A New Synthetic Route to N-Benzyl Carboxamides through the Reverse Reaction of N-Substituted Formamide Deformylase

Abstract

Previously, we isolated a new enzyme, N-substituted formamide deformylase, that catalyzes the hydrolysis of N-substituted formamide to the corresponding amine and formate (H. Fukatsu, Y. Hashimoto, M. Goda, H. Higashibata, and M. Kobayashi, Proc. Natl. Acad. Sci. U. S. A. 101:13726–13731, 2004, doi:10.1073/pnas.0405082101). Here, we discovered that this enzyme catalyzed the reverse reaction, synthesizing N-benzylformamide (NBFA) from benzylamine and formate. The reverse reaction proceeded only in the presence of high substrate concentrations. The effects of pH and inhibitors on the reverse reaction were almost the same as those on the forward reaction, suggesting that the forward and reverse reactions are both catalyzed at the same catalytic site. Bisubstrate kinetic analysis using formate and benzylamine and dead-end inhibition studies using a benzylamine analogue, aniline, revealed that the reverse reaction of this enzyme proceeds via an ordered two-substrate, two-product (bi-bi) mechanism in which formate binds first to the enzyme active site, followed by benzylamine binding and the subsequent release of NBFA. To our knowledge, this is the first report of the reverse reaction of an amine-forming deformylase. Surprisingly, analysis of the substrate specificity for acids demonstrated that not only formate, but also acetate and propionate (namely, acids with numbers of carbon atoms ranging from C₁ to C₃), were active as acid substrates for the reverse reaction. Through this reaction, N-substituted carboxamides, such as NBFA, N-benzylacetamide, and N-benzylpropionamide, were synthesized from benzylamine and the corresponding acid substrates.

INTRODUCTION

We have extensively studied the biological metabolism of compounds with a triple bond between carbon and nitrogen, such as nitriles and isonitriles (1–4). Nitriles are highly toxic and generally nonbiodegradable organic compounds containing a CN moiety. As well as nitriles, isonitriles (more generally called isocyanides) containing an NC functional group are generally highly toxic. The isocyano group, which possesses an unusual valence structure and reactivity, exhibits a dual nucleophilic/electrophilic character, which is often exploited for synthetic applications, e.g., in the synthesis of peptides, coordination chemistry, organometallic reactions, and carbohydrate chemistry (5). On the other hand, many natural isonitriles have been isolated from various organisms, including bacteria, fungi, and marine sponges (5–7). An isocyanide metabolite, xanthocillin, was first isolated from Penicillium notatum (8). Most isonitriles show a wide antibiotic activity spectrum and exhibit potential as possible agents with practical use (9, 10). Some metabolic intermediates of some isonitriles have been elucidated through incorporation experiments (11–15), and the synthetic pathway involved was determined very recently (16–19). However, information on the metabolism of an isonitrile at the protein and gene levels had been quite limited until we discovered an isonitrile-degrading enzyme. We have extensively studied the biological metabolism of isonitriles and revealed that it is quite different from nitrile metabolism (20–23). We initially found isonitrile hydratase, which catalyzes the hydration of an isonitrile (R-NC) to the corresponding N-substituted formamide [R-NH-CH(=O)] (24, 25). We also discovered N-substituted formamide deformylase, which catalyzes the hydrolysis of an N-substituted formamide to the corresponding amine and formate (26). Isonitrile hydratase and N-substituted formamide deformylase have been approved as new enzymes by the International Union of Biochemistry and Molecular Biology (IUBMB), and new EC numbers (EC 4.2.1.103 and EC 3.5.1.91, respectively) have been given to them. Very recently, we discovered the second isonitrile hydratase, which cooperated with N-substituted formamide deformylase in Arthrobacter pascens F164 (27). These two isonitrile hydratases are biochemically, immunologically, and structurally different from each other. However, there have been only three reports (24–27) on the enzymes involved in isonitrile metabolism.

There are several types of amine-forming deformylases involved in the metabolism of N-substituted formamides, including kynurenine formamidase (EC 3.5.1.9) (28–30), formylmethionine deformylase (EC 3.5.1.31) (31, 32), and peptide deformylase (EC 3.5.1.88) (33–35). Although N-substituted formamide deformylase is one of the amine-forming deformylases, it shows no amino acid sequence similarity to any other deformylases known so far. Contrary to expectation, the deduced amino acid sequence of N-substituted formamide deformylase exhibited the highest overall sequence identity (28%) to those of regulatory proteins. Only 15 amino acid residues in the N-terminal region (residues 58 to 72) of N-substituted formamide deformylase showed significant sequence identity (27 to 73%) to those in each member of the amidohydrolase superfamily (36), including imidazolonepropionase (37), atrazine chlorohydrolase (38), cytosine deaminase (39), dihydroorotase (40), and urease (41); except for these 15 amino acid residues in the N-terminal region, there was no similarity in the overall sequences. Like the other enzymes in the amidohydrolase superfamily, N-substituted formamide deformylase would have an (α/β)₈ barrel structure in the N-terminal region that contains five conserved residues (four histidines and one aspartic acid). Moreover, this enzyme contains three zinc ions per subunit (26). Based on these findings, N-substituted formamide deformylase was proposed to be a member of the amidohydrolase superfamily (36).

