AU601420B2 - Process for the production of plasminogen activators in procaryotes - Google Patents

Abstract

In order to produce, in a procaryote cell, a plasminogen activator by expressing a c-DNA coding for this activator, which c-DNA is incorporated in a vector that is suitable for the host cell, a strictly controlled vector is used so that its replication cannot be dissociated, or can only be temporarily dissociated from the replication of the chromosome of the host cell. Its culture is carried out in conditions such that no laxity occurs in its control. Plasmids that are particularly suitable for carrying out this process are disclosed.

Description

i- k'

I

caa Gp

I.

COMMONWEALTH OF AUSTRALIA PATENTS ACT 1952 COMPLETE SPECIFICATION (Original) FOR OFFICE USE SThis document contains the amendments made under Section 49 and is correct for printing.

Class Application Number: Lodged: Complete Specification Lodged: Accepted: Int. Class J' t Priority: Related Art: 0 e Published: 4 IeL? A.I

-Q

(<~1ii 2 Name of Applicant: Address of Applicant: Actual Inventor(s): Address for Service: BOEHRINGER MANNHEIM GMBH Sandhofer Strasse 112 132, D 6800 Mannheim-Waldhof, FEDERAL REPUBLIC OF GERMANY Ralf MATTES DAVIES COLLISON, Patent Attorneys, 1 Little Collins Street, Melbourne, 3000.

Complete specification for the invention entitled: "PROCESS FOR THE PRODUCTION OF PLASMINOGEN ACTIVATORS IN PROKARYOTES" The following statement is a full description of this invention, including the best method of performing it known to us 1 I -2- Description The invention concerns a process for the preparation of a plasminogen activator in a prokaryote cell.

Plasminogen activators are wideb distributed in human and animal tissues and body fluids. In the case of fibrinolysis, they play a central part and convert C plasminogen into plasmin. Plasmin is, in turn, of importance for the dissolution of coaulum. Therefore, plasminogen activators are of :;reat interest for the -clinical treatment of vascular occlusions due to blood Y) coagulations.

Known plasminogen activators are, for example, pro-urokinase as well.as tissue plasminogen activator (t-PA).

Pro-urokinase is a serine protease in one-chained form which is activated by plasmin (Eur. J. Biochem.

150 (1985), 183-188, EP-Al 0 092 182).

t-PA is also a serine protease which occurs in different modifications (Arteriosclerosis 4 (1984), 579 585)..

t-PA isolated from tissue or cell cultures is present glycosilated in low active (zymogenic) form as single-chain t-PA (molecular weight about 70 000 Dalton).

By limited action of plasmin, trypsin or kallikrein on the zymogen,h s is converted into the fully active state (two-chained t-PA). The binding between Arg 275 and Ile 276 is thereby cleaved. There results a heavy chain (A chain), as well as a light chain (B chain, r n'

*W

4.

-3protease domain) oth chains remain bound via a disulphide bridge Biol. Chem. 256 (1981), 7035 7041).

Furthermore, slightly differing natural derivatives of t-PA are known. Thus, about 50t' of the molecules begin with Gly-Ala-Arg-Ser-Tyr-Gln (L chain), whereas )0 in the case of the other about 50% of t-PA. Gly-Ala-Arg is missing and the amino acid sequence begins with O Gly/Ser-T'yr-Gln (S chain). Finally, on the position )L-4 or 3-1, an exchange of Gly/Ser can take place (FEBS Letters 156 (1985), 47-50). However, these slight modifications have no influence on the biological effectiveness of t-PA.

Since natural plasminogen activators only occur in nature in extremely small amounts (t-PA e.g. about 6 ng./mlb in plasma), the amounts of these activators needed for clinical use and scientific purposes are advantageously prepared gene technologically. For example, for t-PA there have been described not only methods for the cloning and expression in prokaryotes (Pennica et al, Nature 501 (1985), 214-221) but also i I. for the expression in eukaryotes (Biotechnology, December 84, 1058-1062). Whereas the plasminogen activators expressed in eukaryotes are glycosilated, the glycosilation is absent in the caseof the expression in prokaryotes.

It is to be assumed that the plasminogen activators expressed in prokaryotes are just as biologically .*umm -4active as the glycosilated, naturally-occurring activators because, for example, in the case of t-PA, after splitting off of the carbohydrate chain (Biochemistry 23 (1984), 6191-6195) or inhibition of the glycosilation (Thrombosis and Haemostasis 54 (1985), S788-791), a so produced non-glycosilated t-PA displays properties identical with the glycosilated natural t-PA.

Consequently, because of the non-problematical fermentation of prokaryotes and of the high yields to be expected, a great interest exists for a process for the expression of plasminogen activators in prokaryotes.

However, the previously known processes (cf. for t-PA for example Nature 301 (1985), 214-221 and EP 00 93 619) have the disadvantage that preponderantly a protein is expressed which does not have the desired chain length.

Thus, it was ascertained that in the case of the expression of t-PA c-DNA in E. coli with pBR 322 as vector, more than 80% of the expressed protein displays a molecular weight of only 32 36 000 d and, consequently, consists of decomposition products of t-PA.

