate synthetase component I contains no tryptophan residues. J. Mol. Biol. 146:45–54.
82. Nichols, B. P., and C. Yanofsky. 1979. Nucleotide sequences of trpA of Salmonella typhimurium and Escherichia coli: an evolutionary comparison. Proc. Natl. Acad. Sci. USA 76:5244–5248.
83. Olsen, G. J., C. R. Woese, and R. Overbeek. 1994. The winds of (evolutionary) change: breathing new life into microbiology. J. Bacteriol. 176:1–6.
84. Ozenberger, B. A., T. J. Brickman, and M. A. McIntosh. 1989. Nucleotide sequence of Escherichia coli isochorismate synthetase gene entC and evolutionary relationship of isochorismate synthetase and other chorismate-utilizing enzymes. J. Bacteriol. 171:775–783.
85. Paluh, J. L., H. Zalkin, D. Betsch, and H. L. Weith. 1985. Study of anthranilate synthase function by replacement of cysteine 84 using site directed mutagenesis. J. Biol. Chem. 260:1889–1894.
86. Piette, J., H. Nyunoya, C. J. Lusty, R. Cunin, G. Weyens, M. Crabeel, D. Charlier, N. Glansdorff, and A. Pierard. 1984. DNA sequence of the carA gene and the control region of carAB: tandem promoters, respectively controlled by arginine and the pyrimidines, regulate the synthesis of carbamoyl-phosphate synthetase in Escherichia coli K-12. Proc. Natl. Acad. Sci. USA 81:4134–4138.
87. Priestle, J. P., M. G. Grutter, J. L. White, M. G. Vincent, M. Kania, E. Wilson, T. S. Jardetzky, K. Kirschner, and J. N. Jansonius. 1987. Three-dimensional structure of the bifunctional enzyme N-(5'-phosphoribosyl)anthranilate isomerase-indole-3-glycerol-phosphate synthase from Escherichia coli. Proc. Natl. Acad. Sci. USA 84:5690–5694.
88. Rivas, M. V., E. D. Jarvis, and R. Rudner. 1990. The structure of the trpE, trpD and 5'trpC genes of Bacillus pumilus. Gene 87:71–78.
89. Ross, C. M., J. B. Kaplan, M. E. Winkler, and B. P. Nichols. 1990. An evolutionary comparison of Acinetobacter calcoaceticus trpF with trpF genes of several organisms. Mol. Biol. Evol. 7:74–81.
90. Ross, C. M., and M. E. Winkler. 1988. Structure of the Caulobacter crescentus trpFBA operon. J. Bacteriol. 170:757–768.
91. Roux, B., and C. T. Walsh. 1993. p-Aminobenzoate synthesis in Escherichia coli: mutational analysis of three conserved amino acid residues of the amidotransferase PabA. Biochemistry 32:3763–3768.
92. Sato, S., Y. Nakada, K. Hon-nami, K. Yasui, and A. Shiratsuchi. 1989. Molecular cloning and the nucleotide sequence of the Clostridium thermocellum trpE gene. J. Biochem. 105:362–366.
93. Sato, S., Y. Nakada, S. Kanaya, and T. Tanaka. 1988. Molecular cloning and nucleotide sequence of Thermus thermophilus HB8 trpE and trpG. Biochim. Biophys. Acta 950:303–312.
94. Sawula, R. V., and I. P. Crawford. 1972. Mapping of the tryptophan genes of Acinetobacter calcoaceticus by transformation. J. Bacteriol. 112:797–805.
95. Schendel, F. J., E. Mueller, J. Stubbe, A. Shiau, and J. M. Smith. 1989. Formylglycineamide ribonucleotide synthetase from Escherichia coli: cloning, sequencing, overproduction, isolation, and characterization. Biochemistry 28:2459–2471.
96. Shimotsu, H., and D. J. Henner. 1984. Characterization of the Bacillus subtilis tryptophan promoter region. Proc. Natl. Acad. Sci. USA 81:6315–6319.
97. Shimotsu, H., M. I. Kroda, C. Yanofsky, and D. J. Henner. 1986. Novel form of transcription attenuation regulates expression of the Bacillus subtilis tryptophan operon. J. Bacteriol. 166:461–471.
98. Shiratsuchi, A., and S. Sato. 1991. Nucleotide sequence of trpE, anthranilate synthase I gene, of Bacillus caldotenax. Biochim. Biophys. Acta 1090:348–350.
99. Sibold, L., and M. Henriquet. 1988. Cloning of the trp genes from the archaebacterium Methanococcus voltae: nucleotide sequence of the trpAB genes. Mol. Gen. Genet. 214:439–450.
100. Slock, J., D. P. Stahl, C.-Y. Han, E. W. Six, and I. P. Crawford. 1990. An apparent Bacillus subtilis folic acid biosynthetic operon containing pab, an amphibolic trpG gene, a third gene required for synthesis of para-aminobenzoic acid, and the dihydropteroate synthase gene. J. Bacteriol. 172:7211–7226.
101. Strambini, G. B., P. Cioni, A. Peracchi, and A. Mozzarelli. 1992. Conformational changes and subunit communication in tryptophan synthase: effect of substrates and substrate analogs. Biochemistry 31:7535–7542.
