Enhancement of thermostability of fungal deglycating enzymes by directed evolution

Fructosyl peptide oxidases are valuable for the determination of glycoproteins such as hemoglobin A1c. For practical use in clinical diagnosis, we applied directed evolution to improve the thermostability of these enzymes. After two rounds of random mutagenesis and high-throughput screening, six thermostabilizing amino acid substitutions were identified. Therefore, site-directed and cassette mutageneses were applied to combine these six stabilizing mutations. The simultaneous mutants showed that the stabilizing effect of the amino acid replacement was cumulative. The sextuple mutant enzyme, R94K/G184D/F265L/N272D/H302R/H388Y, had a half-life of thermal inactivation at 50°C that was 79.8-fold longer than that of the parental fructosyl peptide oxidase. The thermostable variants also showed increased tolerance to digestion by a protease. The sextuple mutant enzyme did not lose its activity on incubation with neutral protease, while the wild-type enzyme almost completely lost its activity. Furthermore, three amino acid substitutions were introduced into another fructosyl peptide oxidase with a different substrate specificity. The half-life of inactivation at 50°C was 3.61-fold longer than that of the parent enzyme. These engineered fructosyl peptide oxidases will be useful for industrial application to clinical diagnosis.     

Recombinant agrobacterium AgaE-like protein with fructosyl amino acid oxidase activity

Agrobacterium tumefaciens AgaE-like protein had a similar sequence to that of a fructosyl amino acid oxidase from Corynebacterium sp. strain 2-4-1. To characterize the AgaE-like protein, we produced the enzyme in Escherichia coli, and purified it to homogeneity. The molecular mass of recombinant AgaE-like protein was 42 kDa on SDS-PAGE and 85 kDa on gel filtration. The protein acted on N-fructosyl valine and N-fructosyl glycine as substrates, but not on glycated protein or N(epsilon)-fructosyl lysine. Apparent Km for N-fructosyl valine and N-fructosyl glycine were 1.64 and 0.31 mM, respectively. The AgaE-like protein had maximum activity at pH 7.8 and 35 degrees C in 0.1 M potassium phosphate, but more than 80% of its activity was lost at 40 degrees C or more. In contrast to eukaryotic fructosyl amino acid oxidases, the AgaE-like protein contained noncovalently bound FAD as a cofactor and was inactive against N(epsilon)-fructosyl N(alpha)-Z(benzyloxycarbonyl)-lysine. These characteristics were similar to a fructosyl amino acid oxidase from Corynebacterium sp. strain 2-4-1, suggesting that these prokaryotic enzymes comprise a new family of fructosyl amino acid oxidases.     

Enzymes used for the determination of HbA1C

To develop an enzymatic measurement of HbA(1C), two key enzymes, i.e., fructosyl peptide oxidase and Aspergillus protease were characterized. Fructosyl peptide oxidase from Eupenicillium terrenum was a flavoenzyme that could catalyze the oxidation of N-(1-deoxyfructosyl)-Val-His. The enzyme showed high specificity toward alpha-glycated molecules, therefore it seemed suitable for the HbA(1C) assay. Since high levels of FPOX expression seemed toxic to host cells, we applied a gene expression system using a bacteriophage vector and achieved high levels of expression in Escherichia coli. Next, we found that Aspergillus protease was able to digest N-(1-deoxyfructosyl)-hexapeptide, a glycated peptide that was released from the beta-chain of HbA(1C) by Glu-C endoproteinase. We showed that the N-(1-deoxyfructosyl)-Val-His released from N-(1-deoxyfructosyl)-hexapeptide by Aspergillus protease could be assayed enzymatically using fructosyl peptide oxidase, therefore these enzymes could be applied to the enzymatic measurement of HbA(1C).     

