A gene (pueA, polyurethane esterase A) encoding an extracellular polyurethanase (PueA) was cloned from Pseudomonas chlororaphis into Escherichia coli. The enzyme secreted from E. coli showed esterase activity when assayed with p-nitrophenyl acetate. Subcloning of a 3.2-kb SalI–EcoRI fragment into a T7 RNA polymerase expression vector (pT7-6) produced a 35S-labeled protein of 65 kDa. Nucleotide sequencing of pueA showed an open reading frame encoding a 65-kDa protein of 617 amino acid residues, with the serine hydrolase consensus sequence GXSXG. PueA was over-expressed using the pT7-6 vector transformed into E. coli BL21(DE3) and was purified in one step using Sephadex G-75.
As polyester-based polyurethane (PU) waste accumulates, methods are being sought to degrade and recycle this man-made, recalcitrant polymer. Biodegradation is an attractive avenue of investigation, since it has been discovered that microbial sources (bacteria [1–3] and fungi [4,5]) can utilize PU for growth. Pseudomonas chlororaphis can grow using PU as its sole carbon and energy source . PU is used in the manufacturing of adhesives, thermal insulation, foam padding and paint , and is best described as a repeating carbamate polymer, -(R1-NH-CO2-R2)-, which is susceptible to hydrolysis by esterase or protease attack.
Extracellular esterase activity is the primary activity found in both bacteria and fungi, which allows them to grow on PU. Six bacterial proteins [1,2,7–9] and one fungal protein  have been purified and five of these enzymes, polyurethanases, demonstrate esterase activity with little or no proteolytic activity. Recently, two of the bacterial polyurethanases have been cloned and expressed in Escherichia coli[10,11]. One of these enzymes is secreted from Pseudomonas fluorescens, while the other is found in Comamonas acidovorans and is membrane-associated. The amino acid sequence derived from the nucleotide sequence of the Comamonas polyurethanase has revealed the serine hydrolyase motif, G-X-S-X-G, and showed a maximum amino acid sequence identity of 34% with carboxylesterases and acetylcholinesterases in GenBank. In this study, a gene (pueA) encoding a secreted polyurethanase from P. chlororaphis has been cloned, sequenced and its encoded protein over-expressed.
Materials and methods
Chemicals and bacterial strains
Amplify™ fluorographic reagent, [35S]methionine (18.5 MBq, 500 μCi per 50 μl) and an ECL nucleic acid labeling kit were obtained from Amersham (Arlington Heights, IL, USA). λZAPII and Gigapack III Gold packaging extract were obtained from Stratagene (La Jolla, CA, USA). Enzyme substrates used were p-nitrophenyl acetate (ICN, Costa Mesa, CA, USA) and hide powder azure (Calbiochem, La Jolla, CA, USA). The nylon membrane for blotting was S&S Nytran from Schleicher and Schuell (Keene, NH, USA). Restriction enzymes and T4 DNA ligase were obtained from Sigma (St. Louis, MO, USA) or Gibco BRL (Grand Island, NY, USA). RPMI medium 1640 (with glutamine, without methionine) and 1-kb DNA ladder were obtained from Gibco BRL (Grand Island, NY, USA). All other buffers and salts were reagent grade or better. P. chlororaphis was isolated from a microbial consortium obtained from the Naval Research Laboratory, Washington, DC, USA, by using its ability to degrade a water dispersible PU, Impranil DLN. P. chlororaphis was grown in Luria–Bertani (LB) medium at 30°C with constant shaking at 180 rpm. E. coli XL1-Blue MRF′ cells were obtained as part of the λZAPII kit and were grown in LB medium at 37°C with shaking at 220 rpm. E. coli BL21(DE3) and plasmids pT7-5 and pT7-6 were donated by Dr. Jessup Shively of Clemsen University.
Construction and screening of the genomic library
Chromosomal DNA was isolated by the method of Marmur . The DNA was extracted two times with ether to remove trace organics . λZAPII phage library was constructed as previously described . Recombinant phages were screened by plating in soft agar overlays containing Impranil as previously described . Plasmids containing pueA were excised and rescued from the λ library using the Stratagene protocol. The E. coli XL1-Blue were plated on LB agar plates supplemented with 1.5% (v/v) Impranil DLN dispersion and 60 μg ml−1 ampicillin. Plasmid DNA from the positive clones was isolated using alkaline lysis  and the restriction endonucleases EcoRI, KpnI, SalI, SstI, ClaI, BglII, BamHI, NotI, HindII, XbaI, XhoI and EcoRV were tested. Results were detected using agarose gel electrophoresis with Tris–acetate buffer. Southern blot experiments were conducted as previously described .