Less than 20% of the expressed protein is single chained t-PA, the molecular eight of which should amount to 57 000 d (corresponding to the sequence given in Nature, see above). However, the purification of a mixture of such similar compounds is uneconomic and laborious.

The observed formation of t-PA derivatives has a similarity with a long since known phenomenon in the case of the expression of recombinant proteins in pro-

'I

ii ItI

A

ii S1~A

I

ii :1

U

A

Li karyotes. For various kinds of proteins, it was, namely, reported that the protein was admittedly completely expressed but its half life time in the host cell is only small because it is subject to a proteolysis (Trends in Biotechnology, 1 (1985), 109-113). In order to avoid this proteolytic breakdown, various ways have been suggested. One such way is the use of host cells in which the proteolytic system responsible for the cleavage is missing.

For example a lon mutant of E. coli can be used as host cell. Such mutants are deficient with regard to a protease occurring in the wild type. Since, however, at least seven further proteases are present in E. coli (J.

Bacteriol. 149 (1982), 1027-1033), the process would then only be suitable when the protein to be expressed is not cleaved by these other proteases. Furthermore, the choice of suitable host cells is very limited.

A protein cleavage can also be avoided in -that in the protein the protease cleavage position is removed by amino acid exchange (on c-DNA However, for this purpose, this cleavage position would first have to be determined by laborious processes. Furthermore, the amiho acid exchange can drastically change the activity and the immunological properties of the protein.

It is further known to incorporate into the cloning vector an antiprotease gene of the phage T 4 (Proc. Natl.

Acad. Sci., 80 (1983), 2059-2062). This process is also only suitable for certain recombinant proteins and, in

I

/t 7 SPp Z/2/ addition, only leads to an unsatisfactory suppression of the proteolysis.

It was further suggested to increase the expression by the of ti copy number V_ 6 2 e LU -evteA to such an extent that the proteases are inundated by an e i'rmous amount of recombinant protein and, consequently, manifest their action in the relationship only in the case of a small percentage of the recombinant protein. However, the disadvantage of this method is that the expression must be immensely increased within only 1-2 generations and cannot be carried out for a comparatively long time (Trends in Biotechnology loc.cit.).

A further process for the avoidance of the proteolysis consists in that a recombinant protein already on the c-DNA paae is protected by fusion with a further protein. This process was described, for example, for the peptide hormones insulin and somatostatin, whereby, as fusion partner, there is used p-galactosidase (Science 198 (1977), 1056-1065). However, this process has the disadvantage that the protection protein is coexpressed and, in the case of the purification, must first again be split off and removed, which means a high additional expense. Therefore, none of the previously suggested processes for the prevention of the breakdown of gene technologically- produced proteins appears suitable for solving the problem in the case of the formation of protein mixtures of so produced Pa (plasminogen activators).

IT° iNTR Vio S6 44i' 5/4eC 7 Therefore, the task of the invention consists in the making available of a process which permits the expression of complete, substantially uniform plasminogen activators in prokaryotes and does not display the disadvantages of the above-mentioned methods for the avoidance of a produce cleavage.

It has now been found, that in the case of the use of expression vectors, the replication of said vectors cannot be decoupled or can only be decoupled transiently from the replication of the chromosome of the host cell so, in the case of the expression in prokaryotes, single-chained plasminogen activator with said vectors results in high purity.

Therefore, the process according to the invention for the production of a plasminogen activator by expression of a c-DNA, which codes this activator, which activator is incorporated in a vector appropriate for the host cell, in a prokaryote cell, comprising using a vector which is subject either to strict control and the replication of which thus cannot be decoupled or can only be transiently decoupled from the replication of the chromosome of the host cell and carrying out the culturing under conditions under which no relaxed control occurs, or the promoter of which is regulated in such a way that the m-RNA formation in the case of the decoupled plasmids is not greater than in the case of the plasmids subject to strict control.

wrmu~- p~Ls~ *~K11 -8greater than in the case of plasmids subject to strict control.

The invention depends upon the surprising ascertainment that the undesired formation of a protein mixtures is to be attributed to the appearance of a mixture of m-RNA chains of different compositions and can be prevented by adjustment of a correspondingly low transcription rate.

Vectors which are subject to strict control increase, in the early stationary phase of the growth of the host cell, their copy number by not more than the factor preferably by not more than the factor 3. They are referred to as "low copy" plasmids, in contradistinction to the relaxed controlled multicopy plasmids usually employed in gene technology.

Such vectors suitable for the process of the invention can be found in thqt one tests whether they require active protein biosynthesis or a DNA polymerase I for their reproduction.

From J. Bacteriol. 110 (1972), 667-676, it is known that, by addition of inhibiting materials of the protein biosynthesis, such as chloramphenicol or spectinomycin, the plasmid replication can be decoupled from the replication of the chromosomes in the relaxed controlled vectors vectors of t;he Col E I type, pBR 522) and the copy number of the vectors can be amplified, for example, to more than 100 to 1000. This property is a substantial reason for the preferred use of these

I

9 vectors for synthesis purposes. However, the vectors suitable for the use according to the invention cannot, by means of this process, be amplified by more than the factor 10, preferably by not more than factor 3, to a copy number of max. about Before beginning of the growth, the vectors are present in a copy number of 1 50 copies per cell, preferably 2 to 20, especially preferably 2 to 10 copies.

