Cyclic-Imide-Hydrolyzing Activity of d-Hydantoinase from Blastobacter sp. Strain A17p-4

Abstract

The cyclic-imide-hydrolyzing activity of a prokaryotic cyclic-ureide-hydrolyzing enzyme, d-hydantoinase, was investigated. The enzyme hydrolyzed cyclic imides with bulky substituents such as 2-methylsuccinimide, 2-phenylsuccinimide, phthalimide, and 3,4-pyridine dicarboximide to the corresponding half-amides. However, simple cyclic imides without substituents, which are substrates of imidase (i.e., succinimide, glutarimide, and sulfur-containing cyclic imides such as 2,4-thiazolidinedione and rhodanine), were not hydrolyzed. The combined catalytic actions of bacterial d-hydantoinase and imidase can cover the function of a single mammalian enzyme, dihydropyrimidinase. Prokaryotic d-hydantoinase also catalyzed the dehydrative cyclization of the half-amide phthalamidic acid to the corresponding cyclic imide, phthalimide. The reversible hydrolysis of cyclic imides shown by prokaryotic d-hydantoinase suggested that, in addition to pyrimidine metabolism, it may also function in cyclic-imide metabolism.

Microbial d-hydantoinase has been applied to the industrial production of optically pure d-amino acids that serve as chiral synthons for the synthesis of antibiotics (d-p-hydroxyphenylglycine), pesticides (d-valine), sweeteners (d-alanine), and therapeutic amino acids (3, 14, 26, 27). This enzyme catalyzes the reversible hydrolysis of cyclic ureides such as dihydropyrimidines and 5-monosubstituted hydantoins to N-carbamoyl amino acids. This enzyme was reported to be identical to dihydropyrimidinase (EC 3.5.2.2), which is involved in pyrimidine metabolism (23, 24). Mammalian dihydropyrimidinases are known to hydrolyze the cyclic imides of some antiepileptic agents, in addition to the cyclic ureides of dihydropyrimidines (2, 9).

During a study of the transformation of cyclic amides, including 5-monosubstituted hydantoins, by the bacterium Blastobacter sp. strain A17p-4, which shows high d-hydantoinase activity (15), we found that this bacterium also metabolizes some cyclic imides (16). Investigation of the cyclic imide transformation in this bacterium led to the finding of a novel cyclic-imide-transforming enzyme, imidase, which is different from cyclic-ureide-transforming enzymes and which specifically hydrolyzes the simple cyclic imides succinimide and glutarimide and sulfur-containing cyclic imides (17) and catalyzes the first step of cyclic imide metabolism (18). Imidase activity was found in a variety of microorganisms (22). Generally found in bacteria, this activity was not correlated with cyclic-ureide-transforming activity, suggesting that these two transformations involve different enzyme systems (22).

Further study revealed that some complex cyclic imides, which are not the substrates of imidase, are metabolized by Blastobacter sp. and that the d-hydantoinase from Blastobacter sp. acts on such complex cyclic imides. d-Hydantoinase of this strain, therefore, catalyzes the transformation of a different class of cyclic imides than the substrates of imidase and functions in cyclic-imide metabolism.

We present here the physicochemical properties and novel catalytic function of d-hydantoinase from Blastobacter sp., especially its cyclic-imide-hydrolyzing activity, and describe the differences between this enzyme and imidase with regard to cyclic-imide hydrolysis compared with mammalian dihydropyrimidinases.

MATERIALS AND METHODS

Chemicals.

All hydantoin derivatives and N-carbamoyl-β-alanine were kind gifts from Kanegafuchi Chemical Co. (Takasago, Japan). N-Carbamoyl-d-amino acids were prepared from the corresponding d-amino acids and potassium cyanate according to the method of Nyc and Mitchell (13). 2-Methylsuccinimide was chemically synthesized according to the method described previously (1). All other chemicals used were of analytical grade and were obtained commercially.

Microorganism and cultivation.

