FIELD OF INVENTION

The present invention relates to a novel tissue plas ino- gen activator (t-PA) analogue, a DNA construct encoding the analogue, a pharmaceutical composition containing the analogue and a method of preparing the analogue.

BACKGROUND OF THE INVENTION

Delicate enzymatic mechanisms control the fluidity of blood. In response to injury, clotting factors ensure rapid coagulation of blood by conversion of plasma fibri- nogen to insoluble fibrin. Fibrinolytic enzymes are re¬ sponsible for the orderly removal of already formed fibrin clots. If these mechanisms become deranged clots may form inappropriately inside the vascular bed. This can impair the local flow of blood and result in tissue malfunction, followed later by necrosis and replacement with scar tis¬ sue. Depending on the site of damage the injured individ¬ ual will suffer from one of a variety of diseases, in¬ cluding myocardial infarction, deep vein thrombosis, pul¬ monary embolism, and stroke, all of which are serious pub- lie health problems.

Tissue damage does not occur instantaneously upon obstruc¬ tion of the blood flow, but rather develops over some hours. Therefore, rapid restoration of the blood supply to the endangered area will minimize permanent damage. It is commonly recognized that therapies ensuring the reopening of occluded vessels within a few hours would be of great benefit and have the potential of appreciably reducing the number of permanently disabled persons.

One approach to the problem has been to stimulate forma-

REP ACEMEMTSHEEΓ

tion of natural fibrinolytic enzymes. The major fibrino¬ lytic enzyme of the body is plasmin. It circulates in blood in the form of the proenzyme plasminogen, from which it is formed by proteolytic cleavage of a single peptide bond between Arg(561) and Val(562). This cleavage is cata¬ lyzed by the action of plasminogen activators, e.g. tissue plasminogen activator (t-PA) . Plasminogen binds to fibrin clots, and degrades fibrin to soluble fragments when con¬ verted to plasmin.

However, plasmin is a nonspecific protease, and if it is formed in circulating blood it will degrade a variety of plasma proteins _r including fibrinogen, and the coagulation factors V and VIII. Initially, these recations are con- trolled by α2~ ^ant iP ^lasm i ⁿ r ^a highly efficient inhibitor of plasmin present in plasma. As the inhibitor is depleted the systemic effects induced by plasmin become pronounced. Degradation of fibrinogen and clotting factors and accumu¬ lation of inactive fibrinogen cleavage products make blood coagulation inefficient. This might lead to serious epi¬ sodes of haemorrhage. A general activation of plasminogen is therefore hazardous and should be avoided. The central issue of fibrinolytic therapy is thus to obtain an effi¬ cient plasminogen activation and fibrin removal localized at the site of occlusion, and at the same time avoiding the side effects of systemic fibrinogenolytic plasminogen activation.

Several plasminogen activators have been tested clinically for their utility as fibrinolytic agents. Streptokinase, which is a protein derived from 0-hemolytic streptococci, binds to plasminogen. The complex formed is an active en¬ zyme which converts plasminogen into plasmin. In other words, streptokinase turns plasminogen into an autoacti- vating enzyme. Streptokinase does not bind to fibrin clots and the systemic side effects of streptokinase ac-

REPLACEMENTSHEET

tion are pronounced. Moreover, because streptokinase is a bacterial protein it induces a prominent immune response. Some patients possess antibodies to it, either as a conse¬ quence of previous therapy or because of past strepto- coccal infections. This complicates the determination of an effective therapeutic dosage.

Urokinase is a serine protease occurring naturally in hu¬ man urine. It is an efficient activator of plasminogen. However, urokinase activity is not dependent on the bind¬ ing to fibrin and specificity for clot resolution in vivo or in vitro has not been demonstrated. It seems to repro¬ duce many of the systemic side effects known from strepto¬ kinase.

Prourokinase is the proenzyme form of urokinase, in which the two chains of urokinase are still joined by a peptide bond. This molecule is essentially devoid of enzymatic ac¬ tivity but it may have increased clot specificity relative to urokinase itself. Presumably, this arises because the plasmin catalyzed conversion to active urokinase occurs primarily on the clot. The mechanism is unclear, since prourokinase per se seems to have a low affinity for fi¬ brin. Any manifest clinical benefit of the molecule has not been demonstrated.

Tissue plasminogen activator (t-PA) is also a serine pro¬ tease, which under natural circumstances activates plasmi¬ nogen. This activator differs from streptokinase and uro- kinase by its much stronger affinity to fibrin. Further¬ more, the presence of fibrin will cause an up to 1000-fold increase in the plasminogen activation rate indicating the formation of a ternary activation complex between t-PA, plasminogen and fibrin. t-PA is a 65,000 dalton glyco- protein, which exists in two major forms: a one-chain mol¬ ecule, and a two-chain molecule in which the two chains

REPLACEMENTSHEET

are held together by a single disulfide bond. These chains are designated the A-chain for the polypeptide originating from the N-terminal part of the enzyme, and the B-chain for the polypeptide from the C-terminal part of the en- zyme. "The cDNA derived amino acid sequence was first re¬ ported by Pennica et al. (Nature 301 _f 1983, pp. 214-221). The A-chain (the heavy chain) contains a so-called finger domain, growth factor-like domain and two kringle struc¬ tures (Kl and K2) and the B-chain (the light chain) con- tains the active site for serine protease activity. N-ter¬ minal processing differences have been reported ( allen et al., Eur. J. Biochem. 132 _f 1983, pp. 681-686) differing by the absence or presence of three additional residues. In the following, the longer of these sequences is referred to for the numbering of full-length t-PA (cf. Fig. 5A-D below) .