Insofar as the initial copy number of the vector lies in the lower range (below or equal to 10), those vectors are also preferred, the amplification factor of which lies on the upper limit of the range (near 10). Insofar as the initial copy number already lies at about 20 50, the amplification factor should advantageously be as small as possible. Amplification factors equal to or below 3 are then especially suitable.

As vectors are suitable the preferred prokaryotic plasmid vectors, phase vectors and combinations of both.

Preferred plasmid vectors which are subject to strict control are pRSF 1010, pKN 402 and pACYC 177, as well as plasmids derived herefrom which advantageously carry their characteristic sequence as ori. Especially preferred are the plaKs'ds pKN 402 and pACYC 177. pKN 402 is a temperature-sensitive plasmid. The replication of pKN 402 can be decoupled from the replication of the chromosome by temperature increase. When the growth temperature of the host cell is increased, increased plasmid replication occurs and the copy numbers can increase several hundred fold. The copy number can thereby be increased in a simple way. However, there is hereby then shown a drastic increase of non-single-chain Pa. Therefore, for the culturing, a z i' d,.

-a.x^ temperature must be chosen at which no decoupling takes place. Tables. 1 and 2 show for some vectors, which are subject to strict or relaxed control, copy numbers and formation of single-chain Pa in of the proteins expressed by these vectors.

Alternqtively, the desired low formation of m-RNA copies of the t-PA chromosome, when using vectors which are present in high copy numbers (high copy vectors), can be achieved in that one so regulates their promoter that the m-RNA formation (transcription) leads to a reduced m-RNA formation in the same order of magnitude as in the case of the vectors discussed above, which are subject to stricter control. In the case of use of such high copy vectors with a strong promoter, according to the invention, the promoter is, therefore, only weakly induced so that only a low transcription of the t-PA gene takes place in the m-RNA. For example, for this purpose, host cells can be used which form a repressor which correspondingly inhibits the promoter of the high copy vector, whereby, however, this inhibition can be overcome by addition of an inductor in correspondingly smaller amount. For this, there are suitable, for example, E. coli mutants which do not produce any lac permease and which can be so transformed that they overproduce lac repressors, Therefore, if one uses them in combination with a high copy plasmid with the strong lac promotor, then this is so strongly repressed that the m-RNA formation is suppressed. By addition of

-M

-11an inductor, a desired low m-RNA formation can be achieved. A typical example for such a mutant is the E. coli K12 strain, DSM 2102 or DSM 2093. If this is transSdrmed with the plasmid pePa 119 (DSM 3691P), then it is a question of a lac repressor overproducer. In the case of a high copy plasmid, the t-PAgene of which is under the control of the lac promoter, it can be used for the desired m-RNA formation in that the requisite inductors are added in corresponding underdosing, for example IPTG (isopropyl-o-D-thiogalactoside) or, in the case of a lac permease defective mutant, lactose. In the latter case, the underdosing of the inductor lactose takes place by its greatly slowed down penetration into the cell.

If one uses promoters with lower efficiency than tac, 1 in the case of fidll induction, a correspondikclv higher i i copy number of the vector can be allowed. The promoter strength can be selected inversely proportional to the copy number or the use of the promoter must be limited by correspondingly low induction.

The effect attainable by limitation of the copy.

number of the expression Vector that no t-PA fragments are produced can be achieved in the following way: If one uses a high copy expression vector with a strong promoter (tec) in a strain with lac rppressor overproduction (lac I q then, by addition of inductors (IPTG, lactose or the like), the t-PA synthesis must be indiced. This is completely successful with amounts of U b

U

-12to 5 mM of IPTG in the medium. If, on the other hand, one uses low copy vectors in the same system, one admittedly observes the formation of refractile bodies but no incorprration of fragments of the t-PA in the refractile bodies in the case of adding equal amounts of IPTG.

If one reduces the IPTG addition, in the case of the high copy vectors, to 0.01 mM, then one obtains the same pure refractile bodies as in the low copy system in the case of full IPTG induction. This effect is achieved analogously with poorer promoters than taCo The copy number of the vector can be increased by the value by which the promoter is less efficient than the tac promoter. I.e. in the case of 1/lOth promoter efficiency possible increase of the plasmid copy number by the factor 10 in comparison with the low copy plasmids used.

Amongst the two variants of the process according to the invention, the use of the low copy system is preferred. The advantage of the low copy system with strong promoter according to the invention is to be found in the easier handling of the system in fermentation. Here, the induction can be carried out less expensively because the amount of inductor must not be sharply controlled by regulation but, as described, a single pot system functions. This induction via the lac system is also to be preferred to a control via the tp system because, by use of lactose, the growth of the cells can also be controlled which, in the trp impoverishment technique (gene tech)is less easily possible.

I -i ,l I /2 -13 The molecular weight of the vectors usually amounts to between 10 6 and 10 8 Dalton. Apart from the plasminogen activator -c-DNA, e.g. t-PA-c-DNA, the vectors can advantageously also contain regulation and termination sequences, as well as selection markers.

As promoters, the tac and trp promoter have proved to be especially suitable. The incorporation of the c-DNA into these vectors can take place in any desired orientation.

