Construction and evaluation of a novel bifunctional N-carbamylase-D-hydantoinase fusion enzyme

A fully enzymatic process employing two sequential enzymes, D-hydantoinase and N-carbamylase, is a typical case requiring combined enzyme activity for the production of D-amino acids. To test the possibility of generating a bifunctional fusion enzyme, we constructed a fusion protein via end-to-end fusion of a whole gene that encodes an intact protein at the N terminus of the D-hydantoinase. Firstly, maltose-binding protein (MBP) gene of E. coli was fused with D-hydantoinase gene from Bacillus stearothermophilus SD1, and the properties of the resulting fusion protein (MBP-HYD) were compared with those of native D-hydantoinase. Gel filtration and kinetic analyses clearly demonstrated that the typical characteristics of D-hydantoinase are maintained even in a fusion state. Based on this result, we constructed an artificial fusion enzyme composed of the whole length of N-carbamylase (304 amino acids [aa]) from Agrobacterim radiobacter NRRL B11291 and D-hydantoinase (471 aa). The fusion enzyme (CAB-HYD) was functionally expressed with an expected molecular mass of 86 kDa and efficiently converted exogenous hydantoin derivatives to the D-amino acids. A related D-hydantoinase (HYD1) gene from Bacillus thermocatenulatus GH2 was also fused with the N-carbamylase gene at its N terminus. The resulting enzyme (CAB-HYD1) was bifunctional as expected and showed better performance than the CAB-HYD fusion enzyme. The conversion of hydantoin derivatives to corresponding amino acids by the fusion enzymes was much higher than that by the separately expressed enzymes, and comparable to that by the coexpressed enzymes. Thus, the fusion enzyme might be useful as a potential biocatalyst for the production of nonnatural amino acids.     

Bifunctional xylanases and their potential use in biotechnology

Plant cell walls are comprised of cellulose, hemicellulose and other polymers that are intertwined. This complex structure acts as a barrier to degradation by single enzyme. Thus, a cocktail consisting of bi and multifunctional xylanases and xylan debranching enzymes is most desired combination for the efficient utilization of these complex materials. Xylanases have prospective applications in the food, animal feed, and paper and pulp industries. Furthermore, in order to enhance feed nutrient digestibility and to improve wheat flour quality xylanase along with other glycohydrolases are often used. For these applications, a bifunctional enzyme is undoubtedly much more valuable as compared to monofunctional enzyme. The natural diversity of enzymes provides some candidates with evolved bifunctional activity. Nevertheless most resulted from the in vitro fusion of individual enzymes. Here we present bifunctional xylanases, their evolution, occurrence, molecular biology and potential uses in biotechnology.     

Crystal structure of D-hydantoinase from Bacillus stearothermophilus: insight into the stereochemistry of enantioselectivity

Industrial production of antibiotics, such as semisynthetic penicillins and cephalosporins, requires optically pure D-p-hydroxylphenylglycine and its derivatives as important side-chain precursors. To produce optically pure D-amino acids, microbial D-hydantoinase (E.C. 3.5.2.2) is used for stereospecific hydrolysis of chemically synthesized cyclic hydantoins. We report the apo-crystal structure of D-hydantoinase from B. stearothermophilus SD1 at 3.0 A resolution. The structure has a classic TIM barrel fold. Despite an undetectable similarity in sequence, D-hydantoinase shares a striking structural similarity with the recently solved structure of dihydroorotase. A structural comparison of hydantoinase with dihydroorotase revealed that the catalytic chemistry is conserved, while the substrate recognition is not. This structure provides insight into the stereochemistry of enantioselectivity in hydrolysis and illustrates how the enzyme recognizes stereospecific exocyclic substituents and hydrolyzes hydantoins. It should also provide a rationale for further directed evolution of this enzyme for hydrolysis of new hydantoins with novel exocyclic substituents.     

