GB2246133A - Amidated human pro-urokinase enzymes and their precursors - Google Patents

Abstract

Amidated fibrinolytic enzymes are provided which are -COOH-terminal amidated derivatives of human pro-urokinase. These derivatives have the general formula P-CONH2 (I) in which P represents the portion of human pro-urokinase, or a modification thereof, preceding the carboxy terminus. The modifications may be by way of deletion, insertion or substitution. A precursor of such derivatives has the formula P-Gly-Yn (I) wherein Y is any basic amino acid residue and n is 0 or an integer of 1-4. Processes for preparation of formula II via recombinant DNA techniques are detailed. Compounds of formula II provide precursors for in vitro or in vivo enzymatic amidation. Pharmaceutical compositions comprising compounds defined in formula (I) or (II) are provided for the treatment of acute myocardial infarction, pulmonary embolism or deep venous thrombosis.

Description

--i: C_. -1., AMIDATED FIBRINOLYTIC E2MY2S M THEIR PRECURSORS, AND

PROCESSES FOR THEIR PREPARATIOX Simr=ary The present invention relates to C-terminal _;-,ddated derivat-3-ves off thie hxn.an fibrinolytic enz--ne Pro-urokinase, to Drecursors thereof %_ - A and to processes for their preparation, including reco=i.nant DNA technologies for the nre:aration of the said precursors. As C-terminal amidated derivatives of human prourokinase either the natural protein or modifications %thereof, such as deletion, insertion and substitution mutants thereof having an a:nidated carboxy-terminal in the polypept-idic chain are included in the invention. Such amidated derivatives may be obtained by enzymatic treatment of an appropriate precursor where an additional glycine residue, Optionally followed by a short stretch of basic amino-acid residues, is present at the COOH-end. The invention also comprises the structural genes coding for said precursors, vectors containing said genes and suitable hosts transformed with said vectors.

introduction

The increasing knowledge of the molecular interactions that regulate physiological fibrinolysis has lead to important implications in the understanding of the mechanisms that dissolve blood clots, and in the development of new thrombolytic agents. In the human fibrinolytic system, a proenzyme, plasminogen, can be activated to the active enzyme, plasmin, by several types of plasminogen activators (Collen D. et al. (1986) CRC Critical Reviews in oncology/hematology, 4 (3) 249; Verstraete, M. et al. (1986) Blood, 67 (6) 1529). Plasmin is the major protease responsible for the degradation of the fibrin component of a blood clot (Rakoczi 1. et al. (1978) Biochim. Biophys. Acta. 540, 295; Robbins, K.C. et al (1967) 1 i i 1 1 i 1 1 1 i 1 i i 1 1 1 J.Biol.Chem. 242, 2333; Wimian B. (1977) --Eur. J.Biochem. 76, 129). However, plasmin can also exert its proteolytic effect on several plasr-na proteins among which It-he components of the coagulation pathway fibrinogen, factor V and VIII,and may therefore predispose to bleeding (Collen D. et al. (1986) CRC Critical Reviews in oncology/hematology, A (3) 249; Verstraete, M. et al. (1986) Blood, 67 (6) 1529; Wiman B. al (1979) Biochim.Biophys.Acta 579, 142). Activation of plasminogen may occur at the systemic level, leading to circulating plasmin that is rapidly neutralized by alpha2-antiplasmin and thus not available for fibrinolysis (Collen D. et al. (1986) CRC Critical Reviews in oncology/hematology, 4 (3) 249; Verstraete, M. et al. (1986) Blood, 67 (6) 1529). When "C'he alpha2-antiplasmin level is markedly reduced, plasmin is less rapidly neutralized and can exert its proteolytic effects not only on fibrin, but also on the blood coagul'ation proteins as described previously. Excessive lowering in the plasma concentrations of fibrinogen, factor V and VIII, together with the inhiLitory effects exerted by some of the fibrinogen degradation products on the coagulation process, on r)latelet aggregation and on fibrin polymerisation leads to haemostatic deficiency and subsequently to an increased bleeding risk (Latallo Z.S. et al (1973) Thrombos-Diath.Haemouh 56, 253; Totty W.G. et al (1982) RI adiology 143, 59). on the other hand, activation of plasminogen may occur at the fibrin level (fibrin-bound plasminogen activation) leading to fibrin-bound plasmin (Collen D. et al. (1986) CRC Critical Reviews in oncology/hematology, 4 (3) 249; Verstraete, M. et al. (1986) Blood, 67 (6) 1529) which is, instead, neutralized to a much lesser extent by alpha2-antiplasmin and cannot induce systemic fibrinogenolysis. Urokinase and streptokinase, the most commonly used plasminogen activators in conventional thrombolytic therapy f in man, have no specific affinity for fibrin. Both compounds activate relatively indiiscriminately either circulating or fibrin-bound plasminogen (Zamarron C. et al. (1984) 71hromb.liaemostas. 52,!9; Sajr.a:- ,.a M. et al. (1985) Sem.' H.Op. Paris 61 (20), 14,23). Therefore, the systemic haemostatic breakdown often encountered during treatment with streptokinase and urokinase and, consequently, the elevated bleeding risk have often hampered the widespread clinical use of these thrombolytic agents, despite their demons trated clinical eff-ficacy (Samama M. et al. (1985) Sem.Hop.Paris 61 (20), 1423; Maizel, A.S. elt a!. (1986) Card4ovasc.in%--erve,-i%--.R;-=d--ol, 9, 236; Bell W.R. (1976) Thromb. Haemiostas. 35, 57; ",car J. et al. (1987) Seminars in Thromb. and Haen:ost. 13 (2) 186: Gruppo Italiano per lo studio delia Streptochinas-i nell'infarto miocardico (GISSI) (1986) Lancel: 1, 397). On the contrary, tissue-ty-pe plasminogen activator (t.-PA) (Hoylaerts M. et al (1982) J. Biol.Chem. 257 (6) 2912), and more recently pro-urokinase (proUK) (Husain S.S. et al (lP83) Arch. Biochem. Biop1hys. 220, 31), both natural proteins, were shown to be weak activators of the circulating plasminogen and, conversely, strong activators of the fibrin -bound plasminogen. Systemic plasminogen and alpha-2-anti-plasmin consumDtions as well as fibrinogen degradation were therefore shown to occur to a low extent within -- "in vitro" and "in vivo', ex-Deriments In laboratorv animals. Both nroteins were therefore designed to induce in patients with thrombotic disease a more efficient thrombolytic effect over streptokinase and urokinase with a lower bleeding incidence. The fibrin-s- Decific thrombolytic activity of t-PA has been explained by its ability to bind fibrin through specific lysine binding sites, located in the triple-disulfide-bonded 'kringle domains" of the molecule. Consequently, f ibrin-bound Dlasminogen could be activated without significant haemostatic breakdown (Collen D. et al (1986) j n i 1 1 1 i 1 j 1 1.

