Characterization of a Recombinant Chimeric Plasminogen Activator Composed of Gly-Pro-Arg-Pro Tetrapeptide and Truncated Urokinase-Type Plasminogen Activator Expressed inEscherichia coli

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS ARTICLE NO.

222, 576–583 (1996)

0786

Characterization of a Recombinant Chimeric Plasminogen Activator Composed of Gly-Pro-Arg-Pro Tetrapeptide and Truncated Urokinase-Type Plasminogen Activator Expressed in Escherichia coli Zi-Chun Hua,1 Xiao-Chun Chen, Chen Dong, and De-Xu Zhu Pharmaceutic Biotechnology Key Laboratory, Department of Biochemistry, Nanjing University, Nanjing 210093, People’s Republic of China Received April 12, 1996 A chimeric plasminogen activator, GPRP-u-PA (144–411), consisting of the Gly-Pro-Arg-Pro tetrapeptide fused to the N-terminal of a truncated urokinase-type plasminogen activator (comprising Leu 144 through Leu 411), was produced by expression of the corresponding chimeric cDNA in Escherichia coli cells. After renaturation, the chimera was purified to homogeneity with specific amidolytic activity of 100,000 IU/mg protein. The chimera showed 6-fold greater affinity for fibrin clots than native low molecular weight urokinase (LUK) and 1.5-fold greater affinity than a chemical conjugate, GPRP-LUK, generating via coupling Gly-Pro-Arg-Pro tetrapeptide to native low molecular weight urokinase. The chimera had 2 to 3 fold greater fibrinolytic potency than native LUK in vitro. Fibrinogen had no influence on fibrinolysis of the chimera. The chimera consumed much less fibrinogen than native LUK. © 1996 Academic Press, Inc.

Plasminogen activators (PAs) are serine proteases which play an important role in fibrinolysis by converting the proenzyme plasminogen to proteolytic plasmin. Plasmin, in turn, degrades fibrin into soluble products. Urokinase-type plasminogen activator (u-PA), one of the plasminogen activators, is initially synthesized as a single-chain form (scu-PA) and can be converted into an active two-chain form (tcu-PA) by a single cleavage between Lys158 and Ile159 by plasmin (1,2). A low molecular weight form of scu-PA (scu-PA 144–411), lacking the epidermal growth factor domain and the kringle domain, has been obtained from monkey kidney cell cultures (3) and from the human lung carcinoma cell line CALU-3 (4). The enzyme is proteolytically derived from scu-PA by the metalloprotease pump-1 which cleaves between Glu143 and Glu144 (5). scu-PA 144–411 has been reported with intact enzymic activity, fibrin specificity and thrombolytic potential as scu-PA (3,6,7,8). Human fibrinogen is transformed into fibrin by the thrombin-catalyzed release of small polar peptides (fibrinopeptides A and B) from the amino terminal of the a and b chains. Upon release of these peptides, polymerization of the fibrin monomer units occurs spontaneously to form a noncovalently bonded gel. The polymer can be covalently cross linked by another enzyme, factor XIII, which activated by thrombin (9). Short peptides beginning with the sequence glycyl-L-prolylL-arginine (Gly-Pro-Arg)… which corresponds to the amino-terminal segment of the fibrin a chain after the release of the fibrinopeptide A, can prevent the polymerization of fibrin monomers. Further these peptides bind to fibrinogen and the plasmin-generated fragment D (10). Although Gly-Pro-Arg itself was found to be effective inhibitor of polymerization, the addition of a fourth residue, particularly proline or sarcosine, significantly increased both the binding and the inhibitory activity (11). When chemically conjugated to a native low molecular weight two-chain urokinase (LUK), tetrapeptide Gly-Pro-Arg-Pro could impart fibrin affinity to native LUK and increase its fibrinolytic potency (12). 1

To whom correspondence should be addressed. Present address: 17 Bradford Street, Albany, NY 12206. After May 25, 1996, please contact Dr. Zi-chun Hua at Room 301, 2 Baijishancun, Tianshan Road, Nanjing 210008, People’s Republic of China. 576 0006-291X/96 $18.00 Copyright © 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.

