Directed evolution to re-adapt a co-evolved network within an enzyme

We have previously used targeted active-site saturation mutagenesis to identify a number of transketolase single mutants that improved activity towards either glycolaldehyde (GA), or the non-natural substrate propionaldehyde (PA). Here, all attempts to recombine the singles into double mutants led to unexpected losses of specific activity towards both substrates. A typical trade-off occurred between soluble expression levels and specific activity for all single mutants, but many double mutants decreased both properties more severely suggesting a critical loss of protein stability or native folding. Statistical coupling analysis (SCA) of a large multiple sequence alignment revealed a network of nine co-evolved residues that affected all but one double mutant. Such networks maintain important functional properties such as activity, specificity, folding, stability, and solubility and may be rapidly disrupted by introducing one or more non-naturally occurring mutations. To identify variants of this network that would accept and improve upon our best D469 mutants for activity towards PA, we created a library of random single, double and triple mutants across seven of the co-evolved residues, combining our D469 variants with only naturally occurring mutations at the remaining sites. A triple mutant cluster at D469, E498 and R520 was found to behave synergistically for the specific activity towards PA. Protein expression was severely reduced by E498D and improved by R520Q, yet variants containing both mutations led to improved specific activity and enzyme expression, but with loss of solubility and the formation of inclusion bodies. D469S and R520Q combined synergistically to improve k_cat 20-fold for PA, more than for any previous transketolase mutant. R520Q also doubled the specific activity of the previously identified D469T to create our most active transketolase mutant to date. Our results show that recombining active-site mutants obtained by saturation mutagenesis can rapidly destabilise critical networks of co-evolved residues, whereas beneficial single mutants can be retained and improved upon by randomly recombining them with natural variants at other positions in the network.     

Directed evolution of transketolase activity on non-phosphorylated substrates

We have used active-site targeted directed evolution by saturation mutagenesis to improve the activity of E. coli transketolase towards non-phosphorylated substrates. Residues were selected for each set based on either structural proximity to substrate, or on phylogenetic variation. Each library was screened towards the reaction between hydroxypyruvate (HPA) and glycolaldehyde (GA) to form L-erythrulose, and the location of improved mutants related to the natural sequence entropy at each residue. A number of mutants from the phylogenetically defined library were found to outperform the wild-type with up to 3-fold specific activity under biocatalytically relevant conditions, though interestingly with substituted residues that differed from those found in nature. Conserved residues which interact with the phosphate group in natural substrates also yielded mutants with almost 5-fold improved specific activity on the non-phosphorylated substrates. These results suggest that phylogenetically variant active-site residues are useful for modulating activity on natural or structurally-homologous substrates, and that conserved residues which no longer interact with modified target substrates are useful sites to apply saturation mutagenesis for improvement of activity.     

The evolution of catalytic efficiency and substrate promiscuity in human theta class 1-1 glutathione transferase

Theta class glutathione transferases (GST) from various species exhibit markedly different catalytic activities in conjugating the tripeptide glutathione (GSH) to a variety of electrophilic substrates. For example, the human theta 1-1 enzyme (hGSTT1-1) is 440-fold less efficient than the rat theta 2-2 enzyme (rGSTT2-2) with the fluorogenic substrate 7-amino-4-chloromethyl coumarin (CMAC). Large libraries of hGSTT1-1 constructed by error-prone PCR, DNA shuffling, or saturation mutagenesis were screened for improved catalytic activity towards CMAC in a quantitative fashion using flow cytometry. An iterative directed evolution approach employing random mutagenesis in conjunction with homologous recombination gave rise to enzymes exhibiting up to a 20,000-fold increase in k(cat)/K(M) compared to hGSTT1-1. All highly active clones encoded one or more mutations at residues 32, 176, or 234. Combinatorial saturation mutagenesis was used to evaluate the full complement of natural amino acids at these positions, and resulted in the isolation of enzymes with catalytic rates comparable to those exhibited by the fastest mutants obtained via directed evolution. The substrate selectivities of enzymes resulting from random mutagenesis, DNA shuffling, and combinatorial saturation mutagenesis were evaluated using a series of distinct electrophiles. The results revealed that promiscuous substrate activities arose in a stochastic manner, as they did not correlate with catalytic efficiency towards the CMAC selection substrate. In contrast, chimeric enzymes previously constructed by homology-independent recombination of hGSTT-1 and rGSTT2-2 exhibited very different substrate promiscuity profiles, and showed a more defined relationship between evolved and promiscuous activities.     

