Introduction

Thermococcus litoralis DSM 5473 is a hyperthermophilic archaeon and is able to grow under anaerobic conditions at temperatures between 55 and 98 °C, growing optimally at 88 °C (Neuner et al. 1990). In general, enzymes from hyperthermophiles are highly thermostable and have been used for the production of various industrial materials, analytical reagents, and biosensors. The nucleic acid- and carbohydrate-metabolizing enzymes from various strains of Thermococcus have been extensively studied. For example, a DNA polymerase (EC 2.7.7.7) from Thermococcus kodakarensis, KOD1, is a thermostable and highly accurate enzyme that has been used in PCR (Nishioka et al. 2001). 4-Alpha-glucanotransferase (EC 2.4.1.56) from T. kodakarensis catalyzes the intermolecular transglycosylation of starch and is useful for industrial starch processing (Tachibana et al. 2000).

Amino acid oxidase is an important amino acid-metabolizing enzyme in various organisms and catalyzes the oxidation of amino acids to 2-oxo acids, ammonia, and hydrogen peroxide in the presence of FAD, which is used as a cofactor. Several types of amino acid oxidases, organized by their substrate specificities, have been reported. d-Amino acid oxidase (DAO, EC 1.4.3.3) specifically acts on d-amino acids and has been isolated from Rhodotorula gracilis (Pilone et al. 1989), Arthrobacter protophormia (Geueke et al. 2007), Rubrobacter xylanophilus (Takahashi et al. 2014), and other organisms. The physiological function of DAO is to catabolize d-amino acids in various organisms (Nishikawa 2005). d-Aspartate oxidase (DDO, EC 1.4.3.1) acts mainly on d-aspartic acid and was isolated from Bos taurus, Homo sapiens, Octopus vulgaris, and others. DDO is known to control the d-aspartic acid concentration in mammalian cells (Schell et al. 1997). l-Amino acid oxidase (LAAO, EC 1.4.3.2) acts on various l-amino acids such as l-alanine, l-arginine, l-lysine, l-histidine, and l-leucine and has been isolated from Crotalus adamanteus (Wellner and Meister 1960), Rhizoctonia solani (Hahn et al. 2017), Pseudomonas sp. AIU 813 (Isobe et al. 2012) and others. Snake venom LAAO (and its function) is the most studied LAAO (Mot et al. 2017, Stabeli et al. 2004; Braga et al. 2008; Vargas et al. 2013); the LAAO from king cobra (Ophiophagus hannah) venom showed antibacterial activity. l-Glutamate oxidase (LGOX, EC 1.4.3.11) mainly acts on l-glutamic acid and has specifically been isolated from several strains of Streptomyces, including Streptomyces endus and Streptomyces violascens. LGOX is considered to be an important enzyme in l-glutamate metabolism in Streptomyces species (Bohmer et al. 1989). l-Aspartate oxidase (LASPO, EC 1.4.3.16) acts mainly on l-aspartate and has been isolated from Arabidopsis thaliana (Katoh et al. 2006), Escherichia coli (Flachmann et al. 1988), Sulfolobus tokodaii (Bifulco et al. 2013), and others. LASPO is physiologically involved in the first reaction of NAD+ de novo biosynthesis in various organisms. However, studies on LASPO have rarely been compared with those of DDO, DAO, LAAO and LGOX.

We found that an l-aspartate oxidase homolog (OCC_06611) exists in the T. litoralis DSM 5473 genome. The gene product of OCC_06611 is expected to be a thermostable l-aspartate oxidase and to be applicable for measurement of l-aspartic acid. In this study, we described the expression of the l-aspartate oxidase homolog gene of T. litoralis DSM 5473 in Escherichia coli and characterized the enzymological properties of its gene product.

Materials and methods

Materials

Escherichia coli NovaBlue and Escherichia coli BL21 (DE3) cells and the pT7blue-2 T-Vector and pET21b vector were purchased from Novagen (Madison, WI, USA). KOD-plus-ver.2 polymerase, DNA ligation kit, and restriction enzymes (NdeI and XhoI) were purchased from Takara bioscience (Kyoto, Japan). All other reagents were of the best commercially available grade and were purchased from Wako chemicals (Osaka, Japan), Kanto chemicals (Tokyo, Japan), Nacalai tesque (Kyoto, Japan), and Sigma Aldrich (St. Louis, USA).

Cloning and expression of l-aspartate oxidase homolog gene from Thermococcus litoralis DSM 5473

