Hydrazidase, a Novel Amidase Signature Enzyme That Hydrolyzes Acylhydrazides

Abstract

The degradation mechanisms of natural and artificial hydrazides have been elucidated. Here we screened and isolated bacteria that utilize the acylhydrazide 4-hydroxybenzoic acid 1-phenylethylidene hydrazide (HBPH) from soils. Physiological and phylogenetic studies identified one bacterium as Microbacterium sp. strain HM58-2, from which we purified intracellular hydrazidase, cloned its gene, and prepared recombinant hydrazidase using an Escherichia coli expression system. The Microbacterium sp. HM58-2 hydrazidase is a 631-amino-acid monomer that was 31% identical to indoleacetamide hydrolase isolated from Bradyrhizobium japonicum. Phylogenetic studies indicated that the Microbacterium sp. HM58-2 hydrazidase constitutes a novel hydrazidase group among amidase signature proteins that are distributed within proteobacteria, actinobacteria, and firmicutes. The hydrazidase stoichiometrically hydrolyzed the acylhydrazide residue of HBPH to the corresponding acid and hydrazine derivative. Steady-state kinetics showed that the enzyme hydrolyzes structurally related 4-hydrozybenzamide to hydroxybenzoic acid at a lower rate than HBPH, indicating that the hydrazidase prefers hydrazide to amide. The hydrazidase contains the catalytic Ser-Ser-Lys motif that is conserved among members of the amidase signature family; it shares a catalytic mechanism with amidases, according to mutagenesis findings, and another hydrazidase-specific mechanism must exist that compensates for the absence of the catalytic Ser residue. The finding that an environmental bacterium produces hydrazidase implies the existence of a novel bacterial mechanism of hydrazide degradation that impacts its ecological role.

INTRODUCTION

Hydrazine (H₂N-NH₂) and its derivatives constitute an important group of popular reducing agents and nucleophiles that serve as precursors of pharmaceuticals, dyes, and herbicides. Hydrazine derivatives containing an unsubstituted -NH₂ moiety are dehydrated and condensed with carbonyl compounds to generate hydrazones. Acylated derivatives of hydrazine are described as R₁R₂N-N(R₃)C(=O)R₄ and are called (acyl)hydrazides. One of the most popular hydrazides is isonicotinic acid hydrazide (INH), an antituberculosis agent that has been clinically applied and marketed under the name isoniazid (1). Isocarboxazid is commonly used as an antidepressant (2). Other hydrazides serve as agricultural herbicides, pesticides, insecticides, fungicides, or plant growth regulators (3). For example, maleic hydrazide has been extensively applied as a plant growth regulator and herbicide, especially in the cultivation of tobacco, potatoes, and onions. The industrial uses of hydrazides include cross-linking acrylic emulsions, curing epoxy resins, and scavenging formaldehyde through their powerful reactivity with ketone and epoxy groups (4).

Over 50 natural compounds possessing a hydrazide moiety have been identified (5). Agaritine from the commercial mushroom Agaricus bisporus and its relatives was the first natural hydrazide to be discovered (6). The false morel Gyromitra esculenta poisons foods and causes Gyromitra syndrome, probably via the toxic hydrazide gyromitrin and its degradation products, N-formyl-N-methylhydrazine and methylhydrazine (7). Other fungi and actinomycetes bacteria produce hydrazides that have potential antibiotic, antiviral, anticancer, and antioxidant properties (5).

In addition to the ecological, agricultural, medical, and industrial importance of hydrazides, a few synthetic and degradative biological mechanisms have been determined. Agaricus bisporus and Escherichia coli generate hydrazides via γ-glutamyltransferases that transfer the acyl residue of glutamate to arylhydrazines (8, 9), and Rhodococcus amidases and Sulfolobus γ-lactamase also transfer acyl residues to hydrazine to generate hydrazides (10–12), although the physiological significance of these reactions remains unclear. With respect to hydrazide degradation, A. bisporus and pig kidney γ-glutamyltransferases hydrolyze agaritine into p-(hydroxymethyl)phenylhydrazine and l-glutamate, but corresponding genes have not been identified (8, 13). Some amidases and chymotrypsin hydrolyze hydrazides, but these reactions seem to be nonphysiological because the reaction rate is very low (10, 14–17). Toida notably discovered an enzyme in Mycobacterium avium that hydrolyzes various hydrazides, including INH as well as nicotinoyl, benzoyl, and pyrazinoyl hydrazides, and named it hydrazidase (18). Although Toida obtained kinetic data about the purified hydrazidase (19), its amino acid sequences and reaction mechanisms remain unknown.

We previously described novel dehydrogenases from the fungus Candida palmioleophila MK883 and the bacterium Pseudomonas aeruginosa PAO1 that catalyze the oxidative hydrolysis of artificial hydrazone compounds at the C=N bond (20, 21). The hydrazone degradation mechanism shared across fungal and bacterial kingdoms indicated the microbial degradation of hydrazine-related compounds. We therefore investigated microorganisms that assimilate hydrazides and isolated a novel hydrazidase that hydrolyzes acylhydrazides from the bacterium Microbacterium sp. strain HM58-2. The deduced amino acid sequence of the hydrazidase was 31% identical to that of the indoleacetamide hydrolase (IAAH) of Bradyrhizobium japonicum and was up to 59% identical to those of hypothetical proteins. Phylogenetic studies indicated that the hydrazidase constitutes a novel “hydrazidase” group among the amidase signature proteins. Our findings shed light on a novel bacterial enzyme involved in hydrazide degradation and on the ecological role of microbial hydrazidases.

MATERIALS AND METHODS

Reagents.

