Cloning and expression of the nitrile hydratase and amidase genes from Bacillus sp. BR449 into Escherichia coli

Enzyme and Microbial Technology 27 (2000) 492–501

www.elsevier.com/locate/enzmictec

Cloning and expression of the nitrile hydratase and amidase genes from Bacillus sp. BR449 into Escherichia coli Sang-Hoon Kim, Patrick Oriel* Department of Microbiology, Michigan State University, East Lansing, MI 48824, USA Received 28 September 1999; received in revised form 2 May 2000; accepted 3 May 2000

Abstract A moderate thermophile, Bacillus sp. BR449 was previously shown to exhibit a high level of nitrile hydratase (NHase) activity when growing on high levels of acrylonitrile at 55°C. In this report, we describe the cloning of a 6.1 kb SalI DNA fragment encoding the NHase gene cluster of BR449 into Escherichia coli. Nucleotide sequencing revealed six ORFs encoding (in order), two unidentified putative proteins, amidase, NHase ␤- and ␣-subunits and a small putative protein of 101 amino acids designated P12K. Spacings and orientation of the coding regions as well as their gene expression in E. coli suggest that the ␤-subunit, ␣-subunit, and P12K genes are co-transcribed. Analysis of deduced amino acid sequences indicate that the amidase (348 aa, MW 38.6 kDa) belongs to the nitrilase-related aliphatic amidase family, and that the NHase ␤- (229 aa, MW 26.5 kDa) and ␣- (214 aa, MW 24.5 kDa) subunits comprise a cobalt-containing member of the NHase family, which includes Rhodococcus rhodochrous J1 and Pseudomonas putida 5B NHases. The amidase/NHase gene cluster differs both in arrangement and composition from those described for other NHase-producing strains. When expressed in Escherichia coli DH5␣, the subcloned NHase genes produced significant levels of active NHase enzyme when cobalt ion was added either to the culture medium or cell extracts. Presence of the P12K gene and addition of amide compounds as inducers were not required for this expression. © 2000 Elsevier Science Inc. All rights reserved.

1. Introduction Nitrile compounds occur widely in nature, and synthetic acrylonitrile is used extensively in the manufacture of acrylamide, an intermediate whose polymers are utilized for papermaking, waste treatment, and oil recovery [10,30]. Although nitriles are generally toxic, some microorganisms can assimilate nitriles as carbon and energy sources. One recognized nitrile-assimilating microbial pathway utilizes nitrile hydratase (EC 4.2.1.84) for conversion of nitriles to the corresponding amides, followed by hydrolysis to the corresponding acids and ammonia utilizing amidase (EC 3.5.1.4) [1]. Nitrile hydratase (NHase) is typically composed of equal amounts of ␣- and ␤-subunits with the unusual utilization of Co3⫹ ion at the active site [13], although Fe3⫹ ion is utilized for this purpose by some microorganisms [9,23].

* Corresponding author. Tel.: ⫹1-517-353-4664; fax: ⫹1-517-3534664. E-mail address: [email protected] (P. Oriel).

NHases from various microorganisms including Pseudomonas, Brevibacter, and Rhodococcus have been successfully used as industrial biocatalysts for the hydration of acrylonitrile to acrylamide (for review, see [30]). Although cloning and expression of NHase has recently been demonstrated in Rhodococcus [14,22], the available genetic tools for Escherichia coli, allowing regulated expression at high levels and gene modification through directed evolution and related methods, continue to make expression of NHase in E. coli a desirable goal. Previously efforts to express nitrile hydratase in E. coli have been disappointing due to formation of inactive and insoluble intracellular enzyme [7,13, 23]. These problems have been partially overcome by lowering the culture temperature and coexpressing an “activator protein” whose function is unclear, but may either modify the NHase active site [24] or facilitate metal ion introduction into the host cell [15] and the enzyme [29]. In this report, we describe the cloning of a Bacillus sp. BR449 gene cluster containing NHase genes, an aliphatic amidase gene, and a gene with unknown function that proved not to be required for the functional expression of the NHase genes in E. coli.

0141-0229/00/$ – see front matter © 2000 Elsevier Science Inc. All rights reserved. PII: S 0 1 4 1 - 0 2 2 9 ( 0 0 ) 0 0 2 4 8 - 9

S.-H. Kim, P. Oriel / Enzyme and Microbial Technology 27 (2000) 492–501

493

Fig. 1. Conserved amino acid sequences of the NHase ␣ subunits from various microorganisms. Amino acid sequences that were used to design primers are shown. Putative cobalt (CTLCSC) or iron (CSLCSC) binding motifs are underlined. B. smi; Bacillus smithii [11], J1-H; R. rhodochrous J1 high molecular-mass NHase [14], J1-L; R. rhodochrous J1 low molecular-mass NHase [15], N-774; Rhodococcus sp. N-774 [7], B23; Pseudomonas chlororaphis B23 [23], R312; Brevibacterium sp. R312 [19]. Identical (*), strongly similar (:), and weakly similar (.).