Conversion of one-chain to two-chain t-PA is catalyzed by plasmin and consists in cleavage of a single peptide bond between the amino acids Arg(278) and lie(279) . This cleav¬ age occurs efficiently on the surface of fibrin where plasmin is present. Both forms of the molecule are enzyma- tically active, but the two-chain form has a somewhat higher specific activity. The factor of activation upon conversion to the two-chain form varies with the assay system.

Due to the ability of fibrin to bind t-PA and to strongly stimulate activation of plasminogen by t-PA it was anti- cipated that the use of t-PA in thrombolytic treatment would eliminate the problems arising from systemic activa¬ tion of plasminogen. However, elaborate clinical trials have shown that the very high dosage needed to obtain sig¬ nificant thrombolysis often results in systemic side ef- fects. These effects which are induced by administration of large amounts of t-PA are presumably associated with an

appreciable activity of the one-chain form even in the ab¬ sence of fibrin. It would accordingly be desirable to re¬ duce this activity and at the same time preserve the full fibrinolytic potential of two-chain t-PA when activated on the clot. The purpose of the present invention is to pro¬ vide a plasminogen activator characterized by substantial¬ ly reduced systemic side effects and a predominantly local lysis effect on fibrin deposits.

As mentioned above, t-PA belongs to the serine protease family. Generally these enzymes are synthesized and se¬ creted by the cells as inactive one-chain zymogen forms. They become active enzymes only when subjected to limited proteolysis at a specific site. As exemplified by chymo- trypsinogen activation, this occurs upon cleavage when the positively charged α-amino group of lie(16) is released to interact with Asp(194) adjacent to Ser(195) of the active site. Such salt bridge interaction appears to be necessary for establishing a proper conformation of the active site for proteolytic activity. The activation of most serine proteases applies to this general scheme and the one-chain forms of most serine proteases are genuine proenzymes. t- PA, however, possessing significant activity in the one- chain form is an exception to this rule.

It has been suggested that the active conformation of one- chain t-PA could be established as a result of an alterna¬ tive salt bridge involving a properly placed amino acid residue which could provide the positively charged group even when the Arg(278)-Ile(279) bond remains intact. Whether this interaction involves the e-amino group of Lys(280) as proposed by Wallen (Wallen, P., Pohl, G. , Bergsdorf, W. , Rhanby, M. , NY, T., and Jόrnvall, H. (1983) Eur.J.Biochem 132, 681-686) or it involves still another residue remains to be seen.

REPLACEMENTSHEET

The present invention is based on the recognition that the activity of one-chain t-PA may result from a salt bridge interaction between the e-amino group of the lysine resi¬ due in position 419 and the -carboxy group of the aspar- tic acid residue in position 480. The putative presence of a salt bridge between these lysine and aspartic acid resi¬ dues, predicted on the basis of a computer-graphical model of the protease domain of the t-PA molecule, is tentative¬ ly suggested as the reason for the activity of one-chain t-PA (A. Heckel and K.M. Hasselbach, J. Computer-Aided Mol. Design 2., 1988, pp. 7-14) . There is, however, no in¬ dication that the one-chain form of t-PA may be inacti¬ vated by preventing salt bridge formation at this site, while substantially retaining the activity of the two- chain form of t-PA.

SUMMARY OF THE INVENTION

The present invention relates to a tissue plasminogen ac- tivator (t-PA) analogue in substantially inactivated one- chain form capable of being converted to the active two- chain form, which analogue is modified in the region from position 414 to position -419 or from position 419 to posi¬ tion 426 or from position 426 to position 433.

It is currently believed that a salt bridge may be formed between the aspartic acid residue in position 480 and an amino acid residue within a distance of about 8 A from Asp(480) in the three-dimensional structure of t-PA. This is the case with the amino acid residues from position 414 to position 433 in the t-PA molecule. Theoretically, therefore, each of these residues represent a possible site of salt bridge interaction with Asp(480) .

More specifically, however, the invention relates to a t- PA analogue in which the interaction of a positively

charged amino acid residue in position 419 and optionally also in position 420 with a negatively charged amino acid residue in position 480 is inhibited.

Thus, the present invention surprisingly provides a t-PA analogue which, in the one-chain form, exhibits the prop¬ erties of a proenzyme (i.e. it is substantially inactive), whereas the activity of the two-chain form is fully re¬ tained on plasmin-catalysed cleavage of the one-chain t-PA analogue. Compared to authentic (native) t-PA, this re¬ sults in a fibrinolytic agent with a higher fibrin selec¬ tivity as the fibrinogenolytic activity induced by the t- PA analogue of the invention is substantially reduced re¬ lative to that induced by native one-chain t-PA.