Table I vector copy number t-PA host (initial) (early stage) (single chain) cell pRSF 1010 pKN 402 300C.

PACYC 177 7r pBR 322 pKN 402 400C.

Table II vector 9-12 9- 12 '15-20 50 10-20 50 30-60 >100 15-20 500 copy number (initial) (early stage)

I

I

90% 90% 90% 20% 10% t-PA derivative (example) 90% 90% 20% E. coli E. coli E. coli host cell E. coli E. coli and P._putida coli/ putida coli pACYC 177 pRSF 1010 pBR 522 10-20 9-12 50-60 50 9-12 100 E. coli r L rul~--- Prokaryote cells are suitable as host cells. The use of E. coli, Klebsiella pneumoniae, Pseudomonas aeruginosa, Pseudomonas putida and Bacillus subtilis has proved advantageous. Of E. coli, there is especially preferred the strain DSM 3689, as well as of Pseudomonas putida the strain DSM 2106. The host cells are expediently so selected according to the vector used that the latter is well replicated in the host. Suitable vector-host combinations are known to the expert. Host cells are expediently used which contain a regulator (repressor) suitable for the particular promoter used, e.g. lac repressor for the lac or tac promoter.

As plasminogen activators,in..the scope of the invention there are understood all proteins showing a plasminogen-activating effectiveness in the abovedefined seqse, preferably t-PA and pro-urokinase, as well as their derivatives.

By derivatives, there are understood not obly the natural derivatives but also artificially produced derivatives from which are absent e.g. one or more domains partly or wholly. The process is equally suitable for derivatives in which one or more amino acids are exchanged.

The process is especially suitable for the expression of natural t-PA, as well as of t-PA derivatives which are derived from the known natural t-PA molecule (cf.

Figure 1) but from which is missing a piece of the L7 7 chain of the complete t-PA molecule beginning with one of the amino axids 45 to 72 and ending with one of the amino acids 179 to 115, as are described in German Patent Application P 36 43 158.3. This nomenclature is analogous to the nomenclature described by Pennica loc. cit.

The introduction of the c-DNA coding the plasminogen i activator into the vector takes place according to methods known herefor to the expert which do not here have to be explained in detail. If, for example, a tPA derivative is to be produced, then one proceeds in such a manner that one cuts out with the corresponding restriction endonuclease -f-rom the tPA-DNA that section which codes the amino acid sequence missing from the derivative in comparison with the complete tPA molecule, which binds the section coding the desired tPA derivative, introduces the so obtained tPA derivative cDNA via a suitable vector according to the invention into prokaryotes and exprbsses therein.

As restriction endonucleases, in this case there are especially suitable those enzymes which only cleave in the region of bp 315 to bp 726 of the cDNA sequence (cf. Pennica, Nature 301 (1983), 214 221).

Especially suitable are the restriction endonuclease combinations Dra III and Mae III for the i L J -16removal of the amino acids 45 179 or DdeI and Nnl I for the removal of the amino acids 45 171, as well as DdeI for the amino acids 45 175 and Fnu 4 II for the removal of amino acids 52 168 and RsaI for the removal of the amino acids 67 162.

For the relinking in the vector, there are used the known methods, preferably '.ith the use of linkers.

Subsequently, these vectors are introduced into prokaryotes according to the known processes and the plasminogen activators, such as e.g. tPA derivatives, expressed.

Furthermore, for the preparation of the finished vectors, one can start from. -the. complete DNA which contains introns. For this purpose, the DNA, e.g, tPA-DNA, is isolated from the Maniatis gene bank, introduced into the selected vector.

Furthermore, from PA-producing cell lines (literature Pennica, loc. cit.), the m-RNA can be isolated and separated by size fractionation. By c-DNA production therefrom and investigation of the resultant clones, the clones can be found which contain the coding region for PA. For this purpose, clones are selected which hybridise with correspondingly constructed oligonucleotides.

According to the invention, especially preferably there is produced a tPA derivative from which the amino acids 45 179 are missing. This can, for example, r be produced on the cDNA p by cleavage with the i I l -17restriction endonucleases Dra III and Mae III and digestion of the remaining individual strand ends with nuclease Sl. Subsequently, it is ligated with ligase.

It is then expressed into prokaryotes with a vector system according to the invention. The tPA derivative thus resulting in the case of the expression into eukaryotes has a molecular weight of about 45 000 D.

According to the invention, one obtains practically pure plasminogen activator mainly in undissolved, inactive form (refractile bodies) which, after isolation, can be converted ihto soluble, active form.

Example 1 Construction of the expression plasmids for tPA: a) Construction of an expression-plasmid with use of a plasmid with the pBR 522 ori: As starting plasmid, there serves the plasmid pBT (DSM 3611 B, DE 35 45 126), from which the plasmid pePa 98.1 was produced in the following way: The plasmid is cleaved with the enzyme Bgl II and post-treated with the nuclease Sl. By further cleavage with the enzyme Sea I, a fragment a can be preparatively obtained which has a size of about 760 base pairs and contains the nucleotide sequence 192-952 coding for tPA (Pennica loc. cit.).

