INTRODUCTION

Top Abstract Introduction Materials & Methods Results & Discussion References

Aerobic degradation of aromatic compounds by bacteria is frequently initiated by non-heme iron-containing dioxygenases (22). These soluble multicomponent enzymes, which introduce two atoms of molecular oxygen into the aromatic ring and thereby activate it for subsequent cleavage, are classified into five groups according to the number of constituent components and the nature of the redox centers (3). Class IIB dioxygenases, such as the TecA chlorobenzene (4) and the TodCBA toluene dioxygenases (39), are comprised of a reductase and ferredoxin, which together serve as a short electron transport chain, and a catalytic terminal dioxygenase composed of a large -subunit and small -subunit with an ()_n configuration (22). Structural information about aromatic ring dioxygenases is very limited, and only recently has the terminal oxygenase component of naphthalene dioxygenase from Pseudomonas sp. strain NCIB 9816-4 been crystallized (21).

-Subunits of class IIB terminal dioxygenases contain a Rieske-type [2Fe-2S] iron-sulfur cluster (11, 28), an active-site non-heme mononuclear Fe(II) center (22), and the substrate binding site, which is assumed to be located in the vicinity of the activating iron (6). By exchanging subunits between different dioxygenase systems, several groups have shown that the -subunit is responsible for substrate specificity (8-10, 26, 32, 34, 36). Further analyses of -subunits subsequently identified a large C-terminal region of nitrotoluene dioxygenase of Pseudomonas sp. strain JS42 (26) and smaller elements of biphenyl dioxygenases from Pseudomonas pseudoalcaligenes KF707 and Pseudomonas sp. strain LB400 (19, 23) as being involved in the determination of substrate specificity.

The -subunits of class IIB enzymes were reported to play a role in subunit association (6, 13, 24) and substrate recognition (12, 13), whereas some investigators excluded a direct involvement of the -subunit in the determination of substrate specificity (26, 27, 34).

The substrate specificities of initial dioxygenases are crucial, because they often limit the range of compounds potentially degradable by the catabolic system. Because substituents often complicate mineralization, removal of one in the first step of a catabolic sequence is an advantageous mechanism that merits special attention. So far only one enzyme, the TecA chlorobenzene dioxygenase of Burkholderia sp. strain PS12, has been shown to dechlorinate a tetrachlorobenzene (4), and no dioxygenase able to transform higher chlorinated benzenes is known to date. Comprehension of the structural requirements for dechlorination is a prerequisite for the improvement of the catalytic properties of biocatalysts.

To identify the structural elements involved in dechlorination, we examined two class IIB enzymes (3) with complementary substrate specificities by exchanging equivalent polypeptide sequences. We report here the construction of an extensive number of chimeric dioxygenases and the analysis of their substrate specificities and transformation rates. This led to the identification of a single amino acid, as well as interacting regions required for dehalogenation of tetrachlorobenzene, and the generation of hybrid dioxygenases with extended substrate ranges.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results & Discussion References

Strain and plasmids. The host strain used in this study was Escherichia coli DH5 from Clontech. The cloning vectors used were pBluescript II KS(+) (Stratagene) and pCR2.1 (Invitrogen). The sources of the tecA1A2A3A4 chlorobenzene dioxygenase genes were plasmids pSTE3 and pSTE7 (4), and the source of the todC1C2BA toluene dioxygenase genes was plasmid pDTG601 (39).