Vol. 222, No. 2, 1996

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

In the present study we constructed, expressed and characterized a recombinant chimeric plasminogen activator, GPRP-u-PA (144–411), composed of u-PA(144–411) and Gly-Pro-Arg-Pro tetrapeptide. MATERIALS AND METHODS Materials. Plasmid pKK233-2 was purchased from Pharmacia. Restriction enzymes, T4 DNA ligase were from New England Biolabs. IPTG and X-gal were from Promega. Fibrinogen (plasminogen free) was purchased from Sigma. Thromin was from Tianjing Biological Product Factory. Reduced (GSH) and oxidized (GSSG) glutathione, protein molecular weight standards were from Shanghai Dongfong Reagent Factory. Monoclonal antibody against human urokinase was from Beijing Academy of Military Medical Sciences. Native LUK and urokinase standard were from Nanjing University Biochemical Pharmaceutical Factory. Construction of expression plasmid pKK233-2-GPRP-u-PA (144–411). The DNA fragment encoding 144–411aa of human scu-PA was obtained as a 800bp fragment by HindIII and ClaI digestion. Oligonucleotides A: GGGGTACCATGGGTCCACGTC and B: CGGACGTGGACCCATGGTACCCC were synthesized. The complementary oligonucleotides encode the initiation methionine (underlined), Gly-Pro-Arg-Pro tetrapeptide and contains one NcoI cohesive end and one ClaI cohesive end after annealing and NcoI digestion. The annealed oligonucleotides A and B, the obtained scu-PA DNA fragment were ligated with pKK233-2 vector digested with NcoI and HindIII to yield expression plasmid pKK2332-GPRP-u-PA (144–411). In the resulting plasmid, the GPRP-u-PA (144–411) gene is under the control of tac promoter. The expression plasmid pKK233-2-GPRP-u-PA (144–411) was transformed into E. coli JA221 for expression. Protein expression and renaturation. Protein expression and renaturation were performed essentially as described by Hua (13). Centrifuged cells from 100ml flask culture were suspended in 0.5ml 0.1M Na phosphate buffer, pH8.0 and disrupted by sonication. The suspension was centrifuged at 10,000rpm for 20 minutes and the collected pellet was solubilized in 8M urea, 50mM b-centrifugation, the supernatant was dialyzed against 8M urea, 1mM b-mercaptoethanol, 0.1M Na phosphate buffer, pH7.5 and stirred at 4°C for 16 hours. After centrifugation, the supernatant was dialyzed against 8M urea, 1mM b-mercaptoethanol, 0.1M Na phosphate buffer, pH7.5 for 2 hours. The centrifugated supernatant was dropped slowly into 150ml 2.5M urea, 50mM Tris-HCl, pH 9.0, 5 mM EDTA, 10 mM NaCl, 0.005% Tween-80, 1.25 mM GSH, 0.25 mM GSSG with vigorous stirring at 4°C for 24–36 hours. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS–PAGE). SDS-PAGE was performed according to Laemmli (14) with 4% stacking gels and 12% separating gels. Cell pellets from 0.1 ml culture were treated by heating at 95°C for 10 minutes in 50mM Tris-HCl, pH6.8, 2% SDS, 10% glycerol, 2% b-mercaptoethanol, 0.02% bromphenol blue before loading on gel. Proteins were visualized by coomassie blue R-250 staining. Western blot analysis. Western blot analysis was performed according to Towbin (15). Protein purification. 7ml CH-Sepharose 4B-Pro-Gly-Arg-H beads were transferred into empty column (0.9 × 10cm) and equilibrated with 0.1M phosphate buffer, pH7.2. The renaturation mixture was passed over the column, washed with 0.1M phosphate buffer, pH7.2, and eluted with 0.1M HAC, 0.05M NaCl, pH3.0. The protein concentration of the elution was measured as described by Bradford (16) with bovine serum albumin as standard. Fibrin clot binding measurement. 1.5 mg fibrin was incubated in solutions containing 300I.U./ml LUK or GPRP-u-PA (144–411) for 15 minutes and then washed with 0.1M phosphate buffer, pH7.4 for 10 minutes. The fibrin clots were placed in a 3ml light path cuvette, which containing 3ml 0.1M phosphate buffer, pH7.4, 0.05M NaCl and 0.1 mg/ml plasminogen. The absorbance changes at 280nm were then plotted versus time. Fibrin clot lysis. 2mg fibrin clots were placed in a 1.5ml light path cuvette containing 0.1M phosphate buffer, pH7.4, then 100I.U. LUK or GPRP-u-PA and 0.1mg plasminogen were added. The increase in absorbance at 280nm at 37°C was measured. Amidolytic activity determination. The amidolytic activity of GPRP-u-PA (144–411) was measured according to Stump (17) using urokinase chromogenic substrate S-2444 at 25°C. Enzyme activity was standardized by comparison with the urokinase standard. Characterization of GPRP-u-PA (144–411). In 0.1 ml 0.09M NaCl, 0.01% Tween-80, 0.05M phosphate buffer, pH7.4 with different S-2444 concentrations of 300mM, 150mM, 75mM or 37.5mM, 20IU GPRP-u-PA (144–411) was added and incubated at 25°C. The increase of absorbance at 405 nm was measured. Fibrinolytic potency assay: Fibrinolytic potency assay was performed according to Chen (12). Influence of fibrinogen on fibrinolysis and fibrinogen consumption: 2mg fibrin clot was placed in 1.5ml light path cuvette containing 0.1M phosphate buffer, pH7.4, 100I.U. GPRP-u-PA (144–411) or LUK, 0.15mg plasminogen, then 1mg fibrinogen were added. The increase in absorbance at 280nm at 37°C was measured. Fibrinogen in supernatant was determined according to Du (18).