The cultivation and genome DNA extraction of T. litoralis DSM 5473 were carried out according to the methods that we previously reported (Washio et al. 2016). PCR amplification of the l-aspartate oxidase homolog gene (Tl-laspo; accession no. EHR78939) was carried out with a pair of primers [Tl-laspo N (with NdeI site): TGTTCGGGTAGACATATGACA, Tl-laspo C (with XhoI site): CTCGAGTAACATACACCTCCC]. PCR was carried out under the following conditions: polymerase activation at 94 °C for 2 min; denaturation at 98 °C for 30 s; annealing at 55 °C for 30 s; and extension at 68 °C for 90 s. The reaction mixture (total volume: 25 μL) contained 1 × buffer for KOD-plus-ver.2, 0.2 mM dNTP, 1.5 mM MgSO4, primer (Tl-laspo N or Tl-laspo C (15 pmol)), KOD-plus-ver.2 (1 U), and ultrapure water. PCR products were purified using agarose gel electrophoresis and treated with 10 × A-attachment mix (Toyobo, Osaka, Japan). The reaction product was ligated into a pT7blue2 T-Vector using a Takara DNA ligation kit. The obtained product, the pT7blue2-Tl-laspo vector, was transformed into Escherichia coli NovaBlue to replicate, and its DNA sequence was analyzed using an Applied Biosystems™ 3130xl DNA analyzer (Applied Biosystems Life Technologies, Foster City, CA, USA). The pT7blue2-Tl-laspo vector was digested with NdeI and XhoI, and the Tl-laspo fragment containing NdeI and XhoI sites was purified using agarose gel electrophoresis. The Tl-laspo fragment that was obtained was ligated to a pET21b vector that had been digested with NdeI and XhoI. The pET21b-Tl-laspo vector that was resulted was transformed into Escherichia coli BL21 (DE3). E. coli BL21 (DE3) harboring pET21b-Tl-laspo was cultivated at 30 °C for 16 h in a test tube (ϕ 18 × 180 mm) containing 5 mL of an LB medium plus 100 µg/mL ampicillin (Medium A) on a reciprocal shaker (Personal-11, Taitec, Saitama, Japan; 140 rpm) under aerobic conditions. Glycerol stocks of E. coli BL21 (DE3) harboring pET21b-Tl-laspo were prepared by mixing 120 µL of its culture with 30 µL of 80% (w/v) glycerol solution that had been sterilized by autoclave (121 °C, 20 min), and the resulting mixture was stored at − 80 °C.

Purification of recombinant l-aspartate oxidase homolog from Thermococcus litoralis DSM 5473

The Escherichia coli BL21 (DE3) cells harboring pET21b-Tl-laspo were cultivated at 37 °C in four Sakaguchi flasks (500 mL) containing 250 mL of Medium A on a reciprocal shaker under aerobic conditions. After 16 h, the cells were collected by centrifugation at 14,500×g and 4 °C for 15 min using an H-923 centrifuge (Kokusan, Tokyo, Japan) and were suspended in 30 mL of 50 mM KPB, pH 7.0, containing 20 mM imidazole and 300 mM NaCl (lysis buffer). This cell suspension was transferred in a conical tube (50 mL) and treated with an ultrasonicator (UD-201, Tomy Seiko, Tokyo, Japan) in ice-cooled water to disrupt the cells (output, 6; duty cycle, 30%). After ultrasonication, the suspension was centrifuged at 14,500×g and 4 °C for 20 min. The supernatant was transferred into an Erlenmeyer flask (500 mL) and then incubated with gentle shaking in a water bath at 80 °C. After 15 min, the suspension was centrifuged again under the same conditions to remove insoluble materials, and the supernatant was used as a cell-free extract. The cell-free extract was then applied to a column of Ni–NTA agarose (ϕ 1.5 × 2.8 cm) (Qiagen, Hilden, Germany) that had been equilibrated with lysis buffer. The unabsorbed proteins were washed from the column with a buffer containing 50 mM KPB (pH 7.0), 300 mM NaCl, and 50 mM imidazole. The absorbed protein was eluted with 15 mL of a buffer containing 50 mM KPB (pH 7.0), 300 mM NaCl, and 250 mM imidazole and was stored as a purified enzyme at − 80 °C after it was dialyzed against 50 mM KPB (pH 7.0).

Standard assay conditions

LASPO activity was measured at 70 °C using an assay mixture (3 mL) containing 50 mM KPB, pH 7.0, 50 µM FAD, 2 mM phenol, 0.4 mM 4-aminoantipyrine (4-AA), 30 mM l-aspartic acid, 9.6 U of horseradish peroxidase (HRP, Wako chemical, Osaka, Japan), and enzyme solution. The enzyme reaction was started by the addition of enzyme solution, and the absorption of the solution derived from an oxidative condensation product produced from hydrogen peroxide reacting with the phenol and 4-AA; the absorbance of HRP was measured at 500 nm using a spectrophotometer (JASCO, Tokyo, Japan). One unit of Tl-LASPO activity was defined as the amount of enzyme that produces 1 µmol of the oxidative condensation product per minute under standard assay conditions. The molar extinction coefficients (ε500; M−1 cm−1) of the oxidative condensation product that were used in various buffers (concentration: 50 mM) were determined experimentally at 70 °C as follows: 7420 in Bis–Tris (pH 6.0); 7440 in Bis–Tris (pH 6.5); 7520 in KPB (pH 6.5); 7600 in KPB (pH 7.0); 7260 in KPB (pH 7.5); 7340 in KPB (pH 8.0); 7300 in Tricine (pH 8.0); 8000 in Tricine (pH 8.5); and 8580 in Tricine (pH 9.0).

Effects of temperature and pH on enzyme activity

The effects of temperature and pH on Tl-LASPO activity were examined by measuring the enzyme activity under various temperature and pH conditions. The reaction temperatures were 30, 35, 40, 45, 50, 55, 60, 65, and 70 °C. The buffers were Bis–Tris (pH 6.0 and 6.5); potassium phosphate (pH 6.5, 7.0, 7.5 and 8.0), Tricine–NaOH (pH 8.0, 8.5 and 9.0).

