Two Rhizobial Strains, Mesorhizobium loti MAFF303099 and Bradyrhizobium japonicum USDA110, Encode Haloalkane Dehalogenases with Novel Structures and Substrate Specificities

Abstract

Haloalkane dehalogenases are key enzymes for the degradation of halogenated aliphatic pollutants. Two rhizobial strains, Mesorhizobium loti MAFF303099 and Bradyrhizobium japonicum USDA110, have open reading frames (ORFs), mlr5434 and blr1087, respectively, that encode putative haloalkane dehalogenase homologues. The crude extracts of Escherichia coli strains expressing mlr5434 and blr1087 showed the ability to dehalogenate 18 halogenated compounds, indicating that these ORFs indeed encode haloalkane dehalogenases. Therefore, these ORFs were referred to as dmlA (dehalogenase from Mesorhizobium loti) and dbjA (dehalogenase from Bradyrhizobium japonicum), respectively. The principal component analysis of the substrate specificities of various haloalkane dehalogenases clearly showed that DbjA and DmlA constitute a novel substrate specificity class with extraordinarily high activity towards β-methylated compounds. Comparison of the circular dichroism spectra of DbjA and other dehalogenases strongly suggested that DbjA contains more α-helices than the other dehalogenases. The dehalogenase activity of resting cells and Northern blot analyses both revealed that the dmlA and dbjA genes were expressed under normal culture conditions in MAFF303099 and USDA110 strain cells, respectively.

Haloalkane dehalogenases are key enzymes for the degradation of halogenated aliphatic compounds that occur as soil pollutants (1, 17, 38). Haloalkane dehalogenases catalyze the hydrolytic cleavage of the carbon-halogen bond(s) and produce the corresponding alcohols, halide ions, and protons. These enzymes belong to the α/β-hydrolase superfamily (39).

Haloalkane dehalogenases are attractive targets for protein-engineering studies aimed at improving catalytic efficiency and at broadening the range of substrate specificity for important environmental pollutants. To date, the three-dimensional structures of three haloalkane dehalogenases have been determined by protein crystallography: DhlA from Xanthobacter autotrophicus GJ10 (48), DhaA from Rhodococcus sp. (25, 35), and LinB from Sphingomonas paucimobilis UT26 (29). The differences in the substrate specificities of these three haloalkane dehalogenases can be accounted for on the basis of their three-dimensional structures (11). Comparison of the kinetic mechanisms of DhlA, DhaA, and LinB showed that the overall reaction mechanisms are similar but that the rate-limiting steps differ, i.e., halide release in the case of DhlA (44), liberation of an alcohol in the case of DhaA (4), and hydrolysis of an alkyl-enzyme intermediate in the case of LinB (41). Partial improvement in the catalytic properties and modification of the substrate specificities of haloalkane dehalogenases by rational design (5, 34) and directed evolution approaches (3, 40) have recently been reported. However, it remains difficult to construct mutant enzymes with entirely new capabilities using only protein-engineering techniques, and therefore, the isolation of new family members is still desirable.

For quite some time, haloalkane dehalogenases have been thought to be present only in soil bacteria that colonize contaminated environments (14). It was recently demonstrated that Mycobacterium tuberculosis H37Rv, the complete genome of which has been sequenced (7), possesses chromosomal genes encoding putative haloalkane dehalogenases (11). Hydrolytic dehalogenation was detected in a number of different mycobacteria (18), and haloalkane dehalogenase DhmA from M. avium N85 has been partially characterized (19). Furthermore, we could find dehalogenase-like open reading frames (ORFs) on the genomes of more than 20 bacterial species. However, experimental confirmation of the dehalogenating activity of these ORF products has not yet been reported.

Rhizobiaceace, the collective name of the genera Rhizobium, Sinorhizobium, Mesorhizobium, and Bradyrhizobium, are soil and rhizosphere bacteria of agronomic importance because they form nitrogen-fixing symbiotic relationships with leguminous plants. In order to gain a more comprehensive understanding of the genetic systems required for the entire process of symbiotic nitrogen fixation, the complete sequences of the genomes of two plant-symbiotic bacteria, M. loti MAFF303099 (21, 22) and Bradyrhizobium japonicum USDA110 (23, 24), were determined. Although these two strains have not been reported as halogenated compound degraders, they have dehalogenase-like ORFs on their genomes. In this study, we demonstrated that these two strains produced functional haloalkane dehalogenases with novel structures and substrate specificities.