Systematic exploration of active site mutations on human deoxycytidine kinase substrate specificity

Human deoxycytidine kinase (dCK) is responsible for the phosphorylation of a number of clinically important nucleoside analogue prodrugs in addition to its natural substrates, 2'-deoxycytidine, 2'-deoxyguanosine, and 2'-deoxyadenosine. To improve the low catalytic activity and tailor the substrate specificity of dCK, we have constructed libraries of mutant enzymes and tested them for thymidine kinase (tk) activity. Random mutagenesis was employed to probe for amino acid positions with an impact on substrate specificity throughout the entire enzyme structure, identifying positions Arg104 and Asp133 in the active site as key residues for substrate specificity. Kinetic analysis indicates that Arg104Gln/Asp133Gly creates a "generalist" kinase with broader specificity and elevated turnover for natural and prodrug substrates. In contrast, the substitutions of Arg104Met/Asp133Thr, obtained via site-saturation mutagenesis, yielded a mutant with reversed substrate specificity, elevating the specific constant for thymidine phosphorylation by over 1000-fold while eliminating activity for dC, dA, and dG under physiological conditions. The results illuminate the key contributions of these two amino acid positions to enzyme function by demonstrating their ability to moderate substrate specificity.     

An amino acid at position 142 in nitrilase from Rhodococcus rhodochrous ATCC 33278 determines the substrate specificity for aliphatic and aromatic nitriles

Nitrilase from Rhodococcus rhodochrous ATCC 33278 hydrolyses both aliphatic and aromatic nitriles. Replacing Tyr-142 in the wild-type enzyme with the aromatic amino acid phenylalanine did not alter specificity for either substrate. However, the mutants containing non-polar aliphatic amino acids (alanine, valine and leucine) at position 142 were specific only for aromatic substrates such as benzonitrile, m-tolunitrile and 2-cyanopyridine, and not for aliphatic substrates. These results suggest that the hydrolysis of substrates probably involves the conjugated pi-electron system of the aromatic ring of substrate or Tyr-142 as an electron acceptor. Moreover, the mutants containing charged amino acids such as aspartate, glutamate, arginine and asparagine at position 142 displayed no activity towards any nitrile, possibly owing to the disruption of hydrophobic interactions with substrates. Thus aromaticity of substrate or amino acid at position 142 in R. rhodochrous nitrilase is required for enzyme activity.     

Identification of Crucial Amino Acids in Mouse Aldehyde Oxidase 3 That Determine Substrate Specificity

In order to elucidate factors that determine substrate specificity and activity of mammalian molybdo-flavoproteins we performed site directed mutagenesis of mouse aldehyde oxidase 3 (mAOX3). The sequence alignment of different aldehyde oxidase (AOX) isoforms identified variations in the active site of mAOX3 in comparison to other AOX proteins and xanthine oxidoreductases (XOR). Based on the structural alignment of mAOX3 and bovine XOR, differences in amino acid residues involved in substrate binding in XORs in comparison to AOXs were identified. We exchanged several residues in the active site to the ones found in other AOX homologues in mouse or to residues present in bovine XOR in order to examine their influence on substrate selectivity and catalytic activity. Additionally we analyzed the influence of the [2Fe-2S] domains of mAOX3 on its kinetic properties and cofactor saturation. We applied UV-VIS and EPR monitored redox-titrations to determine the redox potentials of wild type mAOX3 and mAOX3 variants containing the iron-sulfur centers of mAOX1. In addition, a combination of molecular docking and molecular dynamic simulations (MD) was used to investigate factors that modulate the substrate specificity and activity of wild type and AOX variants. The successful conversion of an AOX enzyme to an XOR enzyme was achieved exchanging eight residues in the active site of mAOX3. It was observed that the absence of the K889H exchange substantially decreased the activity of the enzyme towards all substrates analyzed, revealing that this residue has an important role in catalysis.     