Isonicotinic acid hydrazide, 4-hydroxybenzoic acid hydrazide (HBH), 4-hydroxybenzamide, and indole-3-acetamide were purchased from Tokyo Chemical Industry Co. Ltd. (Tokyo, Japan). Acetophenone and 4-hydroxybenzoic acid were obtained from Wako Pure Chemical Industries Ltd. (Osaka, Japan). Precision Plus protein standards (Bio-Rad Laboratories, Hercules, CA) served as size markers. Acetophenone hydrazone, adipic acid bis-(pentane-2-ylidene hydrazide) (APEYH), and adipic acid bis-(1-phenylethylidene hydrazide) (APEH) were synthesized as described in Supplemental Materials. Other chemicals were standard commercial preparations unless otherwise stated.

Synthesis of HBPH.

Acetic acid (5 ml) and HBH (2.5 g) were mixed, and then methanol (500 ml) was stirred into the mixture until the HBH completely dissolved. A 20-fold molar excess of acetophenone (39 ml) was added to the solution and refluxed at 80°C for 30 min. The methanol was evaporated, and cyclohexane was added to the residual liquid at room temperature to precipitate 4-hydroxybenzoic acid 1-phenylethylidene hydrazide (HBPH). Precipitates collected by filtration were rinsed with water and cyclohexane and then air-dried for a few days to yield HBPH. We recorded ¹H nuclear magnetic resonance (NMR) spectra on an EX-270 NMR spectrometer (Jeol, Tokyo, Japan) with trimethylsilane as the internal standard as follows: δ_H (dimethyl sulfoxide-d₆ [DMSO-d₆]), 7.79 (2H, d, H-1), 6.87 (2H, d, H-2), 2.36 (3H, s, H-1′), 7.82 (2H, d, H-2′), 7.42 (3H, t, H-3′), 10.53 (1H, s, N-H), and 10.11 (1H, br, O-H) (see Fig. S1 in the supplemental material).

Bacterial strains and culture conditions.

Microbacterium sp. HM58-2 in Luria-Bertani (LB) medium or minimal medium (MM) comprising 10 mM HBPH, 10 mM NH₄Cl, 0.052% KCl, 0.052% MgSO₄·7H₂O, 50 mM KH₂PO₄, 0.1% vitamin solution (M6895; Sigma-Aldrich, St. Louis, MO), and 0.2% Hutner's trace element solution (22) (pH 7.2) (MM-HBPH) was shaken at 120 rpm at 30°C unless otherwise stated. Escherichia coli strains JM109 (TaKaRa Bio Inc., Shiga, Japan) and DH10B (Invitrogen, Carlsbad, CA) were used for DNA manipulations, and Escherichia coli strain BL21(DE3) (Novagen, Madison, WI) was used for recombinant enzyme production. Escherichia coli was propagated in LB broth at 37°C with shaking at 120 rpm. Growth was monitored as optical density at 600 nm (OD₆₀₀). Kanamycin (40 μg/ml) was added to LB broth as required.

Screening microorganisms that degrade hydrazide.

We obtained 119 soil samples in or around the campus of the University of Tsukuba. A small amount (20 to 30 mg) of the samples was suspended in 5 ml of MM-HBPH medium and incubated at 30°C for 3 days with shaking. Portions (100 μl) were then subcultured in 5 ml of fresh medium and further incubated for 3 days at 30°C. After three or four subculture procedures, the culture broth was spread on MM-HBPH agar solid medium and incubated at 30°C. We then analyzed the ability of microorganisms that generated colonies on the agar to produce HBPH-degrading enzymes. The 16S rRNA gene of the HM58-2 strain was sequenced with primers listed in Table S1 in the supplemental material.

Hydrazidase assay.

Reaction mixtures comprised 40 mM Tris-HCl buffer (pH 8.0), 0.5 mM HBPH, and 5% methanol in a total volume of 200 μl. The enzyme was added, and the reaction proceeded at 30°C for 15 to 30 min before being stopped by adding 200 μl of ice-cold acetonitrile. The HBPH concentration was determined using a high-performance liquid chromatography (HPLC) system (HP-1100; Agilent Technologies, Santa Clara, CA) equipped with an ODS-4 column (4.6 by 150 mm) containing Inertsil (particle diameter, 3 μl) and by monitoring absorption at 210 nm. The mobile-phase solvent system comprised 50 mM potassium phosphate (pH 7.2) and acetonitrile (6:4 [vol/vol]) at a flow rate of 0.8 ml min⁻¹.

Purification of hydrazidase.

Microbacterium sp. HM58-2 was grown in MM (12 liters) containing 4-hydroxybenzoic acid at 28°C for 20 h. The inoculum (10%) was taken from an overnight culture grown in LB broth at 28°C. The cells were harvested by centrifugation at 5,000 × g for 20 min. All purification steps proceeded at 0 to 4°C in buffer A consisting of 50 mM potassium phosphate (pH 7.2), 10% glycerol, 0.1 mM dithiothreitol (DTT), and 0.1 mM EDTA. Cells suspended in buffer A (70 ml) were lysed by sonication, insoluble material was removed by ultracentrifugation at 100,000 × g for 40 min, and then the cell extract was applied to a 100-ml column containing pDE52 DEAE cellulose (Whatman International Ltd., Maidstone, Kent, United Kingdom) equilibrated with buffer A. The column was washed with 500 ml of buffer A, and then bound proteins were eluted using a 500-ml linear 0-to-1.5 M KCl gradient. The active fractions were combined, and then ammonium sulfate was added to achieve a final concentration of 1.5 M. The resulting enzyme solution was applied to a 60-ml column containing butyl-Sepharose (GE Healthcare, Little Chalfont, Buckinghamshire, United Kingdom) equilibrated with buffer A supplemented with ammonium sulfate to a final concentration of 1.5 M. The column was washed with 300 ml of this buffer, and then bound proteins were eluted in a 300-ml linear gradient of 1.5 to 0 M ammonium sulfate. The active fractions were pooled and dialyzed against buffer A without ammonium sulfate, and a second butyl-Sepharose column chromatography experiment was performed with the same procedure as described above. The active fractions were pooled and dialyzed against buffer A without ammonium sulfate. The dialysate was applied to a Mono Q 5/50 GL column (GE Healthcare) equilibrated with buffer A. The column was washed with 5 ml of buffer A, and then bound proteins were eluted in a 10-ml linear gradient of 0 to 1.5 M KCl. Active fractions were pooled, and then purity was confirmed by SDS-PAGE. Protein concentrations were determined using a Bio-Rad assay (Bio-Rad) with bovine serum albumin as the standard.