DNA manipulations. Standard procedures were performed as described by Sambrook et al. (30). Restriction enzymes were purchased from Amersham, Boehringer Mannheim, MBI Fermentas, New England Biolabs, and U.S. Biochemical Corp. T4 DNA ligase was purchased from New England Biolabs. Isopropyl--D-thiogalactopyranoside was obtained from Roth. Ampicillin was purchased from Sigma. 5-Bromo-4-chloro-3-indolyl--D-galactopyranoside (X-Gal) was obtained from Biomol. Oligonucleotides were obtained from Gibco Life Technologies. Taq and Pfu DNA polymerases were obtained from Boehringer Mannheim and Stratagene, respectively. Blunt-end fragments resulting from Pfu DNA polymerase amplification were A-tailed with Taq polymerase for subsequent cloning into the pCR2.1 T-vector system (Invitrogen) as described previously (31). Elution of DNA from agarose gels was performed with the QiaexII gel extraction kit (Qiagen). Plasmids were purified with a Qiawell 8 plasmid kit or a plasmid midi kit (Qiagen). Sequencing was done with the Applied Biosystems 373A DNA sequencer (Perkin-Elmer, Applied Biosystems) as described previously (18). Site-specific mutations were introduced by using splicing by overlap extension (SOE)-PCR (14) or with the Quikchange site-directed mutagenesis kit (Stratagene). The sequences of all de novo-synthesized DNA molecules and of the commercially obtained oligonucleotide primer sequences were confirmed by sequencing.

Oligonucleotides. The designation, sequence (5'3'), and priming direction of the oligonucleotide primers used for amplification of DNA fragments by PCR (29), SOE-PCR (14), and the Quikchange kit (Stratagene) are as follows: prSTB70 (forward), gcccgcgggctctatgcccattggc (SacII); prSTB71 (reverse), gacgtcggctctcttgacggaatcaagc (AatII); prSTB72 (forward), cgagctcggtgagaagacaatgaatc (SacI); prSTB73 (reverse), cctcggtgcggtcgagcatatggtc (NdeI); prSTB74 (forward), cgaagttctacatggaccatatgctcg (NdeI); prSTB75 (reverse), gtagctggtgacctttggccccatg (BstEII); prSTB76 (reverse), ggatgccatgtccggaccgtgttg (RsrII); prSTB77 (forward), caacacggtccggacatggcatcc (RsrII); prSTB104 (reverse), ctggcaggcctgccaggatgcc (StuI); prSTB105 (forward), ctgtaactggaaactcgccgcagagc (Phe211Leu); prSTB106 (forward), gagcagttttgctgggacatgtaccatg (Ser218Trp); prSTB107 (forward), gttttgcagcgacgcgtaccatgccg (Met220Ala); prSTB108 (forward), gtaccatgccgcgacgacctcgcatc (Gly224Ala); prSTB109 (forward), ccgggacgaccgcgcatctgtctgg (Ser227Ala); prSTB110 (forward), ctgtaactggaaagccgccgcagagc (Phe211Ala); prSTB111 (forward), gagcagttttgcgccgacatgtaccatg (Ser218Ala); prSTB114 (forward), caggcctgccagacggcgttgaactg (StuI); prSTB124 (reverse), gctctgcggcgagtttccagttacag (Phe211Leu); prSTB125 (reverse), catggtacatgtcccagcaaaactgctc (Ser218Trp); prSTB126 (reverse), cggcatggtacgcgtcgctgcaaaac (Met220Ala); prSTB127 (reverse), gatgcgaggtcgtcgcggcatggtac (Gly224Ala); prSTB128 (reverse), ccagacagatgcgcggtcgtcccgg (Ser227Ala); prSTB129 (reverse), gctctgcggcggctttccagttacag (Phe211Ala); prSTB130 (reverse), catggtacatgtcggcgcaaaactgctc (Ser218Ala); prSTB132 (reverse), ggtgacctttggccccatgatggcaagcagcagatcgggttcgccaataaagaagccac (BstEII); prSTB135 (forward), ggcaggcctgccagaagaccttgaaatggccgatc (StuI); prSTB136 (forward), ctggcaggcctgccggacggcgttg (StuI); and prSTB137 (forward), cgaaggtcaccagctactggacc (BstEII). Recognition sequences for the restriction enzymes indicated are underlined, and boldface letters indicate the triplets which were changed by the Quikchange system.

