大肠杆菌纤溶酶基因的克隆和表达

Annals of Microbiology, 58 (1)95-98 (2008)

Gene clone and expression of a fibrinolytic enzyme (FE) in Escherichia coli

Shi-Hua WANG1,2*, Miao DIAO2, Yan-Ling YANG2, Wei-Zhu LIN2, Bi-Fang Huang2

1Key Laboratory of Biopesticide and Chemical Biology, Ministry of Education, 2College of Life Sciences, Fujian Agriculture and Forestry University, Fuzhou 350002, China

Received 2 April 2007 / Accepted 22 January 2008

Abstract -Bacillus subtilis LD-8547, a strong fibrinolytic enzyme (FE) producing strain, was obtained by the mutation from Bacillus subtilis LD-8, isolated in our laboratory. Based on the homology, the DNA fragment encoding the intact FE was amplified from total DNA of Bacillus subtilis LD-8547 by PCR. The vector pET32a (+) was used to construct the N-terminal thioredoxin (Trx) fusion expres-sion plasmid, and the recombinant FE was successfully expressed in Escherichia coli BL21 (DE3). The expressed protein showed high fibrinolytic activity in soluble fraction. This study provided evidences that FE could be actively expressed in E. coli, and large quanti-ties of FE could be produced for pharmacological and clinical research.

Key words: fibrinolytic enzyme (FE); gene clone; expression; activity.

INTRODUCTION

Thrombolytic therapy has been extensively researched as a means of medical treatment of the blood clot, and various fibrinolytic enzymes produced by different microorganisms were discovered (Kim et al.,1997; Choi and Shin, 1998) in recent years. The fibrinolytic activity was also found in douchi, a traditional Chinese fermented soybean food, from which some fibrinolytic enzyme producing strains were subsequently obtained (Peng et al.,2003; Wang et al., 2003; Wang et al.,2007).

Subtilisin fibrinolytic enzyme was purified from the broth by Bacillus amyloliquefaciens DC-4(Peng et al., 2003), and was actively expressed in the protease deficient strain Bacillus subtilis WB600 (Peng et al.,2004). When the native promoter of this subtilisin fibrinolytic enzyme was replaced by that of α-amylase gene, the expressed fibri-nolytic enzyme was increased (Xiao et al.,2004). But these fibrinolytic enzymes were insoluble aggregates. When this fibrinolytic enzyme was expressed as a fusion protein to thioredoxin (Trx), the fibrinolytic enzyme was highly obtained with fibrinolytic activity (Zhang et al.,2005).

Bacillus subt ilis LD-8, a fibrinolytic activity producing strain was isolated in our laboratory. This strain was mutat-ed by UV, NTG and γ-radiation, resulting in a higher fibri-nolytic enzyme (FE) producing strain Bacillus subt ilis LD-8547 (Wang et al.,2007). This purified fibrinolytic enzyme was about 30 kDa, and its optimal temperature and pH val-ues were 50 °C and 8.0, respectively. The enzyme activity was inhibited by phenylmethanesulfonyl fluoride (PMSF), but EDTA and EGTA did not affect the enzyme activity,showing that the enzyme was a serine protease (Wang et al.,2007).

When the DNA fragment encoding the intact FE was cloned into expression vector pET28a (+) and pET29a (+), the FE was highly expressed, but no fibrinolytic activity was detected at all in our laboratory. When pET32a (+) was used as expression vector, we successfully expressed FE in E. coli with fibrinolytic activity. In this paper,we describe the cloning and expression of the intact FE gene from Bacillus subtilis LD-8547 in ET32a (+)-E. coli BL21 (DE3) system.

