Protein Engineering of the 4-Methyl-5-Nitrocatechol Monooxygenase from Burkholderia sp. Strain DNT for Enhanced Degradation of Nitroaromatics

Bacterial strains and growth conditions. Escherichia coli TG1 (24) was used to express the 4M5NC monooxygenase from pBS(Kan)DntB DNT under the control of a lac promoter. The strain was cultured from single colonies in 25 ml of Luria-Bertani (LB) medium (24) with 100 μg/ml kanamycin to maintain the plasmid and was incubated overnight at 37°C at 250 rpm. One milliliter of the overnight culture was used to inoculate 50 ml of LB with 100 μg/ml kanamycin and 1 mM of isopropyl-β-d-thiogalactopyranoside (IPTG), and the cells were grown exponentially to a turbidity at 600 nm of 1.5 to 2. Cultures harvested during the exponential phase were used for product identification and rate determinations. The cells were centrifuged at 13,000 × g for 7 min at 25°C in a Beckman (Palo Alto, CA) J2-HS centrifuge. The collected cells were resuspended in 100 mM sodium phosphate buffer (pH 6.5) at a turbidity of 3 to 5.

Chemicals.1,2,4-Trihydroxybenzene, 2-chloro-4-nitrophenol (2C4NP), and guaiacol were obtained from Aldrich Chemical Co., Inc. (Milwaukee, WI). FAD and NADPH were obtained from Alexis Biochemicals (San Diego, CA). 4NC, 4NP, 3M4NP, hydroquinone, p-benzoquinone, 2-methylbenzoquinone, and methylhydroquinone were obtained from Fisher Scientific Co. (Fairlawn, NJ). 4M5NC was synthesized previously by Spanggord et al. (27).

Construction of pBS(Kan)DntB DNT.The recombinant plasmid pBS(Kan)DntB DNT was constructed for inducible expression of 4M5NC monooxygenase from dntB (1.65 kb from pJS59) using the lac promoter. dntB was cloned into pBS(Kan) (4) using PCR with Pfu (Stratagene, La Jolla, CA) polymerase that consisted of an initial hot start at 96°C for 2 min, after which the polymerase was added, followed by 30 cycles of 94°C for 45 seconds, 55°C for 45 seconds, and 72°C for 2 min (the final elongation was 72°C for 7 min). All the primers used in this work are listed in Table 1. The forward primer DntB DNT front generated the EcoRI site, and the reverse primer DntB DNT rear generated an XbaI site; after double digestion, the PCR fragment was cloned into the multiple cloning site of pBS(Kan). Prior to ligation of the insert and vector fragments, Antarctic phosphatase (New England Biolabs, Inc., Beverly, MA) was used to treat the vector fragment to avoid self-ligation and decrease the vector background. The resulting plasmid was electroporated into E. coli TG1 competent cells using a Bio-Rad (Hercules, CA) GenePulser/Pulse Controller at 15 kV/cm, 25 μF, and 200 Ω. To verify correct plasmid construction, several random colonies were used for plasmid isolation and restriction digests (EcoRI, XbaI, and RsrII), and the dntB insert was sequenced.

Error-prone PCR of dntB.epPCR (3) was performed to introduce random mutations throughout dntB (GenBank accession no. U68411 ). The two primers for cloning, DntAa NotI Front and pBSKan Rev (Table 1), were designed to utilize the two unique restriction sites, BglII and XbaI, that are upstream and downstream, respectively, from the dntB gene. For the low-fidelity reaction, the mixture (100 μl) contained pBS(Kan)DntB DNT template (100 ng), primers DntAa NotI Front and pBSKan Rev (30 pmol each), 0.2 mM (each) of dATP and dGTP, 1 mM (each) of dCTP and dTTP, 2.5 mM of MgCl₂, 0.7 mM of MnCl₂, and 5 U of Taq DNA polymerase (New England Biolabs, Inc., Beverly, MA). PCR (MJ Research [Watertown, MA] minicycler) was performed at 96°C for 2 min, with 30 cycles of 94°C for 1 min, 70°C for 1 min, and 72°C for 2 min, and a final extension of 72°C for 7 min. The resulting randomized PCR product (2,137 bp) was cloned into pBS(Kan) after double digestion of both vector and insert with EcoRI and XbaI and after Antarctic phosphatase treatment of the vector. The resulting plasmid library was transferred into E. coli TG1 competent cells using the electroporation method described above. The error rate was found to be 0.5 to 1.3% based on sequencing 450 bp from three random colonies.