MATERIALS AND METHODS

Bacterial strains and culture conditions.

M. loti MAFF303099 and B. japonicum USDA110 were grown on tryptone-yeast (TY) extract (2), N-2-hydroxyethylpiperazine-N-2-ethanesulfonic acid and 2-(N-morpholino)ethanesulfonic acid (HM) (6), and yeast extract-mannitol (YM) (50) medium plates at 30°C. Escherichia coli strains were grown in Luria broth (LB) (28) at 37°C. Antibiotics were applied at the following final concentrations: ampicillin, 50 μg/ml; carbenicillin, 100 μg/ml; and chloramphenicol, 20 μg/ml.

DNA methodology.

Established methods were employed for the following procedures: preparation of plasmid DNA, digestion of plasmids and PCR-amplified DNA fragments with restriction endonucleases, ligation, agarose gel electrophoresis, and transformation of E. coli cells (28). The genomic DNAs of M. loti MAFF303099 and B. japonicum USDA110 were isolated as described previously (31). The nucleotide sequences were determined by the dideoxy chain termination method with an automated DNA sequencer (ABI PRISM 310 genetic analyzer; Applied Biosystems, Foster City, CA).

Primers and conditions for PCR.

Oligonucleotide primers were designed according to the nucleotide sequences of mlr5354 (GenBank accession no. AB003006; gene ID, 1228693), blr1087 (accession no. AP005939; gene ID, 1051963), and their flanking regions, and these primers are listed in Table 1. The recognition sites for suitable restriction enzymes were added to the forward and reverse primers for cloning. For the overexpression of the protein products in E. coli, a canonical Shine-Dalgarno sequence (TAAGGAGG) (27) was added to the forward primers. For the cloning of mlr5354 into pET-32a(+) (26), the primer 5354BN was used instead of 5354BSD. To add a six-histidyl tail into the C terminus of the protein product of blr1087, we used the primer 1087HHIS instead of the 1087H primer. PCR was carried out using an Expand High-Fidelity PCR system (Roche Diagnostics, Basel, Switzerland) for 35 cycles (30 s at 94°C and 30 s at 56°C for mlr5354 and at 53°C for blr1087, and 1 min at 72°C). PCR-amplified DNA fragments were cloned into the pUC18 vector, and their DNA sequences were confirmed; thereafter, DNA fragments were recloned into the vectors for overexpression.

TABLE 1.

Oligonucleotide primers used in this study^a

Primer	Sequence (5′→3′)	Use
5354BSD (forward)	GCCGGATCCTAAGGAGGAGACCCCATGAGTTCGAA	Amplification of mlr5354 (dmlA)
5354H (reverse)	GCCAAGCTTTCAGGCCAACTCGTCCGTC	Amplification of mlr5354 (dmlA)
5354BN (forward)	GCCGGATCCCATGGCGAGTTCGAAAGCCAACCCG	Amplification of mlr5354 (dmlA)
1087BSD (forward)	GCCGGATCCTAAGGAGGAGCTGACATGAGCAAG	Amplification of blr1087 (dbjA)
1087H (reverse)	GCCAAGCTTTCACGCGGCGAGCTGCGG	Amplification of blr1087 (dbjA)
1087HHIS (reverse)	GCCAAGCTTTCAGTGATGGTGATGGTGATGCGCGGCGAGCTGCGGACG	Amplification of blr1087 (dbjA)
dbjA D103A	CGTGCCCCAGGCCTGCGCGAC	Site-directed mutagenesis of dbjA
dbjA E127A	GCGGATGAATGCCATGAAGGC	Site-directed mutagenesis of dbjA
dbjA H280A	CTGCAGATAGGCCAATCCCGC	Site-directed mutagenesis of dbjA
dbjA 103 down	GCGCTCGCATTTCATCTC	Site-directed mutagenesis of dbjA
dbjA 127 down	CCGATGCCGACCTGGCGAG	Site-directed mutagenesis of dbjA
dbjA 280 down	GAGGACCACGCTGACGCA	Site-directed mutagenesis of dbjA

^aRecognition sites for restriction enzymes used for cloning are underlined. Canocical Shine-Dalgarno sequences (27) are expressed in boldface type. Sequence for the additional six-histidyl tail is expressed in italics.