2. Materials and methods 2.1. Bacterial strains and cultivation The thermophilic Bacillus sp. BR449 [26], and E. coli DH5␣ [5] were grown at 50°C and 37°C, respectively, in LB medium [21]. The recombinant E. coli strains harboring the nitrile hydratase genes and/or amidase gene were grown at 32°C for optimal activity. Ampicillin and cobalt (CoCl2 䡠 6H2O) were added to the medium at 50 ␮g/ml and 20 ␮g/ml, respectively. When used, urea and methacrylamide were added at 2 mg/ml. 2.2. DNA manipulations Chromosomal DNA was isolated by the method of Marmur [18]. Isolation of plasmid, restriction enzyme digestion, DNA ligation, gel electrophoresis, and Southern hybridization were performed using standard procedures [17]. Oligonucleotides that initially had been designed for using as 3⬘-labeled probes were used as primers to obtain a 219 bp DNA probe using PCR; left primer (53 mer), 5⬘-GTA GTG GTT TGC ACT CTA TGT TCA TGT TAT CCT TGG CCG CTG CTT GGT TTA CC-3⬘, and right primer (30 mer), 5⬘-TTC TGT GCC CTC AGG TCT TTG CGG CAA TAC -3⬘. These lengthy primers span the B. smithii active site and proximal downstream sequences [11] that are conserved in NHase ␣ subunits (Fig. 1). The probe was PCR labeled with DIG-dUTP by using a DIG/GeniusTM system (Boehringer Mannheim, Indianapolis, IN, USA) as described by the manufacturer. PCR reactions (100 ␮l final volume) contained: 10 mM Tris pH 8.3, 50 mM KCl, 1.5 mM MgCl2, 0.2 mM dNTPs, 5% DMSO (dimethylsulfoxide), 0.3 ␮m of both primers, 10 ng of template DNA, and 5U Taq DNA polymerase (GIBCO/BRL, Rockville, MD, USA). Amplication was carried out using the GeneAmp 9700 thermocycler (Perkin–Elmer, Norwalk, CT, USA) for 25 cycles of 1 min at 92°C, 1 min at 45°C, and 1.5 min at 72°C. The resulting PCR product was analyzed with agarose gel electrophoresis and purified with a Wizard DNA Cleanup System (Promega, Madison, WI, USA).

2.3. Cloning of the NHase gene from BR449 Chromosomal DNA of Bacillus sp. BR449 was digested with several restriction enzymes, and the DNA fragments were separated on 0.8% agarose gels. Southern hybridization was performed using capillary transfer of the fragmented DNA to a positively charged nylon membrane (Boehringer Mannheim) followed by hybridization with the DIG-labeled 219 bp probe. A hybridizing 6.1 kb SalI band was selected for further study. To isolate this fragment, SalI-cut DNA fragments ranging in size from 5 kb to 7 kb were purified by preparative agarose gel electrophoresis. The recovered DNA was ligated with SalI-cut pBluescriptII SK vector (Stratagene, La Jolla, CA, USA), and transformed into E. coli DH5␣. The resulting ampicillin-resistant transformants were collected, and positive clones were selected by colony hybridization using the labeled 219 bp PCR product as a probe. EC454, thus obtained, was used to subclone and sequence the entire 6.1 kb SalI fragment. 2.4. Construction of the recombinant plasmids Various deletion and mutant plasmids constructed in this study are shown in Fig. 2. pSK456 was constructed by cutting a 2.6 kb PstI-SalI DNA fragment from the 6.1 kb SalI fragment and ligating into the PstI-SalI-cut pBluescriptII SK vector. pSK463 was constructed using a mutagenic PCR procedure in which PCR was carried out with two primers; a left primer (22 mer) containing a unique HindIII site located within the ␤-subunit gene, and a right primer (30 mer) that was designed to create a new SalI site at the intergenic region between the ␣-subunit and P12K genes. The oligonucleotides used for these primers were 5⬘-GAT CCG GCA TTA GTG AAG CTT G-3⬘ (left), and 5⬘-TTT CCT CGT CGA CTA CCG TAA CTT TAG CGG-3⬘ (right). The resulting 1 kb PCR product was trimmed at both ends with HindIII and SalI, and was replaced for the 1.3 kb HindIII-SalI fragment of pSK456. pSK460 containing the intact amidase gene was constructed by subcloning the 1.9 kb SpeI-EcoRI fragment into the pBluescriptII SK vector. pSK461 was constructed by inserting a 3.5 kb SpeI-SalI fragment into SpeI-SalI-cut pBlue-

494

S.-H. Kim, P. Oriel / Enzyme and Microbial Technology 27 (2000) 492–501

Fig. 2. Genetic organization of the NHase gene cluster and construction of a set of plasmids. Restriction sites discussed in the text are shown. S;SalI, Sp;SpeI, P;PstI, H;HindIII, X;XhoI, S*; newly created SalI restriction site. ⍀AGCT denotes four bases insertion (AGCT) at the HindIII site by filling-in reaction. ORFs encoding amidase, NHase ␤-, ␣-subunits, and P12K are shown with arrows indicating the gene directions. Various deletion plasmids are depicted below the restriction map. Direction of the lac promoter (Plac) in the plasmid is indicated.