DNA manipulations.

Restriction endonucleases, DNA polymerase, and T4 DNA ligase were purchased from TaKaRa Bio Inc. Nucleotides were sequenced by dideoxy-chain termination using a CEQ2000 automated DNA sequencer (Beckman Coulter, Brea, CA) according to the manufacturer's instructions. Unless otherwise stated, DNA manipulations proceeded essentially according to the method of Sambrook et al. (23).

Cloning the hydrazidase gene.

Purified hydrazidase (10 μg) was digested with either trypsin or the endoproteinase Asp-N (0.1 μg each) in 20-μl reaction mixtures at 37°C for 3 h. The resulting peptides were resolved by SDS-PAGE and then electronically transferred to polyvinylidene difluoride membranes. Blots were stained with Coomassie brilliant blue R-250, peptides were excised, and then amino-terminal amino acids were determined using a Model Procise 492 automated protein sequencer (Applied Biosystems). The degenerate primers Try_F1 and Asp-N_R1 (see Table S1 in the supplemental material) were used in PCR amplification of the gene fragment of hydA. Inverse PCR proceeded as described previously (24) using primers Inv1 and Inv2 (see Table S1). Genomic DNA of the HM58-2 strain was digested with PstI, and the resulting 2.5-to-3.5-kb fragments purified by agarose gel electrophoresis served as a template. We prepared a digoxigenin (DIG)-labeled DNA probe from 290-bp fragments using a DIG DNA labeling kit and an alkaline phosphatase-conjugated anti-DIG antibody (Roche Diagnostics, Basel, Switzerland). Southern blots were analyzed according to the manufacturer's instructions.

Preparation of recombinant hydrazidase.

The hydrazidase gene was amplified by PCR with hyd_NheI_F and hyd_HindIII_R (see Table S1 in the supplemental material) as primers and Microbacterium sp. HM58-2 total DNA as the template. The product was digested with NheI and HindIII and ligated into NheI-HindIII-digested pET28a(+) (Novagen) to form pET-hyd. Escherichia coli BL21(DE3) was transformed with pET-hyd, propagated in LB broth containing kanamycin at 37°C for 8 h, and inoculated (at 2.5%) into 200 ml of LB broth containing kanamycin at 40 μg/ml. The cells were cultured at 30°C for 3 h, 0.2 mM isopropyl-β-d-thiogalactoside was added, and the cultivation was continued for 6 h. Cells were harvested by centrifugation at 10,000 × g for 20 min and stored at −80°C. Buffer A (10 ml) was added, and the cells were thawed on ice and lysed by sonication. Centrifugation at 30,000 × g for 40 min removed insoluble material, and a recombinant 6×His-tagged hydrazidase (rHyd) in the cleared cell extract was applied to a column containing chelating Sepharose FF (GE Healthcare, Waukesha, WI) preloaded with one column volume of 0.2 M NiCl₂. The column was washed with five column volumes of buffer A, and then rHyd was eluted in buffer A containing 200 mM imidazole.

Preparation of hydrazidase mutants.

The plasmid pET-hyd was mutagenized for single amino acid substitutions of K80A, S155A, and S179A of hydrazidase using a QuikChange site-directed mutagenesis kit (Agilent Technologies, Palo Alto, CA) according to the manufacturer's instructions and the primers listed in Table S1 in the supplemental material. Hydrazidase mutants were expressed and purified as described for wild-type rHyd.

Quantitative PCR.

Microbacterium sp. HM58-2 was cultured in MM containing glucose and succinate (100 mM each) and/or 10 mM HBPH as the sole carbon source at 30°C. When the OD₆₀₀ reached 0.6, 1 ml of culture was mixed with 0.5 ml of RNAprotect Bacteria Reagent (Qiagen, Valencia, CA) and incubated for 5 min at 25°C. After centrifugation at 10,000 × g for 10 min, the pellet was stored at −80°C. Cellular RNA was extracted using an RNeasy minikit (Qiagen) according to the manufacturer's instructions. First-strand cDNA was synthesized using a QuantiTect reverse transcription kit (Qiagen) and 250 ng of cellular RNA. Quantitative PCR proceeded using iQ SYBR green Supermix (Bio-Rad) and MiniOpticon version 3.1 (Bio-Rad) according to the manufacturer's instructions with hyd58-2QF and hyd58-2QR primers (see Table S1 in the supplemental material). Expression of the hydrazidase gene was normalized against that of the 16S rRNA gene amplified with 16S58-2QF and 16S58-2QR primers (see Table S1). Data are shown as relative expression levels.

Gas chromatography.

Hydrazide compounds were analyzed by gas chromatography-mass spectrometry (GC-MS) using a GCMS-QP2010 Plus instrument (Shimadzu) equipped with a DB-5 capillary column (0.25-µm film thickness; 0.32-mm inside diameter [I.D.] by 30 m) (J & W Scientific Inc., Folsom, CA) and the NIST 05 MS Library under the following conditions. The column oven temperature was held at 60°C for 2 min; the temperature was increased in increments of 8°C/min to 180°C and held for 5 min; and the temperature was increased at 40°C/min to 220°C and held for 5 min. The injection, interface, and ion source temperatures were 200°C, 230°C, and 250°C, respectively. Helium (30 ml/min) was the carrier gas.

Nucleotide sequence accession numbers.

The nucleotide sequence of the 16S rRNA gene of the HM58-2 strain has been deposited in the DDBJ/EMBL/GenBank databases under accession number LC005742. The nucleotide sequence of the hydrazidase gene has been deposited in the DDBJ/EMBL/GenBank databases under accession number LC005743.