Evaluation of a novel bifunctional xylanase–cellulase constructed by gene fusion

An artificial bifunctional enzyme, xylanase–cellulase, has been prepared by gene fusion. Three chimeric genes were constructed that encoded fusion proteins of different lengths. The fusion proteins exhibited both xylanase (XynX) and cellulase (Cel5Z::Ω) activity when cel5Z::Ω was fused downstream of xynX, but not when xynX was fused downstream of cel5Z::Ω. Activities of bifunctional enzymes decreased when a shorter xylanase peptide was fused. Three fusion enzymes were purified, and the molecular weights of the enzymes were estimated by CMC-SDS-PAGE and XYN-SDS-PAGE to be 149, 129, and 87 kDa, respectively. The fusion enzymes displayed optimum cellulase activity at pH 8.0 and 50 °C and optimum xylanase activity at pH 8.0 and 70 °C.     

Simultaneous optimization of enzyme activity and quaternary structure by directed evolution

Natural evolution has produced efficient enzymes of enormous structural diversity. We imitated this natural process in the laboratory to augment the efficiency of an engineered chorismate mutase with low activity and an unusual hexameric topology. By applying two rounds of DNA shuffling and genetic selection, we obtained a 400-fold more efficient enzyme, containing three non-active-site mutations. Detailed biophysical characterization of the evolved variant suggests that it exists predominantly as a trimer in solution, but is otherwise similarly stable as the parent hexamer. The dramatic structural and functional effects achieved by a small number of seemingly innocuous substitutions highlights the utility of directed evolution for modifying protein-protein interactions to produce novel quaternary states with optimized activities.     

Thermostable Artificial Enzyme Isolated by In Vitro Selection

Artificial enzymes hold the potential to catalyze valuable reactions not observed in nature. One approach to build artificial enzymes introduces mutations into an existing protein scaffold to enable a new catalytic activity. This process commonly results in a simultaneous reduction of protein stability as an undesired side effect. While protein stability can be increased through techniques like directed evolution, care needs to be taken that added stability, conversely, does not sacrifice the desired activity of the enzyme. Ideally, enzymatic activity and protein stability are engineered simultaneously to ensure that stable enzymes with the desired catalytic properties are isolated. Here, we present the use of the in vitro selection technique mRNA display to isolate enzymes with improved stability and activity in a single step. Starting with a library of artificial RNA ligase enzymes that were previously isolated at ambient temperature and were therefore mostly mesophilic, we selected for thermostable active enzyme variants by performing the selection step at 65°C. The most efficient enzyme, ligase 10C, was not only active at 65°C, but was also an order of magnitude more active at room temperature compared to related enzymes previously isolated at ambient temperature. Concurrently, the melting temperature of ligase 10C increased by 35 degrees compared to these related enzymes. While low stability and solubility of the previously selected enzymes prevented a structural characterization, the improved properties of the heat-stable ligase 10C finally allowed us to solve the three-dimensional structure by NMR. This artificial enzyme adopted an entirely novel fold that has not been seen in nature, which was published elsewhere. These results highlight the versatility of the in vitro selection technique mRNA display as a powerful method for the isolation of thermostable novel enzymes.     

Molecular characterization of a bifunctional glyoxylate cycle enzyme, malate synthase/isocitrate lyase, in Euglena gracilis

Euglena gracilis induced glyoxylate cycle enzymes when ethanol was fed as a sole carbon source. We purified, cloned and characterized a bifunctional glyoxylate cycle enzyme from E. gracilis (EgGCE). This enzyme consists of an N-terminal malate synthase (MS) domain fused to a C-terminal isocitrate lyase (ICL) domain in a single polypeptide chain. This domain order is inverted compared to the bifunctional glyoxylate cycle enzyme in Caenorhabditis elegans, an N-terminal ICL domain fused to a C-terminal MS domain. Purified EgGCE catalyzed the sequential ICL and MS reactions. ICL activity of purified EgGCE increased in the existence of acetyl-CoA at a concentration of micro-molar order. We discussed the physiological roles of the bifunctional glyoxylate cycle enzyme in these organisms as well as its molecular evolution.     