Here, we discovered that N-substituted formamide deformylase involved in the degradation of toxic isonitriles is able to catalyze a unique reverse reaction: N-benzylformamide (NBFA) synthesis from formate and benzylamine. When acetate and propionate were each used as an acid substrate instead of formate, N-benzylacetamide and N-benzylpropionamide, respectively, were also formed. To our knowledge, this is the first report of the reverse reaction of amine-forming deformylase and the enzymatic production of N-benzyl carboxamides through the reverse reaction of amine-forming deformylase.

MATERIALS AND METHODS

Materials.

Benzylamine was purchased from Tokyo Kasei Kogyo Co., Ltd. (Tokyo, Japan). Formate was obtained from Kishida Chemical Co., Ltd. (Osaka, Japan). NBFA was purchased from Sigma. Propionic acid was obtained from Nakalai Tesque (Kyoto, Japan). N-Benzylacetamide was purchased from Wako Pure Chemical Industries, Ltd. (Osaka, Japan). N-Benzylpropionamide was obtained from Chemical Technologies & Investigations, Ltd. (Moscow, Russia). DEAE-Sephacel, Resource ISO, Superose 12 10/30, and a low-molecular-weight standard kit were obtained from GE Healthcare UK Ltd. (Buckinghamshire, United Kingdom). All other biochemicals were standard commercial preparations.

Purification of N-substituted formamide deformylase.

All purification procedures were performed at 0 to 4°C. Potassium phosphate buffer (KPB) (pH 7.5) containing 10% (vol/vol) glycerol was used throughout the purification unless otherwise noted. Centrifugation was carried out for 15 min at 13,000 × g.

Cells were harvested by centrifugation, washed twice with 10 mM KPB, and then disrupted by sonication at 180 W for 20 min with an Insonator model 201 M (Kubota, Tokyo, Japan) to prepare a cell extract. The cell debris was removed by centrifugation. The resultant supernatant was fractionated with ammonium sulfate (40 to 45%), followed by dialysis against 10 mM KPB. The dialyzed solution was applied to a DEAE-Sephacel column (5 by 40 cm; GE Healthcare UK Ltd.) equilibrated with 10 mM KPB containing 0.25 M KCl. Proteins were eluted by increasing the ionic strength of KCl from 0.25 to 0.5 M in a linear manner in the same KPB. The active fractions were collected, and then ammonium sulfate was added to give 70% saturation. After centrifugation of the suspension, the precipitate was dissolved in 10 mM KPB, followed by dialysis against 10 mM KPB. The enzyme solution obtained was brought to 25% ammonium sulfate saturation and then placed on a Resource ISO column (1.6 by 3 cm; GE Healthcare UK Ltd.) equilibrated with 10 mM KPB containing 25% saturated ammonium sulfate. The enzyme was eluted with a linear gradient of ammonium sulfate (from 25 to 15% saturation) in 10 mM KPB. The active fractions were combined and concentrated with a Centricon-10 microconcentrator (Amicon Inc., Beverly, MA) in 50 mM KPB containing 0.15 M NaCl. The enzyme solution obtained was then applied to a Superose 12 10/30 column (GE Healthcare UK Ltd.) equilibrated with 10 mM KPB containing 0.15 M NaCl. Protein was eluted from the column with the same buffer. The active fractions were pooled and then precipitated with ammonium sulfate at 70% saturation. After centrifugation of the suspension, the precipitate was dissolved in 10 mM KPB, followed by dialysis against 10 mM KPB. The homogeneity of the purified protein was confirmed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE).

Assay of the reverse reaction of N-substituted formamide deformylase.

All of the reactions were performed under linear conditions for protein (0.1 mg/ml) and time (5 min). The standard assay mixture comprised 0.1 M KPB (pH 7.5), 100 mM benzylamine, 2 M formate, and 0.1 mg/ml enzyme (N-substituted formamide deformylase) in a total volume of 100 μl. The reaction was started by the addition of the enzyme and carried out at 25°C for 5 min. To stop the reaction, 10 μl of the reaction mixture was taken and added to 190 μl of 1 M citrate-Na₂PO₄ buffer (pH 4.0). A supernatant was obtained by centrifugation (15,000 × g; 5 min). The amount of NBFA formed was determined by high-performance liquid chromatography (HPLC) with a Shimadzu (Kyoto, Japan) LC-10AD system equipped with a Cosmosil 5C₁₈-AR-II column (reversed phase; 4.6 by 150 mm; Nacalai Tesque, Kyoto, Japan). The following solvent system was used at a flow rate of 1.0 ml/min and 40°C: 10 mM KH₂PO₄-H₃PO₄ buffer (pH 2.7)–acetonitrile, 2:1 (vol/vol). The absorbance was measured at 198 nm.