In the present context, the term "substantially inacti¬ vated" is intended to indicate that the t-PA analogue of the invention exhibits little or no catalytic activity in the one-chain form with respect to the amidolytic activity on simple synthetic substrates as well as the plasminogen activating activity.

The amidolytic activity, i.e. the general catalytic acti¬ vity, may for instance be measured by means of an assay involving the use of a chromogenic low molecular weight substrate such as D-Ile-Pro-Arg-pNA (S-2288) (Kabi, Stock¬ holm, Sweden) which is cleaved directly by t-PA and there¬ fore offers a convenient colorimetric determination of the amidolytic activity (Petersen et al., Biochim. Biophvs. Acta 952. 1988, pp. 245-254).

The plasminogen activating activity may for instance be determined indirectly in an assay involving the use of plasminogen and a chromogenic substrate such as D-Val-Phe- Lys-pNA (S-2390) (Kabi, Stockholm, Sweden) which is cleaved by plasmin. In this assay, colour development fol-

REPLACEMENTSHEET

lows an accelerating parabolic curve as plasmin accumu¬ lates. The acceleration constant is directly proportional to the activity of t-PA. As this assay determines the pro¬ teolytic activity of t-PA on the natural substrate plasmi- nogen, it is more closely related to the fibrinolytic ef¬ fect (Petersen et al. , op.cit.) . In both types of assay, it is possible to include fibrin so as to study its stimu¬ lating effect on the t-PA enzymatic activity.

DETAILED DISCLOSURE OF THE INVENTION

In one preferred embodiment of the present invention, the t-PA analogue is one which has no positively charged amino acid in position 419. This may be effected by deleting Lys(419) from the native t-PA molecule. However, as dele¬ tion may possibly lead to incorrect folding of the result¬ ing protein which in turn may lead to inactivation of the two-chain form of t-PA as well, it is currently preferred to replace Lys(419) of native t-PA by a negatively charged or uncharged amino acid residue. Examples of suitable ami¬ no acid residues are Ser, Thr, Glu, Val, Ala, Leu, lie, Tyr, Asn, Asp and Gin. Substitution by for instance Ser or Lys may possibly be preferred in order to avoid too bulky groups in the molecule resulting in insufficient or in- correct folding.

The t-PA analogue carrying a substitution in position 419 may further be modified in position 420, preferably by re¬ placing a positively charged amino acid residue in posi- tion 420 by a negatively charged or uncharged amino acid residue. Thus, in a particular embodiment, His(420) may be replaced by an amino acid residue selected from the group consisting of Ser, Thr, Glu, Val, Ala, Leu, lie, Tyr, Asn, Asp and Gin.

Although it may theoretically be possible to perform a

similar substitution of the negatively charged amino acid residue in position 480 with a positively charged or un¬ charged amino acid residue rather than substituting Lys(419) and optionally His(420), this is currently be- lieved to be less desirable as such substitution may in¬ terfere with the serine protease active site of t-PA and thereby result in inactivation of the two-chain form of t- PA.

T-PA analogues according to the present invention may ad¬ ditionally comprise a modification in the A-chain, always provided that such a further modification has no negative effect on the characteristics of the t-PA analogue of the invention, especially with respect to its plasminogen ac- tivating capability in the two-chain form. Such modifica¬ tions are well known and include deletions or modifica¬ tions of the finger domain and/or growth factor sequence, deletion or modification of kringle 1 or kringle 2 and/or replacement of kringle 1 of t-PA by kringle 1 of plasmi- nogen; Cf. EP 196 920, EP 207 589, EP 231 624, DK 3022/88 and DK 3023/88.

In another aspect, the present invention relates to a DNA construct which comprises a DNA sequence encoding a t-PA analogue which is modified in the region from position 414 to position 419 or from position 419 to position 426 or from position 426 to position 433.

More preferably, a codon specifying a positively charged amino acid in position 419 of the t-PA analogue encoded by the DNA sequence has been replaced by another codon spec¬ ifying a negatively charged or uncharged amino acid. Op¬ tionally, a codon specifying a positively charged amino acid in position 420 of the t-PA analogue encoded by the DNA sequence has additionally been replaced by another codon specifying a negatively charged or uncharged amino

REPLACEMENTSHEET

acid.

The DNA construct of the invention encoding the t-PA ana¬ logue of the invention may be prepared synthetically by established standard methods, e.g. the phosphoamidite method described by S.L. Beaucage and M.H. Caruthers, Te¬ trahedron Letters 22. 1981, pp. 1859-1869, or the method described by Matthes et al., EMBO Journal 3. 1984, pp. 801-805. According to the phosphoamidite method, oligo- nucleotides are synthesized, e.g. in an automatic DNA syn¬ thesizer, purified, duplexed and annealed.