A fragment b is also obtained from pBT 95 by cleavage with Sca I and Hind III; one thereby obtains 4 fragment of about 1200 base pairs which contains the tPA nucleotide sequence 955-2165. A partner c is prepared synthetically as linker and contains the following sequence: 5'GAATTCTTATGTC3'

L

-18- As partner d in the construction, the plasmid pKK 223-3 (DSM 3694 P) is cleaved with the enzymes Eco RI and Hind III in the polylinker. The partners a-d are ligated together. The ligation batch is treated according to usual techniques with T4 ligase and subsequently transformed into E. coli cells (DSM 3689). The transformed cells are cultutra on medium with the addition of 50 p-g./ml. ampicillin. From the clones obtained, those are selected which carry the plasmid pePa 98.1 which differs from the starting plasmids pBT 95 and pKK 223-3 in that in the polylinker of the plasmid pKK 223-3, between the Eco RI and Hind III cleavage points-.---it contains the tPA nucleotide sequence 192 2165 in the correct order.

b) Construction of an expression plasmid for tPA with temperature-sensitive replication control: From the plasmid pela 98.1 prepared in la) there can be obtained, by cleavage with Xho II, an about 1.85 kb sized fragment which codes for tPA and contains the tac promotor. The nucleotide sequence 192 1809 coding for tPA is present unchanged in it.

By treatment with the Klenow enzyme in- the presence of all 4 desoxynucleoside triphosphates, the ends of the Xho II cleavage points are filled. The receptor plasmid pREM 2334 (DSM 3690 the replication and thus the copy numb.er of which can be influenced by temperature control, is linearised with the enzyme Cla I and the 5'-overlapping ends are filled with the -19- Klenow enzyme in the presence of all 4 desoxynucleoside triphosphates. The so prepared partner molecules are combined in a ligation batch. After transformation into E. coli cells (DSM 3689), the cells are cultured on nutrient medium with 25 ijg./ml. chloramphenicol. The temperature for the culturing may only amount to 300C0.

From the so obtained clones, those are selected which contain the plasmid pePa 100.1. This differs from the starting plasmid pREM 2553 in that it contains the Xho II fragment which codes for tPA. This is tested in that one isolates the plasmids of these clones and cleaves with Bam HI. pePa 100.1 molecules can be so divided up into 2 fragmenis "fTat these have a size of about 6.85 and 2.15 kb.

c) Construction of an expression vector for tPA which can be used not only in E. coli but also in Pseudomonas putida.

The Xho II fragment described in Example 1 b), which has been treated with Klenow enzyme, is again used.

As receptor'plasmid there serves the plasmid pREM 5061 (DSM 3692 P)which is linearised with Hpa I. Both partner molecules are treated together with T4 ligase in a ligation batch. After transformation into E. coli cells (DS 3689) and cultilring of the successfully transformed cells on nutrient medium with 25 Lg./ml. kanamycin, clones are obtained. From these are selected those ci c s. which contain the plasmid pePa 107.8. This differs from the starting vector pREM 3061 in that it L YIU~ _I YI^-I-^IU~1-li( I j2 contains the described Xho II fragment. The plasmid is obtained from the cells and cleaved analytically with the enzyme Bst E II. One thereby obtains 2 fragments of the approximate size of 10.8 and 1.5 kb. This plasmid can be introduced by transformation into cells of the strain Pseudomonas putida.

For this purpose, cells of the strain Pseudomonas putida (DSM 2106) are first transformed with the plasmid pePa 119 (DSM 3691 P) and selectioned on nutrientmedium with 25 Lg./ml. kanamycin. These Pseudomonas putida cells now contain a plasmid which is able to produce lac repressor in large amounts.

The pePa 119-contaidirgn-Pseudomonas putida cells are transformed with the plasmid pePa 107.8. The transformants are selectioned on nutrient medium with 50 g./ ml. streptomycin and thereafter contain not only the plasmid pePa 119 but also pePa 107.8. The presende of the plasmid pePa 119 suppresses the production of tPA in the cells by production of the lac repressor protein.

This can reprime the transcription of the tac promotor also in Ps. putida.

d) Construction of an expression plasmid for tPA with the ori of pACYC .177.

There is again used the approximately 1.85 kb sized Xho II fragment which has been treated with Klenow enzyme (from Example Ib). This is called fragment a. A fragment b is prepared from pKK 225-3, namely, by cleavage with Bam HI and treatment with SI nuclease.

-21- Subsequently, it is post-cleaved with the enzyme Pvu I, one thus obtains an approximately 1.05 kb sized fragment which contains transcription terminators and the portion of the 0-lactamase gene. One obtains a fragment c by cleavage of the plasmid pACYC 177 (DSM 3695 P) with Bam HI and subsequent Sl nuclease digestion. This batch is then partially post-cleaved with Pvu I. An approximatley 2.9 kb sized fragment can be obtained preparatively.

The fragments a, b and c are combined in a ligation batch.

After treatment with T4 ligase, this ligation batch is transformed into cells of E. coli (DSM 3689). The cells are subsequently plated on nutrient medium with 25 I.g./ ml. kanamycin and 50 Fg.-/~7mJl Tpicillin. From the resultant clones, those are selected which carry the plasmid pePa 155. This differs from the starting plasmid pACYC 177,'in that it contains unchanged the tPA-coding sequence from position 192 1809 and additionally the transcription terminators from the plasmid pKK 223-3.