RESULTS The recombinant expression plasmid pKK233-2-GPRP-u-PA (144–411) was transformed into E. coli JA221. After IPTG induction, there was a distinguishable extra band about 30KDa in cells 577

Vol. 222, No. 2, 1996

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

harboring pKK233-2-GPRP-u-PA (144–411) as shown in Fig. 1. The band was proven to be recombinant human u-PA by western blotting analysis as shown in Fig. 2. The densitometric scanning results revealed that the expression level of GPRP-u-PA (144–411) was about 5% of total cellular proteins. GPRP-u-PA (144–411) is expressed essentially in an insoluble denaturated form as indicated by observation that there was nearly no u-PA activity in the bacterial lysate. After denaturation and renaturation, 180,000IU fibrinolytically active GPRP-u-PA (144–411) was obtained from 1 liter flask culture as measured using the classical fibrin plate technique. It was reported that when human u-PA was expressed in E. coli, some products were in two-chain form cleft between Lys159-Ile159 (19,20). Considering the very low amidolytic activity of single chain form u-PA (3,6) and the relative difficulty of purifying it, we tried to purify recombinant two-chain form GPRP-u-PA (144–411) for fibrinolytic property analysis. The renaturated GPRP-u-PA (144–411) was purified to homogeneity by one step CH-Sepharose 4B-ProGly-Arg-H affinity chromatography as shown in Fig. 3. CH-Sepharose 4B-Pro-Gly-Arg-H beads have high affinity for native human two chain u-PA and were used to purify urokinase from human urine efficiently. Specific activity of final preparation of GPRP-u-PA (144–411), measured as amidolytic activity, was 100,000IU/mg protein. No aggregates were observed by silver-stained SDS-PAGE under non-reducing conditions, indicating that all the cysteines of the renatured proteins have been reoxidized into internal disulfides. The apparent molecular mass of GPRP-u-PA (144–411) was approximately 30KDa as determined by SDS-PAGE under reducing or nonreducing conditions (Fig. 3). The molecular mass of reducing GPRP-u-PA (144–411) was slightly smaller than that of the non-reducing one, indicating that the purified product is in two chain form generating by cleavage of the Lys158-Ile159 peptide bond (19,20) and resulting in removal of a peptide of 20 residues following reduction of the interchain disulfide bond. The specific activity of 100,000IU/mg protein, measured as amidolytic activity, of the purified products also demonstrated the purified GPRP-u-PA (144–411) was not in single-chain form, considering the very low amidolytic activity of scu-PA 144–411 previously reported (3,6,7,8). Kinetic analysis of GPRP-u-PA (144–411) revealed Michaelis Menten kinetics demonstrated by a linear Lineweaver-Burk plot (Fig. 4). The kinetic parameter was Km440 mM. To determine the affinity of recombinant GPRP-u-PA (144–411) for fibrin clot, fibrin clots were soaked in solutions containing native LUK, chemically conjugated GPRP-LUK or recombinant GPRP-u-PA (144–411) and then after washing were placed in cuvettes containing plasminogen to determine the degradation of fibrin by plasmin, which activated by native LUK, chemically conjugated GPRP-LUK or recombinant GPRP-u-PA (144–411) absorbed on the surface of fibrin clots. The activation of plasminogen induced by native LUK may be attributed to the non-specific absorbance of LUK to fibrin clot surface. Evidently, GPRP-u-PA (144–411) has better affinity for