The DNA construct of the invention may also be of genomic or cDNA origin, for instance obtained by preparing a ge- nomic or cDNA library and screening for DNA sequences coding for all or part of the polypeptide of the invention by hybridization using synthetic oligonucleotide probes in accordance with standard techniques (cf. T. Maniatis et al. , Molecular Cloning: A Laboratory Manual, Cold Spring Harbor, 1982) . In this case, a genomic or cDNA sequence encoding native human t-PA may be modified at a site cor¬ responding to the site(s) at which it is desired to intro¬ duce amino acid substitutions, e.g. by site-directed muta- genesis using synthetic oligonucleotides encoding the de- sired amino acid sequence for homologous recombination in accordance with well-known procedures.

Finally, the DNA construct may be of mixed synthetic and genomic, mixed synthetic and cDNA or mixed genomic and cDNA origin prepared by annealing fragments of synthetic, genomic or cDNA origin (as appropriate) , the fragments corresponding to various parts of the entire DNA con¬ struct, in accordance with standard techniques.

A preferred DNA construct of the invention has a DNA se¬ quence substantially as shown in the appended Fig. 5 A-D,

or a suitable modification thereof. The modification may constitute any one of the known modifications of t-PA, such as one of the modifications mentioned above. The mo¬ dification may also be one which involves the substitution of one or more nucleotides in the DNA sequence by one or more others which do not give rise to an altered amino acid sequence of the t-PA analogue, but which, for in¬ stance, provide a convenient restriction enzyme cleavage site in the DNA sequence encoding the analogue.

In a further aspect, the present invention relates to a recombinant expression vector carrying a DNA sequence en¬ coding the t-PA analogue of the invention may be any vec¬ tor which may conveniently be subjected to recombinant DNA procedures, and the choice of vector will often depend on the host cell into which it is to be introduced. Thus, the vector may be an autonomously replicating vector, i.e. a vector a vector which exists as an extrachromosomal en¬ tity, the replication of which is independent of chromo- somal replication, e.g. a plasmid. Alternatively, the vec¬ tor may be one which, when introduced into a host cell, is integrated into the host cell genome and replicated to¬ gether with the chromosome(s) into which it has been inte¬ grated.

In the vector, the DNA sequence encoding the t-PA analogue of the invention should be operably connected to a suit¬ able promoter sequence. The promoter may be any DNA se¬ quence which shows transcriptional activity in the host cell of choice and may be derived from genes encoding pro¬ teins either homologous or heterologous to the host cell. Examples of suitable promoters for directing the tran¬ scription of the DNA encoding the t-PA analogue of the in¬ vention are the SV 40 promoter (Subramani et al., Mol. Cell Biol. 1 , 1981, pp. 854-864), the MT-1 (metallo- thionein gene) promoter (Palmiter et al., Science 222.

REPLACEMENTSHEET

1983, pp. 809-814) or the adenovirus 2 major late pro¬ moter.

The DNA sequence encoding the t-PA analogue of the inven- tion should also be operably connected to a suitable tran- scriptional terminator, such as the human growth hormone gene terminator (Palmiter et al., op.cit.). The vector should further comprise elements such as polyadenylation signals (e.g. from SV 40 or the adenovirus 5 Elb region) , transcriptional enhancer . sequences (e.g. the SV 40 en¬ hancer) and translational enhancer sequences (e.g. the ones encoding adenovirus VA RNAs) .

The recombinant expression vector of the invention further comprises a DNA sequence enabling the vector to replicate in the host cell in question. An examples of such a se¬ quence (when the host cell is a mammalian cell) is the SV 40 origin of replication. The vector may also comprise a selectable marker, e.g. a gene the product of which com- plements a defect in the host cell, such as the gene coding for dihydrofolate reductase (DHFR) or one which confers resistance to a drug, e.g. neomycin, hygromycin or methotrexate.

The procedures used to ligate the DNA sequences coding for the t-PA analogue of the invention, the promoter and the terminator, respectively, and to insert them into suitable vectors containing the information necessary for replica¬ tion, are well known to persons skilled in the art (cf. , for instance, Maniatis et al., op.cit.) .

In a still further aspect, the present invention relates to a host cell transformed or transfected with the recom¬ binant expression vector described above. In the present context, the term "transformed" is used when the vector is a plasmid vector functional in a bacterial, yeast or fun-

gal cell, while the term "transfected" is used when the vector is a viral vector, i.e. one containing a viral ori¬ gin of replication.

The invention also relates to a method of preparing a t-PA analogue in inactivated one-chain form, the method com¬ prising

(a) modifying a DNA sequence encoding native t-PA by re- placing a codon specifying a positively charged amino acid in position 419 of native t-PA by a negatively charged or uncharged amino acid,

(b) inserting the DNA sequence prepared in step (a) into a suitable replicable expression vector,

(c) transforming or transfecting a suitable host cell with the expression vector prepared in step (b) , and

(d) growing the transformed host cell in a suitable growth medium under conditions conducive to the expression of the t-PA analogue, and recovering the analogue from the cul¬ ture.

Optionally, a codon specifying a positively charged amino acid in position 420 of the native protein may additional¬ ly be replaced by a codon specifying a negatively charged or uncharged/neutral amino acid, in step (a) of the method of the invention.