scriptII SK vector. pSK464 and pSK465 that were constructed using pBluescriptII KS vector contain the same inserts as pSK456 and pSK460, respectively, but differ in orientation with respect to the lac promoter present in the vector. pSK473 was constructed by cutting the unique HindIII site located within the ␤ subunit gene of pSK463, filling in the 3⬘ recessed ends with dNTPs using DNA polymerase Klenow fragments, and rejoining the resulting blunt ends with T4 DNA ligase. pSK474 was constructed from pSK463 by deleting DNA sequence flanked by two XhoI sites that are located at the near end of the ␣ subunit gene (44 bases from 3⬘ end) and in the vector sequence. 2.5. Preparation of cell-free extracts Recombinant cells collected at stationary phase were washed once with 50 mM potassium phosphate buffer (pH 7.6) and resuspended with 1/10 culture volume of the same buffer. The cell suspension was disrupted by sonication (Ultrsonic Homogenizer, Cole–Parmer Instrument Co., Vernon Hills, IL, USA) with 6 ⫻ 30 s bursts and centrifuged at 12 000 ⫻ g for 10 min. The resulting supernatant was used as the cell-free extract.

2.6. Enzyme assays NHase was measured using cell-free extracts in 1 ml assays containing 0.1 mg total protein, 0.8 M acrylonitrile, and potassium phosphate buffer (50 mM, pH 7.6). The reaction mixture was incubated at 50°C for 5 min and stopped by adding 0.1 ml of 2 N HCl. Acrylamide formed was measured using HPLC (Waters 501 pump, 486 detector, 746 data module, Millipore, Bedfore, MA, USA) equipped with Novapak C-18 reverse phase column and 1:12 mixture of acetonitrile and 5 mM potassium phosphate buffer (50 mM, pH 2.5) was used as an elution buffer. Peaks were detected and analyzed at 200 nm with a Waters variable wavelength detector and Waters 746 data module. Amidase activity was measured in the same manner except that the formation of acrylic acid was measured using acrylamide as substrate. Concentrations of acrylamide and acrylic acid were determined using reference standards, and one unit of NHase or amidase activity was defined as one ␮mole acrylamide or acrylic acid formed in one minute, respectively. Protein concentrations were determined using the Bio-Rad Bradford protein determination kit with BSA as the protein standard.

S.-H. Kim, P. Oriel / Enzyme and Microbial Technology 27 (2000) 492–501

495

2.7. N-terminal amino acid sequencing

3.2. Nucleotide sequence analysis

Cell-free extracts of EC460 (for amidase) and EC463 (for NHase ␣ and ␤ subunits) were electrophoresed on a 12% SDS polyacrylamide gel, and proteins were transferred to a PVDF (polyvinylidene fluoride) transfer membrane (BioTrace®, Pall Gelman Laboratory, Ann Arbor, MI, USA) using an XCellII™ Blot Module (Invitrogen Corp., San Diego, CA, USA). Membrane was stained with Coomassie Blue and the corresponding protein bands were cut out and sequenced at the Macromolecular Structure, Sequencing and Synthesis Facility of Michigan State University.

We determined the nucleotide sequence of the entire 6.1-kb SalI DNA fragment of Bacillus sp. BR449. DNA sequencing analysis revealed the presence of six open reading frames (ORFs) with the same gene polarity except the second ORF (Fig. 2). ORF1 and ORF2 coded for putative proteins of 281 and 145 amino acids, respectively, and a search of the protein sequence databases yielded no closely related proteins. Distantly separated from ORF2, ORF3 encoded a 348 amino acid protein with a molecular weight of 38.6 kDa. The deduced amino acid sequence of the protein showed a strong similarity to the nitrilase/aliphatic amidase enzyme family [25], and shared 89% identity with an aliphatic amidase from B. stearothermophilus BR388 [2]. One hundred twenty seven nucleotides downstream from this amidase gene, ORF4 encoding a 229 amino acid protein (MW 26.5 kDa) was identified. Its translated amino acid sequence showed 83% and 37% identity with the B. smithii NHase ␤-subunit and R. rhodochrous J1 low molecular-mass NHase (L-NHase) ␤-subunit, respectively [11,13]. Further, 25 nucleotides downstream of the ␤-subunit gene, ORF5 encoding NHase ␣-subunit (214 aa, MW 24.6 kDa) was identified with 81% and 56% sequence identity to the B. smithii NHase ␣-subunit and R. rhodochrous J1 liter-NHase ␣-subunit. The amino acid sequence of the BR449 ␣-subunit also revealed that it contained a characteristic cobaltbinding motif, CTLCSC (residue 116 –121, Fig. 3), which differs from the proposed iron-binding motif of other NHase ␣-subunits, CSLCSC (Fig. 1) [9]. ORF6 was located 15 nucleotides downstream of the ␣-subunit gene, encoding a 101-aa protein (MW 12 kDa) designated P12K. Analysis of its amino acid sequence showed a weak but significant homology to R. rhodochrous J1 NhhG (36% identity) [13] and NhlE (27% identity) [13] whose coding genes are located immediate downstream of H-NHase (high molecularmass NHase) and L-NHase ␣ subunit genes, and whose functions are unknown (Fig. 5A, 5B). All of these BR449 genes have the similar Shine-Dalgarno (SD, Fig. 3) sequence, (A or G)GGAGG, 6 – 8 nucleotides upstream of their corresponding ATG start codons. The short spacings between the three downstream genes suggest that the ␤-subunit, ␣-subunit and P12K proteins are translated from a single mRNA whose transcription starts upstream of the ␤-subunit gene.