Expression and purification of the polyurethanase
The 3.2-kb SalI–EcoRI fragment of the insert from pPU2 was subcloned into the expression vectors pT7-5 and pT7-6, each with a T7 RNA promoter but at opposite ends of the inserted DNA. These plasmids were then used to transform competent E. coli C600 cells that contained the pGP1-2 plasmid encoding the T7 RNA polymerase. The procedure for the 35S-labeling was previously described . BL21(DE3) cells were made competent with dimethyl sulfoxide and transformed with the expression vector pT7-6 containing pueA as a 3.2-kb SalI–EcoRI insert. These cells (25-ml culture) were grown at 37°C to an OD600=0.6 in LB medium containing 60 μg ml−1 ampicillin and 60 μg ml−1 chloramphenicol. At this time, isopropyl-β-D-thiogalactoside (IPTG, 200 mg ml−1 stock) was added to a final concentration of 1 mM. After 2 h, the cells were harvested by centrifugation and resuspended in 7 ml of 50 mM potassium phosphate buffer pH 7.0. The cells were lysed by adding 10% sodium dodecyl sulfate (SDS) to a final concentration of 1% SDS and warming the solution briefly at 50°C to solubilize the SDS. The lysate was treated with 0.1 mg DNase and allowed to stand overnight at 4°C to precipitate the SDS. One ml of the cell lysate was applied to a Sephadex G-75 column (45.0×1.0 cm) at a flow rate of 0.3 ml min−1 and 1-ml fractions were collected. One main peak was observed after the void volume as detected by SDS–polyacrylamide gel electrophoresis (SDS–PAGE) and this was the purified PueA. Polyurethanase activity was detected as a zone of clearing using a radial diffusion assay (RDA). The gel for the assay was made using 1.5% (w/v) agarose and 1.5% (v/v) Impranil DLN dispersion in 10 mM potassium phosphate pH 7.0. The wells held a maximum volume of 100 μl, and usually a zone of clearing would be evident in 6 h at 23°C which was dependent on enzyme concentration. For protease activity , insoluble hide powder azure (25 mg) was used as the substrate in 3 ml of 50 mM potassium phosphate pH 7.0, which was monitored at 595 nm. For esterase activity , 10 μl of p-nitrophenyl acetate (60 μg per 100 ml acetonitrile) was used as the substrate in a reaction mixture of 2 ml 50 mM potassium phosphate pH 7.0 and 1 ml acetonitrile. The reaction was monitored at 420 nm. Protein concentrations were estimated by the method of Lowry  using the Bio-Rad DC Protein Assay. SDS–PAGE was performed as described by Laemmli  using a 10% gel. Prestained standards (Gibco BRL, Sigma) or Dalton Mark VII-L standards (Sigma) were used. Following electrophoresis, gels were stained with Coomassie blue R-250. For the 35S-labeling experiment, the gel was fixed for 30 min, soaked in Amplify™ for 30 min, dried and autoradiographed with Kodak X-OMAT XAR-5, X-ray film for 45 h at −80°C.
Nucleotide analysis and N-terminal analysis
The nucleotide sequence and N-terminal protein sequence were determined by automated microsequencing performed in the Sequencing Laboratory of the Biotechnology Center at the University of Illinois at Urbana-Champaign. Both DNA strands were sequenced and the alignment was deduced using Sequencher 3.1™ (Gene Codes Corporation, Ann Arbor, MI, USA). The open reading frame (ORF) and the amino acid sequence were also determined using Sequencher 3.1™. Protein blotting for N-terminal analysis was performed using 25 mM Tris, 192 mM glycine, 20% methanol (v/v), pH 8.3, transfer buffer and a mini transblot electrophoretic transfer cell (Bio-Rad, Hercules, CA, USA). Sequi-Blot PVDF membrane was used (Bio-Rad).
Screening the λ library for pueA
The genomic library prepared by ligating 4–10-kb EcoRI fragments of P. chlororaphis chromosomal DNA into the EcoRI site of λZAPII was screened for polyurethanase activity based on the ability of the enzyme to hydrolyze PU. The hydrolysis was evident as a readily distinguishable plaque which was more translucent than neighboring plaques since the milky-white PU, Impranil, was hydrolyzed. A total of 4000 recombinants were screened, and 20 polyurethanase-positive clones were isolated. Although plaques were evident after an overnight incubation at 37°C, the polyurethanase activity did not appear until the plates were transferred to 23°C for several hours.
Cloning pueA in E. coli
Using the bacteriophage from several representative clones, plasmid excision and rescue were carried out and the resulting pBluescript plasmids were harbored in E. coli XL1-Blue. E. coli XL1-Blue containing pPU2 (5.9-kb insert of chromosomal DNA in pBluescript) were plated along with P. chlororaphis and a control of E. coli XL1-Blue cells containing pBluescript on an LB/Impranil plate (Fig. 1A). There were strong zones of clearing for the E. coli containing pPU2 and P. chlororaphis, while the E. coli containing pBluescript showed no clearing. Using pPU2, the restriction endonucleases KpnI, SalI and SstI were found to cut the 5.9-kb insert within a narrow region that divided the insert approximately in half. ClaI, BglII, BamHI, NotI, HindIII, XbaI, XhoI and EcoRV did not cut the 5.9-kb insert. To insure that the insert in pPU2 was from the P. chlororaphis chromosome, Southern blot hybridization was conducted (Fig. 1B). A chemi-luminescent probe was constructed from the 5.9-kb insert by labeling it with horseradish peroxidase. The probe hybridized strongly to the 5.9-kb insert and to the same-sized fragments from the chromosomal DNA indicating that the insert had originated from P. chlororaphis and that only one copy of the gene was present on the chromosome.