102. Swift, S., J. Kuhn, and G. S. Stewart. 1992. Selection and analysis of non-interactive mutants in the Escherichia coli tryptophan synthase α subunit. Mol. Gen. Genet. 233:129–135.
103. Swofford, D. L. 1990. Phylogenetic Analysis Using Parsimony: Program and Documentation. Illinois Natural History Survey, Champaign-Urbana.
104. Tiedeman, A. A., J. M. Smith, and H. Zalkin. 1985. Nucleotide sequence of the guaA gene encoding GMP synthetase of Escherichia coli K12. J. Biol. Chem. 260:8676–8679.
105. Tschumper, G., and J. Carbon. 1980. Sequence of yeast DNA fragment containing a chromosomal replicator and the TRP1 gene. Gene 10:157–166.
106. Tutino, M. L., G. Scarano, G. Marino, G. Sannia, and M. V. Cubellis. 1993. Tryptophan biosynthesis genes trpEGC in the thermoacidophilic archaebacterium Sulfolobus solfataricus. J. Bacteriol. 175:299–302.
107. Walsh, C. T., J. Liu, F. Rusnak, and M. Sakaitani. 1990. Molecular studies on enzymes in chorismate metabolism and the enterobactin biosynthetic pathway. Chem. Rev. 90:1105–1129.
108. Weng, M., C. A. Makaroff, and H. Zalkin. 1986. Nucleotide sequence of Escherichia coli pyrG encoding CTP synthetase. J. Biol. Chem. 261:5568–5574.
109. Wilmanns, M., and D. Eisenberg. 1993. Three-dimensional profiles from residue-pair preferences: identification of sequences with β/α-barrel fold. Proc. Natl. Acad. Sci. USA 90:1379–1383.
110. Wilmanns, M., C. C. Hyde, D. R. Davies, K. Kirschner, and J. N. Jansonius. 1991. Structural conservation in parallel β/α-barrel enzymes that catalyze three sequential reactions in the pathway of tryptophan biosynthesis. Biochemistry 30:9161–9169.
111. Wilmanns, M., J. P. Priestle, T. Niermann, and J. N. Jansonius. 1992. Three-dimensional structure of the bifunctional enzyme phosphoribosylanthranilate isomerase: indoleglycerolphosphate synthase from Escherichia coli refined at 2.0 Å resolution. J. Mol. Biol. 223:477–507.
112. Yanofsky, C. 1984. Comparison of regulatory and structural regions of genes of tryptophan metabolism. Mol. Biol. Evol. 1:143–161.
113. Yanofsky, C., and I. P. Crawford. 1972. Tryptophan synthetase. Enzymes 7:1–31.
114. Yanofsky, C., and I. P. Crawford. 1987. The tryptophan operon, p. 1453–1473. In F. C. Neidhardt, J. L. Ingraham, K. B. Low, B. Magasanik, M. Schaechter, and H. E. Umbarger (ed.), Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology. American Society for Microbiology, Washington, D.C.
115. Yanofsky, C., and R. Kolter. 1982. Attenuation in amino acid biosynthetic operons. Annu. Rev. Genet. 16:113–134.
116. Yanofsky, C., T. Platt, I. P. Crawford, B. P. Nichols, G. E. Christie, H. Horowitz, M. van Cleemput, and A. M. Wu. 1981. The complete nucleotide sequence of the tryptophan operon of Escherichia coli. Nucleic Acids Res. 9:6647–6668.
117. Yanofsky, C., and M. van Cleemput. 1982. Nucleotide sequence of trpE of Salmonella typhimurium and its homology with the corresponding sequence of Escherichia coli. J. Mol. Biol. 155:235–246.
118. Ye, Q.-Z., J. Liu, and C. T. Walsh. 1990. p-Aminobenzoate synthesis in Escherichia coli: purification and characterization of PabB as aminodeoxychorismate synthase and enzyme X as aminodeoxychorismate lyase. Proc. Natl. Acad. Sci. USA 87:9391–9395.
119. Yelton, D. B., and S. L. Peng. 1989. Identification and nucleotide sequence of the Leptospira biflexa serovar patoc trpE and trpG genes. J. Bacteriol. 171:2083–2089.
120. Zalkin, H. 1980. Anthranilate synthase: relationships between bifunctional and monofunctional enzymes, p. 123–149. In H. Bisswanger and E. Schminke-Ott (ed.), Multifunctional Proteins. Wiley, New York.
121. Zalkin, H. 1993. The amidotransferases. Adv. Enzymol. Relat. Areas Mol. Biol. 66:203–309.
122. Zalkin, H., J. L. Paluh, M. van Cleemput, W. S. Moye, and C. Yanofsky. 1984. Nucleotide sequence of Saccharomyces cerevisiae genes TRP2 and TRP3 encoding bifunctional anthranilate synthase:indole-3-glycerol phosphate synthase. J. Biol. Chem. 259:3985–3992.
123. Zalkin, H., and C. Yanofsky. 1982. Yeast gene TRP5: structure, function, regulation. J. Biol. Chem. 257:1491–1500.
124. Zimmer, W., C. Aparicio, and C. Elmerich. 1991. Relationship between tryptophan biosynthesis and indole-3-acetic acid production in Azospirillum: identification and sequencing of a trpGDC cluster. Mol. Gen. Genet. 229:41–51.
Chapter143