RESULTS

Isolation of hydrazide-degrading Microbacterium sp. HM58-2.

We synthesized three potential hydrazide compounds for microbial degradation: 4-hydroxybenzoic acid 1-phenylethylidene hydrazide (HBPH) (Fig. 1A), adipic acid bis-(1-phenylethylidene hydrazide) (APEH), and adipic acid bis-(pentane-2-ylidene hydrazide) (APEYH) (see Supplemental Materials). The compounds were designed to include one or two acylhydrazide and hydrazone moieties since our initial screen aimed at isolating microorganisms that could degrade either of them. Since all of them except HBPH were either unstable or difficult to dissolve in water, we further investigated HBPH. We screened 119 soil samples using enrichment culture in medium containing HBPH as the sole carbon source (MM-HBPH) and isolated 16 microorganisms that proliferated under these conditions. Microbial cell extracts prepared from 14 of the 16 strains decreased the amount of HBPH when incubated in MM-HBPH. Among them, we selected strain HM58-2, the cell extract of which exhibited the highest HBPH-degrading activity. Strain HM58-2 proliferated in MM-HBPH medium with a concomitant decrease of HBPH and the generation of a stoichiometric amount of acetophenone hydrazone, which is a product of microbial HBPH degradation (Fig. 2). The amount of HBA produced was below the detection limit of HPLC (1 μM) after cultivation. These findings are consistent with the notion that HM58-2 hydrolyzes the hydrazide group of HBPH to generate acetophenone hydrazone and HBA (Fig. 1B), the latter of which the strain immediately consumes as a carbon source for growth.

FIG 1.

Microbial targets of acylhydrazide degradation. (A) Structures of chemicals used in this study. (B) Proposed hydrolysis reaction of HBPH performed by Microbacterium sp. HM58-2. HBA, 4-hydroxybenzoic acid; HBH, 4-hydroxybenzoic acid hydrazide; HBPH, 4-hydroxybenzoic acid 1-phenylethylidene hydrazide.

FIG 2.

HBPH-dependent growth of Microbacterium sp. HM58-2. Microbacterium sp. HM58-2 was cultured in MM-HBPH medium at 30°C. Concentrations of HBPH (filled circles) and acetophenone hydrazone (open circles) were monitored by HPLC. The concentration of HBA remained <1 µM throughout cultivation. Cell growth was monitored using the optical density at 600 nm (OD₆₀₀) of culture broth (triangles). Data are typical of the results of triplicate experiments.

We analyzed the phylogenetic, morphological, and physiological properties of the HM58-2 strain. The 16S rRNA gene sequence of HM58-2 was highly similar to those of known Microbacterium strains, including M. hydrocarbonoxydans DSM16089 (99.5%) (25), M. paraoxydans CF36 (99.1%), and M. foliorum DSM12966 (98.6%). A phylogenetic analysis conducted with the 16S rRNA gene sequences located HM58-2 in the same cluster as M. hydrocarbonoxydans DSM16089 (see Fig. S2 in the supplemental material). The HM58-2 cells grown on LB agar plates at 30°C for 48 h were Gram-positive nonmotile rods that formed yellow colonies. The strain did not oxidize or ferment glucose in the O/F test and was catalase positive and negative in oxidase tests. These properties are consistent with general properties of Microbacterium species (26). Table 1 shows the physiological properties determined using API Coryne kits (bioMérieux, Marcy l'Etoile, France) and other standard methods. Most of these properties are similar to those of M. hydrocarbonoxydans except for the negative motility, acid production from d-xylose, l-rhamnose, and sucrose, and utilization of α-d-melibiose and l-arabinose as the carbon source. Based on these findings, we propose that the isolate represents a species related to M. hydrocarbonoxydans and designated it Microbacterium sp. HM58-2.

TABLE 1.

Summary and comparison of characteristics of Microbacterium sp. HM58-2 and its phylogenetically close relatives^a

Characteristic	Result for taxon:
	1	2	3	4
Color of colonies	Y	Y	O	GW/Y
Shape of cells	Rod	Rod, irregular	Rod, irregular	Rod, V-form
Gram staining	+	+	+	+
Motility	−	+	−	−
Glucose oxidation/fermentation test	−/−	ND	ND	ND
Catalase	+	+	+	+
Oxidase	−	−	−	ND
Nitrate reduction	−	ND	ND	+
Pyrazinamidase	−	ND	ND	ND
β-Glucuronidase	−	−	−	ND
Urease	−	−	−	−
β-Galactosidase	+	+	+	ND
α-Glucosidase	+	+	+	ND
Esculin hydrolysis	+	ND	+	ND
Acid production from:
d-Glucose	−	−	−	+
d-Ribose	−	ND	ND	−
d-Xylose	−	+	+	−
l-Rhamnose	−	+	−	−
Inositol	−	−	−	−
d-Mannitol	−	ND	ND	+
Maltose	−	ND	ND	+
Lactose	−	ND	ND	+
Sucrose	−	+	+	−
Glycogen	−	ND	−	ND
Growth under anaerobic conditions	−	−	−	−
Utilization as a carbon source:
α-d-Melibiose	+	−	+	ND
Acetate	+	+	−	+
Citrate	+	+	−	+
l-Arabinose	−	+	+	ND

^aCells grown on LB agar plates were subjected to each of the tests. Taxa: 1, Microbacterium sp. HM58-2; 2, M. hydrocarbonoxydans DSM16089^T (25); 3, M. oleivorans DSM16091^T (25); 4, M. lacticum DSM20427^T (26). Symbols and abbreviations: Y, yellow; O, orange; GW, gray-white; +, positive; −, negative; ND, no data available.

Gene cloning of the hydrazidase responsible for HBPH degradation.