One unit of reverse activity of N-substituted formamide deformylase was defined as the amount of enzyme that catalyzed the formation of 1 μmol NBFA per min from benzylamine and formate under the above-mentioned conditions. Various amines and acids were tested for substrate specificity. The assay was carried out in a reaction mixture (100 μl) consisting of 100 mM amine substrate, 2 M acid substrate, and 0.1 mg/ml enzyme at 25°C for an appropriate time. The amount of reaction product was determined by monitoring the column effluent at 198 nm with authentic standards, using the HPLC system described above. One unit of reverse activity of N-substituted formamide deformylase was defined as the amount of enzyme that catalyzed the formation of 1 μmol product per min from the amine substrate and acid substrate under the above-mentioned conditions. Protein concentrations were determined as described by Bradford (42). Specific activity was expressed as units per mg of protein. k_cat values were calculated using an M_r value of 58,694 for the N-substituted formamide deformylase monomer (26).

Kinetic analysis.

Bisubstrate kinetic analysis was performed with formate concentrations in the range of 0.6 to 1.5 M and benzylamine concentrations in the range of 10 to 50 mM. The equations used for kinetics analysis were as described by Cleland (43–45) and Segel (46). To discriminate sequential and ping-pong mechanisms for two-substrate kinetics, data were fitted to the following two equations: v = V_max/(K_a/[A] + K_b/[B] + K_ia × K_ia/[A][B] + 1) (sequential) and v = V_max/(K_a/[A] + K_b/[B] + 1) (ping-pong).

Enzyme kinetic data conforming to linear inhibition, as determined by secondary replotting of the slopes and/or intercepts of the initial double-reciprocal plots (1/ν versus 1/A) versus inhibitor concentration, were fitted to the following three equations, which correspond to competitive, noncompetitive, and uncompetitive inhibition models (47, 48): v = V_max × [S]/{K_m(1 + [I]/K_i) + [S]} (competitive), v = V_max × [S]/{K_m(1 + [I]/K_i) + [S](1 + [I]/K_i)} (noncompetitive), and v = V_max × [S]/{K_m + [S](1 + [I]/K_i)} (uncompetitive).

Under these conditions, v represents the measured velocity; V_max the maximum velocity; A, B, and S the substrates; I the inhibitor; K_a, K_b, and K_m the Michaelis constants for A, B, and S; K_ia the dissociation constant for A; K_ib the dissociation constant for B; and K_i the inhibitor constant.

LC-ESI-MS analysis.

Because liquid chromatography-mass spectrometry (LC-MS) analysis was prevented by phosphate in the solvent used for the HPLC analysis, the peak fraction of the reaction product obtained on HPLC was taken and again applied to the above-described HPLC system with 33% acetonitrile instead of 10 mM KH₂PO₄-H₃PO₄ buffer (pH 2.7)–acetonitrile, 2:1 (vol/vol), in order to exclude phosphate from the sample. The peak fraction was collected and concentrated by evaporation. In order to determine whether the concentrated products remained stable during the second HPLC process and evaporation, the purified products were applied to the above-described HPLC system and confirmed to show the same retention times as authentic standards. The fraction was then subjected to LC-MS analysis. LC-MS was performed on a Waters Micromass ZQ coupled to a Waters Alliance HPLC system (2690 Separations Module and Waters 996 photodiodoarray detector) employing a Symmetry C₁₈ column (2.1 by 150 mm; 3.5 μm). The column was eluted at 30°C with 20% (vol/vol) acetonitrile in water at a flow rate of 0.2 ml/min. The sample was ionized with an electrospray ionization (ESI) probe in the positive-ion mode under the following source conditions: source temperature, 120°C; desolvation temperature, 300°C; capillary potential, 3.75 kV; sampling-cone potential, 35 V; extractor, 2 V; and nitrogen flow rate, 300 liters/h.

NMR analysis.

The reaction mixture obtained at 25°C for 20 h was purified with a Sep-Pak C₁₈ cartridge (Waters). The reaction product was eluted with 10% (vol/vol) methanol in water and concentrated by evaporation. The reaction product dissolved in water was further purified by HPLC (TSK-gel ODS-80Ts [7.8 by 300 mm; Tosoh Co., Tokyo, Japan], 25% [vol/vol] acetonitrile in water). The peak fraction was collected and concentrated to dryness (19 mg). Nuclear magnetic resonance (NMR) spectra were measured with a DPX-300 (Bruker, Ettlingen, Germany). Samples were prepared by dissolving in CDCl₃.