FIG. 1. SDS–PAGE analysis of GPRP-u-PA (144–411) expression stained by coomassie blue R-250. Lane 1: molecular weight standards, from top to bottom: 94KDa, 67KDa, 43KDa (very weak), 30KDa, 17.5KDa; Lane 2: E. coli JA221 cells containing plasmid pKK233-2-GPRP-u-PA (144–411), no IPTG induction; Lane 3: E. coli JA221 cells containing plasmid pKK233-2-GPRP-u-PA (144–411), with IPTG induction; Lane 4: E. coli JA221 cells containing plasmid pKK233-2. 578

Vol. 222, No. 2, 1996

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

FIG. 2. Western blot analysis of GPRP-u-PA (144–411) expression. Lane 1: E. coli JA221 cells containing plasmid pKK233-2; Lane 2: E. coli JA221 cells containing plasmid pKK233-2-GPRP-u-PA (144–411).

the fibrin clots, thus converted more plasminogen to plasmin to degrade fibrin more rapidly (Fig. 5). Comparison of affinity of GPRP-u-PA (144–411) for fibrin clots with that of GPRP-LUK or native LUK revealed that GPRP-u-PA (144–411) has 6-fold greater affinity for fibrin clots than native LUK and 1.5-fold greater than GPRP-LUK. For evaluation of fibrinolytic potency of GPRP-u-PA (144–411), GPRP-u-PA (144–411) or LUK of equal amount of activity was used to activate plasminogen to degrade fibrin. The absorbance when full fibrinolysis occurred was defined as 100% and the percentage of clot lysis calculated from the absorbance after 1.5 or 2 hour incubation was used to represent the fibrinolytic potency. As shown in Fig. 6, when 200IU GPRP-u-PA (144–411) or LUK was used, GPRP-u-PA (144–411) has 2-fold greater fibrinolytic potency than native LUK after 1.5 hour incubation. When 100IU GPRP-u-PA (144–411) or LUK was used, GPRP-u-PA (144–411) has 3-fold greater fibrinolytic potency than native LUK after 2 hour incubation. The presence or absence of fibrinogen had no influence on GPRP-u-PA (144–411) activating plasminogen to degrade fibrin as shown in Fig. 7. This indicated that GPRP-u-PA (144–411) only has affinity for fibrin clot and nearly has no affinity for fibrinogen. In contrast, after 1.5 hour incubation, the fibrinogen level decreased to 70% in the presence of GPRP-u-PA (144–411) while the fibrinogen level decreased to only 30% in the presence of native LUK (Fig. 8). Owing to its better selective affinity for fibrin, GPRP-u-PA (144–411) consumes less fibrinogen in the reaction solution than native LUK which has no fibrin affinity.