Cell extracts of strain HM58-2 grown in MM-HBPH medium for 48 h hydrolyzed HBPH at a rate of 0.61 μmol min⁻¹ mg⁻¹. The bacterium also proliferated in MM containing 10 mM HBA, and the cell extract produced up to 0.23 μmol min⁻¹ mg⁻¹ HBPH-hydrolyzing activity after 24 h of cultivation. To purify the bacterial HBPH-degrading enzyme, we cultured the bacterium in MM containing commercially available HBA, because the amount of available HBPH was insufficient for large-scale culture. Cell extracts prepared from the culture served as starting material, and then we purified the enzyme 31-fold with a recovery rate of 15% after four steps of column chromatography (Table 2). The HBPH-degrading activity of the purified preparation was 6.2 μmol min⁻¹ mg⁻¹, and the reaction generated acetophenone hydrazone and HBA, indicating that the enzyme hydrolyzes HBPH to acetophenone hydrazone and HBA. The purified enzyme migrated to a position corresponding to 50 kDa in SDS-PAGE (Fig. 3A, lane 2).

TABLE 2.

Summary of hydrazidase purification from Microbacterium sp. HM58-2

Purification step	Total protein (mg)	Total activity (μmol min−1)	Sp act (μmol min−1 mg−1)	Yield (%)	Fold
Crude extract	899	177	0.20	100	1.0
DEAE-cellulose	188	164	0.87	92	4.4
Butyl-Sepharose I	33	76	2.3	43	12
Butyl-Sepharose II	17	65	3.7	37	19
Mono Q	4	27	6.2	15	31

FIG 3.

Purification of Microbacterium sp. HM58-2 hydrazidase and its mutants. (A) Purified enzymes (1 μg each) resolved by SDS-PAGE. Lanes: 1, molecular mass markers; 2, hydrazidase purified from Microbacterium sp. HM58-2; 3, rHyd; 4, rHyd^K80A; 5, rHyd^S155A; 6, rHyd^S179A. (B) Alignment of amino acid sequences of hydrazidase and other related proteins from Pseudonocardia acaciae (WP_028921404), Pusillimonas noertemannii (WP_017522754), Bradyrhizobium japonicum (P19922/HYIN_BRAJP), Pseudomonas putida (O69768/AMID_PSEPU), and T. thermophilus (Q9LCX3/GATA_THET8). Numbers represent HydA sequences. *, conserved residues. Catalytic Ser-Ser-Lys residues are boxed.

Trypsin digestion of the purified hydrazidase yielded two major fragments with molecular masses of 36 and 33 kDa, and endoproteinase Asp-N digestion produced a 25-kDa fragment. We then sequenced the amino-terminal amino acids of these fragments (see Fig. S3 in the supplemental material). A BLAST search of databases revealed that these sequences were similar to parts of the predicted amino acid sequences of putative amidases from Burkholderia cenocepacia and B. cepacia. A comparison of these sequences predicted that the amino acid sequences obtained from the 36-, 33-, and 25-kDa fragments would appear in that order when aligned from the amino terminus. This prediction enabled the design of the degenerate oligonucleotide primers Try_F1 and Asp-N_R1 and PCR amplification of a specific 290-bp DNA fragment that hybridized with a 3-kb fragment of PstI-digested genomic DNA of the bacterium. The entire hydrazidase gene was cloned by inverse PCR using a size-fractionated PstI digest of the bacterial genomic DNA and primers Inv1 and Inv2 (see Table S1). The cloned gene comprised 1,380 bp and encoded a polypeptide comprising 460 amino acid residues, which included the amino acid sequences of the tryptic and Asp-N digests. The calculated molecular mass of the gene product was 49.3 kDa, which agreed with that of the purified hydrazidase. We designated the cloned gene encoding the bacterial hydrazidase hydA on the basis of these findings.

HM58-2 hydrazidase constitutes a novel family of proteins.

A homology search of published databases revealed that uncharacterized proteins from Pseudonocardia acaciae (WP_028921404, 59%) and Pusillimonas noertemannii (WP_017522754, 48%) were the only two proteins with >40% amino acid sequence identity. Proteins with known function had <40% identity to the hydrazidase, and IAAH of Bradyrhizobium japonicum (accession number P19922/HYIN_BRAJP) (27), the amidase of Pseudomonas putida (O69768/AMID_PSEPU) (28), and glutamyl-tRNA amidotransferase subunit A (GatA) of Thermus thermophilus (Q9LCX3/GATA_THET8) (29) were the closest in identity at 30% to 31%. All of these proteins belong to the amidase signature family, which is characterized by the consensus sequence G-[GAV]-S-[GS]₂-G-X-[GSAE]-[GSAVYCT]-X-[LIVMT]-[GSA]-X₆-[GSAT]-X-[GA]-X-[DE]-X-[GA]-X-S-[LIVM]-R-X-P-[GSACTL] (PROSITE documentation no. PDOC00494). The deduced amino acid sequence of the hydrazidase contains this consensus sequence. Enzymes in the amidase signature family possess a catalytic Ser-Ser-Lys triad (30), and these residues are conserved in the bacterial hydrazidase at Ser155, Ser179, and Lys80 (Fig. 3B), indicating that the Microbacterium sp. HM58-2 hydrazidase belongs to the amidase signature family. The Cys-X₃-Cys motif, which appears in some bacterial amidases in this family (31), was not found in the Microbacterium sp. HM58-2 hydrazidase.

We collected 631 amino acid sequences containing the amidase signature sequences (PS00571) from the manually curated PROSITE protein database and analyzed the phylogenetic relationships of Microbacterium sp. HM58-2 hydrazidase. The proteins in the family were grouped under IAAH, GatA, fatty acid amide hydrolase (FAAH), bacterial and archaeal amidase, and fungal acetamidase (AmdS) (see Fig. S4 in the supplemental material). The largest group was GatA, which contained 581 sequences. Hydrazidase was not in any of these groups but resided on the branch related to FAAH and AmdS, indicating a closer relationship with proteins in these groups. We phylogenetically analyzed Microbacterium sp. HM58-2 hydrazidase using the 631 sequences mentioned above and an additional 100 sequences from the databases with high similarity to the hydrazidase sequence (Fig. 4). The resulting phylogenetic tree comprised the above-mentioned five and two new groups. One of the new groups mostly contained hypothetical amidases from actinobacteria. The other contained the hydrazidase, Bradyrhizobium IAAH, and predicted amidases of the phyla Proteobacteria, Actinobacteria, and Firmicutes, suggesting an evolutionally close relationship. These hypothetical amidases were 32% to 59% identical to the hydrazidase. These results indicated that the hydrazidase constitutes a novel group within the amidase signature family.

