AT394208B - Process for the preparation of vectors - Google Patents

Abstract

Processes for the preparation of replicable vectors which bring about in microorganisms the expression of polypeptides with the amino acid sequence of a mature human leukocyte interferon are described and comprise linking a first DNA sequence which encodes these polypeptides operably to a second DNA sequence which contains the codons necessary for expression.

Description

AT 394 208 B

The present invention relates to the field of recombinant DNA technology, i. H. Processes which, in the broadest sense, lead to the production of recombinant proteins by microorganisms.

More specifically, the present invention relates to the field of producing mature human leukocyte interferons by means of recombinant DNA technology, and in particular to a method for producing expression vectors (hereinafter referred to as vectors for short) which end one coding for mature human leukocyte interferon DNA sequence included.

Background of the Invention

Human leukocyte interferon (LelF) was first discovered by Isaacs and Lindenmann (Proc, R. Soc. B 147,258-267 [1957]; U.S.P. 3,699,222) and prepared in the form of the crude precipitates. Efforts to clean and characterize the material since then have led to the production of relatively homogeneous leukocyte interferons from normal leukocytes or from leukocytes from leukemic donors (e.g. DE-OS 2,947,134). These interferons belong to a family of proteins which have the known property of imparting a virus-resistant state to certain cells. In addition, interferon can inhibit cell proliferation and modulate the immune response. These properties have led to leukocyte interferon being used clinically as a therapeutic agent for the treatment of viral infections and diseases.

In the past, leukocyte interferons have been substantially purified to homogeneity (Rubinstein et al., Proc. Natl. Acad. Sci. USA 26, 640-644 [1979]; Zoon et al., Ibid. 26, 5601-5605 [ 1979]). The reported molecular weights range from about 17,500 to about 21,000. The specific activities of these preparations are remarkably high at 2 x 10 to 1 x 10 units per mg protein, but the yields obtained from the cell cultures have so far been discouragingly low. Nonetheless, refined protein sequencing methods succeeded in determining amino acid partial sequences (Zoon et al., Science 207,527 [1980]; Levy et al., Proc. Natl. Acad. Sei. U.S.A. 22, 5102-5104 [1980]). The question of glycosylation of the various leukocyte interferons has not yet been completely clarified, but it is so clear that differences in molecular weights are not caused by glycosylation alone. It is also clear that the leukocyte interferons are clearly distinguished from one another by a different amino acid composition and sequence and the amino acid homology is in some cases lower than 80%. While the homogeneous leukocyte interferon isolated from donor leukocytes was sufficient for partial characterization and limited clinical studies, this source is by no means sufficient for obtaining those amounts of interferon that are suitable for wide clinical trials and the subsequent prophylactic and / or therapeutic use will be necessary. In fact, previous clinical trials to evaluate human leukocyte interferon in antitumor and antiviral testing used primarily raw material (purity <1%) and the long times required to obtain sufficient amounts of pure material led to a large delay in large-scale investigations

However, recombinant DNA technology has meanwhile succeeded in producing a large number of valuable polypeptides in a controlled microbial manner. Bacteria, which have been modified by this technology to allow the production of polypeptides such as somatostatin, the A and B chains of human insulin and human growth hormone (Itakura et al., Science 198.1056-1063 [1977] ; Goeddel et al., Nature 281.544-548 [1979]). Recently, the recombinant DNA technique was used to produce proinsulin and thymosin alpha 1 bacterially, and various authors have reported that they have succeeded in obtaining DNA encoding human leukocyte interferon and corresponding proteins with leukocyte interferon activity (Nagata et al ., Nature 2M> 316-320 [1980]; Mantei et al., Gene 10.1-10 [1980]; Taniguchi et al., Nature 285.547-549 [1980]).