Mitochondrial aldehyde dehydrogenase from higher plants

Aldehyde dehydrogenase has been purified to homogeneity from mitochondria of potato tubers and pea epicotyls. Although the enzyme had a high affinity for glycolaldehyde it also had a high affinity for a number of other aliphatic and arylaldehydes. It is proposed that the codification glycolaldehyde dehydrogenase (EC 1.2.1.22) should be abandoned in favour of mitochondrial aldehyde dehydrogenase (EC 1.2.1.3). The purified enzyme showed esterase activity and had properties similar to those reported for the mammalian mitochondrial aldehyde dehydrogenase. Although the natural substrate(s) for the enzyme is not known, the kinetic properties of the enzyme are consistent with it playing a role in the oxidation of acetaldehyde, glycolaldehyde and indoleacetaldehyde.     

Construction of cellobiose phosphorylase variants with broadened acceptor specificity towards anomerically substituted glucosides

The general application of glycoside phosphorylases such as cellobiose phosphorylase (CP) for glycoside synthesis is hindered by their relatively narrow substrate specificity. We have previously reported on the creation of Cellulomonas uda CP enzyme variants with either modified donor or acceptor specificity. Remarkably, in this study it was found that the donor mutant also displays broadened acceptor specificity towards several β‐glucosides. Triple mutants containing donor (T508I/N667A) as well as acceptor mutations (E649C or E649G) also display a broader acceptor specificity than any of the parent enzymes. Moreover, further broadening of the acceptor specificity has been achieved by site‐saturation mutagenesis of residues near the active site entrance. The best enzyme variant contains the additional N156D and N163D mutations and is active towards various alkyl β‐glucosides, methyl α‐glucoside and cellobiose. In comparison with the wild‐type C. uda CP enzyme, which cannot accept anomerically substituted glucosides at all, the obtained increase in substrate specificity is significant. The described CP enzyme variants should be useful for the synthesis of cellobiosides and other glycosides with prebiotic and pharmaceutical properties. Biotechnol. Bioeng. 2010;107: 413–420. © 2010 Wiley Periodicals, Inc.     

Probing the molecular basis of substrate specificity, stereospecificity, and catalysis in the class II pyruvate aldolase, BphI

BphI, a pyruvate-specific class II aldolase found in the polychlorinated biphenyls (PCBs) degradation pathway, catalyzes the reversible C-C bond cleavage of (4S)-hydroxy-2-oxoacids to form pyruvate and an aldehyde. Mutations were introduced into bphI to probe the contribution of active site residues to substrate recognition and catalysis. In contrast to the wild-type enzyme that has similar specificities for acetaldehyde and propionaldehyde, the L87A variant exhibited a 40-fold preference for propionaldehyde over acetaldehyde. The specificity constant of the L89A variant in the aldol addition reaction using pentaldehyde is increased ～50-fold, making it more catalytically efficient for pentaldehyde utilization compared to the wild-type utilization of the natural substrate, acetaldehyde. Replacement of Tyr-290 with phenylalanine or serine resulted in a loss of stereochemical control as the variants were able to utilize substrates with both R and S configurations at C4 with similar kinetic parameters. Aldol cleavage and pyruvate α-proton exchange activity were undetectable in the R16A variant, supporting the role of Arg-16 in stabilizing a pyruvate enolate intermediate. The pH dependence of the enzyme is consistent with a single deprotonation by a catalytic base with pK(a) values of approximately 7. In H20A and H20S variants, pH profiles show the dependence of enzyme activity on hydroxide concentration. On the basis of these results, a catalytic mechanism is proposed.     