Haemostasis, 16 (3), 25). On the other hand, proUK (also denominated single chain urokinase type plasminogen activator, scu-PA) does not bind to fibrin, however it displays fibrin-specific thrombolytic activity without systemic baemostatic consumption (Pannell R. et al. (1986) Blood 67, 1215; Gurewich V. et al (1987) Seminars in Thromb. and Haemost. 13 (2), 146; L-ijnen H.R. et al. (1986) J.Biol.Chem. 261, 1253). Extensive research has been carried out on the isolation and characterization of the human genes encoding these proteins and subseauent nroduction bv recombinant DNA technology (Collen D. et al. (1-984) J. Pharmacol.'--"xp. Tber. 231 (1), 146; Holmes, W.E. et al. (1985) Biotechnology 3, 923). Recombinant t-PA was submitted to multicenter clinical trials in patients with acute myocardial infarction and was shown to be significantly more effective than streptokinase in the recanalization of obstructed coronary arteries (The European Cooperative Study Group for Recombinant Tissuetype Plasminogen Activator (1985) Lancet, 1, 842; Sheehan F.H. et al. (1987) Circulation 75, (4), 817). Pro-arokinc-se- is at present in phase II clinical trials and is thought to be, at least, as ef f ective as t_PA in terms of thrombolytic activity and safety (Van de Werf.F. et al. (1986) Annals of Internal Medecine, 104, 345; Van de Werf F. et. al (1986) Circulation 74 (5), 1066).