Resting cell assays. E. coli strains were cultured (1% inoculum) in Luria-Bertani medium (30) containing 0.1 mg of ampicillin per ml and 1.0 mM isopropyl-thio--galactopyranoside in baffled round-bottom flasks at 30°C on a rotary shaker operated at 160 rpm, and cells were prepared as described previously (4). Briefly, after washing twice with assay buffer (10 mM glucose in 0.1 × M9 mineral medium), cell suspensions were concentrated to an A₆₀₀ of 50, and then 1.0-ml samples were removed from the concentrate, shock-frozen in liquid nitrogen, and stored at 20°C for Western blot analysis and determination of whole-cell protein. The resting cell assay was started by dilution of concentrated cell suspension to an A₆₀₀ of 2.0 in 10 ml of prewarmed assay buffer containing 0.5 mM substrate. To monitor product formation, samples were taken at regular intervals between 0 and 180 min and immediately shock-frozen in liquid nitrogen for subsequent analysis. None of the transformation products indicated in Table 1 was detected in control experiments with E. coli resting cells carrying pBluescript II KS(+). Standard deviations were below 50%, as determined from three independent transformation experiments each with E. coli (pSTO4), E. coli (pSTE7), and E. coli (pSTE81) cells.

Analysis of transformation products. Dihydroxy compounds were analyzed with a reverse-phase high-performance liquid chromatography (HPLC) system equipped with autosampler and a sample cooler unit operated at 4°C (Shimadzu) on an SC 125 by 4.6-mm Lichrospher 100 RP8 5.0-µm column (Bischoff). The aqueous solvent system contained 0.1% ortho-phosphoric acid and 18 to 63% methanol at a 1.0-ml/min flow rate (Table 1). Shock-frozen samples were thawed and centrifuged (20 min, 4°C, 14,000 × g), and 20 µl of cell-free supernatant fluid was analyzed. Absorbance was monitored between 200 and 400 nm. The identities of cis-1,2-dihydroxy-1,2-dihydrocyclohexa-3,5-diene and 3,4,6-trichlorocatechol were confirmed by comparison with authentic standards, which were also used to calibrate HPLC analyses. Spectral data of 3,4-dichloro-cis-1,2-dihydroxy-1,2-dihydrocyclohexa-3,5-diene were in agreement with those published previously (4).

Western blot analysis. Immunologically detectable soluble -subunits from individual constructs used for transformation experiments were visualized by Western blot analysis. Cell suspensions were thawed and diluted as required in the resting cell assays, ruptured on ice by ultrasonic pulses (six times at 10 s each, 100 W; Labsonic U; B. Braun) in the presence of 40 µg of DNAse I per ml and 20 µg of RNase A per ml, and centrifuged (40 min, 4°C, 130,000 × g). Supernatant fluid was subjected to electrophoresis on a sodium dodecyl sulfate-10% polyacrylamide gel electrophoresis gel (20). Extract of E. coli cells carrying pBluescript II KS(+) obtained in a similar fashion was used as negative control, and different volumes of a similar extract of E. coli (pSTE7) cells producing the wild-type TecA dioxygenase were used as internal standards on each gel.

Substrate transformation rates of 1,2-dichlorobenzene, benzene, and 1,2,4,5-tetrachlorobenzene. Substrate transformation rates were expressed as the amount of product formed per unit of time and amount of whole-cell protein (1,000 area units [AU] min¹ µg¹) measured at the wavelength of the corresponding absorption maximum after HPLC separation (Table 1). Rates were usually constant for at least 30 min (Fig. 1). The concentration of whole-cell protein was determined by boiling cell suspensions for 10 min in 100 mM NaOH according to the method of Bradford (5) with bovine serum albumin as the standard. The use of the same bacterial host and assay conditions allowed a direct comparison of the transformation rates of different dioxygenase systems for each substrate.