Alternatively, the t-PA analogue may be prepared by pro¬ ducing authentic t-PA by recombinant DNA techniques as de¬ scribed above and subjecting the t-PA produced to suitable chemical treatment to convert it to the t-PA analogue of the invention.

REPLACEMENTSHEET

The host cell may be any cell which is capable of pro¬ ducing t-PA, and is typically a mammalian cell. Examples of suitable mammalian cell lines are the COS (ATCC CRL 1650), BHK (ATCC CRL 1632, ATCC CCL 10) or CHO (ATCC CCL 61) cell lines. Methods of transfecting mammalian cells and expressing DNA sequences introduced in the cells are described in e.g. Kaufman and Sharp, J. Mol. Biol. 159. 1982, pp. 601-621; Southern and Berg, J. Mol. Appl. Genet.

1, 1982, pp. 327-341; Loyter et al. , Proc. Natl. Acad. Sci. USA 79. 1982, pp. 422-426; Wigler et al., Cell 14.

1978, p. 725; Corsaro and Pearson, Somatic Cell Genetics

2, 1981, p. 603, Graham and van der Eb, Virology 52, 1973, p. 456; and Neumann et al., EMBO J. 1., 1982, pp. 841-845. The medium used to cultivate the cells may be any conven- tional medium suitable for growing mammalian cells, such as a serum-containing or serum-free medium containing ap¬ propriate supplements. Suitable media are available from commercial suppliers or may be prepared according to published recipes (e.g. catalogues of the American Type Culture Collection) .

The t-PA analogue produced by the cells will typically be secreted into the growth medium and may be recovered from the medium by conventional procedures including separating the host cells from the medium by centrifugation or fil¬ tration, precipitating the proteinaceous components of the supernatant or filtrate by means of a salt, e.g. ammonium sulphate, followed by purification by a variety of chro a- tographic procedures, e.g. ion exchange chromatography, affinity chromatography, or the like.

In a yet further aspect, the present invention relates to a pharmaceutical composition which comprises a t-PA ana¬ logue of the invention and a pharmaceutically acceptable diluent or vehicle. Before use, the t-PA analogue may be stored in dry, e.g. lyophilized, form and be formulated

for parenteral administration (e.g. for injection or in¬ fusion) by being dissolved or suspended in an appropriate liquid vehicle such as sterile isotonic saline. The dosage of the t-PA analogue to be administered is expected to be in the same range as that employed for other t-PA-con- taining compositions, i.e. about 10-200 mg.

The invention also relates to the use of a t-PA analogue of the invention for the preparation of a medicament for the treatment of diseases or disorders associated with the formation of thrombi in blood vessels. Examples of such diseases or disorders are infarctions, thrombosis and embolism.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention is further described with reference to the drawings in which

Fig. 1A-C shows the step-wise construction of plasmids pt-PABam and pt-PA-937,

Fig. 2 shows the construction of plasmid pt-PA- [K419S,H420T] (formerly pBoel-1079) from plasmid pt-PABam and pt-PA-937 and the synthetic oligonucleotide NOR 518,

Fig. 3 shows the construction of plasmid pt-PA[K419S] from plasmids pt-PABam and pt-PA-937 and the synthetic oligo¬ nucleotide NOR 517,

Fig. 4 shows the construction of plasmid pt-PA[C87S,K419S] from plasmids Zem99-9200 and pt-PA[K419S] ,

Fig. 5 A-D shows the DNA sequence together with the en- coded amino acid sequence of a 1.78 kb t-PA BamHI fragment of plasmid pt-PABam; the 5 ¹ and 3' synthetic adapters are

REPLACEMENTSHEET

indicated with horizontal arrows, and the nucleotide sub¬ stitutions leading to the amino acid substitutions C87S, K419S,H420T, and K419S are indicated below the wild-type amino acid sequence,

Fig. 6 A and B shows the amidolytic activity of authentic t-PA and t-PA[K419S,H420T] before and after plasmin cleavage (Δ denotes native one-chain t-PA, o denotes na¬ tive two-chain t-PA,A denotes one-chain t-PA[K419S,H420T] , • denotes two-chain t-PA[K419S,H420T], and □ denotes one- chain t-PA[K419S,H420T] in the presence of 100 μM trans- aminomethyl cyclohexane 1-carboxylic acid)

Fig. 7 shows the neutralization of α-antiplasmin activity induced by authentic t-PA and t-PA[K419S,H420T] in human plasma (Δ enotes authentic t-PA, and o denotes t-PA [K419S,H420T]) , and

Fig. 8 shows the thrombolytic effect of authentic t-PA (•) and t-PA[K419S,H420T] (o) measured by means of an artery- venous shunt in anesthetized rabbits.

The invention is further described in the following examples which are not in any way intended to limit the scope or spirit of the invention as claimed.

Example 1

I. Construction of a Full-Length t-PA cDNA Clone. The sequence of a native human t-PA cDNA clone has pre¬ viously been reported (Pennica et al., Nature 301: 214- 221, 1983). The sequence encodes a pre-pro peptide of 32- 35 amino acids followed by a 527-530 amino acid mature protein.