Thermal and pH stability

The pH and thermal stabilities of Tl-LASPO were determined by measuring residual activity under the standard assay conditions after pH and heat treatment. The pH treatment was carried out by mixing 100 µL of purified Tl-LASPO (0.6 mg/mL) dissolved in 20 mM KPB (pH 7.0) with an equal amount of 250 mM buffer and incubating the solution at 4 °C for 72 h. The buffers were acetate (pH 4.0, 4.5, 5.0, and 5.5); Bis–Tris (pH 5.5, 6.0, and 6.5); potassium phosphate (pH 6.5, 7.0, 7.5, and 8.0); Tricine–NaOH (pH 8.0, 8.5, and 9.0); and glycine–NaOH (pH 9.0, 9.5, and 10.5). The heat treatment was carried out by incubating 500 µL of purified Tl-LASPO (4.04 U/mg) dissolved in 20 mM KPB (pH 7.0) at 70, 80, 85, or 90 °C for 0, 10, 20, 30, 40, 50, and 60 min.

Substrate specificity

The substrate specificity of Tl-LASPO for various l- and d-amino acids was examined. The tested amino acids used were l-alanine, l-arginine, l-asparagine, l-cysteine, l-cysteic acid, l-cysteine sulfinic acid, l-glutamate, l-glutamine, l-histidine, l-isoleucine, l-leucine, l-lysine, l-methionine l-phenylalanine, l-proline, l-serine, l-threonine, l-tryptophan, l-tyrosine, l-valine, glycine, d-asparagine, d-aspartate, and d-glutamate. Each amino acid replaced l-aspartic acid in the standard assay mixture at a final concentration of 30 mM except for l-tyrosine (final conc. 5 mM), and the enzyme activity was measured under standard assay conditions.

Tl-LASPO specificity for electron acceptor

The specificity of Tl-LASPO for electron acceptor was examined by testing fumaric acid, 2,6-dichlorophenolindophenol (DCIP), and ferricyanide as acceptors. The enzyme reaction was carried out anaerobically at 70 °C in a glove box (As one, Osaka, Japan) that had been filled with argon gas (purity: 99.99%). The reaction mixture (total volume: 0.3 mL) contained 50 mM KPB (pH 7.0), 50 µM FAD, 30 mM l-aspartic acid, and purified Tl-LASPO (28 µg), and 30 mM fumaric acid, 42 μM DCIP or 1 mM ferricyanide was added as an electron acceptor. The reaction was stopped by the addition of 60% (w/w) perchloric acid (for fumaric acid) or 10% SDS (w/v) (for DCIP and ferricyanide). In experiments with fumaric acid, the ammonium that was produced in the solution was quantified using Nessler’s reagent. An aliquot of the solution (10 μL) was mixed with Nessler’s reagent (10 μL) and deionized water (180 μL), and the absorption of the resulting solution was measured at 450 nm. The amount of ammonium was determined based on the calibration curve that was generated using ammonium sulfate as a standard. One unit of Tl-LASPO activity was defined as the amount of enzyme that produces 1 µmol of NH4 + per minute. In experiments with DCIP or ferricyanide, the absorption changes were measured at 600 nm for DCIP (ε600 was calculated to be 17,970 M−1 cm−1) and at 420 nm for ferricyanide (ε420 = 970 M−1 cm−1). One unit of Tl-LASPO activity was defined the amount of enzyme that produces 1 µmol of the reduced form of DCIP or ferricyanide per minute at 70 °C.

Kinetic analysis

Tl-LASPO activities were measured with various concentrations of l-aspartic acid (5, 10, 15, 30 and 60 mM) at 37, 50, 60, and 70 °C under standard assay conditions. The kinetic parameters (K m and k cat) of the reaction of Tl-LASPO with l-aspartic acid were determined by analyzing the enzyme activities at each reaction temperature with a Lineweaver–Burk plot (Lineweaver and Burk 1934).

Molecular mass and quaternary structure

The subunit molecular mass of Tl-LASPO was determined using SDS-PAGE with a 12% T or 10% T polyacrylamide gel. The purified Tl-LASPO (50 μL) was mixed with 2 × sample buffer (50 μL) containing 125 mM Tris–HCl buffer (pH 6.8), 4% (w/v) SDS, 10% (w/v) sucrose, 0.01% bromophenol blue, and 5 or 0.05% (v/v) 2-mercaptoethanol, and the sample was then incubated at 100 °C for 5 min.

The molecular mass of purified Tl-LASPO was estimated by gel filtration. The standard proteins were apoferritin (443 kDa), β-amylase (200 kDa), albumin (66 kDa), carbonic anhydrase (29 kDa), and cytochrome c (12.4 kDa). The standard proteins (1.0 mg) were dissolved in 0.1 mL of 50 mM KPB (pH 7.0) and this solution was applied to a column of Superdex 200 (ϕ 10 × 300 cm, GE Healthcare, Tokyo, Japan) that had been equilibrated with 50 mM KPB (pH 7.0) containing 150 mM NaCl. The column was eluted with the same buffer at a flow rate of 0.5 mL/min. After the standard calibration curve had been prepared, purified Tl-LASPO (2.0 mg/mL, 0.1 mL) was applied to the column under the same conditions.

