AU609829B2

AU609829B2 - A factor viii:c-like molecule with coagulating activity

Info

Publication number: AU609829B2
Application number: AU18103/88A
Authority: AU
Inventors: Egon Amann; Bernhard Huber; Hans Kupper; Klaus-Dieter Langner; Gerd Zettlmeissl
Original assignee: Behringwerke AG
Current assignee: Siemens Healthcare Diagnostics GmbH Germany
Priority date: 1987-06-19
Filing date: 1988-06-17
Publication date: 1991-05-09
Anticipated expiration: 2008-06-17
Also published as: FI882890A0; FI882890A; DK332888D0; DK332888A; EP0295597A3; PT87752A; DE3720246A1; KR890000666A; EP0295597A2; PT87752B; JPS6425799A; AU1810388A

Description

11111lIII~- I068L99VC6L 1.2 [ZAXMAliSN dONW1)4F1H edIG DqV 'Id 01 1.1 5 .4 11111.61

O.

U1.25 1.4 111.

Form COMMONWEALTH OF AUSTRALIA PATENTS ACT 1952-69 COMPLETE SPECIFICATION

IORIGINAL)

Class I t. Class ApAication Number: Lodged: rdmplae Specification Lodged: 0 4vtority: 00 T96ated Art: Accepted: Published; dth ,i c cn t co nt i 1- ;ecti)n 49 arid is C-rrect fol Name of Applicant: BEHRINGWERKE AKTIENGESELLSCHAFT Address of Applicant: D-3550 Marburg, Federal Republic of Germany 0 0 00. 0 Aual Inventor: KLAUS-DIETER LANGNER, EGON AMANN, HANS KUPPER, GERD ZETTLMEISSL and BERNARD HUBER EDWD. WATERS SONS, 50 QUEEN STREET, MIELBOURNE, AUSTRALIA, 3000.

Ad dress f or Servica Complete Spe~cification for the invention entitled: A FACTOR VIII:C-LIKE MOLECULE WITH COAGULATING ACTIVITY The following statement is a full description of this invention, Including the best mexhod of performing it known to US Company and Signatures of itsonicers B. Mischlewski prescribed by its Articles of Aptoclaticn.

Reg.st.ered. a ent To: THE COMMISSIONER OF PATENTS.

Edwd. Waters Sons.

Melloulne.

1-.

BEHRINGWERKE AKTIENGESELLSCHAFT HOE 87/8 021 Dr.LP/rh Specification A factor VIII:C-like molecule with coagulating activity The invention relates to a mutant of factor VIII:C (F VIII:C) which has the feature that the new molecule lacks the amino acids Ser 7 4 1 to Arg1689, where the numbering of the amino acids relates to the mature F VIII:C molecule (after elimination of the signal peptide). The latter molecule is composed of 2332 amino 1 2332 acids (Ala -Tyr2332). Despite drastic changes in struc- 10 tural parameters (for example the absence of several SH groups and glycosylation sites) and in the number of thrombin cleavage sites, the new protein has a pro- .o coagulating activity very similar to that of the authen- 0 0 tic protein.

oo o15 In the blood coagulation cascade, F VIII:C occupies a central position, serving as cofactor in the activation 0 o of factor X by activated factor IX. F VIII:C is the 0o e protein which in people suffering from classical hemophilia A is defective or entirely absent. Hemophilia A is a disease inherited through the X chromosore and 0 oo occurs in male individuals with a frequency of about 00.: 0.01%.

Both cDNA coding for human F VIII:C and the complete r VIII:C gene have been isolated and characterized (Gitschier et at., 1984, Nature 312, 326-330; Wood et al.

1984, Nature 312, 330-337; Vehar et al., 1984, Nature 312, 337-342; Tcole et al., 1984, Nature 312, 342-347).

The cDNA codes for 2332 amino acids of the mature protein, whici corresponds to a molecular weight of 264,763d.