Molecular cloning and expression of novel fructosyl peptide oxidases and their application for the measurement of glycated protein

Fructosyl peptide oxidases, enzymes that are active against a model compound of glycated hemoglobin, N(alpha)-fructosyl valyl-histidine, were characterized. To identify the primary structure of fructosyl peptide oxidases, we have prepared cDNA libraries from Eupenicillium terrenum ATCC18547 and Coniochaeta sp. NISL9330. The coding regions, both fungal fructosyl peptide oxidases consisting of 1314-bp, were obtained with degenerated primers based on the amino acid sequences and specific primers by 3(') and 5(') RACE (rapid amplification of cDNA ends). By their sequence similarities and substrate specificities, fructosyl peptide oxidases and their homologs could be categorized into two groups: (A) enzymes that preferably oxidize alpha-glycated molecules and (B) enzymes that preferably oxidize epsilon-glycated molecules. We showed that recombinant fructosyl peptide oxidases could be used to detect protease-treated fructosyl-hexapeptide, a glycated peptide that is released from HbA(1C) by endoproteinase Glu-C, suggesting these enzymes could be useful for the enzymatic measurement of HbA(1C).     

A continuous enzyme assay and characterisation of fructosyl amine oxidase enzymes (EC 1.5.3)

Enzymatic reversal of the Maillard reaction is a growing area of research. Fructosyl amine oxidase enzymes (EC 1.5.3) have attracted recent attention through demonstration of their ability to deglycate Amadori products, low molecular weight intermediates formed during the early stage of the Maillard reaction. Although stopped assays have been described, a bottleneck in current studies is the lack of continuous kinetic assays. Here, we describe the development of a continuous, coupled enzyme assay and its successful application to determining optimal storage conditions and the steady-state kinetic parameters of an enzyme from this group, amadoriase I. A K(m)(app) of 11 microM and a K(cat)(app) of 3.5s(-1) were determined using this assay using fructosyl propylamine as a substrate, which differ from previous reports. This method was also used to test the activity of two site-directed mutants of amadoriase I, H357N and S370A, which were found to be catalytically inactive.     

Engineering of dye-mediated dehydrogenase property of fructosyl amino acid oxidases by site-directed mutagenesis studies of its putative proton relay system

The flavoenzyme fructosyl amino acid oxidase (FAOD) catalyzes the oxidative deglycation of fructosyl amino acids, model compounds of glycated proteins. The high oxygen reactivity of FAODs limits their potential utility in amperometric enzyme sensors employing artificial electron mediators. To alter their electron acceptor availability, site-directed mutagenesis was carried out on conserved residues predicted to be involved in the proton relay system (PRS) of two eukaryotic FAODs, the FAOD from the marine yeast Pichia sp. N1-1 and amadoriase II from the fungus Aspergillus fumigatus. The substitution of a single conserved Asn residue in the putative PRS, Asn47Ala of N1-1 FAOD and Asn52Ala of amadoriase II, resulted in significant loss in the catalytic ability to employ O₂ as the electron acceptor, while having little effect on the dye-mediated dehydrogenase activity employing artificial electron acceptors instead of O₂.     

An electrochemical biosensor for fructosyl valine for glycosylated hemoglobin detection based on core-shell magnetic bionanoparticles modified gold electrode

A high-performance amperometric fructosyl valine (FV) biosensor was developed, based on immobilization of fructosyl amino-acid oxidase (FAO) on core-shell magnetic bionanoparticles modified gold electrode. Chitosan was used to introduce amino groups onto the surface of core-shell magnetic bionanoparticles (MNPs). With FAO as an enzyme model, a new fructosyl valine biosensor was fabricated. The biosensor showed optimum response, when operated at 50 mVs(-1) in 0.1M potassium phosphate buffer, pH 7.5 and 35°C. The biosensor exhibited excellent sensitivity [the detection limit is down to 0.1mM for FV], fast response time (less than 4s), wide linear range (from 0 to 2mM). Analytical recovery of added FV was 95.00-98.50%. Within batch and between batch coefficients of variation were <2.58% and <5.63%, respectively. The enzyme electrode was used 250 times over 3 months, when stored at 4°C.     