Expression of gene product in E. coli and preparation of crude extract.

To overproduce DmlA, DbjA, and His-tagged DbjA in E. coli, we inserted the corresponding DNA fragments into pAQNM (13), a derivative of pAQN (46), to obtain pYMLA1, pYBJA1, and pYBJA2, respectively. In these plasmids, the inserted genes were transcribed by the tac promoter under the control of lacI^q. E. coli BL21 cells containing one of these plasmids were cultured in LB at 37°C. The induction of protein expression was initiated by the addition of isopropyl-β-d-thiogalactopyranoside (IPTG) to a final concentration of 0.5 mM, when the optical density (600 nm) of the culture reached 0.6. After induction for 3 to 4 h at 30°C, the cells were harvested, disrupted by sonication using a Sonopuls GM70 (Bandelin, Berlin, Germany), and centrifuged at 100,000 × g for 1 h. The supernatant obtained in this manner was used as the crude extract. Because DmlA was not expressed in E. coli using pYMLA1, another plasmid was constructed from pET-32a(+) in order to obtain pYMLA2, in which DmlA was encoded as a fusion protein with thioredoxin (Trx-DmlA). E. coli BL21(DE3) was cotransformed with pYMLA2 and pG-KJE8 (36). Plasmid pG-KJE8 encodes the genes for the DnaK-DnaJ-GrpE and GroEL-GroES chaperones under the control of the araB and Pzt-1 promoters, respectively. E. coli BL21 (pYMLA2)(pG-KJE8) cells were cultivated in LB at 37°C. Inducers for chaperone expression, l-arabinose and tetracycline, were used at final concentrations of 10 mg/ml and 10 ng/ml, respectively. The expression of Trx-DmlA was induced by the addition of IPTG, as described above.

Assay of haloalkane dehalogenase activities.

The haloalkane dehalogenase activity was assayed by the method of Iwasaki et al. (16). The halide ions released were measured spectrophotometrically at 460 nm with mercuric thiocyanate and ferric ammonium sulfate. One unit of enzyme activity was defined as that required for the release of 1 μmol of halide ion per minute. Data are expressed by subtraction of background activities.

Purification of His-tagged DbjA.

The His-tagged DbjA protein was purified using Ni-nitrilotriacetic acid (Ni-NTA) agarose (QIAGEN, Hilden, Germany) as described previously (33). The His-tagged enzyme was bound to the resin in the equilibrating buffer (20 mM potassium phosphate buffer, pH 7.5, containing 0.5 M sodium chloride, 10% glycerol, and 10 mM imidazole). Unbound and weakly bound proteins were washed out with the buffer containing 10 mM imidazole. The target enzyme was eluted by a buffer containing 500 mM imidazole. The active fractions were pooled and dialyzed against a 50 mM potassium phosphate buffer (pH 7.5) containing 10% glycerol. The enzyme was stored in the same buffer. The entire process for purification and storage was performed at 4°C.

Determination of kinetic parameters of His-tagged DbjA.

Michaelis-Menten kinetic constants were determined by initial velocity measurements, as described previously (5). The substrate concentration was assessed by a gas chromatography system equipped with a flame ionization detector (Trace GC 2000; Thermo Finnigan) and a DB-FFAP capillary column (30 m × 0.25 mm × 0.25 μm; J&W Scientific). The method described previously by Iwasaki et al. (16) was used for the determination of the product concentration. The steady-state kinetic constants K_m and k_cat were calculated using the computer program Origin 6.1 (OriginLab).

CD spectroscopy.

Circular dichroism (CD) spectra were recorded at room temperature using a Jasco (Tokyo, Japan) J-810 spectrometer. The data were collected from 185 to 260 nm, at 100 nm/min, with a 1-s response time and 2-nm bandwidth using a 0.1-cm quartz cuvette containing dehalogenating enzyme in 50 mM potassium phosphate buffer (pH 7.5). Each spectrum shown is the average of 10 individual scans and was corrected for absorbance caused by the buffer. The CD data were expressed in terms of the mean residue ellipticity (Θ_MRE) using the following equation:

(1)

where Θ_obs is the observed ellipticity in degrees, M_w is the protein molecular weight (33,930.07 g/mol for LinB, 34,068.32 g/mol for DhaA, 35,966.21 g/mol for DhlA, and 34,911.32 g/mol for DbjA), n is the number of residues (302 in LinB, 299 in DhaA, 316 in DhlA, and 316 in DbjA), l is the cell path length (0.1 cm), c is the protein concentration (0.0863 mg/ml for LinB, 0.1165 mg/ml for DhaA, 0.1005 mg/ml for DhlA, and 0.0968 mg/ml for DbjA), and the factor 100 originates from the conversion of the molecular weight to milligrams per decimole. The secondary structure content was calculated from the spectra using self-consistent methods (45) implemented in the program DICROPROT (http://dicroprot-pbil.ibcp.fr).