Directed evolution of enantioselective enzymes for organic chemistry

The production of enantiopure compounds is of steadily increasing importance to the chemical and biotechnological industries. In principal, the application of directed evolution in combination with newly developed screening methods enables the generation of enzymes with improved enantioselectivity. The first and most advanced example relates to a bacterial lipase from Pseudomonas aeruginosa. This enzyme was evolved towards a model substrate to yield in a lipase mutant showing > 90% enantiomeric excess as compared to 2% for the wild-type lipase. The creation of enantioselective enzymes by directed evolution will become an important technology in the near future.     

Structure and Evolution of a Bifunctional Enzyme of the Tryptophan Operon

Multifunctional enzymes composed of a single polypeptide chain could conceivably have arisen through the fusion of contiguous genes coding for distinct proteins. An example of this may be tryptophan synthetase, which is a single bifunctional protein in fungi but a complex of two distinct polypeptide chains in bacteria such as Salmonella typhimurium.     

The urea carboxylase and allophanate hydrolase activities of urea amidolyase are functionally independent

Urea amidolyase (UAL) is a multifunctional biotin‐dependent enzyme that contributes to both bacterial and fungal pathogenicity by catalyzing the ATP‐dependent cleavage of urea into ammonia and CO₂. UAL is comprised of two enzymatic components: urea carboxylase (UC) and allophanate hydrolase (AH). These enzyme activities are encoded on separate but proximally related genes in prokaryotes while, in most fungi, they are encoded by a single gene that produces a fusion enzyme on a single polypeptide chain. It is unclear whether the UC and AH activities are connected through substrate channeling or other forms of direct communication. Here, we use multiple biochemical approaches to demonstrate that there is no substrate channeling or interdomain/intersubunit communication between UC and AH. Neither stable nor transient interactions can be detected between prokaryotic UC and AH and the catalytic efficiencies of UC and AH are independent of one another. Furthermore, an artificial fusion of UC and AH does not significantly alter the AH enzyme activity or catalytic efficiency. These results support the surprising functional independence of AH from UC in both the prokaryotic and fungal UAL enzymes and serve as an important reminder that the evolution of multifunctional enzymes through gene fusion events does not always correlate with enhanced catalytic function.     

Construction of an artificial bifunctional enzyme, beta-galactosidase/galactose dehydrogenase, exhibiting efficient galactose channeling

The in-frame fusion between two oligomeric enzymes, beta-galactosidase and galactose dehydrogenase, is described. The lacZ gene was fused to the 3' end of the galdh gene with a linker encoding only three amino acids. The purified artificial bifunctional enzyme displayed the enzymic activity of both gene products. The hybrid protein was found in two major forms, consisting of four and six subunits, but other forms could also be identified. The molecular weight of each subunit was determined to be 145,000 by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The bifunctional enzyme shows kinetic advantages over the identical native system in conversion of lactose to galactonolactone. A higher steady-state rate and a reduction of the transient time are observed. This phenomenon is especially pronounced at low initial substrate concentrations and when the pH is adjusted to a level at which the galactose dehydrogenase activity is much higher than that of the beta-galactosidase.     

Expression and stabilization of galactose oxidase in Escherichia coli by directed evolution.

We have used directed evolution methods to express a fungal enzyme, galactose oxidase (GOase), in functional form in Escherichia coli. The evolved enzymes retain the activity and substrate specificity of the native fungal oxidase, but are more thermostable, are expressed at a much higher level (up to 10.8 mg/l of purified GOase), and have reduced negative charge compared to wild type, all properties which are expected to facilitate applications and further evolution of the enzyme. Spectroscopic characterization of the recombinant enzymes reveals a tyrosyl radical of comparable stability to the native GOase from Fusarium.     

Cloning, sequencing, and expression in Escherichia coli of the D-hydantoinase gene from Pseudomonas putida and distribution of homologous genes in other microorganisms.