MATERIALS AND METHODS

Bacterial strains and chemicals. Bacillus subt ilis LD-8547, a fibrinolytic enzyme producing strain, was mutated from isolated strain Bacillus subtilis LD-8 in our laboratory (Wang et al.,2007). Plasmid pET32a (+), purchased from Novagen, San Diego, CA, USA, was used as expression vector. Escherichia coli BL21 (DE3), stored in our laborato-ry, was used for expression of FE, and cultured in LB medi-um at 37 °C. Fibrinogen, thrombin, urokinase were pur-chased from Sigma, USA. Other chemicals were analytical grade.

Construction of expression vector. Bacillus subtilis LD-8547 genomic DNA was isolated and used as PCR template. To amplify DNA fragments encoding the intact FE, PCR was performed with forward primer P1 (AGGGAATTCGTGA-GAAGCAAAAAAT) and reverse primer P2 (ACCCTC-GAGTTGTGCAGCTGCT). PCR mixture was prepared in a total volume of 50 μL containing 20 ng of genomic DNA,

200 nmol/L each primer, 0.2 mmol/L dNTP, 1.5 mmol/L

MgCl2, 1x PCR buffer and 2.5 U Taq polymerase. The PCR product was digested with EcoR I and Xho I, and the isolat-ed fragment was inserted into EcoR I / Xho I linearised pET32a (+). The resulting plasmid was named as pET32FE, and transformed into E. coli BL21 (DE3) for recombinant protein expression. There was a His-tag (from vector pET32a) in the C-terminal of the recombinant FE for purifi-cation.

Gene expression of fib rinolytic enzyme (FE). Escherichia coli BL21 (DE3) harbouring the recombinant plasmid pET32FE was grown overnight at 37 °C in liquid LB medium containing 100 μg/mL ampicillin at 37 °C, and grown until an OD600of 0.6. Recombinant FE expression was induced by adding IPTG to a final concentration of 1 mmol/L, and the incubation was extended for additional 12 h at 25 °C. The cells were harvested by centrifugation, and used for SDS-PAGE analysis.

Purification of expressed FE. The recombinant FE pro-ducing strain was scaled up to one litre. After induction for 4 h, the bacteria were pelleted (8000 x g for 10 min at 4°C). The cells were washed and dissolved in PBS buffer. After sonicated, the supernatant was purified using Ni2+-chelation affinity resin (Bi et al.,2003) by His-tag at its C-terminal. The purified FE was identified by activity assay as description followed.

Fib rinolytic activity assay. Fibrinolytic activity was determined by the fibrin plate method with modification (Astrup and Mullertz, 1952; Zhang et al.,2005). In brief, 25 mL of 0.3% bovine fibrinogen solution in 100 mmol/L sodium phosphate buffer (pH 7.4) was mixed with 0.5 mL of thrombin solution (5 U/mL) and 25 mL of 1% agarose gel in a Petri dish. The dish was allowed to stand for 30 min at room temperature to form fibrin clot, and holes were made on the plate by suction using a capillary glass tube. Ten microlitres of sample solutions were added into each hole and the plate was then incubated for 20 h at 37 °C. The fibrinolytic activity was calculated by the lytic area. Urokinase was used as a standard. Amidolytic activity and kinetic constants of the fibrinolytic enzyme were estimated using synthetic substrate (Wang et al.,2007).

Blood clots lytic effect with FE. Blood clots from white rabbit were divided into two groups. One was heated at 80°C for 40 min to inactivate endogenous fibrinolytic factors, and the other group remained untreated. The heated and unheated blood clots were incubated with purified fibri-nolytic enzyme in Petri dish at 37 °C. At different time intervals, the lytic area was used for analysis.