An enzymatic method for the determination of hemoglobinA<Subscript>1C</Subscript>

Fructosyl peptide oxidase is a flavoenzyme that catalyzes the oxidative deglycation of N-(1-deoxyfructosyl)-Val-His, a model compound of hemoglobin (Hb)A_1C. To develop an enzymatic method for the measurement of HbA_1C, we screened for a proper protease using N-(1-deoxyfructosyl)-hexapeptide as a substrate. Several proteases, including Neutral protease from Bacillus polymyxa, were found to release N-(1-deoxyfructosyl)-Val-His efficiently, however no protease was found to release N-(1-deoxyfructosyl)-Val. Neutral protease also digested HbA_1C to release N-(1-deoxyfructosyl)-Val-His, and then the fructosyl peptide was detected using fructosyl peptide oxidase. The linear relationship was observed between the concentration of HbA_1C and the absorbancy of fructosyl peptide oxidase reaction, hence this new method is a practical means for measuring HbA_1C.     

Purification and Characterization of Wheat beta(2-->1) Fructan:Fructan Fructosyl Transferase Activity

Fructans are the major storage carbohydrate in vegetative tissues of wheat (Triticum aestivum L.). Fructan:fructan fructosyl transferase (FFT) catalyzes fructosyl transfer between fructan molecules to elongate the fructan chain. The objective of this research was to isolate this activity in wheat. Wheat (cv Caldwell) plants grown at 25°C for 3 weeks were transferred to 10°C to induce fructan synthesis. From the leaf blades kept at 10°C for 4 days, fructosyl transferase activity was purified using salt precipitation and a series of chromatographic procedures including size exclusion, anion-exchange, and affinity chromatography. The transferase activity was free from invertase and other fructan-metabolizing activities. Fructosyl transferase had a broad pH spectrum with a peak activity at 6.5. The temperature optimum was 30°C. The activity was specific for fructosyl transfer from β(2→1)-linked 1-kestose or fructan to sucrose and β(2→1) fructosyl transfer to other fructans (1-FFT). Fructosyl transfer from oligofructans to sucrose was most efficient when 1-kestose was used as donor molecule and declined as the degree of polymerization of the donor increased from 3 to 5. 1-FFT catalyzed the in vitro synthesis of inulin tetra- and penta-saccharides from 1-kestose; however, formation of the tetrasaccharide was greatly reduced at high sucrose concentration. 6-Kestose could not act as donor molecule, but could accept a fructosyl moiety from 1-kestose to produce bifurcose and a tetrasaccharide having a β(2→1) fructose attached to the terminal fructose of 6-kestose. The role of this FFT activity in the synthesis of fructan in wheat is discussed.     

Glucofructosan biosynthesis in Fusarium oxysporum

Low MW glucofructosans have been detected in the medium of Fusarium oxysporum. A 53-fold purification of fructosyl transferase has been achieved by ethanol precipitation, DEAE-cellulose and Sephadex G-100 column chromatography. Maximum fructosyl transferase activity coincided with maximum glucofructosan concentration in the medium. Invertase showed greatest activity in the later stages of growth when glucofructosans were absent. Fructosyl transferase and invertase have been separated by DEAE-cellulose column chromatography. On the basis of kinetic studies and effect of nucleotides on fructosyl transferase in the presence and absence of MgCl₂, a two site active centre linked through a nucleotide bridge is proposed. Fructosyl transferase and invertase are highly phosphorylated.     

Active site analysis of fructosyl amine oxidase using homology modeling and site-directed mutagenesis

A three-dimensional structural model of fructosyl amine oxidase from the marine yeast Pichia N1-1 was generated using the crystal structure of monomeric sarcosine oxidase from Bacillus sp. B-0618 as template. The putative active site region was investigated by site-directed mutagenesis, identifying several amino acid residues likely playing important roles in the enzyme reaction. Asn354 was identified as a residue that plays an important role in substrate recognition and that can be substituted in order to change substrate specificity while maintaining high catalytic activity. While the Asn354Ala substitution had no effect on the V _max K _m⁻¹ value for fructosyl valine, the V _max K _m⁻¹ value for fructosyl-^ε N-lysine was decreased 3-fold, thus resulting in a 3-fold improvement in specificity for fructosyl valine over fructosyl-^ε N-lysine.     