View larger version (18K): [in this window] [in a new window]   FIG. 1.   HPLC analysis of product formation. E. coli cells carrying dioxygenase genes were incubated with 0.5 mM substrate. At regular time intervals, samples were taken, and the supernatant fluids were analyzed by HPLC (Table 1) for product accumulation. The formation of 3,4-dichloro-cis-1,2-dihydroxy-1,2-dihydrocyclohexa-3,5-diene (DCBDHD []) from 1,2-dichlorobenzene and cis-1,2-dihydroxy-1,2-dihydrocyclohexa-3,5-diene (BDHD []) from benzene by E. coli (pSTO4) and the formation of 3,4,6-trichlorocatechol (Cl3-catechol) from tetrachlorobenzene () by E. coli (pSTE7) are shown as a function of time.

Chemicals. Benzene and 1,2-dichlorobenzene were purchased from Fluka, 1,2,4,5-tetrachlorobenzene was obtained from Aldrich, and cis-1,2-dihydroxy-1,2-dihydrocyclohexa-3,5-diene was from Sigma. All chemicals were of the highest purity available. HPLC-grade methanol was from Baker, and 3,4,6-trichlorocatechol was kindly provided by H.-A. Arfmann.

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials & Methods Results & Discussion References

To define the subunit, region, and amino acids which are responsible for dechlorination of tetrachlorobenzene, genetic elements were exchanged between the tecA and todCBA dioxygenase systems of Burkholderia sp. strain PS12 (4) and Pseudomonas putida F1 (39), respectively (Table 2). A large number of hybrids were prepared and analyzed in order to localize specificity determinants as precisely as possible and to investigate the combined effects of multiple determinants. The resulting hybrid dioxygenases were subsequently expressed in E. coli for analysis of the effects of the substitutions on their catalytic potential (Table 3). Because the final product concentration was not always highest for the substrate that was converted at the highest initial rate (data not shown), initial substrate transformation rates were determined (Fig. 1).

View this table: [in this window] [in a new window]   TABLE 3.   Expression of soluble dioxygenase -subunits and corresponding transformation activities of E. coli cells carrying plasmids with wild-type and hybrid dioxygenase genesa

Expression of -subunits. The concentrations of immunodetectable soluble -subunit in recombinant bacteria differed significantly (Fig. 2). These differences may be caused by lower expression of recombinant -subunit genes (37), or conformational changes in hybrid dioxygenases, leading to precipitation as inclusion bodies, accelerated degradation by cellular proteases, or less efficient transfer of ferrous iron cofactor into the active site by the cellular machinery of E. coli (16). A different response of the antibody to the various constructs analyzed cannot completely be excluded. However, because a polyclonal antibody which generally recognizes different epitopes on a protein was used, such an effect would be insufficient to explain the drastically different signals obtained.

View larger version (76K): [in this window] [in a new window]   FIG. 2.   Western blot analysis of soluble -subunit proteins. After electrophoretic separation of 4 µl of crude extracts of E. coli cells carrying different plasmids encoding dioxygenase systems, soluble -subunits were detected with anti-BedC1 rabbit antibody (38), and bands corresponding to chimeric -subunit proteins (50 to 51 kDa) were visualized on film. Eight microliters of E. coli (pBluescript II KS[+]) was used as a negative control, and 2, 4, and 6 µl of E. coli (pSTE7) crude cell extract served as internal standards on the first four lanes of each gel. Numbers at the top of each lane correspond to the number of the plasmid carried in E. coli according to Tables 2 and 3 and Fig. 3. The sizes of the wild-type proteins produced in E. coli (pSTE7) and E. coli (pSTO4) were 50.1 ± 0.3 (indicated by an arrow) and 51.3 ± 0.3 kDa, respectively, which is in close agreement with the deduced molecular masses of wild-type TecA (50.5 kDa) (4) and TodCBA (50.9 kDa) (39) dioxygenases. Small differences in the relative mobility of -subunit proteins are assumed to be due to differences in size and amino acid composition of individual chimeras. The panel was composed from four different gels with Photoshop software (version 3.0; Adobe).