2.8. DNA sequencing, sequence analysis Determination of the nucleotide sequence of the 6.1 kb BR449 DNA fragment was done at the Michigan State University DNA Sequencing Facility. Primer design, homology searches, sequence analyses, and other characterizations were performed using many molecular biology online services that are listed on the major Molecular Biology Servers, CMS-SDSC (http://www.sdsc.edu/ResTools/ cmshp.html), ExPASy (http://www.expasy.ch/), and NCBI (http://www.ncbi.nlm.nih.gov/). The nucleotide sequences of the amidase, NHase ␤-, ␣-subunit, and P12K genes presented in this study have been assigned GenBank Accession Number, AF257487, AF257488, AF257489, and AF257490, respectively.

3. Results 3.1. Cloning of nitrile hydratase gene of Bacillus sp. BR449 On the basis of the alignment of conserved amino acid sequences of NHase ␣ subunits from Rhodococcus, Pseudomonas, Brevibacterium, and Bacillus (Fig. 1), we initially designed a 53-mer probe that corresponded to the conserved amino acid sequence of the Bacillus smithii NHase ␣ subunit [11]. We also used this DNA segment as a left primer to obtain a larger labeled probe using PCR, together with a similarly designed right primer. The resulting PCR product had the expected size of 219 bp. A Southern blot of Bacillus sp. BR449 DNA cleaved with various restriction enzymes was probed with the 219 bp PCR product that had been labeled with DIG-dUTP. A single 6.1 kb SalI fragment was found to hybridize strongly, and was recovered using preparative agarose gel electrophoresis. The purified DNA was ligated into the pBluescriptII SK vector and transformed into E. coli DH5␣. Among ampicillin-resistant colonies screened, a clone hybridizing with the 219 bp probe was designated EC454.

3.3. NHase activity in the recombinant E. coli strains Attempts to express the NHase genes under the control of an IPTG-inducible lac promoter resulted in a very weak NHase activity (less than 5 U/mg protein). Lowering the growth temperature to 32°C did not improve expression. Instead, we utilized a constitutive expression system in which E. coli DH5␣ and pBluescriptII were used as a host and an expression vector, both of which lack the lacI or lacIq gene required for repression of the lac promoter.

496

S.-H. Kim, P. Oriel / Enzyme and Microbial Technology 27 (2000) 492–501

Fig. 3. Nucleotide sequences and the deduced amino acid sequences for the amidase, NHase ␤-, ␣-subunits, and P12K genes. The deduced amino acid sequences of the structural genes are shown in one-letter code. Potential ribosome binding sequences are indicated as SDs. The cobalt-binding motif (CTLCSC) in the ␣-subunit is underlined. The sequence is numbered from SpeI site upstream of the amidase gene.

S.-H. Kim, P. Oriel / Enzyme and Microbial Technology 27 (2000) 492–501

497

Table 1 Nitrile hydratase and amidase activities of the recombinant E. coli strains

EC454 EC456 EC460 EC461 EC463 EC464 EC465 EC473 EC474

NHase activity (units/mg protein)

Amidase activity (units/mg protein)

20 ⫾ 2.2 42 ⫾ 3.5 ND 22 ⫾ 2.0 38 ⫾ 3.0 12 ⫾ 1.0 NT ND ND

3.2 ⫾ 0.5 ND 16 ⫾ 2.5 4.8 ⫾ 1.1 ND NT ND NT NT

*E. coli recombinant strains were grown in LB-Amp medium supplemented with cobalt (20 ␮g/ml), and the cell extracts were assayed for NHase and amidase activities at 50°C for 5 min. The averaged NHase and amidase activities were obtained each from two independent experiments. ND, not detected, NT, not tested.