Site-directed mutagenesis.

Three DbjA mutants, D103A, E127A, and H280A, were constructed using inverse PCR as described previously (5); the oligonucleotides used here are listed in Table 1.

Multivariate data analysis.

The substrate specificities of the haloalkane dehalogenases were analyzed by principal component analysis (49), a method of data analysis that is intended to extract and visualize systematic patterns or trends in large data matrices. The analyzed data matrix consisted of six haloalkane dehalogenases (DhlA, DhaA, LinB, DhmA, DmlA, and DbjA) and relative activities measured with the substrates listed in Table 2. Only the substrates with activity data available for more than three enzymes (16 substrates) were subjected to the analysis. The data were logarithmically transformed, centered, and scaled to unit variance prior to the analysis. The multivariate data analysis was conducted using the statistical package SIMCA P v10.0 (Umetrics, Umea, Sweden).

TABLE 2.

Substrate specificities of Trx-DmlA and His-tagged DbjA

Substrate	Relative activity (%)a
	Trx-DmlAb	MAFFc	DbjA-Hisd	USDAe
α-Haloalkanes
1-Chloropropanej(C3)	49 ± 15	53 ± 4	44 ± 2	50 ± 2
1-Chlorobutanej(C4)	100f	100g	100h	100i
1-Chloropentanej(C5)	120 ± 4	39 ± 4	140 ± 38	63 ± 2
1-Chlorohexanej(C6)	110 ± 21	38 ± 14	210 ± 11	110 ± 11
1-Chlorodecanej(C10)	110 ± 0	110 ± 18	52 ± 12	400 ± 12
1-bromobutanej	630 ± 8	750 ± 220	180 ± 16	150 ± 82
α,ω-Haloalkanes
1,2-Dichloroethanej	40 ± 3	59 ± 0	44 ± 3	97 ± 18
1,2-Dibromoethanej	210 ± 420	33 ± 20	1,100 ± 190	49 ± 10
1,3-Dibromopropanej	700 ± 320	150 ± 38	420 ± 47	180 ± 17
Non-α-haloalkanes
2-Chloropropanej	61 ± 9	77 ± 6	35 ± 3	76 ± 12
2-Chlorobutanej	22 ± 10	120 ± 16	30 ± 7	49 ± 49
1,2-Dichloropropanej	39 ± 3	100 ± 3	16 ± 3	190 ± 30
1,2-Dibromopropanej	76 ± 13	51 ± 1	130 ± 3	85 ± 3
1,2,3-Trichloropropane	18 ± 4	24 ± 15	25 ± 2	54 ± 17
1-Chloro-2-methylpropanej	46 ± 1	420 ± 94	31 ± 9	110 ± 5
Halocycloalkanes
Chlorocyclohexane	62 ± 2	59 ± 3	30 ± 1	33 ± 7
Bromocyclohexanej	420 ± 110	90 ± 48	84 ± 5	120 ± 9
Haloalkanes
3-Chloro-2-methylpropenej	270 ± 39	1,400 ± 61	320 ± 18	210 ± 58

^aRelative activity was expressed as the rate for 1-chlorobutane.

^bCrude extract of E. coli cells expressing the fusion protein of thioredoxin and DmlA.

^cResting cells of M. loti MAFF303099.

^dPurified His-tagged DbjA.

^eResting cells of B. japonicum USDA110.

^fSpecific activity was 2.6 × 10⁻³ ± 0.2 × 10⁻³ U/mg of protein.

^gSpecific activity was 5.2 × 10⁻⁶ ± 2.4 × 10⁻⁶ U/mg of resting cells.

^hSpecific activity was 1.08 ± 0.07 U/mg of protein.

ⁱSpecific activity was 6.3 × 10⁻⁷ ± 0.7 × 10⁻⁷ U/mg of resting cells.

