Cyclohexadienyl dehydratase from Pseudomonas aeruginosa. Molecular cloning of the gene and characterization of the gene product

Novel exported fusion enzymes with chorismate mutase and cyclohexadienyl dehydratase activity: Shikimate pathway enzymes teamed up in no man's land

Chorismate mutase (CM) and cyclohexadienyl dehydratase (CDT) catalyze two subsequent reactions in the intracellular biosynthesis of l-phenylalanine (Phe). Here, we report the discovery of novel and extremely rare bifunctional fusion enzymes, consisting of fused CM and CDT domains, which are exported from the cytoplasm. Such enzymes were found in only nine bacterial species belonging to non-pathogenic γ- or β-Proteobacteria. In γ-proteobacterial fusion enzymes, the CM domain is N-terminal to the CDT domain, whereas the order is inverted in β-Proteobacteria. The CM domains share 15% to 20% sequence identity with the AroQγ class CM holotype of Mycobacterium tuberculosis (∗MtCM), and the CDT domains 40% to 60% identity with the exported monofunctional enzyme of Pseudomonas aeruginosa (PheC). In vitro kinetics revealed a Km <7 μM, much lower than for ∗MtCM, whereas kinetic parameters are similar for CDT domains and PheC. There is no feedback inhibition of CM or CDT by the pathway's end product Phe, and no catalytic benefit of the domain fusion compared with engineered single-domain constructs. The fusion enzymes of Aequoribacter fuscus, Janthinobacterium sp. HH01, and Duganella sacchari were crystallized and their structures refined to 1.6, 1.7, and 2.4 Å resolution, respectively. Neither the crystal structures nor the size-exclusion chromatography show evidence for substrate channeling or higher oligomeric structure that could account for the cooperation of CM and CDT active sites. The genetic neighborhood with genes encoding transporter and substrate binding proteins suggests that these exported bifunctional fusion enzymes may participate in signaling systems rather than in the biosynthesis of Phe.

The role of oligomerization in the optimization of cyclohexadienyl dehydratase conformational dynamics and catalytic activity

The emergence of oligomers is common during the evolution and diversification of protein families, yet the selective advantage of oligomerization is often cryptic or unclear. Oligomerization can involve the formation of isologous head-to-head interfaces (e.g., in symmetrical dimers) or heterologous head-to-tail interfaces (e.g., in cyclic complexes), the latter of which is less well studied and understood. In this work, we retrace the emergence of the trimeric form of cyclohexadienyl dehydratase from Pseudomonas aeruginosa (PaCDT) by introducing residues that form the PaCDT trimer-interfaces into AncCDT-5 (a monomeric reconstructed ancestor of PaCDT). We find that single interface mutations can switch the oligomeric state of the variants and that trimerization corresponds with a reduction in the KM value of the enzyme from a promiscuous level to the physiologically relevant range. In addition, we find that removal of a C-terminal extension present in PaCDT leads to a variant with reduced catalytic activity, indicating that the C-terminal region has a role in tuning enzymatic activity. We show that these observations can be rationalized at the structural and dynamic levels, with trimerization and C-terminal extension leading to reduced sampling of non-catalytic conformational substates in molecular dynamics simulations. Overall, this work provides insight into how neutral sampling of distinct oligomeric states along an evolutionary trajectory can facilitate the evolution and optimization of enzyme function.

Altered conformational sampling along an evolutionary trajectory changes the catalytic activity of an enzyme

Several enzymes are known to have evolved from non-catalytic proteins such as solute-binding proteins (SBPs). Although attention has been focused on how a binding site can evolve to become catalytic, an equally important question is: how do the structural dynamics of a binding protein change as it becomes an efficient enzyme? Here we performed a variety of experiments, including propargyl-DO3A-Gd(III) tagging and double electron-electron resonance (DEER) to study the rigid body protein dynamics of reconstructed evolutionary intermediates to determine how the conformational sampling of a protein changes along an evolutionary trajectory linking an arginine SBP to a cyclohexadienyl dehydratase (CDT). We observed that primitive dehydratases predominantly populate catalytically unproductive conformations that are vestiges of their ancestral SBP function. Non-productive conformational states, including a wide-open state, are frozen out of the conformational landscape via remote mutations, eventually leading to extant CDT that exclusively samples catalytically relevant compact states. These results show that remote mutations can reshape the global conformational landscape of an enzyme as a mechanism for increasing catalytic activity.