A cDNA clone comprising the coding sequence for mature t-

P

PA was constructed using mRNA from the Bowes melanoma cell line (Rijken and Collen, J. Biol. Chem. 226: 7035 - 7041, 1981) as the starting material. This cDNA was then used to construct the plasmid pDR1296. Escherichia coli strain JM83 transformed with pDR1296 was deposited on 10th Decem¬ ber 1985 in the American Type Culture Collection, Rock- ville, Maryland, USA, with the Accession No. 53347.

The coding region for the complete t-PA pre-pro sequence was not present in the isolated cDNA clone pDR1296, and the DNA encoding this region was therefore constructed from synthetic oligodeoxyribonucleotides and subsequently ligated as an adapter to the cloned cDNA. In this synthe¬ tic adapter, cleavage sites for BamHI. EcoRI and Ncol were introduced immediately 5• to the first codon (ATG) of the pre-pro sequence. The naturally occurring pre-pro sequence lacks a convenient restriction site near the middle, for which reason the sequence GGAGCA (coding for amino acids 20 and 19, Gly-Ala) was altered to GGCGCC to provide a Narl site without changing the amino acid sequence.

Likewise the 3' end of the cloned t-PA cDNA was joined to an adapter sequence (with a BamHI site at the 3 * end) made up of synthetic oligodeoxyribonucleotides. Synthesis of all oligonucleotides was performed on an Applied Bio- systems Model 380-A DNA synthesizer using phosphoramidite chemistry on a controlled pore glass support (Beaucage and Caruthers, Tetrahedron Letters 2_2: 1859 - 1869, 1981). The following 8 oligonucleotides were synthesized:

5' adapter: I: GATCCGAATTCCACCATGGATGCAATGAAGAGAGGGCTCTGCTG

44-mer II: TGTGCTGCTGCTGTGTGGCGCCGTCTTCGTTTCGCCCAGCCAGGA 45-mer

REPLACEMENTSHEET

III: AATCCATGCCCGATTCAGAAGAGGAGCCA

29-mer IV: GATCTGGCTCCTCTTCTGAATCGGGCATGGATTTCCTGGCTGG 43-mer

V: GCGAAACGAAGACGGCGCCACACAGCAGCAGCACACAGCAGAGCCC

46-mer VI: TCTCTTCATTGCATCCATGGTGGAATTCG 29-mer

3* adapter: VII: GATCCGATAAGCTCTAGTCGACCTGCAGCCCAAGCTCG

38-mer VIII:GATCCGAGCTTGGGCTGCAGGTCGACTAGAGCTTATCG 38-mer

20 pmol of each of the corresponding pairs A - D of 5 ¹ - phosphorylated oligonucleotides (A: I and VI; B: II and V; C: III and IV; D: VII and VIII) were heated for 5 min. at 90°C followed by cooling to room temperature over a period of 75 minutes. Pairs A, B and C (5 ¹ adapter) were mixed and treated with T ₄ DNA ligase, and the ligated synthetic adapter was isolated as a 120 bp fragment after electro- phoresis of the ligation mixture on a 2 % agarose gel. This 5* BamHI/XhoII adapter was ligated to the 5' 1.23 kb XhoII/SacI fragment of the t-PA cDNA from plasmid pDR1296; the ligation products were separated on a 1 % agarose gel and a 1.35 kb fragment was eluted and cloned in a BamHI/SacI digested pUC19 vector to generate plasmid pt- PA5'. Restriction enzyme digestions with Bglll on isolated plasmids were used to identify correct DNA constructs. Correct ligation of the two XhoII sites generated a Bglll site at the junction between the 5 ¹ 120 bp adapter and the 5 ¹ 1.23 kb t-PA fragment.

The 3 ¹ 0.39 kb SacI/XhoII t-PA fragment from the t-PA cDNA

clone pDR1296 was ligated to oligonucleotide pair D, and the ligation products were separated on a 5 % polyaery1- a ide gel. A 0.43 kb fragment was eluted and cloned in a BamHI/SacI digested pUC19 vector to generate plasmid pt- PA3 ' . Restriction enzyme digestions on isolated plasmids were used to identify correct DNA constructs.

Finally the full-length prepro-t-PA cDNA with its 5 ^» and 3' flanking synthetic adapters was assembled by ligation of the 1.35 kb 5 ¹ t-PA fragment from pt-PA5 ^f and the 0.43 kb t-PA fragment from pt-PA3 • with a BamHI digested and alkaline phosphatase- treated pUC13 vector to produce plasmid pt-PABam. The sequence of the resulting 1.78 kb t- PA BamHI fragment in ptPABam is shown in figure 6. The DNA sequence of the prepro-t-PA cDNA was confirmed by sub- cloning and sequencing of DNA fragments in M13 vectors. In figure 2 all synthetic DNA (5 ¹ and 3 ¹ adapter frag¬ ments) is indicated with horizontal arrows.