Blastobacter sp. strain A17p-4 (AKU 990; Faculty of Agriculture, Kyoto University) was used as the source of d-hydantoinase and cultured as described previously (16).

Enzyme assay.

The standard enzyme assay mixture was comprised of 10 μmol of Tris-HCl (pH 7.5), 2 μmol of dihydrouracil (the favored substrate of the reported d-hydantoinase), and an appropriate amount of enzyme in 100 μl. After incubation for 30 to 60 min at 30°C, the reaction was stopped with 10 μl of 15% (by volume) perchloric acid, followed by neutralization with 90 μl of 500 mM potassium phosphate (pH 7.0). The reaction mixture was centrifuged at 10,000 × g for 10 min, and the supernatant was analyzed for decreases in the concentration of the substrate, dihydrouracil, and increases in the amount of the product, N-carbamoyl-β-alanine, on a Shimadzu LC-6A high-performance liquid chromatography (HPLC) apparatus at 210 nm, fitted with a Cosmosil 5C₁₈ AR-packed column (4.6 by 250 mm; Nacalai Tesque, Kyoto, Japan) run at a flow rate of 1.0 ml/min, with 250 mM KH₂PO₄ (pH 4.4) as the eluent. Similar HPLC conditions were used for the analysis of the other cyclic imides used (see Table 2), except that H₂O-CH₃CN-trifluoroacetic acid (90:10:0.1) was used as an eluent.

TABLE 2.

Substrate specificity of d-hydantoinase (for hydrolysis reaction) from Blastobacter sp. strain A17p-4^a

Substrate	Km (mM)	Vmax (μmol [min × mg]−1)	Vmax/Km (μmol [min × mg]−1 mM−1)
Cyclic imides
2-Methylsuccinimide	5.7 ± 1.1	0.29 ± 0.05	0.051 ± 0.004
2-Phenylsuccinimide	3.7 ± 0.6	0.12 ± 0.02	0.032 ± 0.001
Phthalimide	2.3 ± 0.2	0.081 ± 0.01	0.035 ± 0.004
3,4-Pyridine dicarboximide	11.0 ± 2.0	1.2 ± 0.2	0.11 ± 0.01
Cyclic ureides
Hydantoin	3.2 ± 0.6	1.2 ± 0.1	0.38 ± 0.04
Dihydrouracil	2.7 ± 0.1	2.5 ± 0.03	0.93 ± 0.03
Dihydrothymine	4.2 ± 1.0	1.5 ± 0.4	0.36 ± 0.02
dl-5-Methylhydantoin	9.3 ± 1.9	1.6 ± 0.3	0.17 ± 0.02
dl-5-(2′-Methylthio-ethyl)hydantoin	12.0 ± 2.5	0.78 ± 0.2	0.065 ± 0.009
dl-5-Phenylhydantoin	6.7 ± 1.2	0.84 ± 0.1	0.12 ± 0.01
dl-5-(p-Hydroxyphenyl)-hydantoin	0.61 ± 0.1	0.019 ± 0.002	0.031 ± 0.005

^aHydrolysis reactions and analysis were carried out under the standard assay conditions except that each test compound was used as the substrate at varied concentrations (0.1 to 50 mM). No spontaneous hydrolysis of compounds other than phthalimide and 3,4-pyridine dicarboximide was observed. Control reactions without enzyme were performed for these two compounds. The kinetic constant values are averages ± the standard deviation of three different determinations. The following compounds were judged to be inactive as substrates for hydrolysis (less than 10⁻³ μmol/min/mg of protein with 20 mM each substrate): succinimide, glutarimide, diacetamide, 2,4-thiazolidinedione, rhodanine, 3-methylglutarimide, 3,3-dimethylglutarimide, 3-ethyl-3-methylglutarimide, α,α-dimethyl-β-methylsuccinimide, cycloheximide, N-aminophthalimide, N-methylmaleimide, N-hydrophthalimide, 1,8-naphthalimide, 1,8-naphthalene-dicarboximide, dl-5-carboxy-ethylhydantoin, dl-5-carboxymethylhydantoin, l-carboxymethylhydantoin, l-5-isopropylhydantoin, 1-methylhydantoin, N-methylhydantoin, 5,5-dimethylhydantoin, dihydro-l-orotate, creatinine, barbiturate, dl-pyroglutamate, dl-pyroglutamate methylester, 2-pyrrolidone, allantoin, succinic anhydride, glutaric anhydride, maleic anhydride, ethyleneurea, ethylenethiourea, benzoylurea, propyleneurea, urazole, alloxan, and alloxazine. 