Error prone PCR. We engaged to enhance the enzyme activity of FE by error prone PCR technology. Random mutations were introduced during the error prone PCR pro-cedure according to Xu et al.(2003).Bacillus subtilis LD-8547 genomic DNA was used as PCR template. PCR was performed with forward primer P1 and reverse primer P2. PCR mixture was prepared in a total volume of 50 μL con-taining 20 ng of genomic DNA, 200 nmol/L each primer, 0.2 mmol/L dATP, 0.2 mmol/L dGTP, 1 mmol/L dCTP, 1 mmol/L dTTP, 5 mmol/L MgCl2, 1 mmol/L MnCl2, 1x PCR buffer (Mg2+free) and 2.5 U Taq polymerase. The first amplified PCR product was used as the PCR template of the second time, and the second amplified product was used as the template of the third time. All the amplified products of the three times were digested with EcoR I and Xho I, followed by the ligation to the same digested vector pET32a (+). Escherichia coli BL21 (DE3) was transformed with the lig-ated mixture to create the mutant library for higher fibri-nolytic enzyme screening.

Bioinformatics analysis of this fib rinolytic enzyme gene. After sequencing, Primer Premier 5.0 was used to deduce the amino acid sequence. Homology analysis was performed using Blastx and SIB BLAST Network Service (https://www.360docs.net/doc/d31803879.html,/tools/blast/). The signal peptide cleavage site was predicted using SignalP 3.0 Server (http:// www.cbs.dtu.dk/services/SignalP/) and ANTHEP-ROT software. The secondary structure prediction and sol-vent ability analysis were carried out through PREDICT PROTEIN database online analysis and ANTHEPROT soft-ware. The prediction of conformation was performed using SWISS MODEL method.

RESULTS

Gene clone of Trx-FE

Bacillus subt ilis LD-8547 genomic DNA was used as PCR template, and PCR was performed with forward primer P1 and reverse primer P2. The intact DNA fragment of FE was 1146 bp, and there was a band of PCR product about 1146 bp. It was showed that gene encoding the intact FE was successfully obtained. The PCR product was digested with EcoR I and Xho I, and inserted into pET32a (+). The result-ing plasmid named pET32FE was used for production of recombinant FE.

Expression of recombinant FE

To obtain a sufficient amount of recombinant enzyme, FE was overproduced in E. coli BL21 (DE3) using the pET32a (+) vector. The expression of FE was induced with IPTG under the control of a T7-phage promoter. The fusion pro-tein Trx-pro-FE (about 54 kDa) was highly expressed (Fig. 1, lane 4). There was a thioredoxin in the N-terminal of the fusion protein, and it could increase the soluble yields to overcome the “inclusion body” problem. The C-terminal His-tag was used to facilitate the purification of the desired protein. Expression of recombinant FE was controlled by T7 promoter, and it was in both soluble (Fig. 1, lanes 3) and inclusion body fraction (Fig. 1, lanes 2). But fibrinolytic activity was only detected in the soluble fraction (Fig. 1, lane 3). This indicated that Trx-pro-FE was cleaved into mature FE (about 30 kDa) and only soluble mature FE had fibrinolytic activity.

Purification and characterisation of recombinant FE The cells were washed and dissolved in PBS buffer. After sonicated, the supernatant was purified using Ni2+-chela-tion affinity resin. To identify whether the expressed FE has the same characteristic as the native one, some character-istics of the purified FE were compared to that of the native FE. The result showed that the fibrinolytic activity was inhibited by 1 mM phenylmethanesulfonyl fluoride (PMSF), but not by EDTA or EGTA, and other results were in Table 1. All the characteristics identified above were almost the same as the native one produced by Bacillus subt ilis LD-

96S.-H. Wang et al.

8547. Specific activity of the enzyme was much higher than those of the fibrinolytic enzyme also isolated from douchi .Therefore, the FE can be easily prepared in large scale for further study.

Blood clots lytic effect with FE in vitro

Blood clots from white rabbit were used to determine the lytic effect of this fibrinolytic enzyme. One group was heat-ed to inactivate endogenous fibrinolytic factors, and the other group remained untreated. The blood clots of two groups were incubated with purified FE in Petri dish at 37°C. The result showed that the heated blood clots dissolved within 2.5 h (Fig. 2 A3), but the non-heated blood clots dis-solved within 40 min (Fig. 2 B2), indicating that the enzyme was able to degrade blood clots in the absence of endogenous fibrinolytic factors.