RESULTS

Reverse activity of N-substituted formamide deformylase.

The ability of the enzyme to catalyze the reverse reaction was examined (Fig. 1). First, when low concentrations of substrates (10 mM formate and 10 mM benzylamine) and the purified enzyme (0.1 mg/ml) were added to the reaction mixture, a new peak was not detected on HPLC. However, when high concentrations of substrates (1 M formate and 1 M benzylamine) and the purified enzyme (0.1 mg/ml) were added to the reaction mixture, a new peak was detected on HPLC. The amount of product was proportional to the reaction time and the enzyme concentration. Without the enzyme, the new product was not detected. The retention time (3.0 min under the experimental conditions used) of the product on HPLC agreed with that of authentic NBFA. Then, the new product was purified, and its molecular weight was determined by LC-ESI-MS. The mass spectrum of the product was consistent with the corresponding estimated molecular weight of NBFA (135 M_r) (Fig. 2A). Additionally, the NMR spectrum of the product [¹H NMR (ppm): 4.46 (2H, d, CH₂), 7.20–7.40 (5H, m), 6.11 (1H, br, NH), 8.23 (1H, s, CHO)] was consistent with that of authentic NBFA. These findings clearly demonstrated that N-substituted formamide deformylase catalyzed the reverse reaction, synthesizing NBFA from benzylamine and formate (Fig. 1).

FIG 1.

Reverse reaction of N-substituted formamide deformylase.

FIG 2.

Mass spectra on LC-ESI-MS of the reverse-reaction products of N-substituted formamide deformylase. The structural formulae show NBFA (A), N-benzylacetamide (B), and N-benzylpropionamide (C). The major mass peaks at m/z 136, 150, and 164 correspond to NBFA, N-benzylacetamide, and N-benzylpropionamide, respectively.

To determine the kinetic parameters of the reverse reaction, the reverse activity was measured with various concentrations of one substrate with a fixed concentration of another one. From linear Lineweaver-Burk plots, the K_m and V_max values for formate were found to be 3,010 ± 110 mM and 33.0 ± 1.0 units/mg, respectively, and those for benzylamine were 53.6 ± 0.5 mM and 22.5 ± 0.2 units/mg, respectively (Table 1). On the other hand, these values for NBFA for the forward reaction were 0.075 mM and 52.7 units/mg, respectively (26). The catalytic efficiencies (k_cat/K_m) of formate (0.011 s⁻¹ · mM⁻¹) and benzylamine (0.411 s⁻¹ · mM⁻¹) for the reverse reaction were significantly lower than that of NBFA for the forward reaction (687 s⁻¹ · mM⁻¹) (Table 1). Judging from the k_cat/K_m ratio, the reverse reaction proceeded only slightly compared with the forward reaction.

TABLE 1.

Kinetic parameters of N-substituted formamide deformylase

Reactiona	Km (mM)	Vmax (units/mg)	kcat/Km (mM−1 · s−1)
Reverse (formate)	3,010 ± 110	33.0 ± 1.0	0.011
Reverse (benzylamine)	53.6 ± 0.5	22.5 ± 0.2	0.411
Forward (NBFA)	0.075	52.7	687

^aNBFA was used as the substrate for the forward reaction. For the reverse reaction, benzylamine and formate were used as the substrates.

Effects of temperature and pH on the activity and stability of the reverse reaction.

The effects of pH on the enzyme activity and stability of the reverse reaction were examined. The enzyme exhibited maximum activity at pH 7.0 (Fig. 3A). The optimum temperature was 25°C. The stability of the enzyme was examined at various temperatures. After the enzyme had been preincubated for 30 min in 10 mM KPB (pH 7.5) containing 10% (vol/vol) glycerol, an aliquot of the enzyme solution was taken and the enzyme activity was assayed. The enzyme solution exhibited the following activities: 60°C, 0%; 55°C, 0%; 50°C, 78%; 45°C, 100%; 40°C, 100%; 35°C, 100%; 30°C, 100%; 25°C, 99%; 20°C, 100%; 15°C, 99%; and 10°C, 99%. The stability of the enzyme was also examined at various pH values. After the enzyme had been incubated at 25°C for 30 min in four buffers at a concentration of 0.1 M (citrate-Na₂HPO₄ buffer, pH 4.0 to 8.0; potassium phosphate buffer, pH 6.0 to 8.0; Tris-HCl buffer, pH 7.5 to 9.0; and NH₄OH-NH₄Cl buffer, pH 8.5 to 10.0), an aliquot of the enzyme solution was taken, and then the enzyme activity was assayed under the standard conditions. N-Substituted formamide deformylase was most stable in the pH range of 7.5 to 8.0 (Fig. 3B).