The cyclization reaction mixture was comprised of 10 μmol of MES (morpholineethanesulfonic acid)-NaOH (pH 6.0), 0.1 to 10 μmol of each compound (half-amides or N-carbamoyl amino acids), and an appropriate amount of enzyme in 100 μl. The reactions were carried out at 30°C for 60 to 90 min and analyzed by the method described above.

One unit of enzyme was defined as the amount of enzyme that catalyzed consumption of the substrate or formation of the product at a rate of 1 μmol/min under the assay conditions described above.

Enzyme purification.

All steps were carried out at 0 to 5°C. The buffer used was 20 mM Tris-HCl (pH 7.5) containing 0.1 mM dithiothreitol.

(i) Step 1.

Cells (20 g [wet weight] from an 8-liter culture) were harvested by centrifugation (10,000 × g at 4°C) and suspended in 20 ml of buffer. The cell suspension was disrupted with glass beads 0.25 to 0.50 mm in diameter (Dyno-Mill KDL; W. A. Bachofen) at 5°C for 30 min. The disrupted cell suspension was centrifuged at 14,000 × g for 60 min at 4°C, and the resultant supernatant was used as the cell-free extract.

(ii) Step 2.

The cell extract was dialyzed against 10 liters of buffer for 12 h. The dialyzed sample was then applied to a DEAE-Sephacel column (2.5 by 40 cm). After the column was washed with 1 liter of buffer, the enzyme was eluted with a linear gradient of 0 to 0.5 M NaCl in 1 liter of buffer. The active fractions were combined and concentrated by ultrafiltration with a 30,000 cutoff membrane.

(iii) Step 3.

The enzyme solution was mixed with solid NaCl to obtain a concentration of 4 M and then applied to a phenyl-Sepharose CL-4B column (2.5 by 20 cm). After the column was washed with buffer containing 4 M NaCl, the enzyme was eluted with a decreasing salt gradient (from 4 to 0 M NaCl) in 500 ml of buffer. Two dihydrouracil-hydrolyzing activities were detected. The activity eluted with the higher salt concentration buffer was that of imidase, while the activity eluted with the lower salt concentration buffer was that of d-hydantoinase (16). Thus, imidase-free d-hydantoinase was obtained by this step. The active fractions containing d-hydantoinase were combined and concentrated by ultrafiltration.

(iv) Step 4.

The concentrated d-hydantoinase was applied to a Sephacryl S-200 HR column (2.0 by 80 cm) equilibrated with buffer containing 0.2 M NaCl and then eluted with the same buffer. The active fractions were combined and dialyzed against the buffer.

(v) Step 5.

The dialyzed enzyme was applied to a MonoQ HR5/5 column equilibrated with the buffer and then eluted with a linear salt gradient (from 0 to 0.5 M NaCl) in 20 ml of buffer. The active fractions were combined and concentrated by ultrafiltration.

(vi) Step 6.

The concentrated enzyme was applied to a Superose-12 HR10/30 column equilibrated with buffer containing 0.2 M NaCl and then eluted with the same buffer. The active fractions were used for characterization.

Analytical methods for d-hydantoinase.