Harnessing Conformational Plasticity to Generate Designer Enzymes

Recent years have witnessed an explosion of interest in understanding the role of conformational dynamics both in the evolution of new enzymatic activities from existing enzymes and in facilitating the emergence of enzymatic activity de novo on scaffolds that were previously non-catalytic. There are also an increasing number of examples in the literature of targeted engineering of conformational dynamics being successfully used to alter enzyme selectivity and activity. Despite the obvious importance of conformational dynamics to both enzyme function and evolvability, many (although not all) computational design approaches still focus either on pure sequence-based approaches or on using structures with limited flexibility to guide the design. However, there exist a wide variety of computational approaches that can be (re)purposed to introduce conformational dynamics as a key consideration in the design process. Coupled with laboratory evolution and more conventional existing sequence- and structure-based approaches, these techniques provide powerful tools for greatly expanding the protein engineering toolkit. This Perspective provides an overview of evolutionary studies that have dissected the role of conformational dynamics in facilitating the emergence of novel enzymes, as well as advances in computational approaches that allow one to target conformational dynamics as part of enzyme design. Harnessing conformational dynamics in engineering studies is a powerful paradigm with which to engineer the next generation of designer biocatalysts.

Evolution of cyclohexadienyl dehydratase from an ancestral solute-binding protein

The emergence of enzymes through the neofunctionalization of noncatalytic proteins is ultimately responsible for the extraordinary range of biological catalysts observed in nature. Although the evolution of some enzymes from binding proteins can be inferred by homology, we have a limited understanding of the nature of the biochemical and biophysical adaptations along these evolutionary trajectories and the sequence in which they occurred. Here we reconstructed and characterized evolutionary intermediate states linking an ancestral solute-binding protein to the extant enzyme cyclohexadienyl dehydratase. We show how the intrinsic reactivity of a desolvated general acid was harnessed by a series of mutations radiating from the active site, which optimized enzyme-substrate complementarity and transition-state stabilization and minimized sampling of noncatalytic conformations. Our work reveals the molecular evolutionary processes that underlie the emergence of enzymes de novo, which are notably mirrored by recent examples of computational enzyme design and directed evolution.

Complementation of the pha2 yeast mutant suggests functional differences for arogenate dehydratases from Arabidopsis thaliana

The final steps of phenylalanine (Phe) biosynthesis in bacteria, fungi and plants can occur via phenylpyruvate or arogenate intermediates. These routes are determined by the presence of prephenate dehydratase (PDT, EC4.2.1.51), which forms phenylpyruvate from prephenate, or arogenate dehydratase (ADT, EC4.2.1.91), which forms phenylalanine directly from arogenate. We compared sequences from select yeast species to those of Arabidopsis thaliana. The in silico analysis showed that plant ADTs and yeast PDTs share many common features allowing them to act as dehydratase/decarboxylases. However, plant and yeast sequences clearly group independently conferring distinct substrate specificities. Complementation of the Saccharomyces cerevisiae pha2 mutant, which lacks PDT activity and cannot grow in the absence of exogenous Phe, was used to test the PDT activity of A. thaliana ADTs in vivo. Previous biochemical characterization showed that all six AtADTs had high catalytic activity with arogenate as a substrate, while AtADT1, AtADT2 and AtADT6 also had limited activity with prephenate. Consistent with these results, the complementation test showed AtADT2 readily recovered the pha2 phenotype after ∼6 days growth at 30 °C, while AtADT1 required ∼13 days to show visible growth. By contrast, AtADT6 (lowest PDT activity) and AtADT3-5 (no PDT activity) were unable to recover the phenotype. These results suggest that only AtADT1 and AtADT2, but not the other four ADTs from Arabidopsis, have functional PDT activity in vivo, showing that there are two functional distinct groups. We hypothesize that plant ADTs have evolved to use the arogenate route for Phe synthesis while keeping some residual PDT activity.