Probing the substrate specificity of aminopyrrolnitrin oxygenase (PrnD) by mutational analysis

Molecular modeling and mutational analysis (site-directed mutagenesis and saturation mutagenesis) were used to probe the molecular determinants of the substrate specificity of aminopyrrolnitrin oxygenase (PrnD) from Pseudomonas fluorescens Pf-5. There are 17 putative substrate-contacting residues, and mutations at two of the positions, positions 312 and 277, could modulate the enzyme substrate specificity separately or in combination. Interestingly, several of the mutants obtained exhibited higher catalytic efficiency (approximately two- to sevenfold higher) with the physiological substrate aminopyrrolnitrin than the wild-type enzyme exhibited.     

Purification and characterization of a rat brain aldehyde dehydrogenase able to metabolize gamma-aminobutyraldehyde to gamma-aminobutyric acid.

An enzyme which catalyses dehydrogenation of gamma-aminobutyraldehyde (ABAL) to gamma-aminobutyric acid (GABA) was purified to homogeneity from rat brain tissues by using DEAE-cellulose and affinity chromatography on 5'-AMP-Sepharose, phosphocellulose and Blue Agarose, followed by gel filtration. Such an enzyme was first purified from mammalian brain tissues, and was identified as an isoenzyme of aldehyde dehydrogenase. It has an Mr of 210,000 determined by polyacrylamide-gradient-gel electrophoresis, and appeared to be composed of subunits of Mr 50,000. The close similarity of substrate specificity toward acetaldehyde, propionaldehyde and glycolaldehyde between the enzyme and other aldehyde dehydrogenases previously reported was observed. But substrate specificity of the enzyme toward ABAL was higher than those of aldehyde dehydrogenases from human liver (E1 and E2), and was lower than those of ABAL dehydrogenases from human liver (E3), Escherichia coli and Pseudomonas species. The Mr and relative amino acid composition of the enzyme are also similar to those of E1 and E2. The existence of this enzyme in mammalian brain seems to be related to a glutamate decarboxylase-independent pathway (alternative pathway) for GABA synthesis from putrescine.     

Designs for a broad substrate specificity keto acid dehydrogenase

Variations have been made to the structure of the nicotinamide adenine dinucleotide (NAD) dependent L-lactate dehydrogenase from Bacillus stearothermophilus at regions of the enzyme that we believe determine specificity toward different alpha-hydroxy acids (RCHOHCOO-, R = CH3, C2H5, etc.). Two regions of LDH that border the active site (but are not involved in the catalytic reaction) were altered in order to accommodate substrates with hydrophobic side chains larger than that of the naturally preferred substrate, pyruvate (R = CH3). The mutations 102-105GlnLysPro----MetValSer and 236-237AlaAla----GlyGly were made to increase the tolerance for large hydrophobic substrate side chains. The triple and double mutants alone gave little improvement for branched-chain-substituted pyruvates. The five changes together produced a broader substrate specificity alpha-hydroxy acid dehydrogenase, with a 55-fold improved kcat for alpha-ketoisocaproate to a value about 1/14 that of the native enzyme for pyruvate. Rational protein engineering enabled coupled changes in enzyme structure to be obtained with greater probability of success than random mutagenesis.     