FIG. 3. Silver-stained SDS–PAGE analysis of purified recombinant GPRP-u-PA (144–411). Lane 1: molecular weight standards, from top to bottom: 94KDa, 67KDa, 43KDa, 30KDa, 17.5KDa; Lane 2: proteins were subjected to electrophoresis after reduction with b-mercaptoethanol; Lane 3: proteins were subjected to electrophoresis without reduction. 579

Vol. 222, No. 2, 1996

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

FIG. 4. Lineweaver–Burk plot of the aminolytic activity of GPRP-u-PA (144–411). 20IU GPRP-u-PA (144–411) was incubated in buffers containing different S-2444 concentrations of 300mM, 150mM, 75mM or 37.5mM or 37.5mM at 25°C. The velocity of increase in absorbance at 405nm (1/Va) was plotted against the concentration of S-2444 (1/[S]).

DISCUSSION Gly-Pro-Arg-Pro tetrapeptide can bind to plasmin-generated fragment D, prevent the polymerization of fibrin monomers (10,11). When chemically conjugated to native LUK, GPRP could impart fibrin affinity to native LUK and increase its fibrinolytic potency (12). Chemical conjugates have several disadvantages including low yields, rendom coupling and, possibly, interference with the functional properties of the constituting moieties (21). Construction of chimeric plasminogen activator molecules by recombinant DNA technology might avoid these problems. In this paper we constructed a recombinant chimeric plasminogen activator, GPRP-u-PA (144–

FIG. 5. Fibrin clot lysis induced by recombinant GPRP-u-PA (144–411), chemical conjugate GPRP-LUK or native LUK. (V) Recombinant GPRP-u-PA (144–411); (n) Chemical conjugate GPRP-LUK; (*) native LUK. 580

Vol. 222, No. 2, 1996

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

FIG. 6. Comparison of fibrin clot lysis potency of native LUK and recombinant GPRP-u-PA (144–411). 1: 200IU recombinant GPRP-u-PA (144–411) was incubated in reaction mixture for 1.5 hour; 2: 200IU native LUK was incubated in reaction mixture for 1.5 hour; 3: 100IU recombinant GPRP-u-PA (144–411) was incubated in reaction mixture for 2 hours; 4: 100IU native LUK was incubated in reaction mixture for 2 hours.

411). The chimera, GPRP-u-PA (144–411), retained both the fibrin binding capacity and u-PA enzymatic properties of its constituting moieties and was found to have a 6-fold higher fibrinolytic potency than native LUK and 1.5-fold higher fibrinolytic potency than a chemical conjugate, GPRP-LUK. GPRP-u-PA (144–411) caused time-dependent lysis of fibrin clots associated with

FIG. 7. Influence of fibrinogen on fibrin clot lysis. The increase in absorbance (A) at 280nm was plotted against the incubation time. (V) without fibrinogen; (n) with fibrinogen. 581

Vol. 222, No. 2, 1996

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

FIG. 8. Fibrinogen consumption during fibrin clot lysis process. Residual fibrinogen levels, expressed as a percentage of the base-line value, were plotted against the incubation time. (n) recombinant GPRP-u-PA (144–411); (V) native LUK.