FIG 4.

Phylogenetic analysis of hydrazidase and related proteins. The data set comprised 631 amino acid sequences obtained from PROSITE pattern signature PS00571 and an additional 100 sequences in the NCBI protein database that showed highest similarity to the hydrazidase. The tree was constructed using the neighbor-joining method with the CLUSTAL W and MEGA4 programs. Bootstrap values were calculated from 1,000 repeats, and those greater than or equal to 50% are shown. NCBI or UniProtKB accession numbers and numbers of sequences compressed in a branch are indicated in parentheses. The bar represents 0.1 substitutions per amino acid position.

Catalytic properties of hydrazidase.

We produced a recombinant hydrazidase fused with a 6×His tag at the amino terminus (rHyd) in E. coli cells and purified it by chelating Sepharose chromatography (Fig. 3A, lane 3). We based the biochemical characterization of the enzyme on rHyd because it was as active (5.6 μmol min⁻¹ mg⁻¹) as the native enzyme. The recombinant hydrazidase was active over broad ranges of pH (5.0 to 10.0) and temperature (15°C to 40°C) (see Fig. S5 in the supplemental material), and the activity was maximal at pH 8 and at 30°C under our experimental conditions. We assessed the inhibitory effects of various compounds on the enzyme activity. Hg²⁺, Zn²⁺, Cu²⁺, and Ni²⁺ at 1 mM were inhibitory, whereas Mn²⁺, Mg²⁺, Fe²⁺, Co²⁺, MoO₄²⁺ and Ca²⁺ affected the rRNA gene activity slightly (see Fig. S6). Reducing agents (2-mercaptoethanol and dithiothreitol), phenylmethylsulfonyl fluoride, chelating agent (EDTA), and nonionic surfactants (Triton X-100, Tween 20, and Tween 80) affected the activity to a lesser extent.

We confirmed the stoichiometry of the hydrazidase reaction. When typical reaction mixtures contained 0.5 mM HBPH and the recombinant hydrazidase was incubated at 30°C for 30 min, 41 ± 9 nmol HBPH was consumed and 38 ± 5 nmol acetophenone hydrazone and 36 ± 11 nmol HBA were generated. This is consistent with the reaction scheme of the hydrazidase (Fig. 1B). We then assayed enzyme activity in a reaction mixture containing HBA and acetophenone hydrazone (0.5 mM each). During incubation at 30°C for 30 min, no HBPH was generated, indicating that the hydrazidase reaction to HBPH was irreversible under these conditions.

Steady-state kinetics.

Steady-state kinetics indicated that the apparent Michaelis constant (K_m) and the kinetic constant (k_cat) of the native hydrazidase for HBPH were 8.1 mM and 77 s⁻¹, respectively (Table 3). These values were comparable to those of other bacterial amidases that hydrolyze amides, which mostly range between 10⁻¹ and 10¹ mM and between 10⁰ and 10² s⁻¹, respectively (17, 32–36), indicating that the hydrazidase hydrolyzes hydrazides as efficiently as the amidases hydrolyze amides. We prepared rHyd proteins that harbored mutations at the conserved Ser-Ser-Lys motif (Fig. 3B). The amino acid residues were replaced with Ala in rHyd^K80A, rHyd^S155A, and rHyd^S179A, and their specific activities against HBPH were 2.1, 0.93, and 0.43 μmol min⁻¹ mg⁻¹, meaning that 38%, 17%, and 7.8% of the wild-type rHyd activity was retained. The mutations increased the K_m values 2-fold to 4-fold and decreased the k_cat values 3-fold to 15-fold (Table 3), indicating that these amino acids are involved in the hydrolysis of HBPH although they are not absolutely essential for hydrazidase activity.

TABLE 3.

Kinetic constants of native and mutant hydrazidases^a

Enzyme	HBPH			HBH			4-Hydroxybenzamide
	Km (mM)	kcat (s−1)	kcat/Km (mM−1 s−1)	Km (mM)	kcat (s−1)	kcat/Km (mM−1 s−1)	Km (mM)	kcat (s−1)	kcat/Km (mM−1 s−1)
rHyd	8.1 ± 1.8	77 ± 18	9.5	8.8 ± 2.5	45 ± 11	5.1	17 ± 4	6.9 ± 0.9	0.41
rHydK80A	19 ± 3	26 ± 7	1.4	25 ± 6	13 ± 3	0.52	37 ± 9	2.1 ± 0.8	0.056
rHydS155A	28 ± 4	12 ± 3	0.42	35 ± 7	5.4 ± 1.9	0.15	ND	ND	ND
rHydS179A	34 ± 5	5.4 ± 2.2	0.16	39 ± 9	2.4 ± 1.3	0.061	ND	ND	ND

^aAbbreviations: HBH, 4-hydroxybenzoic acid hydrazide; HBPH, 4-hydroxybenzoic acid 1-phenylethylidene hydrazide; ND, not detectable. Data represent means ± standard deviations of the results of three experiments.