Spectral analysis of holo- and apoenzyme

The Tl-LASPO apoprotein was prepared by dialyzing the holoprotein of Tl-LASPO (approximately 10 mg) against 500 mL of 50 mM KPB (pH 7.0) containing 3 M KBr and 0.5% (w/v) charcoal at 4 °C for 72 h. The dialysis buffer was changed after 24 and 48 h. After 72 h, the dialysate was corrected and applied to a column of PD-10 (GE Healthcare, Piscataway, NJ, USA) equilibrated with 50 mM KPB (pH 7.0). The UV–VIS spectra of the Tl-LASPO holo- and apoproteins were measured using a spectrophotometer (JASCO, Tokyo, Japan). The sample was diluted with 50 mM KPB (pH 7.0) to a concentration of 1.4 mg/mL.

Determination of FAD dissociation constant

The FAD dissociation constant (K d) of the FAD-Tl-LASPO complex was determined by measuring the quenching of FAD fluorescence that occurred when apo-Tl-LASPO was titrated with FAD. FAD fluorescence was measured using a Hitachi F-2500 spectrofluorometer (Hitachi High-Technologies, Tokyo, Japan). The measurement conditions were excitation wavelength (slit width, 10 nm) 450 nm and emission wavelength (slit width, 20 nm) 522.5 nm. Apo-Tl-LASPO was diluted with 50 mM KPB, pH 7.0, to a final concentration of 7.0 µM, and the FAD concentration was varied from 0 to 24.5 µM using a 50 µM FAD (dissolved in water) stock solution. The K d value was calculated according to the method of Arroyo et al. (2007) using Eq. (1):

$$ 1/(1 - a) = 1/K_{\text{d}} ([L_{0} ]/a - [E_{0} ]). $$
(1)

The parameters in Eq. (1) were as follows: a is the fractional saturation of total concentration of binding site; K d is the dissociation constant; [L 0] is the total ligand concentration; and [E 0] is the total enzyme concentration.

MALDI-TOF-MS analysis

The protein band that had been removed from a SDS-polyacrylamide gel was cut into small pieces (approximately 1 × 1 mm). The gel pieces were put into a microcentrifuge tube and treated with an In-Gel Tryptic Digestion Kit (Thermo Fisher Scientific, Yokohama, Japan) according to its manufacturer’s instructions. The sample that had been digested with trypsin was desalted using a Zip Tip C18 (Millipore, Billerica, MA, USA). The desalted sample (0.5 μL) was mixed with 0.5 μL of matrix solution (2,5-dihydroxybenzoic acid dissolved at a concentration of 1 mg/mL in ultrapure deionized water) containing 40% (v/v) acetonitrile and 0.06% (v/v) trifluoroacetic acid and was put on a MALDI target plate. After the sample was dried naturally with air, mass spectrometry measurements were carried out in reflectron positive ion mode using an AXIMA-CFR-plus MALDI-TOF-MS instrument (Shimadzu, Kyoto, Japan). The mass spectrometry data were analyzed using MASCOT (Matrix science, London, UK) to identify the trypsin-digested peptides and comparing them with peptide sequences in protein databases.

Effects of pH on thiol oxidation during storage of purified enzyme

The purified Tl-LASPO, dissolved in 50 mM KPB, pH 7.0 (2 mL), was passed through a PD-10 (GE Healthcare Japan, Tokyo, Japan) column that had been equilibrated with various buffers (50 mM Bis–Tris, pH 6.0; 50 mM sodium phosphate buffer, pH 7.0 and 8.0; or glycine–NaOH, pH 9.0) to exchange the sample buffer, and the resulting solution was stored at 4 °C. After 168 h had passed, the free thiol content in Tl-LASPO was quantitated using DTNB (Patsoukis and Georgiou, 2004). The sample (30 μL) was mixed with 500 mM sodium phosphate buffer, pH 8.0 (60 μL), 25 mM EDTA (6 μL), and 10 mM DTNB (24 μL), and the solution was allowed to react at 25 °C for 40 min. The absorption of the sample was measured at 412 nm, and the concentration of free thiol groups was determined according to the calibration curve that had been constructed using reduced glutathione (conc.: 0, 5, 10, 20, and 40 µM) as a standard.

Phylogenetic analysis

To clarify the distribution of homologs of Tl-LASPO, we performed the phylogenetic analysis with 71 kinds of the primary structure of l-aspartate oxidase registered already in protein data bank of BLAST program according to the method of Altschul et al. (1990). The phylogenetic tree of l-aspartate oxidase was described using a Clustal W multiple sequence alignment with a MEGA4 program (Thompson et al. 2002) by the method of neighbor-joining (Saitou and Nei 1987).

Accession numbers

The accession number of gene or protein written in this paper was expressed as the number administrated in National Center for Biotechnology Information (NCBI; https://www.ncbi.nlm.nih.gov), and the Reference Sequence (RefSeq) number of NCBI was used only for the protein without accession number.