Background of the invention

Many biologically active peptides are protected at their COOH-terminus by amidation, and, for most of them, amidation is essential for receptor recognition and/or bioactivity (J.F. Rehfeld (1981) Am.J.Physiol. 240, 6251-6266). COOH-terminal amidation (or cL-amidation) confers stability towards most carboxy-pe-ptidases and can possibly give rise to Prolonged half-life, and alteration of biological activity, 1 6 - receptor selectivity, cell permeability, tissue distribution, etc (G.S. Wand et al. (1985) Neuroendocrinol. 41, 482-489). a-Amidation represents one of the post- translational modifications of secretory peptides in eukariotic organisms. It may be carried out 'by an enzyme, named peptidylglycine a-amidating monooxygena-se (PM), which converts CC>OHterminal-glyci-ne-extended peptides into the corresponding des-glycine aamidated peptides, with concomitant formation of glyoxylic acid (A.F. Bradbury et al. (1982), Nature, 289, 686-688). The catalytic reaction occurs in the presence o - oxygen, and requires copper cation and ascorbate as cofactors (B. Eipper et al. (1983), Proc. Natl. Acad. Sci. USA, 80, 5144-5148). In the biosynthetic process, the Gly-extended peptide (substrate for the a-a-midating enzyme) is generated from precursor molecules through a proteolytic cleavage mediated by a specific carboxypeptidase B-like protease. Such precursors present from I to 4 additional basic amino acid residues immediately following the Glyextended end of the peptide (R.E. 14ains elt al. (1983) T.I.N.S. 229-235). It has been suggested that the basic residues might be involved in the recognition by specific a-amidating enzyme.(s) and/or participate in the amidation mechanism (S. Gomez et al. (1984) FEBS Lett. 167, 160-164).

Description of the -invention

The present invention relates to COOH-terminal amidated derivatives of the human fibrinolytic enzyme pro-urokinase, to precursors thereof, and to processes for their preparation, including recombinant DNA technologies for the preparation of the said precursors. Also structural genes coding for said precursors, vectors containing said genes and suitable hosts transformed with said vectors are within the scope of the invention.

i 1 i 1 j 1 A first object of the invention is therefore a COOH-terminal amidated derivative of human pro-urokinase having the formula (I):

P-CM 2 (1) wherein P represents the portion of human pro-urokinase, or a modification thereof, preceding the carboxy terminus of the molecule. A modification of pro-urokinase may be, for example, a deletion, insertion or subs t -4 t_ution mutant thereof.

Another object of the present invention is represented by the Drecursors of the COCH-terminal amidated derivatives of formula (I), having the formula (II):

P-Gly-Yn (II) wherein P is as defined above, Gly is the glycine residue, Y is a basic amino acid residue and n is zero or an integer of I to 4. When n is an integer oil 2 to 4, the Y residues may be equal or different. For example, when n is 2, then Yn may represent, e.g., either Lys-Arg or Arg-Arg.

A further object of the invention is a process- for obtaining a COOHterminal an-idated derivative of formula (I) which comprises the enzymatic conversion of the precursors defined in formula (II). These may be converted into the corresponding amidated polypeptides through the action of an a-midating enzyme, such as, for example, natural or recombinant pept idyl -9 lycine cL-amidating n.onooxygenase (PAM), preceded (when n is different from zero) by a proteolytic cleavage of the Yn-extended precursors with, e.g., carbox-ypeptidase B, in order to eliminate the Yn basic amino acid residues. rurthermore the invention includes a recombinant DNA process for preparing the compounds of formula (II), as potential 1 precursors -for in vivo amidation (G. Gafvelin et al (1984) FEBS.Lett. 184, 347-352). According to this process the said precursors are constructed by oligonucleotide directed mutagenesis of the pro-urok-inase gene and subsequent insertion of the mutated genes in a suitable expression vector which could direct the synthesis of t-he new precursors in a proper host microorganism.

Still another object of the invention is constituted by the genes comprising chemically synthesized polynucleotides coding for the amino acid sequences represented by the formula (I1).

In one more aspect, the present invention relates to the expression vectors carrying the said genes and capable of directing the synthesis oil the pro-urokinase precursors defined in formula (I1).

The invention concerns also the host microorganisms transformed with the said expression vectors and thus producing the-precursors of formula (II).

T 1he pharmaceutical compositions containing'a. compound of formula (1) or a compound of formula (11) ahd one or more pharmaceutically acceptable carriers and/or diluents and/or excipients are incl-dded in the scope of the invention as well. The said compositions may be prepared in a conventional way using conventional ingredients and conventional procedures. Pro-urokinase (proUK) is a serine protease of 411 amino acid residues which, like most proteins, ends with a free carl>oxylic acid (Holmes, W.F,. et al. (1985) Biotechnology 3.L 923). According to the invention the pro-urokinase is preferably the recombinantly produced compound. As the last amino acid residue is leucine, it cannot be amidated by the PAM enzyme. In order to obtain proUK in an amidated form by this procedure, it is therefore necessary to add a glycine residue to the C-terminus.

1 1 1 1 i 1 i Description of the preferred embodiments a) Preparation of the amidation precursors of formula by__recx:).rtD-; nant D.NA methods.