Role of dioxygenase -subunits in substrate specificity. A comparison of recombinant wild-type dioxygenases showed that TecA chlorobenzene dioxygenase dioxygenolytically dechlorinates tetrachlorobenzene but fails to attack benzene (4), whereas the TodCBA toluene dioxygenase (pSTE7 and pSTO4, respectively, in Table 3) displays converse activities. Because 1,2-dichlorobenzene was transformed by both systems, it was used to detect active recombinant enzyme.

Role of region II in dechlorination. To determine which part of the -subunit is responsible for dechlorination, regions I, II, and III of todC1 in the TodC1::TecA2A3A4 hybrid system were sequentially or pairwise replaced by equivalent regions of tecA1 (Table 2 and Fig. 3, sets 2 and 3). Comparison of transformation rates (Table 3) of the resulting active hybrid enzymes showed that the presence of the middle region, TecA1-II, is sufficient for dechlorination of tetrachlorobenzene and that in contrast, the presence of region TodC1-II correlates with benzene transformation (Fig. 3). These results differ from those obtained with biphenyl (8, 19, 23) dioxygenases, in which the C-terminal regions (approximately III and IV in Fig. 4A) of the corresponding -subunits are responsible for substrate specificity.

View larger version (48K): [in this window] [in a new window]   FIG. 3.   Constructs of dioxygenase -subunit proteins. Plasmids pSTE7 (tecA1A2A3A4) and pSTO4 (todC1C2BA) carry wild-type dioxygenase systems. Hybrid systems were constructed with the chimeric -subunits in the tecA2A3A4 background, except for plasmid pSTE86, which contains the todC1C2BA background (Table 2). The regions I of the two enzymes differ by 24 amino acid residues, subregions IIA differ by 5, IIB differ by 4, IIC differ by 1, IID differ by 5, IIE differ by 5, and III differ by 12. The positions at which the amino acids differ between TecA1 and TodC1 are shown on top of pSTE7 -subunit. Black bars indicate fragments of TecA1 origin, grey bars indicate fragments of TodC1 origin, and white boxed amino acids in pSTE84 and pSTE85 were present in neither TecA1 nor in TodC1. Regions I, II, and III and subregions IIA, IIB, IIC, IID, and IID are delineated by black vertical lines. Relevant restriction sites are indicated. The position of the putative Rieske-type [2Fe-2S] iron-sulfur cluster (11, 28) in region I is indicated as an oval. The putative mononuclear iron-coordination motif Glu214-Xaa3-4-Asp219-Xaa2-His222-Xaa4-5-His228 (17) is shown as a boxed area. Solid circles () indicate E. coli cells expressing soluble -subunit protein of TecA1 at a significant level compared to the TecA1 wild-type level. With the exception of E. coli (pSTE18), all constructs transformed 1,2-dichlorobenzene. In addition to 1,2-dichlorobenzene, E. coli resting cells carrying the corresponding plasmid transformed benzene (B), tetrachlorobenzene (T), or all three substrates (B+T) with activities above the detection limit (see Table 3).

Importance of the putative iron ligand region IIA. Although neither the location nor the structure of the substrate binding site of benzene dioxygenases has yet been elucidated, it is assumed to be in the neighbourhood of the non-heme ferrous prosthetic group (6). Mutagenesis studies have suggested that histidines His222 and His228 of the benzene dioxygenase -subunit are iron ligands (6), which is consistent with site-directed mutagenesis studies of TodC1 protein by Jiang (17), who proposed that the motif Glu214-Xaa_3-4-Asp219-Xaa₂-His222-Xaa_4-5-His228, which is also conserved in TecA1 (Fig. 4B), is involved in mononuclear iron coordination.