II. Mutation of t-PA cDNA to substitute Ser for Lvs in position 419 and Thr for His in position 420. From ptPABam a 0.43 kb SacI/BamHI fragment, 'which con¬ tained the 3 ¹ end of the t-PA cDNA was isolated. It was ligated to a SacI/BamHI digested GEM3 vector (Promega Biotec), and the resulting plasmid pt-PA-937 was used in mutagenesis experiments in E. coli as described by Mori- naga, Y., et al (BIO/TECHNOLOGY, 2 : 636 - 639, 1984). Two mutagenic oligonucleotides, the 31'mer (NOR 518):

5' (AGACAAGGCCTCAGTACTGCCGTAGCCGGAG)3•

and the 31'mer (NOR 517):

5 ^» (AGACAAGGCCTCATGGGAGCCGTAGCCGGAG) 3 *

were used. NOR 518 has four mismatches to human wild-type

REPLACEMENTSHEET

t-PA cDNA in the region encoding the amino acid sequence Ser-Gly-Tyr-Gly-Lys-His-Glu-Ala-Leu-Ser (residues no. 447 - 456 of the t-PA precursor) . After in vitro mutagenesis the sequence reads Ser-Gly-Tyr-Gly-Ser-Thr-Glu-Ala-Leu- Ser, and a mutation Lys-His to Ser-Thr has been made (K419S, H420T) . This mutation creates a new Seal restric¬ tion site in the cDNA sequence. NOR 517 has three mis¬ matches to human wild type t-PA cDNA in the region en¬ coding the amino acid sequence Ser-Gly-Tyr-Gly-Lys-His- Glu-Ala-Leu-Ser (residues no. 447 - 456 of the t-PA pre¬ cursor) . After in vitro mutagenesis the sequence reads Ser-Gly-Tyr-Gly-Ser-His-Glu-Ala-Leu-Ser, and a mutation Lys to Ser has been made (K419S) .

The mutant genotypes were identified by colony hybridiza¬ tions with ³² P labelled mutagenic oligonucleotides NOR 518 and NOR 517, and positive clones were sequenced to ve¬ rify the mutations.

From the resulting mutant plasmids: pt-PA-937[K419S,H420T] and pt-PA-937[K419S], the 0.43 kb SacI/BamHI fragments were isolated and ligated to the 5* end of human t-PA cDNA (as a gel-purified 1.35 kb BamHI SacI fragment from pt¬ PABam) in the BamHI digested mammalian expression vector Zem219b (described in DK patent application No. 3023/88) . Two 6.6 kb plasmids pt-PA[K419S,H420T] and pt-PA[K419S] with the mutant t-PA_cDNAs inserted in the correct orien¬ tation, were identified by restriction enzyme digestion.

The 5 ^» end of the t-PA mutant 9200 (described in DK patent application No. 3023/88) was isolated as a BamHI/SacI fragment from the plasmid Zem99-9200 described therein, and ligated to the 0.43 kb SacI/BamHI fragment from pt-PA- 937[K419S] in BamHI digested Zem219b to generate pt- PA[C87S,K419S] . From this construction a t-PA mutant may be produced, in which the longer half-life of the t-PA mu-

tant 9200 is combined with the low fibrinogenolytic acti¬ vity of t-PA[K419S].

III. Expression of mutant t-PA cDNAs in mammalian cells.

For expression of mutant t-PA in cultured BHK cells (Syrian Hamster, thymidine kinase mutant line tk~tsl3, Waechter and Baserga, Proc. Natl. Acad. Sci. USA, 9.- 1106 - 1110, 1982. American Type Culture Collection CRL 1632) expression vectors pt-PA[K419S,H420T] , pt-PA[C87S,K419S] and pt-PA[K419S] were introduced in three separate experi¬ ments into cells by the calcium phosphate-mediated trans- fection procedure (Graham and Van der Eb, Virology, 52: 456 - 467, 1973) . 48 hrs after transfection, cells were trypsinized and diluted into medium containing 400 nM methotrexate (MTX) . After 10 to 12 days, individual colo¬ nies from each of the three experiments were cloned out and expanded separately. Screening for t-PA mutant produc¬ tion was done by an enzyme-linked immunosorbent assay (ELISA) with monoclonal antibodies against wild-type human t-PA.

IV. Purification of t-PA analogues from cell culture supernatants.

The human one-chain tissue plasminogen activator analogues t-PA[K419S], t-PA[K419S,H420T] and t-PA[C87S,K419S] were purified from the cell culture supernatants by immunoaffi- nity chromatography as previously described (Petersen, L.C., Johannessen, M. , Foster D. , Kumar, A. and Mulvihill, E. (1988), Biochim.Biophys.Acta. 952 , 245-254). The speci¬ fic activity was determined by the fibrin clot lysis assay (Gaffney, P.J., Templeman, J. , Curtis, A.D. and Campbell, P.J. (1983) , Thromb. Haemostas ___3, 650-651) . A value of 60 IU/μg was obtained for the one-chain form. This increased to 320 IU/μg when it was converted into the two-chain form

REPLACEMENTSHEET

by incubation with 8.8 nM plasmin for 30 minutes. The pro¬ tein concentration was determined by ELISA, and the clot lysis activity was calibrated against an international re¬ ference t-PA 83/517 obtained from Dr. P.J. Gaffney (National Institute of Biological Standards and Control, London, UK) . The purified t-PA analogues contained less than 2% of the two-chain form as judged from reduced SDS- PAGE. Authentic, human recombinant t-PA expressed by BHK cells was purified by a similar procedure and used for comparison.