Various deletion plasmids were thus constructed (Fig. 2), and their corresponding NHase and amidase activities in E. coli were determined (Table 1). When expressed in E. coli, pSK454 which contains the 6.1 kb BR449 DNA exhibited both NHase (20 U/mg protein) and amidase (3.2U/mg protein) activities, confirming the presence of these genes within the fragment. For NHase activity, cobalt ion was required in the culture medium (Table 3), and substitution of other metal ions such as Fe2⫹, Fe3⫹, Mn2⫹, Cu2⫹, and Ni2⫹ had no effect (data not shown). We also observed that EC454 growing at 37°C exhibited unstable NHase gene expression, but lowering growth temperature to 32°C provided optimal and stable NHase activity (data not shown). pSK456 contains a truncated amidase gene but has the intact genes for ␤-subunit, ␣-subunit, and P12K protein. EC456 strain that contains pSK456, showed no detectable amidase activity, but expressed over 2-fold higher in NHase activity than the parental strain EC454. Because the gene encoding P12K is likely to be co-transcribed with the ␤- and ␣-subunit genes, its function in influencing NHase activity is of interest. pSK463 was thus constructed from pSK456 by deleting the region encoding the putative P12K protein. We found no significant differences in NHase activity (42U/mg protein vs. 38U/mg protein) and enzyme stability (data not shown) between the cell-free extracts of EC456 and EC463. However, mutation in either ␤ or ␣ subunit gene caused a disruption in NHase activity, exemplified by pSK473 and pSK474, both of which were derived from pSK463. pSK473 contains an insertional mutation (four bases, AGCT) at the unique HindIII site located in the middle of the ␤ subunit gene, whereas a deletion plasmid pSK474 lacks the sequence for the C-terminal 15 amino acids of the ␣ subunit. To investigate the outcome of a coupled reaction, we constructed pSK461 in which both amidase and NHase genes are present. When acrylonitrile (0.8 M) was used as a substrate, acrylamide but no detectable amount of acrylic

Fig. 4. Expression of amidase, NHase ␣- and ␤-subunit proteins in the recombinant strains. Cell-free extracts (10 ␮g total protein) were electrophoresed on a 12% SDS-polyacrylamide gel. Lane 1; molecular mass standard, Lane 2; negative control (E. coli DH5␣ containing pBluescriptII SK), Lane 3-Lane 7; EC461, EC456, EC464, EC460, and EC465, respectively.

acid was formed (22U/mg protein of NHase activity). However, when acrylamide was used as a substrate, a significant level of acid was formed (4.8U of amidase activity). This result indicates that EC461 expresses both active NHase and amidase, but the amidase seems to be inhibited by acrylonitrile. Another construct, pSK460, contains the intact amidase gene and the truncated ␤-subunit gene. When expressed in E. coli, its amidase activity was found 3 to 4 fold higher than that of EC461. We tested whether the transcription initiation of the NHase/amidase gene cluster takes place other than the lac promoter. pSK464 and pSK465 contained the same inserts as pSK456 and pSK460, respectively, but differ in orientation of the inserts with respect to the lac promoter. E. coli harboring pSK465 showed an abolished level of the amidase activity, suggesting that most of transcription for the amidase gene initiates from the lac promoter. However, judging from the NHase activity in EC464 (12U), some transcripts are likely to be initiated from the intergenic region between the amidase and ␤-subunit genes. 3.4. Production of amidase and NHase To detect the gene products in the various recombinant E. coli strains, we analyzed cell extracts by SDS-PAGE (Fig. 4). Proteins of 27 kDa and 25 kDa, of the similar molecular mass as the NHase ␤ and ␣ subunit proteins, respectively, were observed from the cell-free extracts of EC461 (lane 3), EC456 (lane 4), and EC464 (lane 5), all of which contained the intact ␤- and ␣-subunit genes. Al-

498

S.-H. Kim, P. Oriel / Enzyme and Microbial Technology 27 (2000) 492–501 Table 3 Effect of cobalt addition on NHase activity

Table 2 Temperature dependence of NHase and amidase activities (min)

5 10 20

Acrylamide formed* (␮moles/mg protein)

Acrylic acid formed** (␮moles/mg protein)

30°C

50°C

30°C

50°C

145 ⫾ 4 270 ⫾ 12 490 ⫾ 18

220 ⫾ 8 235 ⫾ 11 245 ⫾ 6

43 ⫾ 2 79 ⫾ 2 155 ⫾ 7

78 ⫾ 4 140 ⫾ 8 265 ⫾ 6

*EC456 cell extract was preincubated at designated temperatures for 15 min, and acrylonitrile (final conc. 03M) was added to start the reaction. Aliquots were removed and quenched at different time points, and were assayed for acrylamide formation. **EC460 cell extract was used as the amidase source, and amount acrylic acid formed was measured using 0.1M acrylamide as a substrate. The averaged NHase and amidase activities were obtained each from two independent experiments.