The purification of F VIII:C from commercial F VIII:C concentrates has resulted in a number of polypeptides associated with biological activity and having molecular uIu-CLAxu at marDurg, reaerai KepuDiic or uermany this Ist day of June 1988 To the Commissioner of Patents BEHRINGWEE AKTIENGESELLSCHAFT Prokurist okuris ppa. Stein ppa. Bu7 S- 2 2 weights of 80,000-200,000 d, with a protein chain comprising 90,000 d and one comprising 80,000 d representing an important intermediate stage ir the activation process (Andersson et at., 1986, Proc. Natl. Acad. Sci., USA 83, 2979-2983). In the intact molecule these two polypeptide chains are Linked by a Ca 2 bridge and, in this 'orm, represent a partially active form of F VIII:C, which is processed by thrombin.

The 90,000 d chain located at the N terminus is derived from F VIII:C precursor molecules by thrombin cleavage at the position of Arg 7 4 0 (ALal-Arg 7 4 0 whereas the 0 80,000 d protein is produced by thrombin cleavage at the position Arg 164 8 (Glu -Tyr2332). To achieve full activation of F VIII:C, thrombin cleavage is carried out at position Arg 3 7 2 in the 90,000 d fragment located at the N terminus (resulting fragments: 50,000 d and 40,000 d) and at position Arg in the C-terminal 80,000 d fragment (resulting fraqments 73,000 d and 7,000 d) (Eaton et al., 1986, Biochemistry 25, 505-512; Eaton et 20 al., 1986 Biochemistry 25, 8343-8347; Andersson et al., a oo see above).

So a 0 0 Based on a modified F VIII:C cDNA, Toole et at. (1986, Proc. Natl. Acad. Sci., USA 83, 5939-5942) were able to a0o express biologically fully active protein molecules in 25 animal cells, in which the region between amino acids Ser 9 8 2 and Leu 56 2 and between amino acids Thr 7 60 and Asn 16 3 9 is deleted. However, these molecules carry all four thrombin cleavage sites regarded as essential for 372 740 the proteolytic activation (positions Arg Arg Ag 16 4 8 and Arg 16 8 In these experiments the proteins with the deletions were expressed as a continuous chain.

Burke et al. (1986, J. Biol. Chem. 261, 12574-12578) expressed the regions Alal-Arg 7 4 0 and Glu 1 64 9 -Tyr 2 3 2 as separate chains together in one host cell and were able likewise to detect in the medium biologically active F VIII:C molecules containing both chains in associated 1 i 3 form. The two protein chains correspond to the species produced from F VIII:C from plasma by thrombin cleavage from the sites Arg 7 40 and Arg 164 8 with a molecuLar weight of 90,000 d and 80,000 d respectively. As with F VIII:C from plasma, the thrombin cleavage sites at positions A 3 7 2 and Arg 16 8 9 which are necessary for further acti- Arg and Arg which are necessary vation, are present in these species.

Thus, in aLL the mutants of the F VIII:C molecule which have been described hitherto, both of the Last-mentioned thrombin cleavage sites are present in their natural form.

Deletion of amino acids Ser 7 4 1 to Arg 16 89 in the mutant to which this invention relates results in retention of 0o u0 372 o o only the cleavage site at Arg of the four thrombin oo o oo.o cleavage sites which are regarded as essential for the 15 activation of natural F VIII:C (at positions Arg Arg 7 40 Arg 16 4 8 and Arg 16 8 9 The cleavage site at Arg 16 48 is completely absent, and only the adjacent oo.oo 740 1689 amino acids of the cleavage sites at Arg 7 40 and Arg 16 8 9 are present. Owing to the said deletion, amino acids 20 Arg 7 4 0 and Ser 16 9 0 are linked by a peptide bond. This a on o 0 linkage represents a new thrombin cleavage site in this 0 0 0 new mutated molecule which is derived from F VIII:C and is expressed as a continuous chain.

It has been possible to insert and to express the novel 00° 25 F VIII:C gene coding for this new protein into animal o cells. It was possible to detect a molecule having F VIII:C-specific procoagulating activity in the culture supernatant from animal cells modified in this way. The activity of this protein can be increased by an order cf magnitude by thrombin treatment.