FIG 3.

Effects of pH on the reverse activity and stability of N-substituted formamide deformylase. (A) Reactions were carried out for 5 min at 25°C in the following buffers (0.1 mM): citrate-Na₂HPO₄ buffer (○), potassium phosphate buffer (■), Tris-HCl buffer (▲), and NH₄OH-NH₄Cl buffer (×). (B) The reverse activity was assayed after the enzyme had been preincubated for 30 min in the buffers described in the legend to panel A. Relative activity is expressed as a percentage of the maximum activity attained under the experimental conditions used.

Effects of inhibitors on the reverse activity.

Various compounds were investigated for their inhibitory effects on the enzyme activity (Table 2). Each compound was added to the standard reaction mixture without the substrate, and then the reaction was started by adding the substrate. The final concentration of each tested compound was 1 mM unless otherwise stated. The enzyme was very sensitive to HgCl₂, CuCl (at 0.25 mM), CuCl₂, and AgNO₃; the inhibition was 100% in all cases. SnCl₂ and CdCl₂ also showed inhibitory effects on the reverse enzymatic activity (44% and 57%, respectively). The enzyme was inhibited by thiol-specific reagents, such as N-ethylmaleimide and p-chloromercuribenzoate, whereas iodoacetate and 5,5′-dithio-bis-2-nitrobenzoate did not inhibit the reverse activity at all. Carbonyl-specific reagents, e.g., aminoguanidine and semicarbazide, hardly inhibited the enzyme, but phenylhydrazine caused only partial inhibition (23%). Chelating agents, such as α,α′-dipyridyl, KCN, diethyldithiocarbamate, and EDTA, did not influence the reverse activity at all, but o-phenanthroline and 8-hydroxyquinoline caused appreciable inhibition (32% and 84%, respectively). The enzyme was unaffected by oxidizing reagents and serine-modifying reagents, such as H₂O₂, ammonium persulfate, phenylmethanesulfonyl fluoride, and diisopropyl fluorophosphates. However, reducing reagents, such as dithiothreitol and 2-mercaptoethanol, caused remarkable inhibition (100% and 64% inhibition, respectively).

TABLE 2.

Effects of various compounds on the activity of N-substituted formamide deformylase^a

Inhibitor	Relative activity (%)
None	100
LiCl, NaCl, MgCl2, CaCl2, BaCl2, MnCl2, AlCl3, Pb(NO3)2, FeSO4, FeCl3, RbCl, SrCl2, CsCl, Na2MoO4, CoCl2, NiCl2, and ZnCl2	87–111
SnCl2	56
CdCl2	43
CuCl (0.25 mM), CuCl2, AgNO3, and HgCl2	0
Iodoacetate	99
5,5′-Dithio-bis-2-nitrobenzoateb	105
N-Ethylmaleimideb	21
p-Chloromercuribenzoate	0
Semicarbazide	92
Aminoguanidine	81
Phenylhydrazine	77
α,α′-Dipyridyl	85
o-Phenanthroline	68
8-Hydroxyquinolineb	16
EDTA	127
Diethyldithiocarbamate	98
NaN3	101
KCN	108
Dithiothreitol	0
2-Mercaptoethanol	36
H2O2	91
Ammonium persulfate	112
Phenylmethanesulfonyl fluorideb	102
Diisopropyl fluorophosphateb	117

^aEach compound was added to the standard reaction mixture without the substrate, and then assay of the enzyme was performed after adding the substrate. The final concentrations of the tested compounds were 1 mM unless otherwise stated.

^bMethanol was added to the reaction mixture to a final concentration of 5% (vol/vol) to enhance the solubility of the tested compound.

Substrate specificity of the reverse activity.