The precursors of. the amidated pro-urokinase derivatives of formula (71), i.e. the compounds of formula (I!), were constructed by oligonucleotide directed mutagenesis of the pro-urokinase gene. Alternatively, genes coding for deletion, insertion or substitution mutants of pro-urokinase such as, for example, those described in the European Patent Application NO. EP 338409, can be used as starting material. The principle of the mutagenesis method was to subclone the pro-urokinase or the mutant gene into a vector which can be obtained in a single strand form, such as the phage vector M13. The recombinant single strand was amnealed with a complementary synthetic oligodeoxyribonucleotide containing the COOH extended coding sequence. DNA polymerase and ligase were then used to extend the new strand and to ligate it into a circular form. The newly created heteroduplex DNA was used to transform a cell line into which it was replicated and yielded a progeny where the phage bearing the original proUK gene or the gene with the desired insertion was segregated into two different molecular species. The starting mutagenic oligonucleotide was then used as probe to recognize the newly synthesized precursor gene. Such method is widely described in the literature from which more experimental details can be obtained (Zoller M.J. et al (1982) Nucleic Acid Research 101, 6487). The precursor gene was then inserted in an appropriate expression vector which could direct the synthesis of the new proUK derivative in a proper host microorganism such as a prokaryote, e.g. E.coli and B.subtilis, or a eukaryote, e.g. S.cerevisiae, insect or mammalian cell line. The resulting transformants carrying the precursor gene were cultivated thus allowing the expression of the mutant. Using standard technique we have then extracted and purified the proUK mutant which has been used as substrate for the amidation reaction.

1 Bacteria, in particular --;-or example E. coli, are preferred as host microorganiLsm. For the expression in E. coli we used the plasmid pFC44 already described in the Internw-ional Patent Application No. PCT/WO90/04023 and whose construction is reported below under Exa_mple I. This plasmid is shown in Fig. 7. Using traditional gene manipulation techniques (Maniatis T. et al (1982) Molecular cloning A laboratory manual Cold Spring Harbour), we have replaced the proUK wild-type gene inserted in D -_rC44 with a mutated one coding for the COOH extended Drourokinase. in one specific embodi:-nent the invention provides a new plasmid called pPUK-Gly wherein the pro-urokinase wild type gene or a mutant thereof coding for a deletion, insertion or substitution pto-urokinase derivative, has been sub5tituted I - a glycine-extended mutant.

with a new one coding Jo. Following insertion in E. coli host cells and cultivation of the resulting transformants we have obtained high level expression of the mutant. Conditions for transformation and cultivation are equivalent to the ones previously published in 'the International Patent Application NO. PCT/W090/04023 and described below under Example II. Using standard techniques, we have then extracted and purified the proUK mutant which has been used" as -substrate for the amidating enzyme. BY the same techniques used for preparing the Gly-extended pro-urokinase precursor [formula (II), n=zerol also the precursors carrying additional basic residues at the C-terminus [formula (II) n=1-4) could be prepared. For example, the following specific molecules were obtained: proUK-GlyLys-Arg and proUK-Gly-Arg-Arg. Also in this case, for expression in E. coli cells, the genes coding for the said precursors were inserted in the plasmid pFC44 replacing the proUK wild-type gene.

b) Enzymatic conversion of Dro-UK-Gly-Y.- to Dro-UK-Nq,, The pro-UK-Gly-Y n -OH precursor was incubated with a semi-purified protease-free preparation of PAM enzyme, such as thatfrom rat medullary thyroid carcinoma (N.M Mehta et 1 i 1 1 i i 1 i i T.

al. (1988) Arch. Bioc hem. Biophys. 261, 44-54,) or from the conditioned medium. of cells derived from this tumor (J.P. Gilligan et al, (1989) Endocrinol. 124, 2729-2736) in the presence (for n different from zero) or in the absence (for n equal to zero) of carbox-ye-eptidase B (Boehringer Mannhein.)(---/S, 1:2000 by weight). The enzymatic reaction was Performed in an aqueous buffer such as TES, at pli 7, containing copper ion, ascorbate, catalase, potassium -iodide, SDS and Tween 20. Alternatively, aimidation was carried out on S-sulphonated Pro-MK-GlY-Y and no addition of SDS was required. n In all cases, the extent of amidation was evaluated by the Corbett and Corbett's method (J. Org. Chem. (1980), 45, 2834-2839) based on the determination of glyoxylic acid after derivatization with nitrosobenzene. The amidated product was then purified by traditional techniques. The correctness of the C-terminus was verified by carboxypeptidase, Y (CP-Y) treatment (release of Leu, Ala, Gly and Asn). No anino acid release was observed upon treatment with CP-A, thus confirming that all the Product was in the amidated form. By analogous proceduresamidation could be performed on the deletion, insertion and substitution mutants of pro- urokinase mentioned under a) above.