The relative molecular mass was determined by HPLC on a GS-520 column (7.6 by 500 mm; Asahi Kasei, Tokyo, Japan) and a G-3000 SW column (7.5 by 600 mm; Tosoh, Tokyo, Japan) at 0.3 ml/min with an elution buffer of 20 mM Tris-HCl (pH 7.5) containing 0.2 M NaCl and 0.1 mM dithiothreitol. Protein was determined by the Coomassie brilliant blue-R-250 dye-binding method of Bradford with the dye reagent supplied by Bio-Rad and bovine serum albumin as the standard. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on a 12.5% polyacrylamide gel and NH₂-terminal amino acid sequence analysis were performed as described previously (21).

RESULTS

Purification and the criteria for purity.

d-Hydantoinase was purified approximately 30-fold from the soluble cell extract of Blastobacter sp. strain A17p-4 (Table 1). The active fractions from the Superose-12 HR column were ascertained to be pure from the appearance of a single protein band on SDS-PAGE. Further evidence of its purity was provided by gel-permeation HPLC on GS-520 and G-3000 SW columns, giving a quite symmetrical protein absorption peak concomitant with dihydrouracil-hydrolyzing activity.

TABLE 1.

Purification of d-hydantoinase from Blastobacter sp. strain A17p-4

Step	Total vol (ml)	Total protein (mg)	Total activity (U)	Sp act (U/mg)
Cell extract	190	1400	90.0	0.064
DEAE-Sephacel	100	274	79.4	0.290
Phenyl-Sepharose CL-4B	62.0	64.0	25.6	0.400
Sephacryl S-200 HR	17.0	12.6	9.52	0.750
Mono-Q HR5/5	10.0	4.37	5.16	1.18
Superose-12 HR10/30	5.0	1.54	2.76	1.79

Substrate specificity and kinetic properties.

All reactions catalyzed by the enzyme exhibited normal hyperbolic kinetics. The K_m and V_max values were calculated from the double reciprocal of Lineweaver-Burk plots.

The hydrolytic activity of the purified d-hydantoinase toward various cyclic imides is shown in Table 2. Cyclic imides with bulky substituents such as 2-methylsuccinimide, 2-phenylsuccinimide, phthalimide, and 3,4-pyridine dicarboximide were hydrolyzed, while simple cyclic imides and sulfur-containing cyclic imides, which are actively hydrolyzed by imidase, were not hydrolyzed by the purified d-hydantoinase. This showed that the purified d-hydantoinase was not contaminated by imidase and that it has a quite different substrate specificity from imidase with regard to cyclic-imide hydrolysis. N-substituted cyclic imides, disubstituted cyclic imides, and the linear imide (diacetamide) were not hydrolyzed by the purified d-hydantoinase.

With cyclic ureides as substrates, d-hydantoinase exhibited the highest activity and affinity towards dihydropyrimidines such as dihydrouracil and dihydrothymine, and hydantoin (Table 2). Other d-hydantoinases from Pseudomonas striata (23), Pseudomonas fluorescens (12), and Agrobacterium tumefaciens (4) similarly showed the highest activity toward dihydrouracil. The hydrolysis of 5-monosubstituted hydantoin was found to be d-stereospecific on stereochemical analysis of the products (16). N-Substituted cyclic ureides, disubstituted cyclic ureides, and cyclic anhydrides were not hydrolyzed.

d-Hydantoinase also catalyzed the reverse reaction, dehydrative cyclization (Table 3), under acidic conditions (below pH 7.5). Phthalamidic acid (half-amide) was dehydrated to phthalimide (cyclic imide) by d-hydantoinase. However, the simple half-amide succinamic acid, which is a substrate of imidase, was not cyclized by the d-hydantoinase. Some N-carbamoyl α- and β-amino acids were dehydrated to the corresponding cyclic ureides, but at much slower rates than hydrolysis, suggesting that the function of d-hydantoinase is the decomposition of cyclic ureides and cyclic imides, which serve as nutrients and energy sources for cell growth (16, 18). At pH 6.0, d-hydantoinase catalyzed both hydrolysis and cyclization. An example of the equilibrium constant at pH 6.0 was [dihydrouracil]/[β-ureidopropionate] = 0.28.

TABLE 3.