Results

Expression of l-aspartate oxidase homolog gene from Thermococcus litoralis DSM 5473 and purification and characterization of its gene product

DNA sequence analysis showed that the Tl-laspo gene was composed of 1398 bp and encoded 465 amino acid residues. Genes encoding homologs of quinolinate synthetase (EC 2.5.1.72) and nicotinate nucleotide pyrophosphorylase (EC 2.4.2.19) were downstream of the Tl-laspo gene, while two genes encoding hypothetical proteins were upstream. The amino acid sequence of Tl-LASPO was similar to those of other previously reported l-aspartate oxidases, and its similarity to amino acid sequences of LASPOs from Pyrococcus horikoshii OT-3, Sulfolobus tokodaii 7, Bacillus subtilis 168, and Escherichia coli K-12 was calculated to be approximately 65, 38, 35 and 34%, respectively (Fig. 1). The important amino acid residues in the putative FAD binding site were well conserved in the primary structure of Tl-LASPO: Ala12, Gln43, Gly45, Glu331, and Leu348 (Fig. 1). The phylogenetic analysis of l-aspartate oxidase revealed that the homologs of Tl-LASPO existed in three of ten classes of archaea, Thermococci, Thermoprotei, and Halobacteria, especially in the genus of Thermococcus, Pyrococcus, Sulfolobus, and Halobacteria that belong to the families of Thermococcaceae, Sulfobaceae, or Halobacteriaceae (Fig. 2). In domain of Bactria, the homologs of Tl-LASPO existed predominantly in the family of Aquificaceae (Fig. 2), although LASPO was purified from several bacteria such as Escherichia coli (Flachmann et al. 1988), Bacillus subtilis (Marinoni et al. 2008), Pseudomonas putida (Leese et al. 2013), and so on. Tl-LASPO was also homologous to the proteins that were annotated as a fumarate reductase in some archaea (Fig. 2).

Fig. 1
figure 1

Alignment of amino acid sequences of l-aspartate oxidases from Thermococcus litoralis DSM 5473 and other microorganisms. Asterisk conserved amino acid residues for FAD binding. The amino acid sequences used for alignment were as follows: Thermococcus litoralis DSM 5473 (accession no.: EHR78939), Pyrococcus horikoshii OT-3 (accession no.: BAA29083), Sulfolobus tokodaii 7 (accession no.: BAK54476), Bacillus subtilis 168 (accession no.: AIY94113), and Escherichia coli K-12 (accession no.: AMH35960)

Full size image
Fig. 2
figure 2

Phylogenetic analysis of l-aspartate oxidase from Thermococcus litoralis DSM 5473. The blue colored area showed the proteins annotated as l-aspartate oxidase and the pink colored area showed the proteins annotated as fumarate reductase. The amino acid sequences used for phylogenetic analysis were shown as follows: Aquifex aeolicus VF5 (accession no.: AAC06932), Candidatus Methanoperedens sp. BLZ1 (accession no.: KPQ41658), Halalkalicoccus paucihalophilus (accession no.: KYH26718), Haloarcula amylolytica JCM 13557 (accession no.: EMA20236), Halococcus salifodinae (accession no.: EMA49854), Halococcus thailandensis JCM 13552 (accession no.: EMA50543), Haloferax volcanii DS2 (accession no.: ADE03658), Halogeometricum borinquense DSM 11551 (accession no.: ADQ66598), Halomicrobium mukohataei DSM 12286 (accession no.: ACV47611), Halopiger salifodinae (accession no.: SEW21219), Halorubrum tebenquichense DSM 14210 (accession no.: ELZ40236), Halovenus aranensis (accession no.: SDJ44082), Hydrogenobacter thermophilus TK-6 (accession no.: BAI68789), Hydrogenobaculum sp. Y04AAS1 (accession no.: ACG57280), Methanobacterium lacus (accession no.: ADZ08296), Methanocaldococcus infernus ME (accession no.: ADG13203), Methanoculleus sp. MAB1 (accession no.: CVK31361), Methanolacinia petrolearia DSM 11571 (accession no.: ADN37392), Methanothermococcus okinawensis IH1 (accession no.: AEH06252), Methanotorris igneus Kol 5 (accession no.: AEF95579), Natrialba magadii ATCC 43099 (accession no.: ADD06474), Natronomonas pharaonis DSM 2160 (accession no.: CAI49299), Natronorubrum sediminis (accession no.: SEH16664), Palaeococcus ferrophilus (RefSeq no.: WP_048149971), Palaeococcus pacificus DY20341 (accession no.: AIF68566), Pyrococcus abyssi (RefSeq no.: WP_048147208), Pyrococcus furiosus DSM 3638 (accession no.: AAL82100), Pyrococcus horikoshii OT3 (accession no.: BAA29083), Pyrococcus kukulkanii (accession no.: AMM54247), Pyrococcus sp. NA2 (accession no.: AEC51452), Pyrococcus sp. ST04 (accession no.: AFK21679), Pyrococcus yayanosii CH1 (accession no.: AEH25409), Sulfolobus acidocaldarius SUSAZ (accession no.: AHC50960), Sulfolobus islandicus Y.G.57.14 (accession no.: ACP45480), Sulfolobus metallicus (RefSeq no.: WP_054839231), Sulfolobus solfataricus 98/2 (accession no.: AKA74220), Sulfolobus tokodaii 7 (accession no.: BAK54476), Thermococcus barossii (accession no.: ASJ03836), Thermococcus celer (accession no.: ASI98302), Thermococcus celericrescens (accession no.: KUH32063), Thermococcus chitonophagus (accession no.: ASJ15662), Thermococcus cleftensis (accession no.: AFL95005), Thermococcus eurythermalis (accession no.: AIU68991), Thermococcus gammatolerans EJ3 (accession no.: ACS33122), Thermococcus gorgonarius (accession no.: ASJ01095), Thermococcus guaymasensis DSM 11113 (accession no.: AJC71951), Thermococcus kodakarensis KOD1 (accession no.: BAD84486), Thermococcus litoralis DSM 5473 (accession no.: EHR78939), Thermococcus nautili (accession no.: AHL22922), Thermococcus onnurineus NA1 (accession no.: ACJ17376), Thermococcus pacificus (accession no.: ASJ06406), Thermococcus peptonophilus (accession no.: AMQ18168), Thermococcus piezophilus (accession no.: ANF22655), Thermococcus profundus (accession no.: ASJ03423), Thermococcus radiotolerans (accession no.: ASJ15399), Thermococcus sibiricus MM 739 (accession no.: ACS89124), Thermococcus siculi (accession no.: ASJ08395), Thermococcus sp. 2319 × 1 (accession no.: ALV63936), Thermococcus sp. 4557 (accession no.: AEK73068), Thermococcus sp. 5-4 (accession no.: ASA76763), Thermococcus sp. AM4 (accession no.: EEB74599), Thermococcus sp. EP1 (accession no.: KPU63813), Thermococcus sp. PK (RefSeq no.: WP_083965087), Thermococcus thioreducens (accession no.: ASJ11607), Thermococcus zilligii (RefSeq no.: WP_010478909), Thermocrinis albus (RefSeq no.: WP_041434046), Thermocrinis minervae (accession no.: SHK43562), Thermocrinis sp. GBS (RefSeq no.: WP_029551840), Thermoproteus sp. CP80 (RefSeq no.: WP_081226661), Vulcanisaeta distributa DSM 14429 (accession no.: ADN49891)