According to the invention the activation of the mutant by thrombin results, because of the deletion of potential activation cleavage sites, in defined cleavage patterns which are simpler than for the wild-type protein and than 16 Table 3: Factor VIII:C-expressing BHK clones cotransfected with pRMH140 .9 I P st T V T A% for deletion mutants described by other groups. This results in the possibility of a direct experimental approach to the correlation of thrombin cleavage and biological activity. The simplified mechanism of activation may be advantageous when the molecule is used in replacement therapy. Surprisingly, in spite of the lack of two thrombin cleavage sites the new molecule has at least the same biological half-life for inactivation by thrombin compared to the authentic protein. However, the activation of this new molecule is by more than a factor of two faster.

The invention also includes alleles and variants of the mutant F VIII:C. These are functional equivalents of the mutant F VIII:C. The term "allele" can be defined as one of several alternate forms of a gene occupying a given locus on the chromosome. "Variant" is defined as a naturally occurring variation of a gene without any physiological effect.

20 of the 25 potential asparagine glycosylation 0 0 o 0 sites (Vehar et al., see above) have been removed in the new mutant owing to the deletion of the Ser741-Arg 16 8 9 o. polypeptide region. This fact, and the absence of a large 0 00 considerable protein contribution, simplifies the production 0 0° of this molecule in the genetically altered host cell, oo because the host organisms used for this purpose need to 0 25 perform less protein synthesis and glycosylation. It is known that heterologous proteins are, when synthesized in yeasts and in animal cells, not glycocylated in exactly the same way as the corresponding naturally occurring protein (Broker et al. (1987), Biochim. Biophys. Acta 908, 203-213; 30 Conradt et al., 1986, Carbohydrate Res. 149, 443-450). This "incorrect" glycosylation may result in a reduced physiological half-life and undesired immunological reactions. Thus removal of most of the potential N-glycosylation sites drastically reduces the possibility of "incorrect" glycosylation in the host organism.

4a Examples Cloning of V 'VtIE,.C CDNA from human liver tissue was prepared by known methods (Maniatis, Fritsch, E.F. and Sambrook, J. 1982, Molecular Cloning, Cold Spring Harbor Laboratory) 0 o 0 o 0 200 o 0 2 o 0 00 0 o 0 0 0 0 0 00 a000 o 0 4t Co a i: 00 00 oo 0 0 0 0 0 0 0 0 o 00 0 0 0 o oa 0 e 0 00 o 00 0ooo000 5 for the cLoning of F VIII:C-specific DNA sequences. This cDNA was used to establish gene banks in the vectors and Xgtll (Young, R.A. and Davis, R.W. (1983), Proc.

NatL. Acad. Sci. USA 80, 1194-1198). In addition, several human genomic gene banks in the phage vector X Charon 4a (Lawn, et aL. (1978) CeLL 15, 1157-1174) were prepared. These gene banks were examined for F VIII:C-specific DNA sequences. The basis for the isolation of aLL F VIII:C-specific clones were 17 synthetic oligonucLeotides. These oligonucleotides are based on the sequence published by Wood et al. (Wood et al. (1984), Nature 312, 330-337) and correspond to short partial sequences of the coding or of the non-coding strand (Tab.

The said oligonucleotides were used either after 15 terminal radiolabeling as probes for screening the abovementioned gene banks or for the specific priming of new cDNA synthesis using human liver mRNA. Specific hybridization was used to isolate from the cDNA banks four different cLones, and from the genomic gene banks one clone.