Substrate specificity of d-hydantoinase (for cyclization reaction) from Blastobacter sp. strain A17p-4^a

Substrate	Km (mM)	Vmax (μmol [min × mg]−1)	Vmax/Km (μmol [min × mg]−1 mM−1)
Half-amide (phthalamidic acid)	23.0 ± 2.0	0.15 ± 0.02	0.0065 ± 0.0002
N-Carbamoyl amino acid
Hydantoic acid	31.0 ± 0.2	0.11 ± 0.01	0.0035 ± 0.0005
β-Ureidopropionate	11.0 ± 0.7	0.62 ± 0.04	0.056 ± 0.001
N-Carbamoyl-d-alanine	36.0 ± 1.7	4.1 ± 0.3	0.11 ± 0.01

^aCyclization reactions and analysis were carried out as described in Materials and Methods with test compounds at 1 to 100 mM as the substrates. The kinetic constant values are averages ± the standard deviation of three different determinations. 

Relative molecular mass, subunit structure, and NH₂-terminal amino acid sequence.

By high-performance gel permeation liquid chromatography on GS-520 and G-3000 SW columns, the relative molecular weights of d-hydantoinase were estimated to be 180,000 and 200,000, respectively. The relative molecular weight of the subunit was estimated to be 53,000 on SDS-PAGE. Thus, the native enzyme probably consists of four identical subunits. Other prokaryotic d-hydantoinases and mammalian dihydropyrimidinase have also been reported to exist as tetramers (6, 12, 23, 25). Automated Edman degradation with a pulse-liquid phase sequencer revealed that the NH₂-terminal amino acid sequence was Ser-Thr-Val-Ile-Lys-Gly-Gly-Thr-Ile-Val-Ala-Ala-Asp-Arg-Ser-Tyr-Glu-Ala-Asp-Ile-Leu-Ile. This sequence is homologous to bacterial d-hydantoinases, especially d-hydantoinases from Bacillus stearothermophilus and Bacillus sp. strain LU1220 (Fig. 1). A significant homology was also shown to mammalian (rat and human) dihydropyrimidinases (Fig. 1). It was reported that these genes form a gene family related to ureases (5, 11), suggesting that Blastobacter d-hydantoinase is also a member of this superfamily.

FIG. 1.

Comparison of NH₂-terminal amino acid sequence of d-hydantoinase from Blastobacter sp. with those from P. putida (our unpublished data), P. fluorescens (7), Bacillus species (7, 10), and mammalian dihydropyrimidinases from rats (10) and humans (5). Identical amino acids are indicated by black boxes.

Effects of inhibitors and metal ions.

The enzyme was incubated with various compounds (2 mM) at 30°C for 10 min, and then its activity was assayed under standard conditions to evaluate the inhibitory effects of these compounds. Sulfhydryl reagents such as p-chloromercuribenzoate, iodoacetate, and N-ethylmaleimide inhibited the enzyme activity by 76, 23, and 13%, respectively. Among the heavy-metal ions tested, only Hg²⁺ strongly inhibited the enzyme activity by 91%. Metal ion chelators such as EDTA, o-phenanthroline, α,α′-dipyridyl, and 8-hydroxyquinoline showed no significant effects, but the addition of divalent metal ions such as Mg²⁺, Mn²⁺, Co²⁺, Ni²⁺, and Cu²⁺ enhanced the enzyme activity by approximately 149, 219, 187, 270, and 152%, respectively, compared to the initial activity. d-Hydantoinase from Bacillus sp. was also activated by Mg²⁺, Ni²⁺, Co²⁺, and Mn²⁺ ions (20), and the activity of d-hydantoinase from B. stearothermophilus and Agrobacterium sp. was enhanced by Mn²⁺ and Ni²⁺ ions, respectively (8, 19).