Manipulation of the active site loops of D-hydantoinase, a (beta/alpha)8-barrel protein, for modulation of the substrate specificity

We previously proposed that the stereochemistry gate loops (SGLs) constituting the substrate binding pocket of D-hydantoinase, a (beta/alpha)(8)-barrel enzyme, might be major structural determinants of the substrate specificity [Cheon, Y. H., et al. (2002) Biochemistry 41, 9410-9417]. To construct a mutant D-hydantoinase with favorable substrate specificity for the synthesis of commercially important non-natural amino acids, the SGL loops of the enzyme were rationally manipulated on the basis of the structural analysis and sequence alignment of three hydantoinases with distinct substrate specificities. In the SGLs of D-hydantoinase from Bacillus stearothermophilus SD1, mutations of hydrophobic and bulky residues Met 63, Leu 65, Phe 152, and Phe 159, which interact with the exocyclic substituent of the substrate, induced remarkable changes in the substrate specificities. In particular, the substrate specificity of mutant F159A toward aromatic substrate hydroxyphenylhydantoin (HPH) was enhanced by approximately 200-fold compared with that of the wild-type enzyme. Saturation mutagenesis at position 159 revealed that k(cat) for aromatic substrates increased gradually as the size of the amino acid side chain decreased, and this seems to be due to reduced steric hindrance between the bulky exocyclic group of the substrate and the amino acid side chains. When site-directed random mutagenesis of residues 63 and 65 was conducted with the wild type and mutant F159A, the selected enzymes (M63F/L65V and L65F/F159A) exhibited approximately 10-fold higher k(cat) values for HPH than the wild-type counterpart, which is likely to result from reorganization of the active site for efficient turnover. These results indicate that the amino acid residues of SGLs forming the substrate binding pocket are critical for the substrate specificity of D-hydantoinase, and the results also imply that substrate specificities of cyclic amidohydrolase family enzymes can be modulated by rational design of these SGLs.     

A highly active adipyl-cephalosporin acylase obtained via rational randomization

There is strong interest in creating an enzyme that can deacylate natural cephalosporins such as cephalosporin C in order to efficiently acquire the starting compound for the industrial production of semisynthetic cephalosporin antibiotics. In this study, the active site of the glutaryl acylase from Pseudomonas SY-77 was randomized rationally. Several mutations that were found in previous studies to enhance the activity of the enzyme towards adipyl-7-aminodesacetoxycephalosporanic acid (ADCA) and cephalosporin C have now been combined, and libraries have been made in which random amino acid substitutions at these positions are joined. The mutants were expressed in a leucine-deficient Escherichia coli strain and subjected to growth selection with adipyl-leucine or amino-adipyl-leucine as sole leucine source. The mutants growing on these media were selected and purified, and their hydrolysis activities towards adipyl-7-ADCA and cephalosporin C were tested. Several mutants with highly improved activities towards the desired substrates were found in these rationally randomized libraries. The best mutant was selected from a library of totally randomized residues: 178, 266, and 375. This mutant comprises two mutations, Y178F + F375H, which synergistically improve the catalytic efficiency towards adipyl-7-ADCA 36-fold. The activity of this mutant towards adipyl-7-ADCA is 50% of the activity of the wild-type enzyme towards the preferred substrate glutaryl-7-aminocephalosporanic acid, and therefore the characteristics of this mutant approach those needed for industrial application.     

Improvement of thermostable aldehyde dehydrogenase by directed evolution for application in Synthetic Cascade Biomanufacturing

The aldehyde dehydrogenase from Thermoplasma acidophilum, which was previously implemented as a key enzyme in a synthetic cell-free reaction cascade for the production of alcohols, was optimized by directed evolution. Improvements have been made to enhance reaction velocity and solubility. Using a random approach followed by site-directed and saturation mutagenesis, three beneficial amino acid mutations were found after screening of ca. 20,000 variants. Mutation Y399C enhanced the protein solubility after recombinant expression in Escherichia coli 6-fold. Two further mutations, F34M and S405N, enhanced enzyme activity with the cofactor NAD⁺ by a factor of eight. Impacts on enzyme stability and substrate specificity were negligible.     