Loop engineering of amadoriase II and mutational cooperativity

Amadori compounds and their cross-linked products have been implicated in diabetic complications and some age-related diseases. Fructosyl amine oxidases (FAOXs) are a family of enzymes that can cleave the amadori compounds. However, the natural enzymes are only active on small substrates (fructosyl amino acids or dipeptides), which limits the therapeutic and diagnostic applications of these enzymes. In this study, amadoriase II, a member of the FAOX family from Aspergillus fumigatus was engineered to broaden its substrate range using a modified combinatorial active site saturation testing approach. The two loops at the entrance of the substrate channel were targeted. Saturation mutagenesis was carried out to search for hot-spot sites, followed by pairwise mutagenesis and subsequent combination of active mutations. Five sites on the loops were found to be critical for accessibility for two model bulky substrates, fructosyl adamantanamine and fructosyl-polylysine (3–13 lysines). Two best mutants (with three and five mutations, respectively) were obtained, with a specific activity toward the model substrates 20.6-fold and 16.8-fold that of the wild-type, respectively. Deconvolution experiments revealed the cooperativity of the mutations.     

Occurrence of fructosyl-amino acid oxidase-reactive compounds in fungal cells

Fructosyl-amino acid oxidase (FAOD)-reactive fraction (FRY) was found in commercial yeast extract. FRY showed very hydrophilic property and was adsorbed to phenylboronate silica gel, indicating that it contained the Amadori compound. TLC and amino acid analyses revealed that glucosone, lysine, and arginine were produced from FRY after incubation with FAOD. TOF-MS analysis confirmed that FRY is a mixture of fructosyl lysine and fructosyl arginine. These compounds were also detected in mycelial extract of an FAOD-producer, Aspergillus terreus GP1, grown on the minimum medium, suggesting that a glycation reaction occurs in fungal cells and that FAOD acts toward the resultant Amadori compounds.     

Occurrence of Fructosyl-Amino Acid Oxidase-Reactive Compounds in Fungal Cells

Fructosyl-amino acid oxidase (FAOD)-reactive fraction (FRY) was found in commercial yeast extract. FRY showed very hydrophilic property and was adsorbed to phenylboronate silica gel, indicating that it contained the Amadori compound. TLC and amino acid analyses revealed that glucosone, lysine, and arginine were produced from FRY after incubation with FAOD. TOF-MS analysis confirmed that FRY is a mixture of fructosyl lysine and fructosyl arginine. These compounds were also detected in mycelial extract of an FAOD-producer, Aspergillus terreus GP1, grown on the minimum medium, suggesting that a glycation reaction occurs in fungal cells and that FAOD acts toward the resultant Amadori compounds.     

Distribution and properties of novel deglycating enzymes for fructosyl peptide in fungi

Our fungal culture collection was screened for fructosyl peptide oxidase, an enzyme that could be used for the determination of glycated hemoglobin in diabetic subjects with hyperglycemia. Fructosyl peptide oxidases were found in strains of eight genera: Achaetomiella, Achaetomium, Chaetomium, Coniochaeta, Eupenicillium, Gelasinospora, Microascus and Thielavia. By their substrate specificity toward N-fructosyl valyl-histidine (-keto-amine) and N-fructosyl lysine (-keto-amine), fructosyl peptide oxidases could be categorized into two groups: (1) enzymes that oxidize both -keto-amine and -keto-amine, and (2) enzymes that preferably oxidize -keto-amine. A fructosyl peptide oxidase from Achaetomiella virescens ATCC 32393, active toward both N-fructosyl valyl-histidine and N-fructosyl lysine, was purified to homogeneity and characterized. The enzyme was monomeric (M_r=50,000), was most active at 40 °C and pH 8.0, and had a covalently bound flavin as a prosthetic group. Apparent K_m values for N-fructosyl valyl-histidine and N-fructosyl lysine were 2.30 and 1.69 mM, respectively. N-fructosyl valyl-histidine was consumed and the same molar amount of valyl-histidine was produced by the fructosyl peptide oxidase reaction. This enzyme could be useful for the measurement of hemoglobin A_1C, the N-terminal valine residue of the -subunit of which is glycated.Abbreviations HbA_1C Hemoglobin A_1C - FPOX Fructosyl peptide oxidase - FAOX Fructosyl amino acid oxidase - Fru-ValHis N-fructosyl valyl-histidine - Fru-Val N-fructosyl valine - Fru-Lys N-fructosyl lysine - Fru-Gly Fructosyl glycine - TOOS N-ethyl-N-(2-hydroxy-3-sulfopropyl)-3-methylaniline, sodium salt     