limited degrees of fibrinogen degradation whereas native LUK caused fibrinolysis only in association with a more pronounced drop of coagulable fibrinogen. GPRP-u-PA (144–411) caused fibrinolysis with fibrin specificity as shown by relative preservation of fibrinogen levels. This orientation of plasminogen activation towards fibrin occurred in the presence of direct binding of GPRP-u-PA (144–411) to fibrin imparted by GPRP tetrapeptide. Interestingly, it was reported that GPRP tetrapeptide could bind to fibrinogen (10,11), but in this study fibrinolysis was associated with less extensive systemic fibrinogen breakdown with GPRP-u-PA (144–411) than with native LUK. During clot lysis, enzymatically active thrombin, primarily responsible for cleavage of fibrinogen to fibrin, is released rendering the vessel susceptible to prompt rethrombosis, thus resulting in reocclusion. Tetrapeptide GPRP was demonstrated to be potent inhibitor of fibrin polymerization. GPRP-u-PA (144–411) might have the potential of preventing the polymerization of fibrin monomers and reducing the incidences of reocclusion. Thus further in vitro and in vivo studies are needed to evaluate the clot specificity, thrombolytic efficacy, system half-life and prevention of polymerization of fibrin of GPRP-u-PA (144–411), especially its single-chain form. Creation of a GPRP-u-PA (144–411) chimera which retains both thrombolytic and fibrin specific properties, and possible fibrin polymerization inhibition property, opens new alternatives in the establishment of clot-specific thrombolytic therapy. ACKNOWLEDGMENTS This study was supported by the Chinese National High Technology Development Program (863-103-19) to Zi-Chun Hua.

REFERENCES 1. Eaton, D. L., Scott, R. W., and Baker, J. B. (1984) J. Biol. Chem. 259, 6241–6247. 2. White, W. F., Barlow, G. H., and Mozen, M. M. (1986) Biochemistry 25, 2160–2169. 3. Wijngaards, G., Rijken, D. C., van Wezel, A. L., Groeneveld, E., and van der Velden, C. A. M. (1986) Thromb. Res. 42, 749–760. 4. Stump, D. C., Lijnen, H. R., and Collen, D. (1986) 261, 17120–17126. 5. Marcotte, P. A., Kozan, I. M., Dorwin, S. A., and Ryan, J. M. (1992) J. Biol. Chem. 267, 13803–13806. 582

Vol. 222, No. 2, 1996 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

Lijnen, H. R., Nelles, L., Holmes, W. E., and Collen, D. (1988) J. Biol. Chem. 263, 5594–5598. Rijken, D. C., Binnema, D. J., and Los, P. (1986) Thromb. Res. 42, 761–768. Spriggs, D. J., Stassen, J. M., Hashimoto, Y., and Collen, D. (1989) Blood 73, 1207–1212. Bailey, K., Bettelheim, F. R., Lorand, L., and Middlebrook, W. R. (1951) Nature 167, 233–234. Laudano, A. P., and Doolittle, R. F. (1978) Proc. Natl. Acad. Sci. USA 75, 3083–3089. Laudano, A. P., and Doolittle, R. F. (1980) Biochemistry 19, 1013–1019. Chen, X. C., Huan, Z. C., and Zhu, D. X., manuscript submitted for publication. Hua, Z. C., Dong, C., and Zhu, D. X. (1996) Biochem. Biophys. Res. Comm. 220, 131–136. Laemmli, U. K. (1970) Nature 227, 680–685. Towbin, H., Staekelin, T., and Gordon, A. (1979) Proc. Natl. Acad. Sci. USA 76, 4350–4354. Bradford, M. M. (1976) Anal. Biochem. 72, 248–254. Stump, D. C., Thienpout, M., and Collen, D. (1986) J. Biol. Chem. 261, 1267–1273. Du, L., Yu, R. R., Hua, Z. C., Zhu, D. X., Wu, G. X., and Ruan, C. G. (1993) Science In China (Series) 36, 1483–1489. Winkler, M. E., Blaber, M., Bennett, G. L., Holmes, W. E., and Vehar, G. A. (1985) Bio/Technology 3, 990–1000. Winkler, M. E., and Blaber, M. (1986) Biochemistry 25, 4041–4045. Dewerchin, M., and Collen, D. (1991) Bioconj. Chem. 2, 293–300.

583