First, the ability of the enzyme to catalyze the condensation of various amines and formate was examined in a reaction mixture that included the latter as the acid substrate. All of the tested amines, i.e., aniline, phenethylamine, 3-phenyl-1-propylamine, ethylamine, propylamine, isopropylamine, allylamine, and butylamine, were inert as substrates for the reverse reaction. Next, the ability of the enzyme to catalyze the condensation of various acids and benzylamine was examined in a reaction mixture that included the latter as the amine substrate. Among the tested acids, acetate and propionate were also active as substrates for the reverse reaction (Table 3), whereas the other acids, such as butyrate and isobutyrate, were inert. The compounds produced from acetate and propionate through the reverse reaction were examined by HPLC. The retention times of the products (3.5 min for the product derived from acetate and benzylamine and 3.9 min for that from propionate and benzylamine) on HPLC agreed with those of authentic N-benzylacetamide and N-benzylpropionamide. Each of these reaction products was purified and concentrated, and their molecular weights were determined by LC-ESI-MS. The results for the products were consistent with the corresponding estimated molecular weights (N-benzylacetamide, 149 M_r, and N-benzylpropionamide, 163 M_r) (Fig. 2B and C). These findings strongly demonstrated that the products were N-benzylacetamide and N-benzylpropionamide, both of which are nontoxic, unlike NBFA. From linear Lineweaver-Burk plots, the K_m and V_max values for acetate were found to be 1,210 ± 30 mM and 10.0 ± 0.2 units/mg, respectively, with the values for propionate being 3,850 ± 100 mM and 13.6 ± 0.5 units/mg, respectively. The catalytic efficiencies (k_cat/K_m) were 0.008 s⁻¹ · mM⁻¹ and 0.003 s⁻¹ · mM⁻¹ for acetate and propionate, respectively (Table 3).

TABLE 3.

Substrate specificity of the reverse reaction of N-substituted formamide deformylase^a

Substrate	Product	Km (mM)	Vmax (units/mg)	kcat/Km (mM−1 · s−1)
Formate	N-Benzylformamide	3,010 ± 110	33.0 ± 1.0	0.011
Acetate	N-Benzylacetamide	1,210 ± 30	10.0 ± 0.2	0.008
Propionate	N-Benzylpropionamide	3,850 ± 100	13.6 ± 0.5	0.003

^aTo determine kinetic parameters, the reverse activity was measured with various concentrations of acids and a fixed concentration of benzylamine. Apparent K_m and V_max values were obtained from Lineweaver-Burk plots.

Initial velocity patterns.

Distinguishing of the kinetics mechanisms was performed by analyzing the initial velocity patterns when the concentration of one substrate was varied with several fixed concentrations of the second substrate. A classical sequential mechanism is indicated when a family of double-reciprocal plots intersects to the left of the y axis and converges at the x axis, because both the slope and the intercept change as the concentration of the fixed substrate changes. In the case of a ping-pong mechanism, the slope of a series of double-reciprocal plots remains unchanged, i.e., only the intercept changes as the concentration of the fixed substrate changes, giving rise to a series of parallel lines (47). A double-reciprocal plot was constructed by plotting 1/velocity against 1/[benzylamine], with different fixed concentrations of formate (Fig. 4). These data best fitted a classical sequential mechanism, because the family of curves intersected on the x axis. In a sequential mechanism, the enzyme binds to both substrates, and a ternary complex is formed before the first product is released.

FIG 4.

Two-substrate kinetic analysis of N-substituted formamide deformylase. Double-reciprocal plotting of 1/v versus 1/benzylamine was performed using the data obtained in the initial velocity studies. Initial velocities were measured in the presence of 10 to 50 mM benzylamine and 0.6 to 1.5 M formate.

Dead-end inhibitors.

A variety of compounds, comprising analogues of benzylamine and formate, were investigated as possible inhibitors of the reverse activity. Each potential inhibitor was tested over the concentration range of 1 mM to 0.5 M, and the 50% inhibitory concentration (IC₅₀) was determined (Table 4). Amines containing a benzene ring, such as aniline, phenethylamine, and 3-phenylpropylamine, inhibited the reverse reaction. The most potent inhibitor was 3-phenylpropylamine, with an IC₅₀ of 13.3 mM. However, amines containing no benzene ring did not cause 50% inhibition at concentrations as high as 300 mM. On the other hand, acids such as butyrate and isobutyrate inhibited the reverse reaction, with IC₅₀s of 275 mM and 280 mM, respectively. Valerate, isovalerate, and caproate could not be tested because they were insoluble in the reaction buffer.

TABLE 4.

Effects of substrate analogues on the reverse activity^a

Compound	IC50 (mM)
Amines
Aniline	14.5
Phenethylamine	27.8
3-Phenylpropylamine	13.3
Ethylamine	>300
Propylamine	>300
Isopropylamine	>300
Allylamine	>300
Butylamine	>300
Acids
Butyrate	275
Isobutyrate	280
Valerate	NA
Isovalerate	NA
Caproate	NA

^aThe reverse activity was assayed as described in the text, and the IC₅₀ for each inhibitor was determined. NA, not applicable.