c) Bioloaical activity Either the compounds defined in formula (I) or those of formula (11) can find a useful application in the same therapeutic field in which pro- urokinase is active, for instance in the treatment of acute myocardial infarction, pulmonary embolism or deep venous thrombosis. Like natural pro-urokinase, the amidated derivatives of the invention and

1 12 their precursors have a greater affinity for plasminogen bound to fibrin in clots than to circulating plasminogen and are therefore characterised by high thrombus selectivity and reduced bleeding risk associated with the treatment. In addition, the COOH-terminal amidated compounds could show improved properties over the corresponding non-amidated enzymes, for example, an increased stability towards most carboxypeptidases and a prolonged half life. The compounds defined in formula (II), in particular the compounds of formula (II) wherein n is an integer fo 1 to 4, can be potential precursors for in vivo amidation.

Example I -

Construction of plasmid pFC44 In order to obtain the expression plasmid"pFC44 it is necessary to go through the following steps:

1) Cloning of the human cDRA gene coding for pro-urokinase To obtain the cDNA clone coding for human pro-u.rokinase, the authors have utilized the protein sequence data published in the literature (GUnzler, W.A., Steffens, G.J., Oetting, F., Kim, SM.A., Frankus.E., and Floh6, L. Hoppe-Seyler's Z. Physiol. Chem., 3_U, p. 1155, 1982; GUnzler, W.A., Steffens, G.J., Oetting, F., Buze, G., and Flobd, L.: Hoppe- Seyler's Z. Physiol. Chem. 3-a, p. 133, 1982; Steffens, G.J., GiInzler, W. A: Oetting, F., Frankus, E., and Floh6, L.: Hoppe-Seyler's z. Physiol. Chem., M, p.1043, 1982). Accordingly, specific probes have been prepared and an appropriate cDNA library has been screened. Oligonucleotides coding for selected peptides of single-chain urokinase-type plasminogen activator (pro-UK) were chemically synthesized (Caruthers, M.H., Gassen, H.G.

i i 1 i i 1 1 i i i 13 and Lang, J.A. (eds) Verlag-Chemie, Weinheim, Deefield Beach. Basel, p. 71, 1982) to serve as specific probes to monitor enrichment of pro-UK mRNA and to select for clones containing pro-urokinase cDN:"% from an enriched cDNA library. The oligomers were 14 to 17 mer in lengt1li and each one was synthesized either as unique sequence (named p7) or in pools containing two (named pl, p2, p3) or 16 (named p16) oligonucleotides as indicated in Fig. 1. The oligomers were tested for specificity to pro-UK by northern hybridization. For this analysis polyA-containing RNA was extracted from the HEp-3 epidermoid carcinoma (Miskin, R., Haemostasis (Switzerland), 11, No. suppl 1, p. 63, 1982). For each oligomer the temperature of the washing following the hybridization reaction was adjusted so as to be 2 to 5"C below the minimal melting temperature, as calculated according to Suggs et al. for hybridizati6n to southern blots (Suggs, S.V., Hirose, T., Miyake, T., Kawashima, E.G., Johnson M.Y., Itakura, K. and Wallach, R.B.: Developmental Biology Using Purified Genes; Brown D.D. and Fox, C.F. (eds) Academic Press, New York, p. 638, 1981). In this text the five pro-UK probes, shown in Fig. 1, reacted with one common major carcinoma mRNA band about 2.3 kb, which is the size expected for pro-UK mRNA. Cloning took place using enriched mRNA fractions from the HEp-3 epidermoid carcinoma. RNA preparations were extracted and enriched about 3 fold on two successive sucrose gradients. cDNA was synthesized according to published procedures (Efstratiadis, A., Kafatos, F.C., Maxam, A.M. and Maniatis, T.: Cell,:L, p. 279, 1976; Buell, G.N., Wickens, M.P., Payvar, F. and Schimke, R.T.J.: Biol. Chem. 25a, p. 2471, 1978) using oligo-dT as a primer. Longer molecules were isolated using polyacrylamide gel electrophoresis followed by electro-elution of the appropriate gel fractions. The cDNA was then extracted using standard