Taken together these five clones cover the entire F VIII:C sequence (Fig. 1; intron DNA in the genomic DNA clone XK20 is represented by a broken Line. The base numbering in Fig. 1 relates to exon DNA). Isolation of the restriction fragments shown in Fig. 2 from these clones, 25 and ligation of these fragments into the vectors pUC18 or pUC19 (Norrander et al. (1983) Gene 26, 101-106; Yanish- Perron et aL. (1985) Gene 33, 103-119) resulted, via several intermediate steps, in a plasmid which harbors the complete F VIII:C coding sequence. This plasmid, pGX3254 (Fig. is 10,623 base pairs Long and contains the F VIII:C coding region of 7,056 base pairs (corresponding to 2351 amino acids of F VIII:C) as weLL as 71 base pairs of the 5'-untranslated region and 821 base pairs of the 3'-untransLated region. The entire F VIII:C coding sequence in pGX3254 was determined by the dideoxy method (Sanger et aL. (1977) Proc. Natl. Acad. Sci. USA 74, 5463-5467) and is identical to the F VIII:C sequence published by Wood et al. ALL the sequence coordinates 00 0 o a 4 6 indicated hereinafter relate to the F VIII:C sequence published by Wood et al. The vector sequences of pGX3254 have contributions from pUC18 and contributions from pUC19. When the F VIII:C coding sequence was constructed using suitable restriction sites on plasmid vectors, a synthetic XhoI Linker (GCTCGAGC) was inserted at the naturally occurring HpaI site in the 3'-non-coding region (position 7433) of the F VIII:C sequence (pGX3243).

Furthermore, subcloning of a PvuII fragment which carries the F VIII:C sequence from -71 to 1038 from X98 (Fig. 1) into the SmaI site of the vector pGX3237 (polylinker region: ClaI-SphI-XhoI-XbaI-EcoRI-SmaI-BamHI-SaI-PstI- HindIII) resulted in the plasmid pGX3238 which carries a o. o. single XhoI site in position -104 upstream of the 0 0 0 15 F VIII:C ATG start codon). Linki.ge of the 5' F VIII:C o° o regions of pGX3238, the subfragments shown in Fig. 2, and the 3' F VIII:C regions of pGX3243 resulted in pGX3254 o 00 0 which now carries the entire F VIII:C sequence flanked by Xhol sites. In Fig. 2, the coding region is indicated 0 0 by full black Lines, and the non-coding region is outlined by thin lines.

o o 9 0 o 2. Preparation of the deletion mutants pGX3788 and pGX3789 o o Two deletion mutants, relative to the naturally occurring F VIII:C cDNA sequence in pGX3254, were prepared. The site-directed mutagenesis technique of Zoller and Smith 0 oo. (Zoller, M.J. and Smith, M. (1983) in Methods Enzymol. 100, 468-500) was used. The plasmid pGX3788 contains a sequence 0 4 for F VIII:C in which the coding information for amino acids 741-1689 F VIII:C A741-1689) has been deleted.

Plasmid pGX3789 contains a sequence in which the information for amino acids 816-1598 F VIII:C A816-1598) has been deleted. pGX3788 and pGX3789 were prepared in analogy to the construction of pGX3254 described above, Likewise with flanking Xhol sites at the same position.

7 a) Mutagenesis strategy It was suspected that the repetitive sequences in the F V:II:C cDNA sequence might cause a possible occurrence of spontaneous deletions during mutagenesis. For this reason, the mutagenesis was carried out on subclones of the pGX3254 sequence which do not carry repetitive sequences of this type. The subclones were selected such that, after mutagenesis had taken place, it was possible, by choosing suitable restriction sites, to construct a plasmid which codes for a F VIII:C molecule having the desired deletion.

o ob) Preparation of subcLones for the mutagenesis 0 0 onco 00° Two F VIII:C fragments were subcloned for the mutagenesis: o"o a BamHI fragment 2874 bp long (positions 1869-4743; Fig.

o 15 3a) and a SphI fragment 2650 bp long (positions 3931- 6581; Fig 3a). The vector for the subcloning was pGX2627 0 0 (Fig. It carries a polylinker with cleavage sites for the restriction enzymes: 0'0000 ClaI-SphI-PvuII-XbaI-EcoRI-SmaI-BamHI-SaLI-PstI-HindIII o. o 20 as well as the origin of replication of the bacteriophage M13 in addition to one which is plasmid-specific.

0 oo ooo° The former allows the packaging of single strands of the plasmid DNA with the aid if a superinfection with M13 bacteriophages. Vector pGX2683 carries the polylinker o 25 region in the opposite orientation to that in pGX2627.