Cell culture* 2⫹

Co added (mM) NHase activity

0 ND

0.1 23 ⫾ 1

Cell extract** 0.5 45 ⫾ 2

0.1 41 ⫾ 4

0.5 49 ⫾ 2

*EC456 was grown in LB-Amp medium supplemented with 0.1 mM or 0.5 mM of CoCl2 䡠 6H2O, and the cell extracts were assayed for NHase activity at 30°C for 10 min. NHase activity; units per mg protein. **Cell extracts were obtained from EC456 which had been grown in LB-Amp medium supplemented with 0.1 mM cobalt. Additional cobalt (0.1 mM or 0.5 mM) was added to the extracts which had been adjusted to contain 0.1 mg total protein per ml, and after preincubation for 10 min at room temperature, the cell extracts were assayed at 30°C for 10 min. The averaged NHase activities were obtained from two independent experiments. ND, not detected.

3.6. Effects of cobalt and amide inducers though these cells also contained the downstream gene encoding the putative P12K protein, we could not detect a protein band corresponding to that protein. Consistent with its NHase activity, EC464 synthesized a small, but significant amount of NHase. A protein band corresponding to the amidase (38.6 kDa) was observed in EC461 and EC465 (Fig. 4, lane 6). As expected from its NHase activity, EC465 (lane 7) did not produce a detectable level of amidase. EC463, which only differs with EC465 in lacking the P12K gene, expressed similar amount of ␤ and ␣ subunit proteins with identical mobility on a gel (data not shown). These protein bands (EC463 ␤ and ␣ subunits) and amidase band (EC460 amidase) were cut out and determined for the N-terminal amino acid sequencing. The results were in agreement with the deduced sequences except that the Nterminus methionine was not observed for the ␤-subunit, presumably as a result of processing. We were unable to confirm the translational start for the P12K gene because of the difficulty in detecting its gene product on a SDS-PAGE gel. 3.5. Temperature optimum The temperature optimum of wild-type BR449 NHase was previously determined to be 50 to 55°C [26]. We examined the thermostability of NHase and amidase that are expressed in E. coli. Cell-free extracts of EC456 and EC460 were used for the NHase and amidase sources, respectively. As seen in Table 2, NHase activity showed its highest rate at 50°C, but the enzyme became inactivated during reactions longer than 5 min. At 30°C, such an abrupt inactivation did not occur, and linear product accumulation was found. Preincubations at 50°C in the absence of acrylonitrile did not affect the high initial reaction rate shown at 50°C (data not shown), suggesting that the inactivation resulted from enzyme inactivation by acrylonitrile at high temperature. In contrast, amidase in EC460 demonstrated undiminished activity 50°C for extended periods, and was approximately 2-fold higher at 50°C than at 30°C.

We examined stimulation of NHase activity by adding an excess amount of cobalt to the culture medium. When cobalt was added at 5-fold higher concentration (0.5 mM CoCl2 䡠 6H2O) than routinely used, the cells exhibited a higher NHase activity by 2-fold (Table 3). Similarly, direct treatment of additional cobalt ion to the cell-free extract also significantly enhanced NHase activity (Table 3). It was reported that amide inducers such as methacrylamide and urea were required for subunit assembly and gene expression of R. rhodochrous J1 NHase, respectively [14,22]. We tested the effects of such inducers by adding urea or methacrylamide to the growing EC456, followed by assaying for the NHase activity in its cell-free extract. We found no significant difference in activity between cultures with and without inducers (data not shown).

4. Discussion Expression of the nitrile hydratase (NHase) genes from Bacillus sp. BR449 in E. coli to produce active enzyme in significant amounts was accomplished using a suboptimal growth temperature and addition of cobalt ions to the medium. Utilization of a reduced growth temperature is known to help reduce formation of insoluble aggregates of recombinant proteins [6]. Interestingly, inclusion of the two intact ␣- and ␤-subunit genes was sufficient for E. coli to express functional NHase (Table 1. EC463). Requirement of coexpression of the genes in the flanking regions has been reported for the functional expression of NHase in E. coli [24,29]. However, expression of BR449 amidase and NHase in E. coli did not require co-expression of the upstream genes (ORF1 and ORF2), nor was enzyme activity enhanced by their presence. This is consistent with the constitutive expression of NHase and amidase observed in the parental Bacillus sp. BR449 thermophile [26]. A downstream linked gene, P12K, which showed a significant similarity with R. rhodochrous J1 nnhG and nnlE in their amino

S.-H. Kim, P. Oriel / Enzyme and Microbial Technology 27 (2000) 492–501

499

Fig. 5. (A) Genetic organization of the NHase gene clusters from the various NHase producing organisms. The direction and extent of the genes are indicated by arrows. Structural genes for amidase (ami), NHase ␣- and ␤-subunits (␣ and ␤) are shown, and the potential regulatory genes for NHase expression or activation are depicted in gray arrows. BR449; Bacillus sp. BR449, J1-H; R. rhodochrous J1 high molecular-mass NHase [14], J1-L; R. rhodochrous J1 low molecular-mass NHase [15], 5B; Pseudomonas putida 5B [29], N-771; Rhodococcus sp. N-771 [24], B23; Pseudomonas chlororaphis B23 [23].