^jSubstrates used for principal component analysis.

Preparation of resting cells.

The resting cells of M. loti MAFF303099 and B. japonicum USDA110, grown in TY and YM media, respectively, were prepared in glycine buffer (0.2 M glycine, 0.2 M NaOH, pH 8.0) as described previously (31).

Northern blot analysis.

M. loti MAFF303099 and B. japonicum USDA110 were grown in TY, HM, and YM media. Total RNA was isolated using an RNeasy Mini kit (QIAGEN) according to the manufacturer's protocol. Hybridization and detection were performed using digoxigenin-labeled DNA with the CSPD system (Boehringer, Mannheim, Germany), according to the manufacturer's protocol.

RESULTS AND DISCUSSION

Identification of putative haloalkane dehalogenase genes on the genomes of M. loti MAFF303099 and B. japonicum USDA110.

In this study, a detailed comparison of the deduced amino acid sequence of LinB with the translated genomic sequences of M. loti MAFF303099 and B. japonicum USDA110 led to the identification of two putative gene products of the ORFs mlr5434 (Mlr5434) and blr1087 (Blr1087) with significant identity (44 and 41%, respectively). The deduced amino acid sequences of Mlr5434 and Blr1087 also showed significant identity to those of other known haloalkane dehalogenases, DhaA (49% for both) and DhlA (26 and 25%, respectively). The alignment of the deduced amino acid sequences of LinB, Mlr5434, and Blr1087 (Fig. 1) revealed that Mlr5434 and Blr1087 have the conserved catalytic triad Asp-His-Glu of the haloalkane dehalogenases (11), suggesting that these proteins potentially catalyze dehalogenation by a hydrolytic mechanism. However, it remained unclear whether or not these proteins actually dehalogenate any halogenated compounds.

FIG. 1.

Alignment of amino acid sequences of the putative dehalogenases from M. loti MAFF303099 (Mlr5434/DmlA) and B. japonicum USDA110 (Blr1087/DbjA) and haloalkane dehalogenase LinB from S. paucimobilis UT26. Putative protein products of mlr5434 (Mlr5434) and blr1087 (Blr1087) were named DmlA and DbjA, respectively. Sequence alignment was created using CLUSTALW 1.7 (47) and adjusted manually. The secondary structure elements (indicated by lines under the sequence) and the catalytic triad (indicated by triangles above the sequence) of LinB were deduced from the crystal structure (29). Secondary structure elements of DbjA and DmlA are consensus predictions using the programs PHD (43), PSIPRED (20), Jpred (9), SSThread (15), and Network Protein Sequence Analysis (8).

Cloning and expression of mlr5434 (dmlA gene) in E. coli.

Because E. coli(pYMLA1) cells did not express Mlr5434, pET-32a(+) (26) was used to construct pYMLA2 for the expression of Mlr5434 as a fusion protein with thioredoxin. E. coli BL21(DE3)(pYMLA2)(pG-KJE8) cells produced a relatively large amount of the fusion protein in the soluble form (Fig. 2a). Based on the observed activity of the crude extract of E. coli BL21(DE3)(pYMLA2)(pG-KJE8) cells in the presence of the 18 halogenated compounds (Table 2), we concluded that mlr5434 encodes a member of the haloalkane dehalogenase family, and we designated it dmlA (dehalogenase from Mesorhizobium loti). The fusion protein designated Trx-DmlA showed relatively higher levels of activity in the presence of two methylated compounds, 1-chloro-2-methylpropane and 3-chloro-2-methylpropene, than did the other known dehalogenases (see below for more details).

FIG. 2.

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis analyses of Trx-DmlA (a), DbjA (b), and His-tagged DbjA (c). a. E. coli BL21(DE3)(pYMLA2)(pG-KJE8) cells. Lanes: 1, total proteins of noninduced cells; 2, total proteins of IPTG-induced cells; 3, crude extract of IPTG-induced cells. b. E. coli BL21(pYBJA1) cells. Lanes: 1, total proteins of noninduced cells; 2, total proteins of IPTG-induced cells; 3, crude extract of IPTG-induced cells. c. Protein patterns during the His-tagged DbjA purification from E. coli BL21(pYBJA2) cells. Lanes: 1, total proteins of noninduced cells; 2, total proteins of IPTG-induced cells; 3, crude extract of IPTG-induced cells; 4, purified His-tagged DbjA. Arrows indicate Trx-DmlA (a), DbjA (b), and His-tagged DbjA (c), respectively.