View larger version (57K): [in this window] [in a new window]   FIG. 4.   Protein sequence alignment of -subunits of selected class IIB dioxygenases. Amino acid alignment of -subunits of chlorobenzene, toluene, and biphenyl class IIB dioxygenases is shown as follows: TecA1.PS12, chlorobenzene dioxygenase of Burkholderia sp. strain PS12 (4); TodC1.F1, toluene dioxygenase of P. putida F1 (39); BphA.LB400, biphenyl dioxygenase of Pseudomonas sp. LB400 (7); BphA1.KF707, biphenyl dioxygenase of P. pseudoalcaligenes KF707 (33). The assignments of regions, subregions, and restriction sites used in this study are indicated by grey boxes above the alignment, and those used in the study of Kimura et al. (19) are indicated below the alignment by dashed boxes, whereas regions specified by Mondello et al. (23) are shown as black boxes (A). Additional sequences used in the alignment of the putative iron ligand region were as follows: TcbAa.P51, chlorobenzene dioxygenase of Pseudomonas sp. strain P51 (37); BedC1.ML2, benzene dioxygenase of P. putida ML2 (35); BnzA.Ppu, benzene dioxygenase of P. putida (15); BphA1.P6, biphenyl dioxygenase of Rhodococcus globerulus P6 (2); NtdAc.JS42, 2-nitrotoluene dioxygenase of Pseudomonas sp. strain JS42 (25); DxnA1.RW1, dioxin dioxygenase of Sphingomonas sp. strain RW1 (1) (B). Differences between TecA1 and TodC1 and BphA and BphA1, respectively, are indicated by shaded amino acids. The numbering of the position of amino acid residues in the putative active-site iron liganding region IIA refers to the TecA1 sequence. Arrows indicate amino acids and their positions, which are different between TecA1 and TodC1 in subregion IIA. Solid circles show putative active-site mononuclear iron ligands (17). Asterisks indicate the conserved cysteines and histidines, putative ligands of the Rieske-type [2Fe-2S] iron-sulfur cluster (11, 28), with the consensus sequence Cys-Xaa-His16-17-Xaa-Cys-Xaa2-Xaa-His (22).

Single amino acid substitutions. Since region TecA1-IIA is crucially involved in dehalogenation, but insertion into TodC1 results in an inactive enzyme, we decided to make individual amino acid substitutions which may perturb the protein less (Fig. 3, set 7). Sequence comparison revealed that the amino acids in region IIA of TecA1 and TodC1 differ in positions 211, 218, 220, 224, and 227, of which only Ala220 is conserved in the two chlorobenzene dioxygenases TecA (4) and TcbA (37) (Fig. 4B). Of a number of substitutions made, only one, Met220Ala (Fig. 3, pSTE81), led to restoration of dehalogenase activity to wild-type levels (Table 3). Substitution of methionine by alanine, which has a smaller side chain, may facilitate access of tetrachlorobenzene to the active-site iron. Because E. coli (pSTE81) cells additionally transformed benzene (Table 3), a biocatalyst has been generated with an extended substrate range. Replacement of two other large amino acids (Phe211 and Ser218) by alanine in the putative active-site iron ligand region (Fig. 3, pSTE84 and pSTE85, respectively) did not lead to restoration of dehalogenase activity.

The dioxygenase -subunits. Because the -subunits of toluate-1,2-dioxygenase and of toluene and biphenyl dioxygenases have been suggested to be involved in substrate specificity (12, 13), we investigated the influence of the -subunit on transformation specificity and efficiency by introduction of the Met220Ala substitution into the TodCBA wild-type system (Fig. 3 [pSTE86]). E. coli (pSTE86) cells were capable of dechlorinating tetrachlorobenzene (Table 3). Thus, the -subunits of the (chloro)benzene dioxygenases studied here are not directly involved in the control of substrate specificity, which is consistent with the indications of other investigators that the -subunits of 2-nitrotoluene, 2,4-dinitrotoluene, and biphenyl dioxygenases are not determinants of substrate specificity (26, 27, 34).