Saturation mutagenesis at M22 and L380 of DntB.Saturation mutagenesis was performed at codons M22 and L380 of DntB. Thirty-two codons for all 20 amino acids were created at the target position by replacing the codon with NNS via overlap extension PCR (5).

The degenerate primers dntB DNT M22 front and dntB DNT M22 rear were designed to randomize position M22 in DntB. The two additional primers for cloning, DntAa NotI Front and pBSKan Rev, were chosen to utilize the two unique restriction sites, BglII and XbaI, that are upstream and downstream, respectively, from position M22. pBS(Kan)DntB DNT (100 ng) was used as the template in the initial PCR, and Pfu DNA polymerase was used in the PCR to minimize random point mutations. A 373-bp degenerate DNA fragment was amplified using primers DntAa NotI Front and dntB DNT M22 rear, and a 1,828-bp degenerate DNA fragment was amplified using dntB DNT M22 front and pBSKan Rev. After being purified from a 0.6% agarose gel, the two PCR fragments were combined at a 1:1 ratio as templates to obtain the full-length PCR product by using primers DntAa NotI Front and pBSKan Rev. PCR was performed at 96°C for 2 min, with 30 cycles of 94°C for 45 seconds, 80°C for 45 seconds, and 72°C for 2 min, and a final extension of 72°C for 7 min. The resulting randomized PCR product (2,137 bp) was cloned into pBS(Kan) after double digestion of both vector and insert with BglII and XbaI and after Antarctic phosphatase treatment of the vector. The resulting plasmid library was transferred into competent E. coli TG1 cells as described above.

A similar approach was used for saturation mutagenesis at L380 of DntB. The degenerate primers dntB DNT L380 front and dntB DNT L380 rear were designed to mutagenize position L380 in DntB. A 1,445-bp degenerate DNA fragment was amplified using primers DntAa NotI Front and dntB DNT L380 rear, and a 739-bp degenerate DNA fragment was amplified using dntB DNT L380 front and pBSKan Rev.

Saturation mutagenesis was also performed simultaneously at M22 and L380 of DntB. To combine the fragments containing the randomized positions M22 and L380, two internal primers, DntBDNT M+L front and DntBDNT M+L rear, were designed that allowed the individual saturation mutagenesis PCR products of both positions (1,111 bp for M22 and 1,069 bp for L380) to be combined due to overlap of 44 bases; these two primers replaced pBSKan Rev and DntAa NotI Front, which were used for saturation mutagenesis for M22 and L380, respectively. The two additional primers for cloning, DntAa NotI Front and pBSKan Rev, were chosen to utilize the two unique restriction sites, BglII and XbaI, that are upstream and downstream, respectively, from positions M22 and L380.

His-tagged protein purification.To purify the 4M5NC monooxygenases (the wild type and its variant M22L/L380I) using a polyhistidine tag, genes encoding both 4M5NC monooxygenases were amplified from the pBS(Kan) vectors by PCR using primers DntB DNT pET front and DntB DNT pET rear (Table 1) under the same thermal amplification profile used for pBS(Kan)DntB DNT. The PCR products were cleaved with NdeI and HindIII and inserted into pET28b(+) (Novagen, Madison, WI) to generate pET-DntB and pET-M22L/L380I, which yielded a 6-amino-acid His tag, as well as a thrombin cleavage site at the N termini of the 4M5NC monooxygenases (in total, 20 amino acids, NH₃⁺-MGSSHHHHHHSSGLVPRGSH, were added to the complete wild-type amino acid sequence). Both genes were sequenced using primer SEQ PET Front upstream (Table 1) to confirm that there was no amino acid change in the recombinant plasmids at the N termini of the 4M5NC monooxygenases. To overexpress the 4M5NC monooxygenases, pET-DntB and pET-M22L/L380I were transformed into E. coli BL21(DE3) (Novagen, Madison, WI) by electroporation.