Pseudomonas putida DSM 84 produces N-carbamyl-D-amino acids from the corresponding D-5-monosubstituted hydantoins. The gene encoding this D-hydantoinase enzyme was cloned and expressed in Escherichia coli. The nucleotide sequence of the 1.8-kb insert of subclone pGES19 was determined. One open reading frame of 1,104 bp was found and was predicted to encode a polypeptide with a molecular size of 40.5 kDa. Local regions of identity between the predicted amino acid sequence and that of other known amidohydrolases (two other D-hydantoinases, allantionase and dihydroorotase) were found. The D-hydantoinase gene was used as a probe to screen DNA isolated from diverse organisms. Within Pseudomonas strains of rRNA group I, the probe was specific. The probe did not detect D-hydantoinase genes in pseudomonads not in rRNA group I, other bacteria, or plants known to express D-hydantoinase activity.     

Domain dissection and characterization of the aminoglycoside resistance enzyme ANT(3″)-Ii/AAC(6′)-IId from Serratia marcescens

Aminoglycosides (AGs) are broad-spectrum antibiotics whose constant use and presence in growth environments has led bacteria to develop resistance mechanisms to aid in their survival. A common mechanism of resistance to AGs is their chemical modification (nucleotidylation, phosphorylation, or acetylation) by AG-modifying enzymes (AMEs). Through evolution, fusion of two AME-encoding genes has resulted in bifunctional enzymes with broader spectrum of activity. Serratia marcescens, a human enteropathogen, contains such a bifunctional enzyme, ANT(3″)-Ii/AAC(6′)-IId. To gain insight into the role, effect, and importance of the union of ANT(3″)-Ii and AAC(6′)-IId in this bifunctional enzyme, we separated the two domains and compared their activity to that of the full-length enzyme. We performed a thorough comparison of the substrate and cosubstrate profiles as well as kinetic characterization of the bifunctional ANT(3″)-Ii/AAC(6′)-IId and its individually expressed components.     

A novel composite locus of Arabidopsis encoding two polypeptides with metabolically related but distinct functions in lysine catabolism

Both plants and animals catabolize lysine via saccharopine by two consecutive enzymes, lysine-ketoglutarate reductase (LKR) and saccharopine dehydrogenase (SDH), which are linked on a single polypeptide. We recently demonstrated that Arabidopsis plants possess not only a bifunctional LKR/SDH but in addition a monofunctional SDH enzyme. We also speculated that these two enzymes may be controlled by a single gene (G. Tang et al. Plant Cell, 1997, 9, 1305-1316). By expressing several epitope-tagged and GUS reporter constructs, we demonstrate in the present study that the Arabidopsis monofunctional SDH is encoded by a distinct gene, which is, however, nested entirely within the coding and 3' non-coding regions of the larger bifunctional LKR/SDH gene. The entire open reading frame of the monofunctional SDH gene, as well as some components of its promoter, are also parts of the translated coding sequence of the bifunctional LKR/SDH gene. These special structural characteristics, combined with the fact that the two genes encode simultaneously two metabolically related but distinct enzymes, render the LKR/SDH locus a novel type of a composite locus. Not all plant species possess an active monofunctional SDH gene and the production of this enzyme is correlated with an increased flux of lysine catabolism. Taken together, our results suggest that the composite LKR/SDH locus serves to control an efficient, highly regulated flux of lysine catabolism     

Evolution of enzymes in metabolism: a network perspective

Several models have been proposed to explain the origin and evolution of enzymes in metabolic pathways. Initially, the retro-evolution model proposed that, as enzymes at the end of pathways depleted their substrates in the primordial soup, there was a pressure for earlier enzymes in pathways to be created, using the later ones as initial template, in order to replenish the pools of depleted metabolites. Later, the recruitment model proposed that initial templates from other pathways could be used as long as those enzymes were similar in chemistry or substrate specificity. These two models have dominated recent studies of enzyme evolution. These studies are constrained by either the small scale of the study or the artificial restrictions imposed by pathway definitions. Here, a network approach is used to study enzyme evolution in fully sequenced genomes, thus removing both constraints. We find that homologous pairs of enzymes are roughly twice as likely to have evolved from enzymes that are less than three steps away from each other in the reaction network than pairs of non-homologous enzymes. These results, together with the conservation of the type of chemical reaction catalyzed by evolutionarily related enzymes, suggest that functional blocks of similar chemistry have evolved within metabolic networks. One possible explanation for these observations is that this local evolution phenomenon is likely to cause less global physiological disruptions in metabolism than evolution of enzymes from other enzymes that are distant from them in the metabolic network.     