Evolution strategy and screening of mutant library The strategy of FE evolution towards higher activity in this study was to use error prone PCR technology. The fibri-nolytic enzyme gene of LD-8547 was used as the parent template to generate the first generation error PCR prod-uct. Then the following two generations were successively created using the PCR products from the previous genera-tion as the new parents. All these error PCR products of the three generations were used to construct mutant library.After screening, about 100 colonies were random screened

and 35 clones gave positive signal. One clone giving the strongest activity was obtained, about 32% higher fibri-nolytic activity than that of the original recombinant FE (Detail data not shown).

Bioinformatics analysis

The results of DNA sequencing showed that the full length of the FE gene covered 1146 bp, with the GC contents of 46.25%, encoding a mature FE of 275 amino acids and a pro-peptide of 106 amino acids. ANTHEPROT analysis indi-cated that the isoelectric point of FE was 6.325, and the molecular weight was about 30 kDa.

Online analysis using PREDICE PROTEIN database indi-cated that the secondary structure was composed of 17.45% α-helix, 24.73% β-pleated sheet, and 57.82% ran-dom coil. Thus, the FE was a mixed protein. The prediction results of solvent accessibility composition indicated that the content of composition e (amino acid residues exposed to the surface of protein) was 41%, while the composition b (other residues) was 59%. ANTHEPROT and structural analysis revealed that the forth and the fifth amino acid residues were positive charged, and there was a distinct hydrophobic region from residue 6 to residue 25. SWISS-MODEL was applied to predict the conformation of FE, and the result showed that the peptide was composed of 7 α-helix connected by small peptides, and moved together to each other to form a spherical molecule (Fig. 3).

Ann. Microbiol., 58 (1), 95-98 (2008)

97

FIG. 1 - SDS-PAGE analysis of expression and purification of fib-rinolytic enzyme (FE). Lane 1: the purification result of the mature FE after induction; lane 2: the precipitation of the pET32FE after sonication and centrifugation; lane 3: the supernatant of the pET32FE after sonication and centrifugation; lane 4: the expression products of the whole pET32FE cells; lane 5: the control of empty vec-tor ET32a (+); lane 6: mid-range protein molecular

weight markers.

FIG. 2 - Blood clots from white rabbit were used to determine

the lytic effect of this fibrinolytic enzyme. A: all the three blood clot plats (A1, A2, A3) were heated at 80 °C for 30 min to inactivate endogenous fibrinolytic factors and incubated with purified fibrinolytic enzyme at 37 °C. B:all the three plats (B1, B2, B3), without heating, were incubated with fibrinolytic enzyme at 37 °C. A1 and B1:0 min incubation; A2 and B2: 40 min incubation; A3 and B3: 2.5 h incubation.

TABLE 1- Screening of plant growth promoting traits in Bacillus thuringiensis isolates Fibrinolytic Total protein

Total activity

Specific activity

Optimal Optimal temperature K m

V max

enzyme (mg)

(U)(U/mg)

pH (°C )(mmol/L )

(mmol/min)Native FE 3.088375862719448.0500.5210.049Expressed FE

2.56

696050

271895

8.0

50

0.527

0.047

DISCUSSION

When the DNA fragment encoding the intact FE was cloned into expression vector pET28a (+) and pET29a (+), target FE was expressed, but no fibrinolytic activity was detected at all. The possible reason was that the expression target protein formed “inclusion body”, resulting in no enzyme activity. So we turned to soluble expression vector pET32a (+). In this paper, recombinant plasmid pET32FE was con-structed to express FE in E. coli. The recombinant protein was expressed, and became more soluble when fused with the N-terminal Trx. It might yield soluble, active target pro-tein to overcome the “inclusion body” problem (Bi et al., 2003). C-terminal having a His-tag was used to facilitate the purification of the desired protein.