Aniline and butyrate, which are analogues of benzylamine and formate substrates, respectively, were used as inhibitors of both substrates in order to elucidate the substrate-binding order. The results are summarized in Table 5. Aniline was found to be a competitive inhibitor of benzylamine (K_is [the dissociation constant for the enzyme-inhibitor complex] = 2.87 mM) and also an uncompetitive inhibitor of formate (K_ii [the dissociation constant for the inhibitor from the enzyme-substrate-inhibitor complex] = 4.56 mM) (Fig. 5A and B and Table 5). Butyrate was found to be not only a noncompetitive inhibitor of benzylamine (K_is = 57.5 mM and K_ii = 59.7 mM), but also a competitive inhibitor of formate (K_is = 273 mM) (Fig. 5C and D and Table 5). This is consistent with ordered two-substrate, two-product (bi-bi) substrate addition, with formate being the obligate first substrate, followed by benzylamine addition (Fig. 6).

TABLE 5.

Kinetic constants of N-substituted formamide deformylase and inhibition patterns with aniline and butyrate

Inhibitor	Variable substrate	Fixed substrate	Inhibition patterna	Kis (mM)	Kii (mM)
Aniline	Benzylamine	Formate	C	2.87
	Formate	Benzylamine	UC		4.56
Butyrate	Benzylamine	Formate	NC	57.5	59.7
	Formate	Benzylamine	C	273

^aC, competitive; NC, noncompetitive; UC, uncompetitive.

FIG 5.

Dead-end inhibition of N-substituted formamide deformylase by aniline or butyrate. Shown are double-reciprocal plots of 1/v versus 1/benzylamine with 5 to 25 mM benzylamine and 2 M formate (A), 1/v versus 1/formate with 0.6 to 1.5 M formate and 0.1 M benzylamine (B), 1/v versus 1/benzylamine with 5 to 25 mM benzylamine and 2 M formate (C), and 1/v versus 1/formate with 0.3 to 1.2 M formate and 0.1 M benzylamine (D). The aniline and butyrate concentrations were as indicated in the insets.

FIG 6.

Proposed kinetic mechanism of N-substituted formamide deformylase. F, formate; B, benzylamine; Enz, enzyme.

Enzymatic production of N-benzyl carboxamides.

Various reaction conditions for the efficient production of N-benzyl carboxamides were studied using the purified N-substituted formamide deformylase. The effects of the reaction time and substrate concentration on the production of NBFA were investigated. At first, the concentration of formate in the reaction mixture containing 0.2 M benzylamine was varied. The production of NBFA with 1 M formate was high compared with that with 0.5 M, 1.5 M, and 2 M formate (Fig. 7A). Thereafter, the amounts of NBFA were measured in the presence of 1 M formate and various concentrations of benzylamine. The maximum production of NBFA in the 24-h reaction was 16 mM when the concentration of benzylamine used was 0.2 M (Fig. 7B). With regard to the production of N-benzylacetamide and N-benzylpropionamide, the effects of the reaction time and substrate concentration were also investigated. The maximum production of N-benzylacetamide and N-benzylpropionamide in the 24-h reaction amounted to 22 mM and 18 mM, respectively, when the concentrations of each acid and benzylamine initially used were 1.0 M and 0.2 M, respectively (Fig. 7C to F).

FIG 7.

Effects of the concentrations of the substrates on N-benzyl carboxamide synthesis. The amounts of NBFA were measured in the presence of various concentrations of formate and benzylamine: 0.5 to 2 M formate and 0.2 M benzylamine (A) and 0.1 to 0.4 M benzylamine and 1 M formate (B). The amounts of N-benzylacetamide were measured in the presence of various concentrations of acetate and benzylamine: 0.5 to 2 M acetate and 0.2 M benzylamine (C) and 0.1 to 0.4 M benzylamine and 1 M acetate (D). The amounts of N-benzylpropionamide were measured in the presence of various concentrations of propionate and benzylamine: 0.5 to 2 M propionate and 0.2 M benzylamine (E) and 0.1 to 0.4 M benzylamine and 1 M propionate (F).

DISCUSSION

We are interested in how C-N hydrolases evolved (27, 49, 50). Because N-substituted formamide contains a nitrogen-carbon bond, an N-substituted formamide deformylase belongs to the C-N hydrolases. Functional and kinetic analyses of N-substituted formamide deformylase would contribute to clarification of the metabolism of isonitriles in nature and provide us with new knowledge about C-N hydrolases, which might facilitate elucidation of their functional and structural evolution. Previously the reverse activity of an N-substituted formamide deformylase had never been reported. In this work, we clarified the biochemical characteristics of the reverse activity of the purified N-substituted formamide deformylase. It is well known that some lipases (51, 52), proteases (53), and glycosidases (54, 55) catalyze reverse-hydrolysis (condensation) reactions or exchange reactions. However, these reactions hardly occur in an aqueous solution and require the presence of substrates at almost saturated concentrations. N-Substituted formamide deformylase was able to catalyze a condensation reaction efficiently even in an aqueous solution with high concentrations of substrates. To our knowledge, this is the first report of the reverse reaction of not only an amine-forming deformylase, but also an enzyme in the amidohydrolase superfamily, to which N-substituted formamide deformylase belongs (36).