Cloning and expression of blr1087 (dbjA gene) in E. coli.

E. coli BL21(pYBJA1) cells expressed Blr1087 in the soluble form (Fig. 2b, lane 3). Since the crude extract containing Blr1087 showed obvious haloalkane dehalogenase activity toward 1,3-dibromopropane and other substrates (data not shown), we concluded that blr1087 encodes a haloalkane dehalogenases, and we designated it dbjA (dehalogenase from Bradyrhizobium japonicum).

Purification and characterization of His-tagged DbjA.

Efficient expression of DbjA in E. coli enabled its detailed biochemical characterization. A six-histidyl tail was added to the C terminus of DbjA, and the resultant protein (His-tagged DbjA) was purified to homogeneity in a single step using Ni-NTA resin (Fig. 2c). The specific activity of the purified His-tagged DbjA toward 1,2-dibromoethane was 14.7 U/mg protein, indicating that the enzyme was purified approximately 5.4-fold to a crude extract (2.7 U/mg). CD spectra were recorded for His-tagged DbjA and its structurally related haloalkane dehalogenases, DhlA, DhaA, and LinB, which are His-tagged variants. The CD spectra of all tested enzymes showed two negative maxima at 210 and 222 nm that are characteristic for α-helical content (Fig. 3a). Compared to other haloalkane dehalogenases, DbjA exhibited more intensely negative maxima than the other enzymes. This finding suggests an increased number of amino acid residues in an α-helical conformation (Fig. 3b). This result was in good agreement with the secondary structure prediction presented in Fig. 1. The catalytic efficiency of DbjA was assessed by determination of the steady-state kinetic constants for the conversion of eight substrates (Table 3). In general, the catalytic rates and binding affinities of DbjA with particular substrates are comparable (in the same order of magnitude) with other biochemically characterized haloalkane dehalogenases.

FIG. 3.

CD spectra of haloalkane dehalogenases. a. Far-UV CD spectra of four haloalkane dehalogenases, LinB, DhaA, DhlA, and DbjA. b. Comparison of α-helical content of haloalkane dehalogenases estimated by the self-consistent method (45). The predictions made using this method corresponded well with the secondary structure content deduced from the crystal structure (LinB, 40.3% of α-helical content; DhaA, 43% of α-helical content; DhlA, 40.9% of α-helical content) (35, 37, 42).

TABLE 3.

Kinetics parameters of His-tagged DbjA

Substrate	Km (mM)	kcat (s−1)	kcat/Km (mM−1 · s−1)
1-Chlorobutane	4.0 ± 1.8	1.4 ± 0.42	3.5 × 10−1
1-Chlorohexane	0.05 ± 0.01	1.4 ± 0.04	2.8 × 10
1,3-Dibromopropanea	0.22 ± 0.07	3.6 ± 0.49	1.6 × 10
1,2-Dibromopropane	1.8 ± 0.35	0.96 ± 0.06	5.3 × 10−1
1,2,3-Trichloropropane	0.28 ± 0.09	0.24 ± 0.02	8.6 × 10−1
Chlorocyclohexane	0.04 ± 0.02	0.25 ± 0.001	6.3
Bromocyclohexane	0.11 ± 0.02	0.59 ± 0.02	5.4
3-Chloro-2-methylpropenea	0.47 ± 0.29	3.5 ± 1.1	7.4

^aSubstrate inhibition was detected for 1,3-dibromopropane (substance inhibition constant [K_si] = 6.98 ± 2.91 mM) and for 3-chloro-2-methylpropene (K_si = 5.12 ± 3.47 mM).

Site-directed mutagenesis of His-tagged DbjA.

For further confirmation that DbjA is a member of haloalkane dehalogenases, we constructed three mutants. The Asp103, Glu127, and His280 residues, which form a putative catalytic triad in DbjA, were changed to alanine, and the mutants were designated D103A, E127A, and H280A, respectively. The D103A and H280A mutant proteins were purified to homogeneity using Ni-NTA resin (data not shown), but the E127A mutant could not be purified, because only a small amount of the protein was detected in the soluble form in E. coli. The purified D103A and H280A mutant proteins and the crude extract of E. coli cells expressing E127A did not show any activity toward 1,3-dibromopropane and 1-chlorobutane, indicating that these amino acid residues are essential for the activity of DbjA. This result strongly suggested that DbjA catalyzes dehalogenation by a two-step hydrolytic mechanism known in the case of other haloalkane dehalogenases (11).