Full size image

We succeeded in expressing Tl-LASPO in soluble fractions of E. coli BL21 (DE3) cells harboring pET21b-Tl-laspo cells. Approximately, 60% of the Tl-LASPO protein was detected in soluble fractions using SDS-PAGE analysis. Tl-LASPO was purified to homogeneity from cell-free extracts of E. coli BL21 (DE3) cells harboring pET21b-Tl-laspo using heat treatment and Ni–NTA and hydroxyapatite columns (Fig. 3), and approximately 20 mg of purified Tl-LASPO was obtained from cell-free extracts that were prepared from 1 L of culture. The specific activity of purified Tl-LASPO was 3.86 U/mg. The molecular mass of the Tl-LASPO was calculated to be approximately 140 kDa by gel-filtration (Fig. 3c). The subunit molecular mass of Tl-LASPO was calculated to be approximately 52 kDa by SDS-PAGE. Accordingly, the quaternary structure of Tl-LASPO was determined to be homotrimeric. We found that after storage of the purified Tl-LASPO at 4 °C for more than 96 h, a protein band corresponding to a molecular mass of 100 kDa, in addition to the band at 52 kDa, appeared in SDS-PAGE analysis (Fig. 3a); however, the specific activity of the purified Tl-LASPO did not change during the storage period. The MALDI-TOF-MS analysis showed that the protein bands at both 52 and 100 kDa were derived from the polypeptide chain of Tl-LASPO and did not contain peptides from other proteins (Table 1). We examined the effects of storage pH (pH 6–9) on the appearance of the protein band of 52 kDa and found that the intensity of the protein band gradually increased at acidic pH values and peaked at pH 6.0 (Fig. 3b). The number of free thiol groups in purified Tl-LASPO during storage also decreased at acidic pH and was lowest at pH 6.0 (Table 2). These results suggested that the protein band at 100 kDa might be composed of two polypeptide chains of Tl-LASPO.

Fig. 3
figure 3

Molecular mass and subunit molecular mass of purified l-aspartate oxidase from Thermococcus litoralis DSM 5473. a Effects of storage time on the appearance of 100-kDa band. Lanes 1, 3: 5 μg of purified Tl-LASPO stored at 4 °C for 0 and 192 h, respectively. Lanes 2,4: Molecular mass marker (4 µL): 14.4 kDa (Lysozyme), 18.4 kDa (β-Lactoglobulin), 25.0 kDa (REase Bsp98 l), 35.0 kDa (Lactate dehydrogenase), 45.0 kDa (Ovalbumin), 66.2 kDa (Bovine serum albumin), and 116 kDa (β-Galactosidase). SDS-PAGE was carried out using 12%T polyacrylamide gels. The samples were treated with sample buffer containing 5% 2-mercaptoethanol. b Effects of storage pH on the appearance of 100-kDa band. Lane 1: molecular mass marker (4 μL). PageRuler™ Unstained Broad Range Protein Ladder (Thermo fisher scientific, Waltham, MA, USA). Lanes 2, 3, 4, 5: 5 μg of purified Tl-LASPO stored at various pH values (lane 2, pH 6.0; lane 3, pH 7.0; lane 4, pH 8.0; and lane 5, pH 9.0). SDS-PAGE was carried out using 10%T polyacrylamide gels. The samples were treated with sample buffer containing 0.05% 2-mercaptoethanol. c Molecular mass. Filled circle: standard proteins, open circle: Tl-LASPO. 12.4 kDa (Cytochrome c), 29 kDa (Carbonic anhydrase), 66 kDa (Albumin), 200 kDa (β-Amylase), and 443 kDa (Apoferritin)

Full size image
Table 1 MALDI-TOF/MS analysis
Full size table
Table 2 Measurement of free thiol group
Full size table

Effects of pH and temperature on the activity of the l-aspartate oxidase from Thermococcus litoralis DSM 5473

The enzyme activity of Tl-LASPO increased as the reaction temperature increased from 30 to 70 °C (Fig. 4a). The enzyme activity of Tl-LASPO was measured from pH 6.0 to 9.0, and it peaked at pH 8.0 when a potassium phosphate buffer was used in the assay (Fig. 4b). Tl-LASPO was stable at temperatures up to 80 °C when heat treated for 1 h (Fig. 5a). The half-life time of Tl-LASPO at 90 and 85 °C was calculated to be 46 and 170 min, respectively. Approximately, 70% of its initial activity remained after being incubated at 85 °C for 1 h. The thermal stability of Tl-LASPO was similar to that of other LASPOs from hyperthermophiles such as Sulfolobus tokodaii (stable up to 80 °C) (Sakuraba et al. 2008) and Pyrococcus horikoshii (high stability at 80 °C) (Sakuraba et al. 2002). Tl-LASPO was highly stable between pH 5.0 and 10.5 (Fig. 5b).