0o4": The single-stranded DNA obtained in this way was then used as template for the oligonucleotide-induced mutagenesis. Subclone pGX4602 (Fig. 5) harbors the abovementioned BamHI fragment in the BamHI site of pGX2627.

Subclone pGX4604 (Fig. 7) harbors the abovementioned SphI fragment in the SphI site of pGX2627, and subclone pGX4605 (Fig. 9) harbors the same SphI fragment in the opposite orientation in pGX2627.

8 c) Mutagenesis The non-coding strand of the plasmid pGX4602 (Fig. was isolated and, with the aid of the synthetic oligonucleotide 5'-AAA TAT GAG ACG CGT TCT GAT GAT a new Mlul site (position 2500) was inserted at codon position 814/815 (ACT TTT ACG CGT) into the F VIII:C sequence.

The resulting plasmid is pGX3787 (Fig. 6).

The non-coding strand of the plasmid pGX4604 (Fig. 7) was isolated as described above and, with the aid of the synthetic oLigonucLeotide 5'-AAT CAG AGC CCC CGA AGC TTT CAA AAG-3', a new HindIII site (position 5124) was inseroo co ted at codon positions 1689/1691 (CGC AGC TTT CGA o AGC TTT) into the F VIII:C sequence. The resulting plasmid is pGX3782 (Fig. 8).

e 0 o° o 15 The coding strand of the plasmid pGX4605 (Fig. 9) was o o o Likewise isolated and used as template DNA, together with the synthetic oligonucleotide 5'-CAG GGA CAA ACG CGT ATC CTT TTT CTT-3', for the purpose of inserting a new Mlul o o site (position 4846) at codon position 1597/1598 (ACC .o00" 20 ATT ACG CGT) into the F VIII:C sequence. The resulting plasmid is pGX3783 (Fig. d) Construction of the plasmids pGX3788 and p'X3789 S ooo The plasmids pGX3788 F VIII:CA741-1689) and pGX3789 F VIII:CA816-1598) were subsequently prepared by addition of the missing 5' and 3' F VIII:C regions. The assembly was carried out in several steps.

e) pGX3788 The isolated HindIII (position 5124)-SphI (position 6581) fragment of pGX 3782 (Fig. 8, shaded part) was ligated with the isolated BamHI (position 1869)-Hindlil (position 2277) fragment of pGX4602 (Fig. 5) and with the Large 9 BamHI-SphI fragment, which had been treated with alkaline phosphatase, of pGX2683. The mixture was transformed into the E. coli strain JM101, and plasmid DNA from ampicillin-resistant colonies was analyzed by restriction endonucLeolytic digests. PLasmid DNA of a correct clone, pGX3791, was isolated, and the deletion was confirmed by sequencing (Fig. 11B).

pGX3788 Has constructed by Ligation of the fragments shown in Fig. 13. pGX3794 harbors the 5' end of the F VIII:C sequence as Xhol (position 104)-BamHI (position 1869) fragment (Fig. 128), and pGX3793 harbors the 3' end of the F VIII:C sequence as SphI (position 6581)- O.o" XhoI (position 7437) fragment (Fig. 12A). The fragments 0o.o were ligated together in such a way that the transLational o 0 15 reading frame of F VIII:C is retained, and the codons 0 o 740 (Arg) and 1690 (Ser) are fused by means of a HindIII 0 0 site (Fig. 14). Hence the molecule specified by pGX3788 0 harbors a deletion covering amino acids 741-1689 of 0 0 factor VIII:C.

0 00 0oo0 20 f) pGX3789 0 0 0 Initially, plasmid pGX3790 (Fig. 11A) was constructed by o o0 ligation of the following fragments: Mlul (position 4846)- Sphl (position 6581) fragment of pGX3783, BamHI (position 1869)-MLuI (position 2500) fragment of pGX3787, and the oo Soo0.. 25 large BamHI-SphI fragment, which had been treated with alkaline phosphatase, of pGX2683. The plasmid DNA of the clone which was verified by restriction analysis, pGX3790, was isolated, and the deletion was confirmed by sequencing (Fig. 11A). pGX3789 was then constructed by ligation of the fragments shown in Fig. 15. The fragments were ligated together in such a way that the translational reading frame of F VIII:C is retained, and the codons 815 (Thr) and 1599 (Leu) were fused by means of a Mlul site, with an additional amino acid (arginine) being inserted between 815 (Thr) and 1599 (Leu) (Fig. 16). Hence the 9;; :if t t I t I r ite

S

r 0 o 00 0 0ooo 0

C.