We assessed the substrate preference of the hydrazidase for the HBH and INH hydrazides (Fig. 1A). We also tested acetamide, indoleacetamide, and fatty acid amides that are substrates of amidases in the amidase signature family (37–39) and 4-hydroxybenzamide, which is an amide that is structurally similar to HBPH (Fig. 1A). Among these compounds, the hydrazidase could utilize HBH and 4-hydroxybenzamide as substrates and produced a stoichiometric amount of HBA (data not shown). The specific activities for HBH and 4-hydroxybenzamide were 3.4 and 0.47 μmol min⁻¹ mg⁻¹, respectively. None of the other tested compounds was a substrate for the hydrazidase, and the amount of activity was less than 0.01 μmol min⁻¹ mg⁻¹. HBPH, HBH, and 4-hydroxybenzamide share a 4-hydroxybenzoic acid moiety that is important for hydrazidase substrates since the apparent K_m values of rHyd for HBPH and HBH were similar (Table 3).

The specific activity for 4-hydroxybenzamide is 14% of that for HBH, indicating that the hydrazidase is also active as an amidase, the activity of which was lost due to the mutation at Ser155 and Ser179, suggesting that the catalytic mechanism is shared by amidase and hydrazidase reactions. The apparent K_m of rHyd for 4-hydroxybenzamide was 1.9-fold as high as that for HBH, in accordance with their structural similarity. The k_cat value for 4-hydroxybenzamide was much (15%) lower than that for HBH. This indicates that the hydrazidase preferentially hydrolyzes the hydrazide rather than the amide.

Regulation of hydrazidase production.

We cultured Microbacterium sp. HM58-2 with different carbon sources and measured HBPH-hydrolyzing activity in cell extracts (Fig. 5A). The activity was maximal when the cells were incubated in MM-HBPH medium compared with glucose or succinate as the carbon source (0.20 versus <0.001 μmol min⁻¹ mg⁻¹). Adding 100 mM glucose to MM-HBPH medium decreased the hydrazidase activity to 17% of that without glucose. We quantified transcripts of the hydrazidase and analyzed the carbon source-dependent expression of the hydrazidase gene (Fig. 5B). The mRNA level was highest when the bacteria were cultured in MM-HBPH, and the level decreased by 88% when 100 mM glucose was included in the medium. Moreover, mRNA was undetectable when glucose or succinate served as the carbon source. These results were consistent with the cellular enzyme activity described above and indicate that expression of the hydrazidase gene is induced by HBPH, repressed by glucose, and regulated at the transcriptional level depending on the carbon sources available to the bacteria.

FIG 5.

Regulation of hydrazidase production by Microbacterium sp. HM58-2. (A) Ability of cell extracts to degrade HBPH. Extracts were prepared from cells cultured in MM containing 10 mM HBPH, 100 mM glucose, or 100 mM sodium succinate as the carbon source and were harvested at the logarithmic-growth phase. SA, succinic acid. (B) Hydrazidase gene transcript quantified by PCR. Cells were cultured as described for panel A, and total RNA was prepared from cells harvested when the OD₆₀₀ reached 0.6. Data are expressed as relative expression levels after normalizing against 16S rRNA transcripts. Data are means of the results of three experiments. Bars indicate standard errors.

DISCUSSION

The present study identified Microbacterium sp. HM58-2 as a strain that hydrolyzes the HBPH acylhydrazide and a novel hydrazidase that hydrolyzes hydrazides:

$${\mathrm{R}}_1\mathrm{C}(=\mathrm{O})-\mathrm{N}\mathrm{H}\mathrm{N}{\mathrm{R}}_2{\mathrm{R}}_3+{\mathrm{H}}_2\mathrm{O}\hspace{0.17em}\to {\mathrm{R}}_1\mathrm{C}(=O)\mathrm{O}\mathrm{H}+{\mathrm{R}}_2{\mathrm{R}}_3\mathrm{N}\mathrm{N}{\mathrm{H}}_2$$ (1)

where R₁ represents HO-(C₆H₄)- and R₂ and R₃ represent =C (C₆H₅) (CH₃) and -H, -H, respectively.

Hydrazidases have not received much investigative focus since the first hydrazidase was purified in 1963 (19). However, the absence of sequence information prevented determining their evolutional or structural relationships to the Microbacterium hydrazidase here or even whether or not it belongs to the amidase signature family. However, we speculate that the novel enzyme is different because a protein that resides in its hydrazidase group was not found among the published genome sequences of Mycobacterium strains. The present study showed the first genetic basis for the physiological mechanism of hydrazide degradation. We previously discovered and characterized a dehydrogenase from soil microorganisms that degrades hydrazone compounds (20, 21):

$${\mathrm{R}}_1\mathrm{C}\mathrm{H}=\mathrm{N}\mathrm{N}\mathrm{H}\mathrm{C}(=\mathrm{O}){\mathrm{R}}_2+\mathrm{N}\mathrm{A}\mathrm{D}{(\mathrm{P})}^{+}+2{\mathrm{H}}_2\mathrm{O}\to {\mathrm{R}}_1\mathrm{C}(=\mathrm{O})\mathrm{O}\mathrm{H}+{\mathrm{R}}_2\mathrm{C}(=O)\mathrm{N}\mathrm{H}\mathrm{N}{\mathrm{H}}_2+\mathrm{N}\mathrm{A}\mathrm{D}(\mathrm{P})\mathrm{H}+{\mathrm{H}}^{+}$$ (2)

where R₁ represents CH₃- and R₂ represents CH₃(CH₂)₃-, CH₃CH=NNHC(=O)-(CH₂)₄-, or NH₂NHC(=O)-(CH₂)₄-. The present results expanded these findings and deepened understanding of the biological degradation of hydrazide compounds.

The deduced amino acid sequence of the Microbacterium sp. HM58-2 hydrazidase indicated that it has low similarity to proteins of known function. Low identity (∼59%) was found among hypothetical proteins from proteobacteria, firmicutes, and actinobacteria. A bioinformatics study showed that these hypothetical proteins are unambiguously part of the amidase signature family (Fig. 4), whereas the present study showed that the enzymatic properties of this family of proteins are those of a hydrazidase. The highest sequence identity of the hydrazidase to an amidase signature protein was 31%, which was the same as that to B. japonicum IAAH (27), an amidase that shares the Ser-Ser-Lys triad and is active against indoleacetamide and acetamide. This amidase and the hydrazidase are in separate clades in the phylogenetic tree (see Fig. S4 in the supplemental material). These indicate that the hydrazidase is unique in terms of its enzymatic and evolutional properties, and we propose that the novel enzyme is a hydrazidase of the amidase signature family.