Fig. 4
figure 4

Effects of temperature and pH on the activity of the l-aspartate oxidase from Thermococcus litoralis DSM 5473. a Effect of temperature. b Effect of pH. Filled circles, Bis–Tris pH 6.0–6.5; filled squares, potassium phosphate, pH 6.5–8.0; filled triangles, Tricine–NaOH pH 8.0–9.0

Full size image
Fig. 5
figure 5

Thermal and pH stability of the l-aspartate oxidase from Thermococcus litoralis DSM 5473. a Thermal stability. Filled circles, 90 °C; filled squares, 85 °C; filled triangles, 80 °C; and open circles, 70 °C. b pH stability of Tl-LASPO. Filled circles, sodium acetate pH 4.5–5.5; filled squares, Bis–Tris pH 5.5–6.5; filled triangles, potassium phosphate pH 6.5–8.0; open circles, Tricine–NaOH pH 8.0–9.0; open squares, glycine–NaOH pH 9.0–10.5

Full size image

Substrate specificity

Tl-LASPO was highly specific for l-aspartic acid and did not act on other natural amino acids, although it showed weak activity against the unnatural amino acids l-cysteic acid (relative activity to l-aspartic acid: 4.02%) and l-cysteine sulfinic acid (1.09%) (Table 3). Compared with Tl-LASPO, the LASPOs from other microorganisms showed relatively low substrate specificity: the Sulfolobus tokodaii strain 7 enzyme acted on l-aspartic acid and l-asparagine (Bifulco et al. 2013), and the Pseudomonas putida ATCC 47054 enzyme acted on l-asparagine and l-glutamate (Leese et al. 2013).

Table 3 Substrate specificity
Full size table

Tl-LASPO specificity for electron acceptor

Tl-LASPO utilized fumaric acid, DCIP, and ferricyanide individually as cofactors under anaerobic conditions. The specific activities were calculated to be 0.96 ± 0.02, 3.7 ± 0.6, and 15.7 ± 1.2 U/mg for fumaric acid, DCIP, and ferricyanide, respectively.

Kinetic parameters

The kinetic parameters describing Tl-LASPO activity are summarized in Table 4. The K m value for l-aspartic acid decreased with increasing reaction temperature, while the k cat value increased; the highest catalytic efficiency of Tl-LASPO, i.e., k cat/K m value, in the temperature range tested was observed in the assays at 70 °C. The K m value of Tl-LASPO activity for l-aspartic acid at 37 °C was higher than that of the enzyme from S. tokodaii strain 7 (13.3 mM, at 37 °C, pH 8.0) (Bifulco et al. 2013). Accordingly, Tl-LASPO showed the lower affinity to l-aspartic acid under low temperature conditions as compared with the enzyme from S. tokodaii strain 7.

Table 4 Kinetic parameters
Full size table

Spectroscopic properties, FAD content, and FAD dissociation constant

FAD bound tightly to Tl-LASPO and did not dissociate when treated with 3 M KBr. In addition to KBr, we used charcoal and succeeded in preparing apo-Tl-LASPO. The absorption spectrum of holo-Tl-LASPO showed maxima at 380 and 450 nm, and these peaks were not present in apoenzyme (Fig. 6). Titrating apo-Tl-LASPO with FAD exhibited fluorescence at 522.5 nm that demonstrated a break point at 1 mol of FAD/mol of subunit (Fig. 7). Accordingly, the spectral features of holo-Tl-LASPO were derived from FAD, and Tl-LASPO contained 1 mol of FAD per subunit. The FAD dissociation constant, K d, of the FAD-Tl-LASPO complex was calculated to be 5.6 × 10−9 M.

Fig. 6
figure 6

Spectral analysis of holo- and apo-l-aspartate oxidase from Thermococcus litoralis DSM 5473. Solid line, holoprotein; dashed line, apoprotein. Protein concentration: 22.5 μM. Buffer: 50 mM KPB (pH 7.0)

Full size image
Fig. 7
figure 7

Determination of dissociation constant of FAD/l-aspartate oxidase from Thermococcus litoralis DSM 5473 complex

Full size image

Discussion

We succeeded in expressing an l-aspartate oxidase homolog gene (accession no: OCC_06611) from T. litoralis DSM 5473 in the soluble fraction of E. coli BL21 (DE3) using a pET21b vector with a 6X His tag at the C-terminus. We showed that the gene product exhibits l-aspartate oxidase activity in vitro, and this report is the first to detail an l-aspartate oxidase derived from a Thermococcus species. The homologs of Tl-LASPO distributed mainly in restricted classes of archaea and scarcely in bacteria (Fig. 2). This might be one of the reasons why the number of report for l-aspartate oxidase purified from bacteria is few in databases for literatures.