0 t 0 1 00 1 0 10 molecule specified by pGX3789 harbors a deletion which covers amin- acids 816-1598 of F VIII:C. Fig. 3 shows the restriction map of pGX3254, comparing with the deletion mutants pGX3788 and pGX3789.

3. Construction of F VIII:C expression plasmids The Xhol fragment, which is 7540 base pairs in size and harbors the entire F VIII:C coding sequence, of plasmid pGX3254 was isolated, filled-in by using the KLenow fragment of DNA polymerase, and ligated into the filled-in Xbal site of the vector pZET4 or into the filled-in XbaI site of the vector pSVASTOP1. The vector pZET4 uses the promoter for the expression of heterologous genes and has mRNA splice signals (published Patent Application DE 36 21 371A1). The vector pSVASTOP1 likewise uses the 15 SV40 promoter, but carries no mRNA splice signals (pub- Lished Patent Application DE 36 24 453A1). The resulting plasmids which harbor the F VIII:C fragment in the correct orientation relative to the SV40 promoter of pZET4 or of pSVASTOP1 are pZF VIII:C and pSVF VIII:C, respectively.

20 The Xhol fragment, uhich is 4694 base pairs in size and which harbors the coding region of the deletion mutant F VIII:CA741-1689, was isolated, filled in and ligated into the filled-in Xbal site of the vector pZET4 or into the filled-in XbaI site of the vector pSVASTOP1. The 25 resulting plasmids which harbor the F VIII:C fragment in the correct orientation relative to the SV40 promoter in pZET4 or pSVASTOP1 are pZF VIII:CA741-1689 and pSVF VIII: CA741-1689, respectively.

The Xhol fragment, which is 5195 base pairs in size and which harbors the coding region of the deletion mutant F VIII:CA816-1598, was isolated, filled in and ligated into the filled-in XbaI site of the vector pZET4 or into the filled-in XbaI site of the vector pSVASTOP1. The resulting plasmids which harbor the F VIII:C fragment in 1o a o oooO 0 0 11 the correct orientation reLative to the SV40 promoter in pZET4 or in pSVASTOP1 are pZF VIII:CA816-1598 and pSVF VIII:CA816-1598, respectively.

4. Expression of factor VIII:C in animaL cells The factor VIII:C constructions described in section 3 were cotransfected with two different DHFR plasmids (pSVOAdhfr, pubLished Patent Application DE 36 24 453A1, or pSV2dhfr, Lee et at., Nature 294, (1981) 228-232) in DHFR CHO ceLLs (Urtaub and Chasin (1980), Proc. NatL. Acad.

Sci., USA 77, 4216-4220), or with the vector pRMH140 (Hudziak et at., CeLL 31 (1981), 137-146) in BHK ceLLs.

00o The transfected ceLLs were then selected for the DHFR 0 or NEO phenotype. The supernatants from the ceLL cLones 0 0. obtained by selection were examined for the production of Ooocea 0015 F VIII:C.

0 0 0 oogo Three different methods were used for detection: i) ImmunoLogicaL detection by ELISA.