14 phenol/chloroform extraction followed by ethanol precipitation. These cDNA molecules were first ligated to EcoRI linkers and then cloned into the phage lambda-gtlO vector according to a modification of the technique of Davis <Maniatis, T., Fritsh, E.F., Sambrook J.: Molecular cloning. A laboratory manual, Cold Spring Harbour Laboratory, Cold Spring Harbour, NY, 1982). By doing so, a library containing 2xI0 5 pfu (plaque forming units) was constructed. Half of the library was screened on duplicate filters, one filter with 32 P-labelled probe pl, and the counterpart filter with a mixture of probes p3 and p6. A total of 36 positive clones were obtained, seven of which were positive in the duplicate filters, thus indicating cDNA inserts corresponding to a large portion of the pro- UK coding sequence. Recombinant phages that hybridized with the 3 probes were plaques purified using probe pl, and further characterized by restriction mapping with EcoRI and by DNA sequencing. fraction of the positive clones in the total cDNA library indicated that the frequency of pro-urokina. se mRNA in the HEp-3 carcinoma is approximately 0.01%. Sequence analysis of four pro-UK cDNA clones revealed that that three of the clones had deletions or sequences not consistent with the amino acid sequence of the enzyme. Only one clone, lambda-Ucl7, had a sequence with complete concordance with the known amino acid sequence. However, lambda- Ur-17 did not include the entire 3' non-coding end of the mRNA and was missing 30 nucleotides of the coding sequence. A full length pre-pro- urokinase cDNA clone was constructed by ligation of a 1325 bp lambda-Ucl7 SmaI-BamHI fragment, containing the 5' non-coding region and the majority of the coding sequence, with a BamHI-EcoRI fragment containing the remaining missing 3' region from another clone lambda-Uc6 (Fig. 2).

1 i j i 1 1 I 1 1 1 1 i i 1 1 _f is This construct was ligated into the Smaj-EcoRI site of the plasmid vector pUN121 (Nilsson B., Uhlen M., Josephson S., Gatenbeck S. and Philipson L. : Nucleic Acid Research P. 8019, 1983), eliminating, thus, most of the ci gene, and given the name pcUK176 (Fig. 3). The DNA sequence of the complete cDNA clone is depicted in Fig. 4. It consists of 2296 nucleotides, including 69 non-coding nucleotides at the 56 end, 1296 coding nucleotides, and 9331 non-coding nucleotides at the 3' end, followed by a poly(A) tail of more than 80 residues. The coding sequence starts with 60 bp coding for 20 amino acids comprising the "pre- pradrokinase" (Heyneker, H.L., Holmes, W.E. and Vehar, G.A. (1983); European Patent Application Publ. No. 0092182) and is followed by the sequence coding for the entire pro-urokinase protein, which is in complete concordance with the amino'acid sequence. The complete fragment has been checked by sequence and restriction analysis. The sequence coding for mature pro-urokinase has been inserted into the expression vector used for production.

2) Construction of the pro-UK expression plasmid The original full length cDNA, present in pcUK 176, was used to construct the pro-urokinase expression plasmid, pFC44, in which the pro-UK gene is under the transcriptional and the translational control of the promotor Ptrp and of the "Shine-Dalgarno" sequence MS-2, respectively. The plasmid pFC44 is shown in Fig. 7. In order to obtain pFC44, several intermediate plasmids were constructed. Starting with pDS20 (Fig. 5) (Duester, G., Helfard. R.M. and Holmes, W.M.: Cell 3-Q, p. 855, 1982), we have first replaced the EcoRI-HindIII fragment coding for 16 the galactose operon promoter Pgal with the EcoRI-HindIII polylinker sequence from the M13 mp8 vector (vieira, J. and Messing, I.: Gene, 1-2, p. 259, 1982), obtaining a new plasmid, named pAB1 (Fig. 5). The promoter Ptrp has been obtained from the plasmid pDR720 (bought from Pharmacia) as an EcoRI-SalI restriction fragment. This fragment has been inserted in the polylinker region of pAB1 between the EcoRI and SalI site. By doing so, we have obtained a new plasmid, named pFC10 (Fig. 5). pFC10 can be considered as the base vector into which we have inserted the pro-UK gene as well as the "Shine-Dalgarno" sequence from the phage MS-2. To achieve expressi on of mature pro-urokinase, it is necessary to fuse the pro-UK coding sequence, from the first codon of the mature protein to the initiator triplet ATG. This fusion must then be preceded by the "Shine- Dalgarno" sequence. The ribosome binding site (RBS) from the bacterial phage MS-I was known and its nucleotide sequence had already been published (Fiers, W., Contreras, R., Duerinck, F., Haegeman, G., Iserentant, D., Merregaert, J., Min Jou, W,;', Molemans, F., Raeymaekers A., Van den Berghe, A., Volckaert, G. and Ysebaert. M.: Nature, 2_6&, p. 500 (1976). It is thought to be a strong signal for an efficient translation off the mRNA. Therefore, we have chosen this region as translation signal for the production of pro-UK. In order to obtain the correct nucleotide fusion with the Pro-UK gene, we have synthesized a double strand DNA region of the MS-2 RBS directly joined to the beginning of the pro-UK gene. A TaqI site is present on the 25th nucleotide of the mature pro-UK sequence. We have taken advantage of this site and isolated, by chemical synthesis, the following DNA sequence:

9 i t j 1 f 17 HindIII MetSer -AGCTTTA.ATAGACGCCGGCCATTCAAACATGAGGATTACCCATGAGCA 3-AATTATCTGCGGCCGGTAAGTTTGTACTCCTAATGGGTAC = TaqI ATGAACTTCATCAAGTTCCAT-3' TACITGAAGTAGTTCAAGGTAGC-S whidh is f lanked upstream by an HindIII site and downs-L-main by a TaqI site. The initiator codon ATG is shown in bold face. The sequence coding for the beginning of he mature pro-UK sequence is underlined. The synthetic fragment has been used in a ligation reaction with the two following restriction fragments: -the TaqI-BglII fragment from pcU.;(176 (Fig. 3), which carries the pro-UK sequence from nucleotide 155 to nucleotide 392 (see Fig. 4): -the large BamHI-HindIII fragment from pFC 10 (Fig. 5), which carries the antibiotic resistance to ampicillin as well as the promoter Ptrp. Through this construction, we have isolated a new plasmid, named pAB8, whose schematic map is shown in Fig. 6. In this plasmid, the promoter Ptrp and the MS-2 RBS are fused to the first 260 nucleotides of the mature pro-UK gene (corresponding tonucleotides 131- 391 in Fig. 4). In addition, pABB has a unique Ncol site into which we have inserted the rest of the pro-UK sequence through anNcoI-NcoI restriction fragment from pcUK 176 thus obtaining plasmid pFcl6. This ligation has caused the duplication of an NcoI-BglII fragment downstream of the pro-UK gene in the non-coding region. However, this duplication does not affect plasmid stability. Through this construction signals can now 18 direct the synthesis of the complete pro-UK sequence (see Fig. 6). All the plasmids, described so far, were selected in the F,. C21i, K-12 host strain C-600 galK (ATCC 33955), on the basis of ampicillin resistance. Indeed, they carry the gene coding for 13-lactamase, the enzyme responsible for the degradation of the antibiotic ampicillin in the culture medium. Early experiments have shown that pFC16 could be successfully inserted in E. g. type B strains and cause high level production of recombinant pro-UK. However, to comply with international guidelines for the production of recombinant DNA-derived products, we have modified pFC16 to create a tetracycline-resistant plasmid able to express the pro-UK gene at high levels. In particular, from the well- known plasmid pBR322 (Mani atis, T., Fritsch. E.F.,Sambrook, J.: Molecular'cloning, A laboratory manual. Cold Spring Harbour Laboratory. Cold Spring Harbour, KY. 1982) (Fig. 6), we have isolated a EcoRI-AvaI fragmen't where the sticky ends were filled in using the klenow fragment of DNA polymerase 1 (Perbal. B., A. Wiley: Interscience Publication John Wiley and Sons, p. 231, 1984). This fragment was ligated to the larger AatII-PvuI fragment from pFC16, whose ends were made blunt by DNA polymerase 1 (Perbal, B., A. Wiley-Interscience Publication John Wiley and Sons, p. 231, 1984). By doing so, we have replaced the amino terminal portion of the S-lactamase gene and its controlling sequence with the tetracycline-resistance gene. Following ligation the tetracycline resistance gene is in the same orientation as the pro-UK gene. Moreover, a new EcoRI site has been created at the junction between the PvuI and EcoRI sites, previously filled in.

The plasm-id thus obtained is pFC44 (Fig. 7).