Both wild-type 4M5NC monooxygenase and its variant M22L/L380I with a His tag were purified using an Ni-nitrilotriacetic acid column (Invitrogen, Carlsbad, CA). The strains were initially inoculated into 25 ml of LB medium containing kanamycin (100 μg/ml) from single colonies and cultured overnight at 37°C and 250 rpm. Eight milliliters was then inoculated into 1 liter of fresh LB medium with the same antibiotic and incubated at 37°C. IPTG (4 mM) was added at a cell optical density at 600 nm (OD₆₀₀) of ∼0.4 to induce the T7 promoter, and the cultures were incubated at 37°C for 6 h. Cells (1 liter) were harvested by centrifugation at 10,000 × g and 4°C for 10 min and resuspended in 10 ml of the purification buffer (50 mM NaH₂PO₄, pH 8.0, 0.5 M NaCl, 60 μM of FAD). Cell lysis was performed by sonication using the Braunsonic 1510 sonicator at 20 cycles for 10 s at 10 W with 10-s intervals (B. Braun Medical Inc., Bethlehem, PA), and the cells were centrifuged at 20,000 × g and 4°C for 1 h to get rid of the cell debris. The supernatants containing the soluble 4M5NC monooxygenases were loaded into a 10-ml Ni-nitrilotriacetic acid column for purification. His-tagged enzymes were eluted with the purification buffer with imidazole (250 mM) and dialyzed against TE buffer (100 mM Tris-Cl, 0.2 mM EDTA, 60 μM of FAD) overnight and then dialyzed against storage buffer (10 mM Tris-SO₄, 1 mM EDTA, 1 mM β-mercaptoethanol, 0.02% sodium azide, 60 μM of FAD, and 20% glycerol, pH 7.5). FAD was added (3.5× molar excess based on 1 mol FAD/mol protein) to all the buffers to avoid losing FAD during protein purification (18).

The concentrations of the purified enzymes were evaluated by comparing standard protein samples (Precision plus protein standards; Bio-Rad Laboratories, Hercules, CA) of known concentration using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (24). To corroborate these values, a method based on spectrophotometric measurement of the purified enzymes in 6 M guanidine at 280 nm was used (8). The concentrations of His-DntB and His-M22L/L380I were 0.25 μg/μl and 1 μg/μl, respectively, which were used for the calculation of the specific activities of enzymes.

Molecular techniques.Plasmid DNA was isolated using a Midi or Mini Kit (QIAGEN, Inc., Valencia, CA) and digested by the restriction enzymes from New England Biolabs, Inc. (Beverly, MA). DNA fragments were isolated from agarose gels using a QIAquick Gel Extraction Kit (QIAGEN, Inc., Valencia, CA). Ligation reactions were performed at 16°C for 20 h (24) with a 3-to-1 molar ratio (insert-vector). T4 DNA ligase and 10× T4 DNA ligase buffer were from New England Biolabs, Inc. (Beverly, MA).

Colony screening and nitrite detection.A nylon membrane agar colony-screening method (17) was used with whole cells expressing the evolved enzymes with 500 μM of 4NC, 500 μM of 4NP, 1 mM of 3M4NP, 1 mM of 2C4NP, or 1 mM of guaiacol. The screened plates were incubated overnight at 37°C.

Monooxygenase activity was measured indirectly from the nitrite that accumulated due to substrate hydroxylation (Fig. 1). Nitrite concentrations were determined by combining reaction supernatants with sulfanilamide and N-(1-naphthyl)ethyldiamine and then measuring the colored complex that formed spectrophotometrically (7). Purified enzymes (100 μg) were added to 5 ml of sodium phosphate buffer (100 mM; pH 6.5) and were incubated with 300 μM of 4M5NC, 4NC, 4NP, or 3M4NP with 300 μM of NADPH for 5 to 60 min at 250 rpm at 37°C. The sulfanilamide and N-(1-naphthyl)ethyldiamine reagents were added in equal volumes of 100 μl to 1 ml of the sample. The formation of azo dye was detected at 543 nm after 10 min of incubation at room temperature. The nitrite concentration was determined using a linear calibration curve (0.72 to 145 μM). At least three time points from the linear parts of the nitrite release rate curves were used for the calculation of specific initial rates. Two replicates of both wild-type enzyme and the enzyme variants were analyzed. To obtain the apparent V_max, K_m, and k_cat values for 4NP oxidation by both purified wild-type DntB and the M22L/L380I variant at 300 μM of NADPH, 4NP was used at 50, 100, 200, 300, 500, and 1,000 μM. The steady-state kinetic parameters (apparent V_max and K_m) were determined based on a Michaelis-Menten kinetic equation plot.