The quaternary structure of Tl-LASPO was homotrimeric with a subunit molecular mass of 52 kDa. Most of the l-aspartate oxidases reported thus far have been monomeric (Seifert et al. 1990; Marinoni et al. 2008; Bifulco et al. 2013) including the enzyme from S. tokodaii strain 7 with the exception of the LASPO from P. horikoshii OT-3, which was a homotrimer (Sakuraba et al. 2002). Accordingly, the mechanisms of the thermostabilities of the LASPOs between Thermococcales and Sulfolobus might be different, but further X-ray crystallographic study must be needed to clarify the relationship between structure and thermostability of Tl-LASPO.

When the purified Tl-LASPO was stored at 4 °C for more than 96 h, a protein band corresponding to a molecular mass of 100 kDa was detected by SDS-PAGE. This molecular mass (100 kDa) was about twice the subunit molecular mass of Tl-LASPO (52 kDa), and the protein band at 100 kDa was probably derived from two polypeptide chains of Tl-LASPO. This estimate was supported by the pH dependency of the SDS-PAGE protein band intensity and the total number of free thiol groups (Fig. 3b). The pK a value of the thiol group of cysteine is generally 6.16, and the H+ in thiol groups dissociates at alkali pH values. Accordingly, the disulfide bond between two cysteine residues in two polypeptide chains of Tl-LASPO might be more easily formed in acidic pH ranges. This observation is specific for Tl-LASPO and has not been reported for other LASPOs from hyperthermophilic archaea. A greater number of cysteine residues exist in the primary structure of Tl-LASPO (4 cysteine residues: Cys152, Cys351, Cys430, and Cys463) compared with those of LASPOs from Pyrococcus horikoshii OT3 (2 cysteine residues) and Sulfolobus tokodaii (1 cysteine residues). This reason may explain why Tl-LASPO specifically formed a disulfide bond between two polypeptide chains during storage at 4 °C.

Although Tl-LASPO activity was highly specific for l-aspartic acid, l-cysteic acid and l-cysteine sulfinic acid also reacted with the enzyme. This substrate specificity of Tl-LASPO agreed well with that of the aspartate racemase (EC 5.1.1.13) from T. litoralis DSM 5473 (Tl-AspR) that we reported in our previous study (Washio et al. 2016). A genome analysis of T. litoralis DSM 5473 suggests that l-aspartic acid produced from d-aspartic acid by Tl-AspR might be converted to oxaloacetic acid by Tl-LASPO. The produced oxaloacetic acid is then probably converted to l-aspartic acid by aspartate aminotransferase, but interestingly, there is no report of aspartate aminotransferase from T. litoralis DSM 5473 although 5 putative aspartate aminotransferase genes are annotated in its genome (accession no.: OCC_05516, OCC_03517, OCC_11879, OCC_08839, and OCC_10965). Except aspartate aminotransferase, the putative genes of pyruvate carboxylase subunit B (EC 6.4.1.1; accession no. OCC_09596) and oxaloacetic acid decarboxylase subunit β (EC 4.1.1.3; accession no. OCC_03162) exist in the genome of T. litoralis DSM 5473 in relation to oxaloacetic acid metabolism, but these genes seem to encode only a subunit of mature enzymes, and these genes products probably do not function in vivo. Judging from the similarity in substrate specificities in these two enzymes, both d-cysteinic acid and d-cysteine sulfinic acid might convert to sulfopyruvic acid and β-sulfinyl pyruvic acid, respectively, via same enzyme pathway in vivo. The high substrate specificity of Tl-LASPO is considered to be useful for the specific conversion of l-aspartic acid to oxaloacetic acid or the quantification of l-aspartic acid in biological samples. Under anaerobic conditions, Tl-LASPO utilized fumaric acid, DCIP, and ferricyanide as cofactors in vitro similar to LASPO from Pyrococcus horikoshii OT-3 (Sakuraba et al. 2002). These activities are compatible with the enzyme activity of Tl-LASPO under aerobic conditions. Since T. litoralis DSM 5473 is an anaerobe that grows under only anaerobic conditions, Tl-LASPO might catalyze the enzyme reaction in vivo using fumaric acid as a cofactor. The fumaric acid used as a cofactor of Tl-LASPO in vivo might be produced from l-aspartic acid via adenylosuccinate by adenylosuccinate synthetase (EC 6.3.4.4; accession no. OCC_02397) and adenylosuccinate lyase (EC 4.3.2.2; accession no. OCC_01109) whose putative genes were encoded in the genome of T. litoralis DSM 5473. Interestingly, almost all enzyme gene in tricarboxylic acid (TCA) cycle lacked in the genome of T. litoralis DSM 5473 except the putative fumarate hydratase subunit alpha gene (accession no. OCC_11212).

Previous studies on LASPOs in prokaryotes have extensively focused on their role as the first enzyme in the de novo NAD+ biosynthetic pathway. The homologous genes of this pathway were also conserved in the genome of T. litoralis DSM 5473. Accordingly, Tl-LASPO might also be related to the NAD+ biosynthetic pathway in T. litoralis DSM 5473, but further studies are needed to clarify the function of Tl-LASPO in vivo.

We are currently studying the application of Tl-LASPO as an analytical tool for the measurement of l-aspartic acid in various biochemical samples.