Mutation of a rice gene encoding a phenylalanine biosynthetic enzyme results in accumulation of phenylalanine and tryptophan

Two distinct biosynthetic pathways for Phe in plants have been proposed: conversion of prephenate to Phe via phenylpyruvate or arogenate. The reactions catalyzed by prephenate dehydratase (PDT) and arogenate dehydratase (ADT) contribute to these respective pathways. The Mtr1 mutant of rice (Oryza sativa) manifests accumulation of Phe, Trp, and several phenylpropanoids, suggesting a link between the synthesis of Phe and Trp. Here, we show that the Mtr1 mutant gene (mtr1-D) encodes a form of rice PDT with a point mutation in the putative allosteric regulatory region of the protein. Transformed callus lines expressing mtr1-D exhibited all the characteristics of Mtr1 callus tissue. Biochemical analysis revealed that rice PDT possesses both PDT and ADT activities, with a preference for arogenate as substrate, suggesting that it functions primarily as an ADT. The wild-type enzyme is feedback regulated by Phe, whereas the mutant enzyme showed a reduced feedback sensitivity, resulting in Phe accumulation. In addition, these observations indicate that rice PDT is critical for regulating the size of the Phe pool in plant cells. Feeding external Phe to wild-type callus tissue and seedlings resulted in Trp accumulation, demonstrating a connection between Phe accumulation and Trp pool size.

Phenylalanine biosynthesis in Arabidopsis thaliana. Identification and characterization of arogenate dehydratases

There is much uncertainty as to whether plants use arogenate, phenylpyruvate, or both as obligatory intermediates in Phe biosynthesis, an essential dietary amino acid for humans. This is because both prephenate and arogenate have been reported to undergo decarboxylative dehydration in plants via the action of either arogenate (ADT) or prephenate (PDT) dehydratases; however, neither enzyme(s) nor encoding gene(s) have been isolated and/or functionally characterized. An in silico data mining approach was thus undertaken to attempt to identify the dehydratase(s) involved in Phe formation in Arabidopsis, based on sequence similarity of PDT-like and ACT-like domains in bacteria. This data mining approach suggested that there are six PDT-like homologues in Arabidopsis, whose phylogenetic analyses separated them into three distinct subgroups. All six genes were cloned and subsequently established to be expressed in all tissues examined. Each was then expressed as a Nus fusion recombinant protein in Escherichia coli, with their substrate specificities measured in vitro. Three of the resulting recombinant proteins, encoded by ADT1 (At1g11790), ADT2 (At3g07630), and ADT6 (At1g08250), more efficiently utilized arogenate than prephenate, whereas the remaining three, ADT3 (At2g27820), ADT4 (At3g44720), and ADT5 (At5g22630) essentially only employed arogenate. ADT1, ADT2, and ADT6 had k(cat)/Km values of 1050, 7650, and 1560 M(-1) S(-1) for arogenate versus 38, 240, and 16 M(-1) S(-1) for prephenate, respectively. By contrast, the remaining three, ADT3, ADT4, and ADT5, had k(cat)/Km values of 1140, 490, and 620 M(-1) S(-1), with prephenate not serving as a substrate unless excess recombinant protein (>150 microg/assay) was used. All six genes, and their corresponding proteins, are thus provisionally classified as arogenate dehydratases and designated ADT1-ADT6.

A core catalytic domain of the TyrA protein family: arogenate dehydrogenase from Synechocystis

The TyrA protein family includes prephenate dehydrogenases, cyclohexadienyl dehydrogenases and TyrA(a)s (arogenate dehydrogenases). tyrA(a) from Synechocystis sp. PCC 6803, encoding a 30 kDa TyrA(a) protein, was cloned into an overexpression vector in Escherichia coli. TyrA(a) was then purified to apparent homogeneity and characterized. This protein is a model structure for a catalytic core domain in the TyrA superfamily, uncomplicated by allosteric or fused domains. Competitive inhibitors acting at the catalytic core of TyrA proteins are analogues of any accepted cyclohexadienyl substrate. The homodimeric enzyme was specific for L-arogenate (K(m)=331 microM) and NADP+ (K(m)=38 microM), being unable to substitute prephenate or NAD+ respectively. L-Tyrosine was a potent inhibitor of the enzyme (K(i)=70 microM). NADPH had no detectable ability to inhibit the reaction. Although the mechanism is probably steady-state random order, properties of 2',5'-ADP as an inhibitor suggest a high preference for L-arogenate binding first. Comparative enzymology established that both of the arogenate-pathway enzymes, prephenate aminotransferase and TyrA(a), were present in many diverse cyanobacteria and in a variety of eukaryotic red and green algae.