Variants M22L, L380I, M22L/L380M, and M22S/L380V were evaluated using nitrite released from the oxidation of 4NP by whole cells; cells were incubated with 500 μM of 4NP at a cell density (OD₆₀₀) of 3.0 to 5.0 in sodium phosphate buffer (100 mM, pH 6.5) for 5 to 60 min at 250 rpm at 37°C. Rates from whole cells were determined with a protein concentration of 0.22 mg protein/ml/OD₆₀₀ unit (16). Two or three replicates of both wild-type enzymes and the enzyme variants were analyzed.

Product identification by HPLC.Reverse-phase high-performance liquid chromatography (HPLC) was used to identify the products from 4NC, 4NP, and 3M4NP. Samples were prepared in the same way as the samples for nitrite assay. The samples were incubated for 60 min at 250 rpm at 37°C. Samples (20 μl) were collected and analyzed using HPLC with a Zorbax SB-C8 column (Agilent Technologies; 5 μm, 4.6 by 250 mm) and a Waters Corporation (Milford, MA) 515 solvent delivery system. Compounds were detected by a photodiode array detector (Waters 996) and injected by an autosampler (Waters 717 plus). A gradient elution was used with H₂O (0.1% formic acid) and acetonitrile (75:25 from 0 to 3 min, 50:50 at 18 min, and 75:25 at 25 min) as the mobile phase at 1.5 ml/min. The compounds were identified by comparison of the retention time and UV-visible spectra to those of authentic standards, as well as by coelution with the standards. Two or three replicates of both the wild-type enzyme and the enzyme variant were analyzed.

DNA sequencing.A dye terminator chain-sequencing technique based on the dideoxy method of sequencing DNA (25) was used with the ABI Prism BigDye Terminator Cycle Sequencing Ready Reaction Kit (Perkin-Elmer, Wellesley, MA). A PE Biosystems ABI 373 DNA sequencer (Perkin-Elmer) was used to determine the nucleotide sequence of the dntB gene for both wild-type DntB and its variants. Vector NTI software (InfoxMax, Inc., Bethesda, MD) was used to analyze the sequencing data. The primers shown in Table 1 were used to sequence the complete dntB gene; all sequencing was performed at least two times.

Nucleotide sequence accession number.The corrected DNA sequences for dntB have been deposited in GenBank (accession number DQ298257 ).

Protein engineering of DntB.To expand the substrate specificity of 4M5NC monooxygenase, epPCR was performed, and 3,000 colonies were screened for activity on 4NP. The 4M5NC monooxygenase variant M22L/L380I was identified, since the colony expressing this enzyme turned dark brown after overnight incubation at 37°C on the 4NP screening plate whereas the wild-type colony turned only light brown. The variant M22L/L380I was the only colony found that turned dark brown on the 4NP screening plate. The negative control E. coli TG1/pBS(Kan) had no color, so background E. coli oxygenases have no activity on this substrate. The dark-brown color arises from the intracellular dihydroxylated product (in this case, hydroquinone) (Fig. 1), which auto-oxidizes to quinones upon secretion (17). DNA sequencing indicated there were two amino acid codon changes, M22L and L380I, in dntB.

After the variant M22L/L380I was identified, saturation mutagenesis at M22 and L380 of DntB individually was performed to see if better enzymes could be obtained. Variants M22L and M22S and variants L380V, L380I, and L380M were identified after 300 colonies were screened to ensure that all possible variants were checked (22) on the 4NP screening plate from each library of saturation mutagenesis at M22 and L380, respectively. Colonies expressing these variants were darker brown than the wild type, but not as brown as cells expressing variant M22L/L380I, which indicated that although both positions M22L and L380I affect enzyme activity, both mutations in these residues are required to obtain activity significantly greater than that of wild-type DntB.

To try all possible amino acids at both positions M22 and L380 of DntB, simultaneous saturation mutagenesis was performed. A saturation mutagenesis library of 650 variant colonies was screened to ensure that all of the possible variants were checked at both positions (22). DntB variants M22L/L380M and M22S/L380V were identified, since the colonies from which they were obtained both had a dark-brown color as intense as that of the DntB variant M22L/L380I on the 4NP screening plates. For the 3M4NP screening plates, the same trend in color formation was seen, although there was lighter color. There was no difference in the color formed between colonies expressing wild-type DntB and its variants using the 4NC, 2C4NP, and guaiacol screening plates.