Soluble supernatant and “inclusion body” pellet of induced pET32FE were both used to test fibrinolytic activi-ty, and the result showed that only soluble supernatant had activity, whereas inclusion body had no activity. Purification result (Fig. 1, lane 1) only gave a 30 kDa FE with strong fibrinolytic activity, indicating that Trx-pro-FE was cleaved into mature FE. The purified FE has the same characteris-tics as native one produced by Bacillus subt ilis LD-8547, which showed that expressed FE can be easily prepared in large scale, and used for further pharmacological and clin-ical research.

Acknowledgement

We are sincerely thankful for the support from the Key Scientific and Technology Project of Fujian Province of China (2007N0094, 2007Y0002), Natural Sciences Foundation of Fujian Province of China (B0610008, X0750005), Postdoctoral Sciences Foundation of China (20060390183), and Program for New Century Excellent Talents in Fujian Province University. Dr. Yong Peng and Prof. Yi-Zheng Zhang (Sichuan University, Chengdu, China) are greatly appreciated for paper donation and technical discussion. REFERENCES

Astrup T., Mullertz S. (1952). The fibrin plate method for esti-mating fibrinolytic activity. Arch. Biochem. Biophys., 40: 346-351.

Bi L.J., Zhou Y.F., Zhang X.E., Deng J.Y., Zhang Z.P., Xie B., Zhang C.G. (2003). A muts-based protein chip for detection of DNA mutations. Anal. Chem., 75 (16): 4113-4119.

Choi H.S., Shin P.H. (1998). Purification and partial characteri-zation of a fibrinolytic protease in Pleuro us os rea us.

Mycologia, 4: 674-679.

Kim H.K., Kim G.T., Kim D.K., Choi W.A., Park S.H., Jeong Y.K., Kong I.S. (1997). Purification and characterization of a novel fibrinolytic enzyme from Bacillus sp. KA38 originated from fermented fish. J. Ferment. Bioeng., 84: 307-312.

Peng Y., Huang Q., Zhang R.H., Zhang Y.Z. (2003). Purification and characterization of a fibrinolytic enzyme produced by Bacillus amyloliquefaciens DC-4 screened from douchi, a tra-ditional Chinese soybean food. Comp. Biochem. Physiol. B., 134: 45-52.

Peng Y., Yang X.J., Xiao L., Zhang Y.Z. (2004). Cloning and expression of a fibrinolytic enzyme (subtilisin DFE) gene from Bacillus amyloliquefaciens DC-4 in Bacillus sub ilis. Res.

Microbiol., 155: 167-173.

Wang C.T., Ji B.P., Li B., Ji H. (2003). Isolation of Douchi fibri-nolytic enzyme producing strain DC33 from Chinese Douchi and primary analysis of the enzyme property. Food Sci.

(China), 6: 34-38.

Wang S.H., Zhang C., Yang Y.L., Diao M., Bai M.F. (2007).

Screening of a high fibrinolytic enzyme producing strain and characterization of the fibrinolytic enzyme produced from Bacillus subtilis LD-8547. World J. Microbiol. Biotechnol., 7: epub.

Xiao L., Zhang R.H., Peng Y., Zhang Y.Z. (2004). Highly efficient gene expression of a fibrinolytic enzyme (subtilisin DFE) in Bacillus subtilis mediated by the promoter of α-amylase gene from Bacillus amyloliquefaciens. Biotechnol. Lett.,26: 1365-1369.

Zhang R.H., Xiao L., Peng Y., Wang H.Y., Bai F, Zhang Y.Z.

(2005). Gene expression and characteristics of a novel fibri-nolytic enzyme (subtilisin DFE) in Escherichia coli. Lett. Appl.

Microbiol., 41: 190-195.

98S.-H. Wang

et al. FIG. 3 - The SWISS-MODEL analysis result of the 3-D structure

of the mature FE. The molecule consists of a-helix

(helix), sheet (arrow) and loops (thread).