Salicylate biosynthesis in Pseudomonas aeruginosa. Purification and characterization of PchB, a novel bifunctional enzyme displaying isochorismate pyruvate-lyase and chorismate mutase activities

Isochorismate pyruvate-lyase (IPL), the second enzyme of pyochelin biosynthesis and the product of the pchB gene, was purified to homogeneity from Pseudomonas aeruginosa. In the reaction catalyzed by this enzyme, isochorismate --> salicylate + pyruvate, no cofactors appear to be required. At the pH optimum (pH 6.8), the enzyme displayed Michaelis-Menten kinetics, with an apparent K(m) of 12.5 microm for isochorismate and a kcat of 106 min(-1), calculated per monomer. The native enzyme behaved as a homodimer, as judged by molecular sieving chromatography, electrophoresis under nondenaturing conditions, and cross-linking experiments. PchB has approximately 20% amino acid sequence identity with AroQ-class chorismate mutases (CMs). Chorismate was shown to be converted to prephenate by purified PchB in vitro, with an apparent K(m) of 150 microm and a kcat of 7.8 min(-1). An oxabicyclic diacid transition state analog and well characterized inhibitor of CMs competitively inhibited both IPL and CM activities of PchB. Moreover, a CM-deficient Escherichia coli mutant, which is auxotrophic for phenylalanine and tyrosine, was functionally complemented by the cloned P. aeruginosa pchB gene for growth in minimal medium. A mutant form of PchB, in which isoleucine 88 was changed to threonine, had no detectable IPL activity, but retained wild-type CM activity. In conclusion, the 11.5-kDa subunit of PchB appears to contain a single active site involved in both IPL and CM activity.

Investigating and Engineering Enzymes by Genetic Selection

Natural enzymes have arisen over millions of years by the gradual process of Darwinian evolution. The fundamental steps of evolution-mutation, selection, and amplification-can also be exploited in the laboratory to create and characterize protein catalysts on a human timescale. In vivo genetic selection strategies enable the exhaustive analysis of protein libraries with 10(10) different members, and even larger ensembles can be studied with in vitro methods. Evolutionary approaches can consequently yield statistically meaningful insight into the complex and often subtle interactions that influence protein folding, structure, and catalytic mechanism. Such methods are also being used increasingly as an adjunct to design, thus providing access to novel proteins with tailored catalytic activities and selectivities.

Cyclohexadienyl dehydrogenase from Pseudomonas stutzeri exemplifies a widespread type of tyrosine-pathway dehydrogenase in the TyrA protein family

The uni-domain cyclohexadienyl dehydrogenases are able to use the alternative intermediates of tyrosine biosynthesis, prephenate or L-arogenate, as substrates. Members of this TyrA protein family have been generally considered to fall into two classes: sensitive or insensitive to feedback inhibition by L-tyrosine. A gene (tyrA(c)) encoding a cyclohexadienyl dehydrogenase from Pseudomonas stutzeri JM300 was cloned, sequenced, and expressed at a high level in Escherichia coli. This is the first molecular-genetic and biochemical characterization of a purified protein representing the feedback-sensitive type of cyclohexadienyl dehydrogenase. The catalytic-efficiency constant k(cat)/K(m) for prephenate (7.0x10(7) M/s) was much better than that of L-arogenate (5.7x10(6) M/s). TyrA(c) was sensitive to feedback inhibition by either L-tyrosine or 4-hydroxyphenylpyruvate, competitively with respect to either prephenate or L-arogenate and non-competitively with respect to NAD(+). A variety of related compounds were tested as inhibitors, and the minimal inhibitor structure was found to require only the aromatic ring and a hydroxyl substituent. Analysis by multiple alignment was used to compare 17 protein sequences representing TyrA family members having catalytic domains that are independent or fused to other catalytic domains, that exhibit broad substrate specificity or narrow substrate specificity, and that possess or lack sensitivity to endproduct inhibitors. We propose that the entire TyrA protein family lacks a discrete allosteric domain and that inhibitors act competitively at the catalytic site of different family members which exhibit individuality in the range and extent of molecules recognized as substrate or inhibitor.