This incLuded the use of the combination of potyclonal o 00 0 0 antiserum against F VIII:C (for example K F VIII:C .0 0 000 20 Cag: 9460-010886; Behringwerke) and of the mono- 0 0 clonal antibodies ESH2, ESH8 (Bioscott) conjugated 0 006 0o 0 with peroxidase. The lower detection Limit was 1-3 ng of F VIII:C/mL.

ii) BioLogical activity assay using the F VIII:C COA 00 00co 25 assay (Kabi Vitrum). The Lower detection limit was o~oo 1-3 ng of F VIII:C/mL.

iii) Coagulation test of Lee et at., Thromb. Res. (1983) 511-519.

a) Expression in CHO ceLLs: The F VIII:C plasmids pSV F VIII:C, pSV F VIII:CA816-1598, pSV F VIII:CA741-1689, pZ F VIII:C, pZ F VIII:CA816-1598 or pZ F VIII:CA741-1689 were each cotransfected with one of the DHFR vectors pSVOAdhfr or pSV2dhfr using the I, IN -N 4. A pharmaceuticaL containing a factor VIII:C derivative as cLaimed in claim 1 or 2, 12 calcium phosphate coprecipitatio'n method (Graham and van der Eb, ViroLogy 52 (1973), 456-467) in CHO ceLLs.

For this purpose, in each case 20 pg of the F VIII:C expression plasmids were mixed with 5 pg of the DHFR vectors (pSV2dhfr or pSVOAdhfr), coprecipitated and then 6 2 used for the transfection (0.5-1 x 10 ceLLs in a 25 cm cuLture dish). After 3 days, the ceLLs were trypsinized, transferred into 60 mm Petri dishes and mixed with selection medium (without glycine, hypoxanthine and thymidine), The onLy ceLLs to survive under these conditions are those which have undergone stabLe transfection with the DHFR gene.

After 1-3 weeks, individuaL cell clones were isolat._ and grown in medium without glycine, hypoxanthine and thymidine. The supernatants from a total of 73 cell clones were tested, and it was possible to detect F VIII:C in 12 supernatants, both immunoogically as antigen and via the biological activity. Tab. 2 represents a summary of the distribution of the F VIII:C-expressing cell cLones related to the transfected constructs in CHO cells.

The baseline expression before amplification with methotrexate was found, with the aid of ELISA detection, to be 2-3 ng/ml F VIII:C antigen. The biological activity of the F VIII:C-expressing clones was between 8 and 14 6 6 25 mU/ml/10 6 cells.

b) Expression of F VIII:C in BHK cells jg of each of the F VIII:C expressiov plasmids pSV F VIII:C, pSV F VIII:CA816-1598, pSV F VIII:CA741-1689, pZ F VIII:C, pZ F VIII:CA816-1598 or pZ F VIII:CA741-1689 were cotransfected with 10 pg of the plasmid pRMH140, which codes for G4i8 resistance, into BHK ce'ls by the calcium phosphate coprecipitation method described in a).

After 3 days, the cells were trypsinized, transferred 13 into three 60 mm Petri dishes and mixed with selection medium which contained 400 ig/ml G418. After 12-14 days, G418-resistant ceLL clones were isolated and grown. In total, the supernatants from 45 cell cLones were examined for antigen and biological activity using the methods described in and 19 supernatants showed a positive reaction for F VIII:C. Tab. 3 summarizes the distribution of the F VIII:C-expressing ceLL cLones relative to the transfected constructs in BHK ceLLs.

The mean biological activity of the F VIII:C-expressing cell cLones was 12 MU/ml/10 6 cells. 2-4 ng/mL F VIII:Z antigen were detectable in the ELISA.

The culture supernatant of a new ceLl clone which had o9,oo been transfected with the construct pZ F VIII:CA741-1689 was examined for the abiLity of the expressed deLetion o mutant to be activated with thrombin (Toole et aL., Nature 312 (1984), 342-347). For this purpose, the 0o,0a0 culture supernatant was mixed with 1 U/mL thrombin and incubated at room temperature. At various times (tl) o' 20 aLiquots were taken, and the bioLogicaL activity was 0 0 determined by a coagulation test. The resuLt of this experinent is shown in Fig. 17. In this, the coagulation °o times (t 2 are represented by and the corresponding MU/mL are indicated by with the effect of athrombin (1 U/mL) on the activity of the F VIII:CA741-1689 molecuLe being measured. It is evident from the experiment that the F VIII:CA741-1689 molecule is activated by a factor of about 10 due to the influence of thrombin.