Purification and characterization of human liver "high Km" aldehyde dehydrogenase and its identification as glutamic gamma-semialdehyde dehydrogenase

The enzyme previously considered as an isozyme (E4, ALDH IV) of human liver aldehyde dehydrogenase (NAD+) (EC 1.2.1.3) has been purified to homogeneity by the use of ion exchange chromatography on CM-Sephadex and affinity chromatography on Blue Sepharose CL-6B and 5'-AMP Sepharose 4B and identified as glutamic gamma-semialdehyde dehydrogenase, or more precisely 1-pyrroline-5-carboxylate dehydrogenase (EC 1.5.1.12). Glutamic gamma-semialdehyde dehydrogenase was never previously purified to homogeneity from any mammalian species. The homogeneous enzyme is seen on isoelectric focusing gels as two fine bands separated by 0.12 pH units: pI = 6.89 and 6.77. In addition, the enzyme also appears as two bands in gradient gels; however, in polyacrylamide gels containing sodium dodecyl sulfate the enzyme migrates as one band, indicating that its subunits are of identical size. Because the enzyme molecule is considerably smaller (Mr approximately 142,000-170,000) than that of aldehyde dehydrogenases (EC 1.2.1.3) (Greenfield, N. J., and Pietruszko, R. (1977) Biochim. Biophys. Acta 483, 35-45; Mr approximately 220,000) and its subunit weight is different (70,600 versus approximately 54,000 for E1 and E2 isozymes), the enzyme is not an isozyme of aldehyde dehydrogenase previously described. The Michaelis constants for glutamic gamma-semialdehyde dehydrogenase with acetaldehyde and propionaldehyde are in the millimolar range. Its substrate specificity within the straight chain aliphatic aldehyde series is essentially confined to that of acetaldehyde and propionaldehyde with butyraldehyde and longer chain length aldehydes being considerably less active. Other substrates include succinic, glutaric, and adipic semialdehydes in addition to glutamic gamma-semialdehyde. The reaction velocity with glutamic gamma-semialdehyde is at least an order of magnitude larger than with carboxylic acid semialdehydes. Aspartic beta-semialdehyde is not a substrate. The reaction catalyzed appears to be irreversible. Although NADP can be used, NAD is the preferred coenzyme. The enzyme also exhibits an unusual property of being subject to substrate inhibition by NAD.     

Engineering delta 9-16:0-acyl carrier protein (ACP) desaturase specificity based on combinatorial saturation mutagenesis and logical redesign of the castor delta 9-18:0-ACP desaturase

Six amino acid locations in the soluble castor Delta(9)-18:0-acyl carrier protein (ACP) desaturase were identified that can affect substrate specificity. Combinatorial saturation mutagenesis of these six amino acids, in conjunction with selection, using an unsaturated fatty acid auxotroph system, led to the isolation of variants with up to 15-fold increased specific activity toward 16-carbon substrates. The most improved mutant, com2, contained two substitutions (T117R/G188L) common to five of the 19 complementing variants subjected to further analysis. These changes, when engineered into otherwise wild-type 18:0-ACP desaturase to make mutant 5.2, produced a 35-fold increase in specific activity with respect to 16-carbon substrates. Kinetic analysis revealed changes in both k(cat) and K(m) that result in an 82-fold improvement in specificity factor for 16-carbon substrate compared with wild-type enzyme. Improved substrate orientation apparently compensated for loss of binding energy that results from the loss of desolvation energy for 16-carbon substrates. Mutant 5.2 had specific activity for 16-carbon substrates 2 orders of magnitude higher than those of known natural 16-carbon specific desaturases. These data support the hypothesis that it should be possible to reengineer archetypal enzymes to achieve substrate specificities characteristic of recently evolved enzymes while retaining the desired stability and/or turnover characteristics of a parental paralog.     