Nucleotide sequence of pueA and the translated ORF
After sequencing pueA, an ORF of 1851 nucleotides encoding a protein, PueA, of 617 amino acids with a molecular mass of 65 kDa was found (Fig. 2). In order to confirm that the correct ORF had been found, the protein was blotted on PVDF membrane and its N-terminus was determined to be MGVFDYKNFTAS, which is an exact match for the putative ORF. Further analysis of the ORF showed two regions of special interest, one a putative active site and the other a region found to be conserved in secreted proteins that lack an N-terminal leader sequence (Fig. 2, underlined). The serine hydrolase motif, G-X-S-X-G, was found at Gly205, while the secretion consensus signal was found at Gly382. The active site and secretion signal sequences align well with conserved regions within extracellular lipases found by using a BLASTP search of GenBank at the National Center for Biotechnology Information (Fig. 3). Not only are the reported sequences highly conserved, but they are also all located within the same region of the lipase. The active site residues are found from residue 202 to 211 and the secretion signal residues are found from residue 363 to 400 for all the lipases (Fig. 3).
Subcloning of pueA and expression
In order to narrow the search for the polyurethanase gene, the 5.9-kb insert was digested with SalI which yielded 2.7-kb and 3.2-kb fragments which were subcloned into the expression vectors pT7-5 and pT7-6. A protein of approximately 65 kDa was expressed by the pT7-6 plasmid as observed by using the radiolabel, [35S]methionine (Fig. 4B, lane 2). No expression was observed for the pT7-5 plasmid (Fig. 4B, lane 1) which indicates that the gene is read from SalI to EcoRI. In order to attempt to classify the extracellular protein PueA as an esterase or a protease, two substrates were tested. No activity was observed for extracellular protein using hide powder azure, which is a known substrate for the protease trypsin. Esterase activity was observed using p-nitrophenyl acetate for the case of extracellular protein from E. coli harboring pPU2 which was 1.5 times greater than the activity for the control cells. Over-expression of the enzyme, by transforming E. coli BL21(DE3) with the pT7-6 expression vector containing the 3.2-kb SalI–EcoRI insert, facilitated obtaining enough protein for conducting N-terminal sequencing. Induction of the protein was achieved with 1 mM IPTG for 2 h. In one chromatographic step (Sephadex G-75), good yields of the purified enzyme were obtained (Fig. 4A, lane 3). To confirm expression of active PueA, a RDA was conducted (Fig. 4A). A RDA using spirit blue agar (Difco) and Wesson oil confirmed that lipase activity was also present (not shown).
Evidence has been presented which indicates the polyurethanase, PueA, from P. chlororaphis is a lipase. Using the deduced amino acid sequence, a BLASTP search of GenBank found five lipases with 60–80% identity to PueA. The active site and secretion signal regions are also highly conserved in these lipases. To further corroborate the conclusion PueA is a lipase, the polyurethanase enzyme from P. fluorescens which had been cloned  has now been sequenced (accession #AF144089, enzyme designated PulA for PU lipase A) and it showed amino acid sequence identity of 50–70% to lipases in GenBank. Comparing the amino acid sequence of PulA to PueA yielded 56% identity.
The ORF of pueA was confirmed by comparison of the deduced N-terminal amino acid sequence and the experimentally determined N-terminus, which yielded an exact match. A search upstream at nucleotides 162–167 revealed a sequence AAGAGG, which is the probable ribosomal binding site. The deduced amino acid sequence yields a molecular mass for PueA of 65 kDa that corresponds exactly with the [35S]methionine-labeled protein observed and the protein observed after IPTG-induced over-expression (Fig. 4). A 63-kDa, native polyurethanase had been purified from P. chlororaphis and PueA probably represents the same enzyme. Of the remaining 19 polyurethanase-active clones, only one was found which had a different restriction enzyme map than pueA and it is being studied.
While the five related lipases varied in total amino acid residues, P. fluorescens SIK W1 (449), Pseudomonas LS107d2 (474), P. fluorescens B52 (476) and the Serratia strains (both 613), the active site and the secretion signal sequences aligned perfectly, indicating this is a well conserved enzyme. One interesting note that may further distinguish PueA is that it lacks the amino acid cysteine, and four of the five related lipases also contain no cysteine.
Another polyurethanase  whose sequence is in GenBank (accession #AB00906) is from C. acidovorans and is membrane-associated. This enzyme is more closely related to carboxylesterases and acetylcholinesterases than to lipases. In conclusion, there appear to be a least two mechanisms for serine hydrolase activity that can degrade polyester-based PU.
This work was supported by the Office of Naval Research Grant N00014-96-1229.