acid sequences and gene location (Fig. 5), also was not required for NHase expression in E. coli. Although we could not detect the P12K protein on a SDS-PAGE gel, its deduced amino acid sequence suggests that it is a soluble protein. The possibility of formation of a fusion protein in E. coli with upstream ␣ subunit protein was excluded because deletion of the P12K gene had no effect on the mobility of ␣ subunit band on a gel (data not shown). It may be required to examine the mRNA and gene product of the P12K gene in the parental Bacillus sp. BR449 before its role, if any, in NHase expression is clarified. The requirement for cobalt in the growth medium for enzyme activity and the presence of a characteristic cobalt binding motif in ␣ subunit are consistent with the identification of BR449 NHase as a cobalt-containing enzyme. NHases from three other moderate Bacillus thermophiles

have been reported [3,27,28]. No information is yet available regarding encoding NHase genes or gene expression in E. coli for these isolates. For BR449 NHase, the requirement for cobalt in the growth medium together with the high enzyme levels observed by SDS-PAGE and cobalt stimulation of cell extracts suggests that intracellular availability of cobalt ion to the NHase during synthesis may be a limiting factor in formation of active enzyme in the E. coli recombinant. We are presently examining new approaches to further increase provision of cobalt ion to the BR449 recombinant NHase. Closely spaced NHase gene clusters with varied arrangement of the NHase subunit genes and containing an adjacent amidase have been observed for Rhodococcus sp. N-774, Pseudomonas choraphnis B23, Brevibacter sp. 316, and an uncharacterized Rhodococcus sp. (reviewed in [30]). These

500

S.-H. Kim, P. Oriel / Enzyme and Microbial Technology 27 (2000) 492–501

clusters have been interpreted as co-transcribed operons [19,20]. Although the amidase/hydratase gene arrangement for Bacillus sp. BR449 resembles that of the undesignated Rhodococcus sp., the BR449 amidase is a member of the nitrilase-related family [25] whereas the amidases associated with NHases in the organisms examined to date belong to the amidase family related to aspartic proteinases [12]. Although the amidase and NHase genes are proximal and seem to have identical Bacillus-type Shine-Dalgarno sequences [4], the significant NHase expression in reverse orientation to the vector promoter in EC464 (Table 1.) and significant spacing between amidase and NHase genes raises the possibility of existence of a promoter upstream to the NHase genes that facilitates independent expression of the NHase genes. The utilization of amidases from different families and the varied arrangement of amidase and NHase ␣ and ␤ subunit genes in different bacteria suggests that like other catabolic pathways [16], the nitrile pathway was assembled utilizing pre-existing genes. Although we cannot at this time rule out the possible existence of additional NHase genes in BR449, we consider the possibility unlikely based on the observation of only one hybridizing fragment. High level expression of the BR449 NHase and amidase genes will allow convenient examination of NHase and amidase structure and function as well as providing an opportunity for improvements in their enzymatic properties using genetic alteration. Important in this regard will be efforts to improve resistance to inactivation by acrylonitrile at elevated temperature.

Acknowledgments We thank SNF Floerger, France, for support of these nitrile hydratase studies.

References [1] Asano Y, Tani Y, Yamada H. A new enzyme, nitrile hydratase, which degrades acetonitrile in combination with amidase. Agric Biol Chem 1980;44:2251–2. [2] Cheong T, Oriel P. Cloning of a wide-spectrum amidase from Bacillus stearothermophilus BR388 in Escherichia coli and marked enhancement of amidase expression using directed evolution. Enzyme Microb Technol 2000;26:152– 8. [3] Cramp R, Cowan DA. Molecular characterization of a novel thermophilic nitrile hydratase. Biochem Biophys Acta 1999;1431:249 – 60. [4] Hager PW, Rabinowitz JC. Translational specificity in Bacillus subtilis. In: Dubnau DA, editor. The molecular biology of Bacilli. New York: Academic Press, Vol. II, 1985, p. 1–30. [5] Hanahan D. Studies on transformation of Escherichia coli with plasmids. J Mol Biol 1983;166:557– 80. [6] Hannig G, Makrides SC. Strategies for optimizing heterologous protein expression in Escherichia coli. Trends Biotechnol 1998;16:54 – 60.