The plasmid is isolated and cleaved analytically with the enzyme Bst E II. One obtains two fragments with the size of about 5.9 and 1.8 kb.

Example 2 Expression of the tPA protein in prokeryote cells-.

a) Plasmids produced in la), c) and d) are transformed into E. coli (DSM 5689) which contains a lac Iq plasmid. The transformants are selectioned on media with 50 pLg./ml. ampicillin for the plasmids from la), c) and d) or 25 chloramphenicol for n m substantial reason for the preferred use of these -22the plasmid from lb). The cells which thus contain the plasmids pePa 98.1, pePa 100.1, pePa 107.8 or pePa 155 are cultured in nutrient broth up to OD 550 nm 0.4 with aeration and then induced with the addition of 100 mM IPTG. The culturing takes place at 3 5C0. for the plasmids pePa 98.1, pePa 107.8 and pePa 155 but at 300C.

for the plasmid pePa 100.1. After 4 hours culturing with aeration in the presence of IPDG, the cells are harvested and digested. For this purpose, they are treated for min. on ice with 0.5 g./ml. lysozyme (cell concehtration 10 OD/ml.). There is used a tris buffer, pH 8.6 (0.1 M/l. Tris-HCl with 100 mM/1. EDTA and 100 mM/1.

NaCl). Thereafter, the cell-s ar-e digested by ultrasonic treatment. The so obtained suspension is centrifuged for 10 min. at 20,000 g and the supernatant discarded.

The .sediment contains the tPA protein in inactive form.

This can be dissolved and reactivated according to the method described in DE 55 37 708. However, the pellet obtained can also be dissolved by addition of 1% SDS and 100 mM/1. mercaptoethanol and separated electrophoretically in an SDS-polyacrylamide gel. By coloration of the proteins after electrophoresis has taken place by means of choomassie blue, the protein-bands can be made visible. The extract from the cells which carry the plasmids pePa 100.1, pePa 107.8 or pePa 133 mainly contain 1 protein of the molecular weight of about 57,000 Dalton, which corresponds to the size of the non-glycosylated tPA protein. The extract from the I -23cells which carry the plasmid pePa 98.1 mainly contains material of the size of magnitude of 32 34,000 Dalton and only a small amount of the size of 57,000 Dalton.

The identified protein bands can be identified as being immunologically equal according to the Western blot process with anti-tPA-PAB conjugate from the goat.

b) Pseudomonas putida cells (DSM 2106), which have been transformed with the plasmids pePa 119 and pePa 107.8 according to Example Ic), are cultured at 3000. in nutrient broth up to oD550nm 0.4 with aeration and then induced with the addition of 100 mM IPTG. After 4 hours culturing at 30 C. with aeration in the presence of IPDG, the cells are harvesbed-and digested. One proceeds further as described under 2a). After gel electrophoresis of the cell extract has taken place and coloration of the proteins by coomassie blue, in the extract there is mainly found a protein of the molecular weight of about 57,000 Dalton. This protein is immunologically detectable as tPA protein according to the Western blot process with anti-tPA-PAB conjugate from the goat. Proteins which are smaller than 57,000 Dalton are, as a rule, not immunologically detectable as tPA in the described way.

Example 3 Comparison of the tPA expression with high and low copy number of the vector plasmid:' There is used an E. coli strain (DSM 3689) which contains the plasmid pePa 100.1. The cells are cultued that the m-RNA formation is suppressed. By addition of li -24in nutrient broth up to oDO nm 0.4 with aeration at 50 0 C. By the addition of 100 mM IPTG, they are subseqyeatly induced. The batch is divi&ed up into equally large parts by volume. In each case, one third is further incubated at 30°C. or at 37°C. or at 42°C.

for 4 hours in the presence of IPTG. The cells are subseujantly harvested and digested as described in Example 2.

After SDS gel electrophoresis of the extracts, the tPA protein bands can be made visible with coomassie blue.

The extract from the 500C. batch mainly contains a protein of the molecular weight of 57,000 Dalton. The extracts from the batches at 37°C. and 42°C. contain only a small-proportion of the protein of the size of 57,000 Dalton, but a very large proportion of the size of magnitude of 32,000 to 34,000 Dalton. These protein bands can' be detected as being immunologically equal according to the Western blot process with anti-tPA-PAB conjugate from the goat.

The plasmid pePa 100.1 is more strongly replicated at higher temperatures, which leads to a higher copy number of the plasmid in the cells. This higher copy number acts disadvantageously on the obtaining of inactive tPA protein of the size of magnitude of 57,000 Dalton.

Example 4 Renaturing of tPA proteins from prokaryotes.

The cells fractions obtained according to Example 2 and 5, which contain tPA protein, can be dissolved ^oouou r une liKe), tie t-pA synthesis must be induced. This is completely successful with amounts of r and reactivated according to the method described in DE 35 37 708. From extracts of the cells with pePa 107.8 or pePa 133 or pePa 100,1 cultured at 300C.

there is thereby obtained about 10 times more active tPA than from cells with pePa 98.1 or, however, pePa 100.1 cultured at 370C. or 42 0 C. The acLive tPA obtained is, in each case, also stimulatablp by fibrin. As a rule, the stimulatability is greater than 10 fold.