Example 2

Functional properties of Lvs419 substitution analogues

I. Intrinsic activity of one-chain Lvs419 substitution analogues

The catalytic activity of one-chain t-PA analogues was studied with chromogenic peptide substrates and with plas¬ minogen. The chromogenic substrates Ile-Pro-Arg-pNA (S 2288) , and Val Phe-Lys-pNA (S 2390) and human fibrinogen - and plasmin were obtained from ^' Kabi (Stockholm, Sweden) . Human Lys ⁷⁷ -plasminogen was purified as previously described (Thorsen, S., Clemmensen, I., Sottrup-Jensen, L. and Magnusson, S. (1981), Biochim.Biophys.Acta. 668. 377- 387) . Thrombin was obtained from Leo Pharmaceutical Pro¬ ducts (Ballerup, Denmark) . The amidolytic activity with S 2288 is shown in Fig. 6A and B. The reaction mixture con- tained 10 nM t-PA, 1 μM aprotinin, 0.6 mM S2288, 0.1 M NaCl, 0.05 M Tris Cl, 0.01% Tween 80, pH 7.4, 25 ^β C. The intrinsic activity of authentic one-chain t-PA with this substrate amounted to about 25% of the two-chain activity, whereas the activity of t-PA[K419S, H420T] was about 10- fold lower. Incubation with 6 nM plasmin for 30 minutes resulted in full conversion of both proteins into their

two-chain forms and restored full two-chain activity also with t-PA[K419S,H420T] . The results indicated that the one-chain t-PA[K419S,H420T] possessed at most a few per¬ cent of the activity of its two-chain counterpart. It could not be excluded that this activity was fully or in part explained by two-chain contamination. In any case, one-chain t-PA[K419S,H420T] was far less active with this substrate than authentic one-chain t-PA. This was also true for the activity with the natural substrate, plasmi- nogen, in the absence of fibrin (results not shown) . Simi¬ lar results were obtained with t-PA[K419S] and t- PA[C87S,K419S].

The activity with plasminogen in the presence of fibrin is shown in Fig. 3B. This was determined as previously de¬ scribed (Petersen et al., Biochim.Biophys.Acta 952. 245- 254, 1988). The reaction mixture contained 0.06 nM t-PA, 0.2 μM Lys ⁷⁷ -plasminogen, 0.6 mM S2390, 0.15 μM fibrino¬ gen, 0.1 NIHU/ml thrombin, 0.2 μM aprotinin. Other con- ditions were as described above. The activity with one- chain authentic t-PA was indistinguishable from that of the two-chain form. Also the plasminogen activation with two-chain t-PA[K419S,H420T] proceeded at a similar rate in this assay. However, the activity of one-chain t- PA[K419S,H420T] was somewhat decreased, although the major effect might well be caused by a delay in the lag phase before the maximum stimulation of plasminogen activation (maximum slope of the progress curve) was reached.

II. Fibrinogenolvtic and thrombolytic activity of t- PAΓK419S.H420T1

The plasminogen activation in plasma induced by authentic t-PA and t-PA[K419S,H420T] was tested by measuring the de- crease in antiplasmin activity. This parameter is an in¬ dication of the systemic effects expected due to plasmin

REPLACEMENTSHEET

generation in the absence of a fibrin clot. Human citrate plasma was incubated for 3 hours with various amounts of t-PA as indicated. Then the plasma was diluted 10-fold in buffer containing 200 μM trans-4-(aminomethyl)cyclohexane- 1-carboxylic acid and analysed for α ₂ -antiplasmin activi¬ ty. The plasmin inhibition capacity of a plasma control without t-PA was set to 100%. The results show (Fig. 7) that non-specific plasmin generation with authentic t-PA was obtained at a much lower dose than with t- PA[K419S,H420T] . A 50% decrease in antiplasmin over 3 hrs was observed with 1.9 μg/ml t-PA whereas a similar de¬ crease was only observed with 7.5 μg/ml t-PA-1079. This indicates that systemic, fibrinogenolytic effects caused by Lys419 substitution analogues are much less pronounced than with authentic t-PA.

The fibrinolytic efficiency .in vivo was tested in a rab¬ bit model with an external shunt from the carotid artery to the jugular vein. A pre-formed ¹²⁵ I-fibrinogen-con- taining whole blood thrombus was placed in the shunt and t-PA infusion was initiated. Various amounts of t-PA per kg body weight was infused via an ear vein over one hour and the ¹²⁵ I-fibrinogen content of the thrombus was monitored with a gamma-counter. As shown in Fig. 8 the t- PA[K419S,H420T] analogue was as efficient as authentic t- PA in dissolving the thrombus.