[7] Hashimoto Y, Nishiyama M, Horinouchi S, Beppu T. Nitrile hydratase gene from Rhodococcus sp. N-774 requirement for its downstream region for efficient expression. Biosci Biotech Biochem 1994; 58:1859 –65. [8] Hashimoto Y, Nishiyama M, Ikehata O, Horinouchi S, Beppu T. Cloning and characterization of an amidase gene from Rhodococcus species N-774 and its expression in Escherichia coli. Biochimica et Biophysica Acta 1991;1088:225–33. [9] Huang W, Jia J, Nelson M, Schneider G, Beppu T. Crystal structure of nitrile hydratase reveals a novel iron center in a novel fold. Structure 1997;5:691–9. [10] Hughes J, Armitage YC, Symes KC. Application of whole cell Rhodococcal biocatalysts in acrylic polymer manufacture. Antonie van Leeuwenhoek 1998;74:107–18. [11] Ishii T, Yamaga Y, Takashima Y. Nitrile hydratase, its gene and its use. Japanese Patent; JP 1997248188-A 1998, Sumitomo Chem. Co. LTD. [12] Kobayashi M, Fujiwara Y, Goda M, Komeda H, Sakayu S. Identification of active sites in amidase: evolutionary relationship between amide bond and peptide bond-cleaving enzymes. Proc Natl Acad Sci USA 1997;94:11986 –91. [13] Kobayashi M, Nishiyama M, Nagasawa T, Horinouchi S, Beppu T, Yamada H. Cloning, nucleotide sequence and expression in Escherichia coli of two cobalt-containing nitrile hydratase genes from Rhodococcus rhodochrous J1. Biochimica et Biophysica Acta 1991; 1129:23–33. [14] Komeda H, Kobayashi M, Shimizu S. Characterization of the gene cluster of high-molecular-mass nitrile hydratase (H-NHase) induced by its reaction product in Rhodococcus rhodochrous J1. Proc Natl Acad Sci USA 1996;93:4267–72. [15] Komeda H, Kobayashi M, Shimizu S. A novel transporter involved in cobalt uptake. Proc Natl Acad Sci USA 1997;94:36 – 41. [16] Maeda M, Ohta Y, Kang S-K, Lee L, Kudo T. The evolution of metacleavage pathway for degradation of polychlorinated biphenyl (PCB)/ biphenyl in gram-negative bacteria. In: Horikoshi K, Fukuda M, Kudo T, editors. Microbial diversity and genetics of biodegradation. Tokyo: Sci Soc Press, 1997. p. 149 – 65. [17] Maniatis T, Fritsch EF, Sambrook J. Molecular cloning: a laboratory manual. 2nd edition. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory. [18] Marmur J. A procedure for the isolation of deoxyribonucleic acid from microorganisms. J Mol Biol 1961;3:208 –11. [19] Mayaux J-F, Cerbelaud E, Soubrier F, Faucher D, Petre D. Purification, cloning, and primary structure of an enantiomer-selective amidase from Brevibacterium sp. strain R312: structural evidence for genetic coupling with nitrile hydratase. J Bacteriol 1990;172:6764 – 73. [20] Mayaux J-F, Cerbelaud E, Soubrier F, Yeh P, Blanche F, Petre D. Purification, cloning, and primary structure of a new enantiomerselective amidase from a Rhodococcus strain: structural evidence for a conserved genetic coupling with nitrile hydratase. J Bacteriol 1991; 173:6694 –704. [21] Miller J. Experiments in molecular genetics. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory. [22] Mizunashi W, Nishiyama M, Horinouchi S, Beppu T. Overexpression of high-molecular-mass nitrile hydratase from Rhodococcus rhodochrous J1 in recombinant Rhodococcus cells. Appl Microbiol Biotechnol 1998;49:568 –72. [23] Nishiyama M, Horinouchi S, Kobayashi M, Nagasawa T, Yamada H, Beppu T. Cloning and characterization of genes responsible for metabolism of nitrile compounds from Pseudomonas chlororaphis B23. J Bacteriol 1991;173:2465–72. [24] Nojiri M, Yohda M, Odaka M, et al. Functional expression of nitrile hydratase in Escherichia coli: requirement of a nitrile hydratase

S.-H. Kim, P. Oriel / Enzyme and Microbial Technology 27 (2000) 492–501 activator and post-translational modification of a ligand cysteine. J Biochem 1999;125:696 –704. [25] Novo C, Tata R, Clemente A, Brown PR. Pseudomonas aeruginosa aliphatic amidase is related to the nitrilase/cyanide hydratase enzyme family and Cys166 is predicted to be the active site nucleophile of the catabolic mechanism. FEBS Lett 1995;367:275–9. [26] Padmakumar R, Oriel P. Bioconversion of acrylonitrile to acrylamide using a thermostable nitrile hydratase. Appl Biochem Biotech 1999: 77–79:671–9. [27] Pereira RA, Graham D, Rainey FA, Cowan DA. Novel thermophilic nitrile hydratase. Extremophiles 1998;2:347–58.

501

[28] Takashima Y, Yamaga Y, Mitsuda S. Nitrile hydratase from a thermophillic Bacillus smithii J. Industrial Microbiol. Biotechnol 1998; 20:220 – 6. [29] Wu S, Fallon RD, Payne MS. Over-production of stereoselective nitrile hydratase from Pseudomonas putida 5B in Escherichia coli: activity requires a novel downstream protein. Appl Microbial Biotechnol 1997;48:704 – 8. [30] Yamada H, Kobayashi M. Nitrile hydratase and its application to industrial production of acrylamide. Biosci Biotech Biochem 1996; 60:1391– 400.