1 1 ^ i 1 i i i f i 1 Example II - Transformation of E. coli type B strains Several type B strains of F,. rg_U are available and can be used for successful expression of the COOH-extended pro-UK gene. Preferred strains are: ATCC 12407, ATCC 11303, NCTC 10537. Belcw is an example of transformation of strain NCTC 10537 with plasmid pPUK-Gly and subsequent cultivation of the transformant. Competent cells of strain NCTC 10537 were prepared using the calcium chloride procedure of Mandel and Higa (Mandel, M. and Higa. A.: J. Mol. Bio. 53-, p. 154, 1970). Approximately 200 / ul of a preparation of these cells at 1 x 10 9 cells per milliliter were transformed with 2 / ul of plasmid DNA (approximate concentration 5 / ug/ml). Transformants were selected on plates of L-agar containing 12.5 / ug/ml tetracycline. Two small colonies were streaked with wooden tooth picks (each as three streaks about 1 cm long) onto L-agar containing the same antibiotic. After 12 hours incubation at 370C, portiuns of the streaks were tested for Gly-e;-,tended pro-urokinase production by inoculation onto 10 ml of LB medium (containing tetracycline at a concentration of 2.5 / ug/ml) and incubated ove rnight at 37C. The following day the cultures were diluted 1:100 in M9 medium, containing the same concentration of tetracycline, and incubated for 6 hours at 370C.

claims 1 A COOH-terminal amidated derivative of human pro-urokinase of formula (I):

P-CONIII 2 (1) wherein P represents the portion of human pro-urokinase, or a modification thereof, preceding the carboxy terminus of the molecule.

A derivative according to claim 1 which is COOH-terminal annidated human pro-urokinase.

3. A derivative according to claim 1 which is a COOH-terminal amidated deletion, insertion or substitution mutant of human pro-urokinase.

4. A precursor of a COOH-terminal amidated derivative of human prourokinase according to claim 1 having the formula (II):

P-Gly-Yn (II) wherein P is as defined in claim 1, Gly is a glycine residue, Y is a basic amino arid residue and n is zero or an integer Of 1 to 4.

A precursor according to claim 4 wherein P represents human pro-urokinase.

A precursor according to claim 4 wherein P represents a deletion, insertion or substitution mutant of human pro-urokinase.

A precursor according to anyone of claims 4, 5 and 6 wherein Y is lysine or arginine.

8. A precursor according to claim 4 wherein Yn is Lys-Arg.

9. A precursor according to claim 4 wherein Yn is Arg-Arg.

1 i 1 1 i i 1 i i j i I 1 i j i i A process for obtaining a COOH-terminal amidated derivative of formula (I) according to claim 1 which comprises the enzy-matic conversion f rom a corresponding precursor of formula (II) according to claim 4.

11. A process according to claim 10 comprising the enzymatic conversion of the compound of formula (II) according to claim 4 wherein n is zero by means of an amidating enzyme.

12. A process according to claim 10 comprising the enzymatic conversion of a compound of formula (II) according to claim 4 wherein n is an integer of 1 to 4 through the combined action of an amidating enzyme and the carboxypeptidase B. 13. A process according to claim 11 or 12 wherein the amidating enzyme is peptidylglycine a-amidating monooxygenase.

14. A recombinant DNA process for preparinj a precursor of formula (II) according to claim 4 comprising oligonucleotide directed mutagenesis of the human Pro:rokirLase gene and subsequent insertion of the mutated genes in a suitable expression vector which could direct the synthesis of the new precursors in a proper host microorganism.

15. A process according to claim 14 wherein the expression vector is the plasmid pPUK-Gly and the host microorganism is E. coli.

16. A gene comprising a chemically synthesized polynucleotide coding for a pro-urokinase precursor as defined in formula (J1) according to claim 4.

- 22 17. An expression vector carrying a gene according to claim 16 and capable of directing the synthesis of a pro-urokinase precursor having the formula (II) according to claim 4.

18. The expression vector pPUK-Gly.

19. A unicellular organism or animal cell transformed with an expression vector according to claim 17.

20. An E. coli host cell transformed with the expression vector pPUK-Gly.

21. A pharmaceutical composition comprising an effective amount of a COORterminal amidated derivative of human Pro-urokinase as defined in claim 1 and a pharmaceutically acceptable carrier and/or diluent.

22. A pharmaceutical composition comprising an effective amount of a precursor as defined in claim 4 and a pharmaceutically acceptable carrier and/or diluent.

23. A COOH-terminal amidated derivative of human pro-urokinase substantially as described herein with reference to Example I.

24. A process for obtaining a COOH-terminal amidated derivative according to claim 23 substantially as described herein with reference to Example I.

Published 1992 at The Patent Office. Concept House. Cardiff Road Newport, Gwent NP9 I RH. Further copies may be obtained from Sales Branch. Unit 6. Nine Mile Point. Cwnifelinfach. Cross Keys. Nevport. NPI 7HZPrinted by Multiple.X techniques lid. ST Mary Cray. Kent- V, 1 i i