Example Delection of the half life-determining domains Df-tPA a) Construction of the tPA mutein gene As starting material, there is used the plasmid pePa 98.1. From this pla-iid7iCa4S obtained the tPA-coding fragment with the tac promoter via the Xho II cleavage.

This fragment is about 1.85 kb sized and is completely filled on the remaining ends by treatment with Klenow enzyme in the presence of all four desoxynucleoside triphosphates. In a step A, this fragment is cleaved with the enzyme Dra III and remaining ends digested with the nuclease S1. The fragment of the size of about 370 base pairs is obtained gel electrophoretically. In a step B, the same Xho II starting fragment is cleaved with...the .enzyme .Mae-III- and also post-digested w-ith- S1- Subsequently, this batch is additionally treated with the enzyme Sac I. A fragment of the size of about 700 base pairs is isolated therefrom gel electrophoretically.

In a batch C, the vector plasmid pePa 98.1 is cleaved with the enzyme BamH I and the remaining ends filled l~' -26with the Klenow enzyme with the use of all four desoxynucleoside triphosphates and post-cleaved with the enzyme Sac I. Subsequently, the largest fragment is obtained from this batch gel electrophoretically.

In a ligation batch, the fragments obtained from the batches A, B and C are combined and ligated with the addition of DNA ligase.

The ligation mixture, which contains a lac I q plasmid, is transformed into E. coli DSM 3689 and the resulting transformants are selectioned on nutrient medium with 50 yg./ml. ampicillin. From the so obtained clones, those are selected which contain the desired plasmid pePa 129 which differs-from the starting plasmid pePa 98.1 in that it no longer contains the DNA sequence between the tPA cDNA sequence pos. 501 to 726. Figure 1 shows thenucleotide sequence of the so obtained mutein gene.

b) Construction of an expression vector for the mutein for the expression into E. coli.

From the plasmid pePa 129 prepared according to la) is obtained, by digestion with the restriction enzymes Bam H I and Pvu I, the tPA-coding fragment.

-T lais-mii pWCYC 177 (DSM 36935 P) i-s also cleaTv7d with Bam H I and limitedly with Pvu. Both fragments are ligated together and transformed into E. coli DSM 3689 which contains a lac I q plasmid. By selection on kanamycin and ampicillin with 25 and 50 respectively, one obtains transformants from which those are o i

:I

I

I

'i

I

ii -27are selected which contain the plasmid pePa 137. This differs from the starting vectors in that it, in each case, it contains completely the mutein gene fragment from pePa 129 and the sequence of the plasmid pACYC 177.

Figure 2 shows a restriction cleavage point chart of the plasmid pePa 137.

c) Construction of an expression vector for the mutein for the expression into Pseudomonas putida.

As vector for Pseudomonas putida is used a derivative of the plasmid pRSF 1010 which additionally contains a kanamycin-resistant gene from pACYC 177.

This vector bears the designation pREH 3061 (DSM 3692 P).

This plasmid is linearised-by-use of the restriction enzyme Hpa I. From the vector pePa 129 prepared according to la) is obtained, by Xho II cleavage, an about 1.46 kb sized fragment which contains the tac promotor and the complete mutein gene sequence. By treatment with Klenow fragment in the presence of all four desoxynucleoside triphosphates, the Xho II cleavage points are completely filled. The so treated fragment is li-ated with the linearised vector pREM 3061 and transformed into E. coli cells (DSM 3689). The transformants are selectioned on medium wi-th-25 P g./ml.-kanamycin.

Transformants which contain the plasmid pePa 143 are selected. This plasmid differs from the starting vector pREM 3061 in that it contains the complete mutein gene sequence with tac promotor. This plasmid is isolated and introduced by transformation into cells of the -28strain Pseudomonas putida, DSM 2106. The transformants are selectioned on medium with 25 xg./ml. kanamycin and subsequently analysed. They contain the plasmid pePa 145. Additionally, into these transformant cells can also be introduced by transformation a plasmid pePa 119 (DSM 3691 P) which is compatible with pePa 143 and contains the lac 1 q gene. This plasmid is a derivative of the plasmid RP 4. The presence of this plasmid suppresses the production of the mutein in the cells by production of the lac repressor protein. This can reprime the transcription from the tac promoter.

Example 6 Expression of intact t-PA in E. coli with high copy vectors with strong promoter A derivative of E. coli K-12 is used which possesses a mutation in the lac- -gene and which cannot produce the lac permease. The strain is ED 8654 (DSM 2102) or C600 (DSM 2095). From this strain, variants can be prepared which overproduce lac repressor (see la) in that they are transformed with pePa 119 (DSM 3691P). If, in the bacteria thus obtained, one transforms the plasmid pePa 98.1 (see la), one obtains bacteria which overproduce lac repressor and here synthesise t-PA.

One cultures these bacteria as described under 2a in nutrient broth at 37 with aeration.

a) On attaining Od 5 50 0.4, 0.2% lactose is added to the medium and the culturing continued for 4 hours. Due to the absence of lac permease, only a slow take up of

I

-29the lactose into the cells takes place. As described under 2a, the cells are subsequently harvested, digested, the t-PA fraction obtained and analysed in the SDS polyacrylamide gel.