Multivariate data analysis of substrate specificity.

The dehalogenase activity of the purified His-tagged DbjA protein in the presence of 18 halogenated compounds was measured spectrophotometrically (Table 2). The relative activity data measured for DbjA with 18 different halogenated substrates were complemented with the data collected for other enzymes from the literature, i.e., DhlA (12), DhaA (12), LinB (12), and DhmA (19), and also with the data obtained using a crude extract of E. coli expressing the Trx-DmlA fusion protein (Table 2). Only the substrates with activity data available for more than three enzymes (16 substrates) were subjected to statistical analysis. Principal component analysis has previously been found to be a suitable method for the comparison of enzyme substrate specificity (5, 10, 30, 32). In the present study, this type of analysis led to one statistically significant principal component accounting for 46% of the data variability, which revealed that DbjA and DmlA constitute a new specificity class of haloalkane dehalogenases. This conclusion was supported by the score plot (Fig. 4a) showing positive values for only DbjA and DmlA and negative values for all of the other enzymes tested. Examination of the loading plot (Fig. 4b) also revealed that DbjA and DmlA exhibited unusually high activity in the presence of the two methylated compounds in the data set: 1-chloro-2-methylpropane and 3-chloro-2-methylpropene. The high activity of the enzymes toward the substrates carrying a bulky substituent in the β position indicated the presence of additional pockets in the active sites. Such pockets can be formed by extended/additional α-helices of the specificity-determining cap domain, as was inferred from the prediction of the secondary structure of both DmlA and DbjA (Fig. 1). Furthermore, the CD spectra of the haloalkane dehalogenases strongly suggested that DbjA contains more α-helices than the other dehalogenases (Fig. 3).

FIG. 4.

Principal component analysis of substrate specificities of DbjA, DmlA, and other haloalkane dehalogenases. (a and b) The score plot (a) and the loading plot (b) of the first principal component from analysis of specific activities determined for 16 halogenated substrates.

Expression of the haloalkane dehalogenase genes in the rhizobial strains.

The dehalogenase activities toward the 18 halogenated compounds of resting MAFF303099 and USDA110 cells grown in TY and YM media, respectively, were measured spectrophotometrically. Both of these cells showed activity toward all 18 compounds tested (Table 2). These results suggested that DmlA and DbjA are expressed in MAFF303099 and USDA110 cells, respectively, under normal culture conditions. Furthermore, Northern blot analyses using these dehalogenase genes as probes revealed that the dmlA and dbjA genes were transcribed in the MAFF303099 and USDA110 cells, respectively, grown in various media (data not shown). Both dehalogenase genes were transcribed at higher levels when the cells were grown in YM medium than in the other two media, TY and HM. Mannitol or its metabolites may thus have an effect in the induction of the expression of dehalogenase genes.

In this study, we demonstrated that strains that have not yet been reported as degraders of halogenated compounds do possess genes encoding the haloalkane dehalogenases of a novel substrate specificity class, thus indicating that the genomic sequence information is useful as a genetic source for bioremediation and industrial purposes. On the other hand, the physiological function of haloalkane dehalogenases in rhizobial strains is still unknown. As far as we tested, the rhizobial strains could not grow on minimal medium to which haloalkane was supplied as a sole carbon source. Furthermore, we confirmed that the rhizobial strains and dehalogenases converted 11 halogenated substrates (1-chlorobutane, 2-chloropropane, chlorocyclohexane, 1-chloropropane, 1-chlorodecane, 2-chlorobutane, 1-chloropentane, 1,2-dichloroethane, 1,2-dichloropropane, 1,2,3-trichloropropane, and 1,3-dibromopropane) to the corresponding alcohols, but we did not observe the degradation of the alcohols in rhizobial strains. These results suggested that the rhizobial dehalogenases are not used for the assimilation of haloalkanes. Recently, we demonstrated that the products of haloalkane dehalogenase-like genes in other rhizobial strains also exhibited the dehalogenating activity (unpublished data). Further study of haloalkane dehalogenases in these rhizobial strains will provide some clues to the origin and physiological function(s) of these enzymes.