Cell death of human polymorphonuclear neutrophils induced by a Pseudomonas aeruginosa cystic fibrosis isolate requires a functional type III secretion system

With a coincubation model incorporating Pseudomonas aeruginosa and human polymorphonuclear neutrophils (PMNs), a cystic fibrosis (CF) P. aeruginosa isolate has been shown to resist the bactericidal action of PMNs and to induce their cellular death. An isogenic mutant of this CF isolate in which the type III secretion system was rendered nonfunctional was unable to induce cellular death of PMNs.

L-Arogenate Is a Chemoattractant Which Can Be Utilized as the Sole Source of Carbon and Nitrogen by Pseudomonas aeruginosa

L-Arogenate is a commonplace amino acid in nature in consideration of its role as a ubiquitous precursor of L-phenylalanine and/or L-tyrosine. However, the questions of whether it serves as a chemoattractant molecule and whether it can serve as a substrate for catabolism have never been studied. We found that Pseudomonas aeruginosa recognizes L-arogenate as a chemoattractant molecule which can be utilized as a source of both carbon and nitrogen. Mutants lacking expression of either cyclohexadienyl dehydratase or phenylalanine hydroxylase exhibited highly reduced growth rates when utilizing L-arogenate as a nitrogen source. Utilization of L-arogenate as a source of either carbon or nitrogen was dependent upon (sigma)(sup54), as revealed by the use of an rpoN null mutant. The evidence suggests that catabolism of L-arogenate proceeds via alternative pathways which converge at 4-hydroxyphenylpyruvate. In one pathway, prephenate formed in the periplasm by deamination of L-arogenate is converted to 4-hydroxyphenylpyruvate by cyclohexadienyl dehydrogenase. The second route depends upon the sequential action of periplasmic cyclohexadienyl dehydratase, phenylalanine hydroxylase, and aromatic aminotransferase.

Synthesis of multiple exoproducts in Pseudomonas aeruginosa is under the control of RhlR-RhlI, another set of regulators in strain PAO1 with homology to the autoinducer-responsive LuxR-LuxI family

Mutants of Pseudomonas aeruginosa PAO1 that were deficient in the ability to produce proteases that degrade casein were detected among the survivors of chemical mutagenesis. One such mutant (PDO31) showed reduced production of elastolytic activity, beta-hemolytic activity, and pyocyanin. A 4.3-kb EcoRI fragment from a gene bank of PAO1 that complemented defects in PDO31 was found. Transposon mutagenesis and deletion derivatives of the clone were used in conjunction with complementation tests to determine the physical location of the gene of interest. Nucleotide sequence analysis revealed an open reading frame (rhlR) encoding a putative 27.6-kDa protein (RhlR) with homology to autoinducer-responsive regulators of quorum sensing systems such as LuxR of Vibrio fischeri and LasR of P. aeruginosa. Further sequence analysis downstream of rhlR revealed an independently transcribed gene (rhlI) that encodes a putative 22.2-kDa protein with homology to members of the family of autoinducer synthetases, such as LuxI of V. fischeri and LasI of P. aeruginosa. The rhlRI sequences were also recently reported by others (U.A. Ochsner and J. Reiser, Proc. Natl. Acad. Sci. USA 92: 6424-6428, 1995) as an autoinducer-mediated regulation mechanism for rhamnolipid biosurfactant synthesis in P. aeruginosa PG201. Mutants with defects in rhlR or rhlI were constructed in PAO1 by gene replacement, using clones modified by Tn501 insertion. Compared with the wild type, the rhlR and rhlI mutants both showed defects in the production of elastase, LasA protease, rhamnolipid, and pyocyanin. Transcription from the gene for elastase, as measured with a lasB-cat fusion, demonstrated that production of elastase was subject to cell density-dependent gene activation in PAO1. However, transcription of lasB-cat in the rhlI mutant, which had lost the presumptive autoinducer synthetase (predicted to activate RhlR), showed low basal activity and had lost all cell density-dependent transcription of lasB. Thus, RhlR-RhlI represent the second autoinducer-responsive regulatory mechanism found in P. aeruginosa that controls expression of multiple virulence factor exoproducts, including elastase.