Furthermore, it was shown by comparison with the natural F VIII:C that the mutant F VIII:CA816-1598 has a comparable inactivation time and a considerably faster activation time.

concentrates ha~s resuLted in a number of poLypeptides associated with bioLogicaL activity and having moLecuLar

I,

14 Tabte 1: F VIII :C-spec ific oL igonucteotides Otigonucteotide No. Position coding strand: 1 572 1609 1610 1613 1611 1573 1612 1614 1 301 317 962 978 1761 1777 3429 3445 4141 4160 4992 5008 6141 6157 00 00 0 0 0 0 0 0 0 0 00 0 0 0 0 00 0 0 0 0 0 000 n'in-cod ing 1492 1493 1494 1495 1531 1496 1497 1498 '14 99 strand: 7198 6465 5685 5092 2400 2339 1179 570 120 7179 6446 5666 5073 2351 2320 1160 551 101 0 0 o 00 0 4 Position in bp based on -the sequence (1934) Nat6ure 312, 330-33? of Wood et at.

Table 2: Factor VIII:C-expressing CHO clones A) Cotransfected with pSV2dhfr: Ptasmid Number of DHFR clones Positive for F VIII:C e~x press ion (ELISA; COA test) 0 1 0 pSV F VIII:C pSV F VIII:CA816-1598 pSV F VIII:CLA741-l689 00 00 w 0 0 j 0 0000 0 0 0000 0 2 ~0000 o 0 0 0 0 0 J o:0 0 00 00 0 0 00 o 0 0 0 00 PZ F VIII:C 2 P2 F VIII:CA816-1598 13 PZ F VIII:CAc741-1689 4 13) Cotransfected with pSVOAdhfr: PSV F VIII:C 14 pSV F VIII:CA816-1598 6 pSV F VIII:CA741-1689 6 p2 F VIII:C 4 P2 F VIII:CA8.6-1598 7 p2 F VIId:CA741-1689 3 1 2 0 12 0 00 00 0 0 00 0 00 0t o 000 000 00 0 jJ which are simpLer than for the wiLd-type protein and than 16 Table 3: Factor VIII:C-expressinig BHK ctones cotransfected with pRMH140 PL'ismid Number of Positive for F ViII:C Neo clones expression (ELISA; COA test) pSV F VIII:C pZ F VIII:CA741-1689 pZ F VIII:CA816-1598 pZV F VIII:CA816-1598 7 6 19 0 0 0 o 0 00 .00 0 0000 0 0 0 0 4 0

Claims

1. A factor VIII:C derivative, wherein the natural throm- bin cleavage site at Arg 3 7 2 is retained, a new one has been produced at Arg 7 4 0 and those present at Arg 16 4 8 and Arg 16 8 9 are absent.

2. A factor VIII:C derivative as claimed in claim 1, wherein amino acids 741 to 1689 of the authentic mol- ecule have been deleted.

3. DNA sequences which code for the factor VIII:C deriva- tives specified in claim 1 or 2, as well as alleles 04 0. o o and variants os here'\n 0 o d 04° a

4. A pharmaceutical containing a factor VIII:C derivative io as claimed in claim 1 or 2. o

5. An injectable solution containing a pharmaceutical as claimed in claim 4. o o 0 0 0 O o

6. The use of the factor VIII:C derivatives as claimed in claim 1 or 2 for the preparation of pharmaceuticals for o"o the therapy of coagulation disturbances. o oo

7. DNA structures and vectors which contain DNA sequences 00o as claimed in claim 3. o oi

8. Pro- or eukaryotic cells which contain DNA structures or vectors as claimed in claim 7.

9. A process for the preparation of factor VIII:C deriva- tives which comprises bringing about the expression of cells as claimed in claim 8. The use of factor VIII:C derivatives as claimed in claim 1 or 2 for the preparation of diagnostic aids. DATED this 16th day of June 1988, BEHRINGWERKE AKTIENGESELLSCHAFT EDWD. WATERS SONS PATENT ATTORNEYS ,1 MELBOURNE. VIC. 3000.