HPLC identification of the 4NC, 4NP, and 3M4NP transformations.HPLC analysis was used to identify the oxidation products of purified wild-type DntB and its variant, M22L/L380I. 1,2,4-Trihydroxybenzene was detected from 4NC oxidation by both wild-type DntB and variant M22L/L380I. Hydroquinone and methylhydroquinone were detected from oxidation of 4NP and 3M4NP, respectively, by the purified M22L/L380I enzyme; hence, activity toward these substrates was achieved for the first time with the purified variant M22L/L380I, since no oxidation products were detected for both reactions with purified wild-type DntB. Hydroquinone and methylhydroquinone were also detected on the time scale of these reactions from the intermediates p-benzoquinone and 2-methylbenzoquinone, respectively, with or without any enzyme in the reaction mixtures, which indicates that p-benzoquinone and 2-methylbenzoquinone are not stable under these conditions.

Nitrite detection from oxidation of nitroaromatics.Nitrite is released from the degradation of 4M5NC by DntB (11), and 4NC was reported to be the only other compound that could serve as a substrate for the wild-type enzyme (11). In this work, we confirmed that nitrite is derived from the degradation of 4M5NC and 4NC using purified wild-type DntB. More importantly, we found that variant M22L/L380I also generates nitrite from the degradation of the two new substrates 4NP and 3M4NP (Fig. 1). The purified variant M22L/L380I oxidized 300 μM 4NC 1.6-fold faster than purified wild-type DntB (Table 2). For 4NP and 3M4NP oxidation, the purified M22L/L380I enzyme oxidized 4NP and 3M4NP 9.8- and 3.9-fold faster, respectively, than purified wild-type DntB at 300 μM (Table 2).

To compare the M22L/L380I variant with the wild-type enzyme, apparent V_max, K_m, and k_cat values for 4NP oxidation were determined (Table 3). For the purified wild-type DntB enzyme, the apparent V_max, K_m, and k_cat values were 14 nmol/min/mg protein, 350 μM, and 0.014 s⁻¹, respectively. For the purified variant M22L/L380I enzyme, the apparent V_max, K_m, and k_cat values were 44 nmol/min/mg protein, 100 μM, and 0.045 s⁻¹, respectively. Therefore, the k_cat/K_m values of the purified wild-type DntB enzyme and purified variant M22L/L380I were 40 and 450 (s⁻¹ M⁻¹), respectively, which shows that the M22L/L380I variant has 11-fold-higher efficiency than the wild-type enzyme for 4NP degradation. It was found that DntB was inhibited at concentrations of 4NP higher than 1 mM.

Since 4M5NC is the natural substrate of DntB (29), its rate of oxidation was also checked via nitrite release for both purified wild-type DntB enzyme and its variant, M22L/L380I. The initial rate of 4M5NC oxidation by the purified variant M22L/L380I enzyme was 1.4-fold higher than that of the wild-type DntB enzyme at 300 μM substrate concentration (Table 2).

The initial nitrite release rates of whole cells expressing the variants M22L, L380I, M22L/L380M, and M22S/L380V were also determined to compare them to the activities of variant M22L/L380I and wild-type DntB using 500 μM of the substrate (Table 4). Variants M22L, L380I, and M22L/L380M had 5-, 2.5-, and 5-fold-lower rates of oxidation of 4NP than the variant M22L/L380I, and variant M22S/L380V had about 20% greater initial rate of nitrite release from the oxidation of 4NP than the variant M22L/L380I. Hence, a new enzyme variant, M22L/L380I, was discovered that has elevated nitrite release for 4NP and 3M4NP, and simultaneous saturation mutagenesis created an even better variant (M22S/L380V). To ensure that the differences in whole-cell activities were not due to differences in expression, sodium dodecyl sulfate-polyacrylamide gel electrophoresis was performed, which indicated that the variants M22L, L380I, M22L/L380I, M22L/L380M, and M22S/L380V all had the same expression level as wild-type DntB (data not shown).

DNA sequencing.We sequenced the entire wild-type dntB gene from our cloning plasmid pBS(Kan)DntB, as well as the entire dntB gene from the original plasmid (pJS59) that was used to determine the DNA sequence from Burkholderia sp. strain DNT. Plasmid pJS59 is derived from pJS53 (29). The dntB DNA sequences from pBS(Kan)DntB and pJS59 are identical, so there were no changes in the DNA sequence due to cloning into pBS(Kan). However, two sequencing errors were identified in the published DNA sequence of dntB, GenBank accession number U68411 (12), that had an impact on the protein sequence. The two nucleotide changes at base pair positions 934 and 935 of dntB in U68411 should be changed from CGC to GCC, which results in an amino acid change in DntB from arginine to alanine (R312A). These changes were verified by sequencing at least five times.