A Chimeric Fusion Protein Engineered with Disparate Functionalities—Enzymatic Activity and Self-assembly

The fusion of protein domains is an important mechanism in molecular evolution and a valuable strategy for protein engineering. We are interested in creating fusion proteins containing both globular and structural domains so that the final chimeric protein can be utilized to create novel bioactive biomaterials. Interactions between fused domains can be desirable in some fusion protein applications, but in this case the optimal configuration will enable the bioactivity to be unaffected by the structural cross-linking. To explore this concept, we have created a fusion consisting of a thermostable aldo-keto reductase, two α-helical leucine zipper domains, and a randomly coiled domain. The resulting protein is bifunctional in that (1) it can self-assemble into a hydrogel material as the terminal leucine zipper domains form interprotein coiled-coil cross-links, and (2) it expresses alcohol dehydrogenase and aldo-keto reductase activity native to AdhD from Pyrococcus furiosus. The kinetic parameters of the enzyme are minimally affected by the addition of the helical appendages, and rheological studies demonstrate that a supramolecular assembly of the bifunctional protein building blocks forms a hydrogel. An active hydrogel is produced at temperatures up to 60 °C, and we demonstrate the functionality of the biomaterial by monitoring the oxidation and reduction of the native substrates by the gel. The design of chimeric fusion proteins with both globular and structural domains is an important advancement for the creation of bioactive biomaterials for biotechnology applications such as tissue engineering, bioelectrocatalysis, and biosensing and for the study of native assembled enzyme structures and clustered enzyme systems such as metabolons.     

Characterization of the alpha-2,8-polysialyltransferase from Neisseria meningitidis with synthetic acceptors, and the development of a self-priming polysialyltransferase fusion enzyme

Glycoconjugates containing polysialic acid have many biological activities and represent target molecules for therapeutic interventions. Enzymatic synthesis of these glycoconjugates should give access to these important molecules to evaluate their potential. The polysialyltransferases from both Neisseria meningitidis and Escherichia coli were cloned and expressed as recombinant proteins in E. coli. We have used synthetic acceptors to probe the acceptor requirement of these enzymes and to examine the basic enzymology. The minimum number of sialic acid residues (Neu5Ac) on the acceptor for activity in vitro was shown to be 2 for both enzymes, but a large increase in activity was seen if the acceptor had three Neu5Ac residues. The polysialyltransferase from N. meningitidis generated longer reaction products than the enzyme from E. coli on FCHASE acceptors. Examination of the products showed them to be a heterogeneous mixture, but products with >50 Neu5Ac residues could be seen using capillary zone electrophoresis analyses. In addition we made fusion proteins of these polysialyltransferase enzymes with the bifunctional alpha-2,3/alpha-2,8-sialyltransferase from Campylobacter jejuni to create self priming polysialyltransferases. These bifunctional sialyltransferases utilized various synthetic disaccharide acceptors with a terminal galactose, and we demonstrate here that the PST enzyme from N. meningitidis and its fusion protein with the C. jejuni sialyltransferase can be used to create polysialic acid on O-linked glycopeptides.     

Strategy and success for the directed evolution of enzymes

Directed evolution is widely used to improve enzymes, particularly for industrial biocatalytic processes. Molecular biology advances present many new strategies for directed evolution. Commonly used techniques have led to many successful examples of enzyme improvement, yet there is still a need to improve both the efficiency and capability of directed evolution. Recent strategies aimed at making directed evolution faster and more efficient take better advantage of available structural and sequence information. The underlying principles that lead to early dead-ends for directed evolution experiments are also discussed along with recent strategies designed to by-pass them. Several emerging methods for creating novel enzymes are also discussed including examples of catalytic activity for which there is no precedent in nature. Finally, the combined use of several strategies is likely to be required in practice to improve multiple target properties of an enzyme, as successfully shown by a recent industrial example.