One obtains qualitatively and quantitatively the same results for the t-PA production as for extracts of fully induced low copy vectors from Example 2.

b) In the same way, by addition of small amounts of ITPG at this time point, a correspondingly low induction rate for t-PA expression can be attained. With amounts of 0.05 to 5 mM IPTG and more, a complete induction is achieved, whereas with amounts between 0.002 and 0.01 mM,A correspondingly 6lowerinduction is attained.

This low induction leads to qualitatively and quantitatively the same result as the full induction of low copy vectors from Example 2.

c) If one uses promoters with lower efficiency than tac, in the case of full induction, a correspondingly higher copy number of the vectors can be permitted. The promoter strength can be chosen inversely proportionally to the copy number or the use of the promoter must be limited by correspondingly lower induction.

The effect attainable by limitation of the copy number of the expression vector that no t-PA fragments are produced, can also be achieved in the followiing manner: 4 *1 If one uses a high copy expression vector with strong promoter (tac) in a strain with lac repressor overproduction (lac I q then, by addition of inductor (IPTG, lactose or thers), the t-PA synthesis must be induced. This is completely effective in amounts of from 0.5 to 5 mM IPTG in the medium. If, on the other hand, one uses low copy vectors in the same system (tac promoter, lac Iq), one admittedly observes the formation of refractile bodies but no incorporation of fragments of the t-PA in the refractile bodies with the same amount of IPTG added.

If one reduces the IPTG addition in the case of high copy vectors to 0.01 mM, one obtains the same pure refractile bodies as in the low copy system in the case of full IPTG induction. The effect should be achieved analogously with worse promoters then tac.

The copy number of the vector can be increased by the value by which the promoter is less efficient than the tac promoter. I e. in the case of 1/lOth promoter efficiency, a possible increase in the plasmid copy number by a factor of 10 in comparison with the low copy plasmids used.

The advantage of the low copy system according to the invention is to be found in the easier handling of the system during the fermentation. Here, the induction can be carried out with lower costs since the amount of inductor does not have to be sharply I 1 ii- 0- -31controlled by regulation but, as described, a single pot system functions. This induction via the lac system is also to be preferred to a control via the type system because, by lactose use, the growth of the cells can also be controlled which, in the case of the type impoverishment technique (genetech) is less easily possible.

Microorganism BMTU 2744 was deposited with the Deutsche Sammlung von Mikroorganismen on Ist October, 1981 and was accorded Accession Number DSM 2093.

Microorganism BMTU 2602 was deposited with the Deutsche Sammlung von Mikroorganismen on 1st October, 1981 and was accorded Accession-Number DSM 2102.

Microorganism BMTU 3123 RM 82 p REM 6677 was deposited with the Deutsche Sammlung von Mikroorganismen on 9th April, 1986 and was accorded Accession Number DSM 3689.

Microorganism BMTU 2749 was deposited with the Deutsche Sammlung von Mikroorganismen on 9th April, 1986 and was accorded Accession Number DSM 2106.

ii

Claims (11)

1. Process according to the invention for the production of a plasminogen activator by expression of a c-DNA, which codes this activator, which activator is incorporated in a vector appropriate for the host cell, in a prokaryote cell, comprising using a vector which is subject either to strict control and the replication of which thus cannot be decoupled or can only be transiently decoupled from the replication of the chromosome of the host cell and carrying out the culturing under conditions under which no relaxed control occurs, or the promoter of which is regulated in such a way that the m-RNA formation in the case of the decoupled plasmids is not greater than in the case of the plasmids subject to strict control.

2. Process according to claim 1, characterised in that one uses a vector, the amplification factor of which is not greater than 3.

3. Process according to claim 1 or 2, characterised in that, as vector, one uses pRSF 1010, pKN 402, pACYC 177 or plasmids derived therefrom which carries its origin of replication (ori).

4. Process according to claim 3, characterised in that one packs the c-DNA of a plasminogen activator into pKN 402, introduces into a prokaryote host cell appropriate for this vector and cultures the host cell at a temperature of 30 0 C. Process according to claim 3 or 4, characterised in that one uses Escherichia coli DSM 3689 as host cell. j\ T R 0 cleavage with Bam HI and treatment with S1 nuclease. -33-

6. Process according to claim 3, characterised in that, in the case of pRSF 1011, one uses Pseudomonas putida DSM 2106 as host cell.

7. Process according to claim 5, characterised in that one cultures Escherichia coli DSM 3689 which contains plasmid pePa 100.1 and/or pePa 107.8 and/or pePa 133 or pePa 137.

8. Process according to claim 6, characterised in that one cultures Pseudomonas putida DSM 2106 which c6ntains the plasmids pePa 107.8 or pePa 147 alone or together with plasmid pePa 119.

9. Plasmid pePa 100.1. Plasmid pePa 107.8.

11. Plasmid pePa 133

12. Plasmid pePa 137.

13. Plasmid pePa 147. Dated this 17th day of December 1987 BOEHRINGER MANNHEIM GMBH By its Patent Attorneys DAVIES COLLISON i 1 l l