Both hydrazidase (scheme 1) and amidases in general (scheme 3) hydrolyze the C-N bond of acylated nitrogen:

$${\mathrm{R}}_1\mathrm{C}(=\mathrm{O})\text{-}\mathrm{N}\mathrm{H}{\mathrm{R}}_2+{\mathrm{H}}_2\mathrm{O}\hspace{0.17em}\to {\mathrm{R}}_1\mathrm{C}(=\mathrm{O})\mathrm{O}\mathrm{H}+{\mathrm{R}}_2\mathrm{N}{\mathrm{H}}_2$$ (3)

Our kinetic studies using structurally similar HBH and 4-hydroxybenzamide (Fig. 1A) showed that the hydrazidase prefers hydrazides to amides as substrates (Table 3). Few kinetic studies have investigated the hydrolysis of hydrazides. Pseudomonas aeruginosa aliphatic amidase hydrolyzes acetohydrazide with a k_cat of 10 s⁻¹, which is much lower than that for acetamide (k_cat = 162 s⁻¹) (17). The respective K_m values were 81 and 0.83 mM, indicating that this amidase prefers amide to hydrazide. This amidase is characterized by a catalytic Glu-Lys-Cys, and it belongs to the nitrilase family, which is not related to the Ser-Ser-Lys-type amidases (40). The specific activity of amidase signature amidases from Rhodococcus sp. J1 (10) and Arabidopsis thaliana IAAH (41) was ≤0.4% for hydrazides in comparison to that of the corresponding amides. Thus, the substrate preference for hydrazides is a unique feature of the Microbacterium sp. HM58-2 hydrazidase.

Substitutions of the conserved Ser-Ser-Lys residues decreased enzymatic activity, but a substantial amount was conserved (Table 3). This is in contrast to corresponding mutations of A. thaliana amidase 1 and other amidases containing the Ser-Ser-Lys triad, which result in almost total inactivation (30, 39, 42, 43). That rHyd^S179A retained 7.8% activity is notable since an amino acid sequence comparison has predicted that Ser179 is the nucleophile that attacks the carbonyl carbon of hydrazides and that corresponding residues (namely, Ser137 in A. thaliana amidase 1 and Ser241 in mammalian FAAH) are essential for producing the activity (39, 42). Such predictions imply that the catalytic roles of the Ser-Ser-Lys residues of the hydrazidase and of these amidases are the same, although their contributions to the hydrolytic mechanism differ. Several bacterial Ser-Ser-Lys signature amidases contain an additional Cys-X₃-Cys motif, and structure modeling has suggested that Sulfolobus solfataricus amidase uses the second Cys as an alternative catalytic nucleophile (31). Although the amino acid sequence of the Microbacterium sp. HM58-2 hydrazidase does not have this motif, Cys129 corresponds to its second Cys residue and it might function as a nucleophile in the absence of Ser179, a notion that is currently being investigated by X-ray crystallography.

Our screening using the HBPH as a carbon source isolated bacteria that hydrolyzed the hydrazide moiety of HBPH and none that cleaved the other functional C=N hydrazone moiety of HBPH (Fig. 1). This is in contrast to previous screens that had routinely isolated bacteria that degrade hydrazones (20). These studies targeted adipic acid bis-(ethylidene hydrazide) (AEH) for degradation, and its hydrazone moiety is described as R₁CH=NNC(=O)R₂. The hydrazone moiety of HBPH described as R₁R₂C=NNC(=O)R₃ substitutes the hydrogen atom of the AEH hydrazone with a carbon atom. The hydrazone dehydrogenase initiates AEH degradation by extracting the hydrogen atom from the hydrazone carbon (20, 21) and is thus unable to degrade HBPH. This implies that biological systems distinguish hydrazones by their chemical properties. Future studies of environmental bacteria should investigate the microbial degradation of carbon-substituted, HBPH-type hydrazones.

That Microbacterium sp. HM58-2 uses HBA as a carbon source impacts the role of this organism in the global carbon cycle since HBA is one of the major phenolic compounds that are released during the degradation of plant-derived lignin. In general, lignin is derivatized to low-molecular-weight aromatic compounds before metabolism by meta cleavage or the β-ketoadipate pathways of soil bacteria (44). The protocatechuate branch of the β-ketoadipate pathway converts protocatechuate, which is generated from lignin-derived vanillate, ferulate, and HBA, to β-ketoadipate, which is then cleaved into the tricarboxylic acid (TCA) cycle intermediates succinyl coenzyme A (succinyl-CoA) and acetyl-CoA. Several actinomycetes strains such as Streptomyces setonii, Rhodococcus opacus, and Corynebacterium glutamicum have this mechanism and catabolize HBA as well as other lignin derivatives (45–49). Microbacterium lacus isolated from deep-sea sediments can metabolize HBA (50), although, to the best of our knowledge, a Microbacterium mechanism for metabolizing HBA has not been reported. Database searches revealed that some Microbacterium strains possess a complete set of genes for the β-ketoadipate pathway (data not shown), indicating that Microbacterium sp. HM58-2 also utilizes this pathway to catabolize lignin-derived HBA in soil. This could afford the bacterium a competitive advantage over other soil microbes.

Recent genome sequencing has predicted numerous amidases produced by both prokaryotes and eukaryotes, which indicates divergent roles of amidases for catalyzing various environmental amides. Current understanding of their functions is limited (51) because speculating about them from sequence information is difficult. The present report is the first to provide information about the amino acid sequence of hydrazidase and will contribute to future studies on the distribution of hydrazidases and their sources in nature. Further studies on structure-function relationships are required to determine a novel amino acid sequence signature that discriminates hydrazidases and amidases.