Novel fluorescent sensing system for alpha-fructosyl amino acids based on engineered fructosyl amino acid binding protein

A novel fluorescent sensing system for alpha-glycated amino acids was created based on fructosyl amino acid binding protein (FABP) from Agrobacterium tumefaciens. The protein was found to bind specifically to the alpha-glycated amino acids fructosyl glutamine (Fru-Gln) and fructosyl valine (Fru-Val) while not binding to epsilon-fructosyl lysine. An Ile166Cys mutant of FABP was created by genetic engineering and modified with the environmentally sensitive fluorophore acrylodan. The acrylodan-conjugated mutant FABP showed eight-fold greater sensitivity to Fru-Val than the unconjugated protein and could detect concentrations as low as 17 nM, making it over 100-fold more sensitive than enzyme-based detection systems. Its high sensitivity and specificity for alpha-substituted fructosyl amino acids makes the new sensing system ideally suited for the measurement of hemoglobin A1c (HbA1c), a major marker of diabetes.     

Molecular characterization of the acyl-coenzyme A:isopenicillin N acyltransferase gene (penDE) from Penicillium chrysogenum and Aspergillus nidulans and activity of recombinant enzyme in Escherichia coli.

The final step in the biosynthesis of beta-lactam antibiotics in Penicillium chrysogenum and Aspergillus nidulans involves removal of the L-alpha-aminoadipyl side chain from isopenicillin N (IPN) and exchange with a nonpolar side chain. The enzyme catalyzing this reaction, acyl-coenzyme A:isopenicillin N acyltransferase (acyltransferase), was purified from P. chrysogenum and A. nidulans. Based on NH2-terminal amino acid sequence information, the acyltransferase gene (penDE) from P. chrysogenum and A. nidulans were cloned. In both organisms, penDE was located immediately downstream from the isopenicillin N synthetase gene (pcbC) and consisted of four exons encoding an enzyme of 357 amino acids (approximately 40 kilodaltons [kDa]). The DNA coding sequences showed approximately 73% identity, while the amino acid sequences were approximately 76% identical. Noncoding DNA regions (including the region between pcbC and penDE) were not conserved. Acyltransferase activity from Escherichia coli producing the 40-kDa protein accepted either 6-aminopenicillanic acid or IPN as the substrate and made a penicillinase-sensitive antibiotic in the presence of phenylacetyl coenzyme A. Therefore, a single gene is responsible for converting IPN to penicillin G. The active form of the enzyme may result from processing of the 40-kDa monomeric precursor to a heterodimer containing subunits of 11 and 29 kDa.     

Molecular cloning of the L-amino-acid oxidase gene from Neurospora crassa

The addition of D-phenylalanine to starved cultures of Neurospora crassa leads to de novo synthesis of L-amino-acid oxidase. Poly(A) RNA from D-phenylalanine-treated mycelium was therefore used to generate a cDNA library which was subsequently screened by hybrid-selected translation. A positive L-amino-acid oxidase clone served as a probe to isolate the complete gene from a genomic library of N. crassa. The nucleotide sequence obtained revealed an open reading frame coding for a protein of 695 amino acids. A comparison of the deduced primary structure with the partial amino-terminal sequence of the isolated enzyme showed that the protein is synthesized as a precursor. The proform exceeds the mature enzyme by 129 amino acids. The presence of a cluster of basic amino acid residues preceding Ala129 in the precursor suggests a post-translational modification brought about by limited proteolysis. N. crassa L-amino-acid oxidase shares a highly conserved region with many well-characterized flavoproteins that is known to constitute part of the flavin-adenine dinucleotide-binding site.