It has also been reported that the reverse-hydrolysis and transfer reactions of several lipases, proteases, and glycosidases proceed when water molecules in the reaction mixture are depleted on the addition of an organic solvent to the reaction mixture (51, 53, 55). Therefore, we examined the effects of various organic solvents on reverse hydrolysis at 12°C. Contrary to our expectations, N,N-dimethyl formamide, dimethyl sulfoxide, ethanol, methanol, and 1,4-butanediol inhibited the reverse reaction of N-substituted formamide deformylase at a concentration of 30% (vol/vol) (data not shown). Because organic solvents are in general known to abolish the activity of enzymes, N-substituted formamide deformylase may also become inactive in the presence of organic solvents.

Investigation of the substrate specificities for various amines revealed that only benzylamine was active as a substrate of the reverse reaction. This finding is consistent with the fact that N-substituted formamide deformylase exhibits a narrow substrate spectrum for the forward reaction (26). When various N-substituted formamides, amides, and other compounds were examined as forward reaction substrates, NBFA was found to be the most suitable substrate for the enzyme. N-Butylformamide (3.4%) was hydrolyzed to N-butylamine and formate at rates significantly lower than those of the activity toward NBFA (100%). However, the product of the forward reaction from butylformamide, N-butylamine, was inactive as an amine substrate for the reverse reaction. On the other hand, when acetate and propionate were each used as the acid substrate instead of formate, the reverse reactions, surprisingly, proceeded (Table 3 and Fig. 1), and N-benzylacetamide and N-benzylpropionamide were detected as reverse-reaction products, respectively (Fig. 2B and C). These findings demonstrated that acids with numbers of carbon atoms ranging from C₁ to C₃ were active as acid substrates for the reverse reaction of N-substituted formamide deformylase, which produces formate (a C₁ acid) as one of the forward reaction products. This unique substrate specificity of the reverse reaction has not previously been reported for any other deformylases known so far. If the substrate specificity for an amine is improved so as to be broad, various N-substituted acetamides and N-substituted propionamides could be obtained enzymatically. The effects of pH (Fig. 3) and chemical modification (Table 2) on the reverse reaction were similar to those on the forward reaction, suggesting that the same catalytic site on the enzyme is involved in both the reverse and forward reactions.

Elucidation of the kinetic mechanism of N-substituted formamide deformylase required the combination of several techniques. The enzyme kinetic mechanism of the two substrate reactions can be directly proved by simultaneously varying both substrates (46). It is well established that double-reciprocal plot analysis that generates an intersecting-line pattern suggests a sequential mechanism and that a parallel-line pattern is characteristic of a ping-pong mechanism (47). Here, the bisubstrate intersecting-line pattern obtained with N-substituted formamide deformylase (Fig. 4) is consistent with a sequential mechanism. In order to distinguish an ordered sequential mechanism from a random sequential mechanism, dead-end inhibition analyses were carried out. Aniline acted as a competitive inhibitor for benzylamine, suggesting that both compounds bind to the same enzyme form, but was uncompetitive when formate was the substrate (Fig. 5), suggesting formate binding generates the enzyme form to which aniline can bind. These data are consistent with an ordered sequential mechanism wherein formate binds to the enzyme first, followed by benzylamine association (Fig. 6).

In conclusion, we initially demonstrated the reverse activity of N-substituted formamide deformylase: enzymatic NBFA synthesis from formate and benzylamine. We found this unique characteristic only when the substrate concentration was high, although no reverse reaction was observed with low substrate concentrations. Interestingly, when acetate and propionate were each used as the acid substrate instead of formate, N-benzylacetamide and N-benzylpropionamide, respectively, were formed enzymatically. The maximum production of NBFA, N-benzylacetamide, and N-benzylpropionamide in the 24-h reaction amounted to 16 mM, 22 mM, and 18 mM, respectively, when the concentrations of each acid and benzylamine initially used were 1.0 M and 0.2 M, respectively. Kinetic studies revealed that the reverse reaction follows an ordered sequential mechanism. Additional characterization, including mutant analysis and determination of the enzyme structure, could provide information for further clarification of the reaction mechanism.

Our discovery of the reverse reaction of N-substituted formamide deformylase raises the possibility that other deformylases, which play various physiologically important roles, can also catalyze a reverse reaction to produce N-substituted carboxamides. The active site of N-substituted formamide deformylase has never been identified experimentally. Further studies on N-substituted formamide deformylase from the standpoint of its three-dimensional structure could provide information about the evolutionary relationship of this enzyme and other enzymes involved in the cleavage and synthesis of a C-N bond, such as nitrilase (56, 57), nitrile hydratase (58), amidase (49, 50), and aldoxime dehydratase (59).