No significant inhibition was observed with substrate-like compounds such as succinimide, glutarimide, 2,4-thiazolidinedione, rhodanine, 4-methylphthalimide, N-methylphthalimide, urazole, dl-5-carboxyethylene hydantoin, l-5-methylhydantoin, l-5-hydroxymethylene hydantoin, l-5-carboxymethylene hydantoin, propyleneurea, N-methylhydantoin, or dihydro-l-orotic acid. d-Hydantoinase activity was also examined in the presence of product-like compounds such as succinamic acid and hydrolysis products such as hydantoic acid, β-ureidopropionate, and N-carbomyl-d-alanine. At 5 mM, β-ureidopropionate and N-carbomyl-d-alanine inhibited the enzyme activity by 38 and 25%, respectively. Similar results were reported for rat dihydropyrimidinase (6, 25).

Effects of pH and temperature.

The enzyme activity and stability were assayed in MES-NaOH, MOPS (morpholinepropanesulfonic acid)-NaOH, Tris-HCl, and NaHCO₃-Na₂CO₃ buffer systems at pH 4.5 to 5.1, 6.2 to 8.2, 8.1 to 9.0, and 9.4 to 10.0, respectively. Under the standard assay conditions, the pH optima for dihydrouracil hydrolysis and β-ureidopropionate cyclization were pH 10.0 and 5.0, respectively. When the enzyme was incubated at 30°C for 30 min, more than 70% of the initial activity was retained at pH 5.0 to 10.0.

The initial velocity of hydrolysis increased with increasing temperature, reaching a maximum at 60°C. More than 95% of the initial activity remained after incubation at 60°C for 30 min, and 50% remained at 65°C for 30 min.

DISCUSSION

Earlier studies of rat dihydropyrimidinase indicated its role in the transformation of cyclic imides of some antiepileptic agents (2, 9, 25). The catalytic function similarity of bacterial d-hydantoinases and mammalian dihydropyrimidinases was expected from the amino acid sequence homology of these enzymes (10). This similarity was first revealed in the present study, showing that Blastobacter d-hydantoinase is able to hydrolyze cyclic imides.

The Blastobacter d-hydantoinase has a structure, physicochemical properties, and an NH₂-terminal amino acid sequence similar to other reported bacterial d-hydantoinases. As with other bacterial d-hydantoinases, the enzyme exhibited the highest catalytic efficiency toward dihydropyrimidines and was induced by dihydropyrimidines (16). These results suggest that the newly found cyclic-imide-hydrolyzing activity of Blastobacter d-hydantoinase might also be a general property of bacterial d-hydantoinases that are identical to dihydropyrimidinase. However, these should be clarified in further investigations.

As shown in Fig. 2, the rat dihydropyrimidinase is able to hydrolyze cyclic ureides and both simple and bulky cyclic imides, while the Blastobacter d-hydantoinase hydrolyzes cyclic ureides and only bulky cyclic imides. Blastobacter sp. also produces another cyclic-imide-hydrolyzing enzyme, imidase (17), which has properties distinct from rat dihydropyrimidinase and only hydrolyzes simple but not bulky cyclic imides (Fig. 2). The catalytic action of rat dihydropyrimidinase is fully complemented by the combined actions of d-hydantoinase and imidase in the Blastobacter system.

FIG. 2.

An overview of the substrate range exhibited by d-hydantoinase and imidase from Blastobacter sp. strain A17p-4 and mammalian dihydropyrimidinases.

The different substrate specificities of Blastobacter d-hydantoinase and imidase provide some insight into how these enzymes recognize their substrates. Despite the absence of a nitrogen on the α-position of the imide functional group compared to ureide, the electron density of the α-position provided by the substitution of an aromatic moiety, methyl, or phenyl group facilitates the ring-opening reaction by d-hydantoinase. Thus, electron density at the α-position of a substrate is important for d-hydantoinase. Furthermore, the planar structure of the substrate, such as with phthalimide, also facilitates d-hydantoinase reaction. A similar mechanism of action has been proposed for rat dihydropyrimidinase (25). Conversely, imidase is tolerant of electron density at the α-position but sensitive to steric hindrance caused by bulky substitution (17).

The present observation of novel d-hydantoinase function toward complex cyclic imides will provide further opportunities for practical applications.