Probing the molecular determinants of aniline dioxygenase substrate specificity by saturation mutagenesis

Aniline dioxygenase is a multicomponent Rieske nonheme-iron dioxygenase enzyme isolated from Acinetobacter sp. strain YAA. Saturation mutagenesis of the substrate-binding pocket residues, which were identified using a homology model of the alpha subunit of the terminal dioxygenase (AtdA3), was used to probe the molecular determinants of AtdA substrate specificity. The V205A mutation widened the substrate specificity of aniline dioxygenase to include 2-isopropylaniline, for which the wild-type enzyme has no activity. The V205A mutation also made 2-isopropylaniline a better substrate for the enzyme than 2,4-dimethylaniline, a native substrate of the wild-type enzyme. The I248L mutation improved the activity of aniline dioxygenase against aniline and 2,4-dimethylaniline approximately 1.7-fold and 2.1-fold, respectively. Thus, it is shown that the alpha subunit of the terminal dioxygenase indeed plays a part in the substrate specificity as well as the activity of aniline dioxygenase. Interestingly, the equivalent residues of V205 and I248 have not been previously reported to influence the substrate specificity of other Rieske dioxygenases. These results should facilitate future engineering of the enzyme for bioremediation and industrial applications.     

Virtual screening of mandelate racemase mutants with enhanced activity based on binding energy in the transition state

Mandelate racemase (MR) is a promising candidate for the dynamic kinetic resolution of racemates. However, the poor activity of MR towards most of its non-natural substrates limits its widespread application. In this work, a virtual screening method based on the binding energy in the transition state was established to assist in the screening of MR mutants with enhanced catalytic efficiency. Using R-3-chloromandelic acid as a model substrate, a total of 53 mutants were constructed based on rational design in the two rounds of screening. The number of mutants for experimental validation was brought down to 17 by the virtual screening method, among which 14 variants turned out to possess improved catalytic efficiency. The variant V26I/Y54V showed 5.2-fold higher catalytic efficiency (k_cat/K_m) towards R-3-chloromandelic acid than that observed for the wild-type enzyme. Using this strategy, mutants were successfully obtained for two other substrates, R-mandelamide and R-2-naphthylglycolate (V26I and V29L, respectively), both with a 2-fold improvement in catalytic efficiency. These results demonstrated that this method could effectively predict the trend of mutational effects on catalysis. Analysis from the energetic and structural assays indicated that the enhanced interactions between the active sites and the substrate in the transition state led to improved catalytic efficiency. It was concluded that this virtual screening method based on the binding energy in the transition state was beneficial in enzyme rational redesign and helped to better understand the catalytic properties of the enzyme.     

Kinetic analysis of phenylalanine dehydrogenase mutants designed for aliphatic amino acid dehydrogenase activity with guidance from homology-based modelling.

Through comparison with the high-resolution structure of Clostridium symbiosum glutamate dehydrogenase, the different substrate specificities of the homologous enzymes phenylalanine dehydrogenase and leucine dehydrogenase were attributed to two residues, glycine 124 and leucine 307, in Bacillus sphaericus phenylalanine dehydrogenase, which are replaced with alanine and valine in leucine dehydrogenases. As predicted, making these substitutions in phenylalanine dehydrogenase decreased the specific activity towards aromatic substrates and enhanced the activity towards some aliphatic amino acids in standard assays with fixed concentrations of both substrates. This study did not, however, distinguish effects on affinity from those on maximum catalytic rate. A fuller kinetic characterization of the single- and double-mutant enzymes now reveals that the extent of the shift in specificity was underestimated in the earlier study. The maximum catalytic rates for aromatic substrates are reduced for all the mutants, but, in addition, the apparent Km values are higher for the single-mutant G124A and double-mutant G124A/L307V compared with the wild-type enzyme. Conversely, specificity constants (kcat/Km) for the nonpolar aliphatic amino acids and the corresponding 2-oxoacids for the mutants are all markedly higher than for the wild type, with up to a 40-fold increase for l-norvaline and a 100-fold increase for its 2-oxoacid in the double mutant. In some cases a favourable change in Km was found to outweigh a smaller negative change in kcat. These results emphasize the risk of misjudging the outcome of protein engineering experiments through too superficial an analysis. Overall, however, the success of the predictions from molecular modelling indicates the usefulness of this strategy for engineering new specificities, even in advance of more detailed 3D structural information.