The driving force behind recombinant DNA technology is the plasmid, a non-chromosomal loop of double-stranded DNA that is often found in many copies per cell in bacteria and other microbes. The information contained in the plasmid DNA includes that relating to the reproduction of the plasmid in daughter cells (ie a &quot; replicon &quot;) and usually one or more selection characteristics such as, in the case of bacteria, resistance to antibiotics which allow clones to be cloned Recognize host cells containing the plasmid of interest and selectively grow in a selected medium. The usefulness of the plasmids lies in the fact that they can be cleaved specifically by one or the other restriction endonuclease (also called restriction enzyme), each endonuclease recognizing a different site of the plasmid DNA. Heterologous genes or gene fragments can be inserted into the plasmid at the cleavage sites or subsequently reconstructed ends at the cleavage sites. DNA recombination is carried out outside the cell, but the recombinant plasmid obtained is introduced into the cell by a process known as transformation, and recombinant plasmids containing large amounts of heterologous genes can be obtained by culturing the transformed cells. If the heterologous gene is inserted correctly (ie operably) with respect to those parts of the plasmid which are responsible for the transcription and translation of the encoded DNA message, the recombinant plasmid obtained (expression vector, vector) can be used to produce the polypeptide, which is encoded by the heterologous gene. -2- ΑΤ 394 208 Β

This process is called expression.

Expression is triggered in the promoter region which is recognized and bound by RNA polymerase. In some cases, such as B. in the tryptophan or "trp" promoter, which is preferred in connection with the present invention, promoter regions are overlapped by so-called operator regions and then form a combined promoter operator. Operators are DNA sequences that are recognized by so-called repressor proteins, which in turn regulate the frequency of the start of transcription on a special promoter. The polymerase migrates along the DNA and overwrites the information contained in the coding strand from the 5 'to the 3' end in mRNA, which in turn is then translated into the polypeptide with the amino acid sequence encoded by the DNA. Each amino acid of a polypeptide is encoded within the so-called structural gene by a nucleotide triplet or codon. After binding to the promoter, the RNA polymerase first transcribes nucleotides that encode a ribosome binding site, then a translation start signal (usually ATG, which is in the mRNA becomes AUG) and finally the nucleotide sequences of the structural gene itself. At the end of the structural gene, so-called stop codons are transcribed, according to which the polymerase can form further mRNA sequences, which, however, are no longer translated by the ribosomes due to the existing stop signal. The ribosomes attach to the intended binding site of the mRNA, usually in bacteria as it arises, and then in turn produce the encoded polypeptide, starting at the translation start signal and ending with the stop signal already mentioned. The desired product is made when the sequences encoding the ribosome binding site are correct with respect to the AUG start signal and when all other codons are in phase with the start signal. The desired polypeptide can be obtained by lysis of the host cell, separation from other bacterial proteins and purification in a conventional manner.

The present invention was based on the idea that the use of recombinant DNA technology (ie the insertion of interferon genes in microbial vectors and their expression under the control of microbial gene regulatory elements) would be the most effective way of obtaining large amounts of leukocyte interferon, which despite the The fact that it is not glycosylated like perhaps the natural one could be used for the clinical treatment of many viral and neoplastic diseases.

A first leukocyte interferon gene was obtained according to the invention by the following measures: (1) amino acid partial sequences of homogeneous human leukocyte interferon were used to construct synthetic DNA probes, the codons of which represented all possible nucleotide combinations by which the amino acid partial sequences can be encoded. (2) Banks of bacterial colonies were prepared that contained complementary DNA (cDNA) from induced mRNA. Other induced and radioactively marinated mRNA was hybridized with plasmid cDNA from this library. Hybridizing mRNA was eluted and tested for translation into interferon in the oocyte test. Plasmid DNA from colonies which had interferon activity in this test were also tested for hybridization with probes obtained according to (1). (3) In parallel to the procedure under (2), cDNA obtained by means of induced mRNA was used in plasmids to to form an independent bank of transformed colonies. The probes obtained in (1) were used as starters for the synthesis of radiolabeled single-stranded cDNA, which in turn were used as hybridization probes. The synthetic probes hybridized with induced mRNA as a template and were also used by reverse transcription to form induced, radiolabeled cDNA. Clones of the colony bank obtained in this way, which hybridized with radiolabeled cDNA, were further examined for the presence of a gene fully coding for interferon. Any gene segment obtained according to (1) or (2), possibly encoding a partial sequence of the interferon, can be used as a probe for an unabridged gene. (4) The unabridged gene thus obtained was tailored using synthetic DNA to exclude any leader sequence that could prevent microbial expression of the mature polypeptide and to be able to position it in a vector in relation to the start signals and the Ribosome binding site of a microbial promoter. The expressed interferon was purified to the extent that it could be characterized and its activity determined. (5) The interferon gene fragment obtained in this way could in turn be used as a probe for the isolation (by means of hybridization) of further genes for other, partially homologous leukocyte interferons.

By using the recombinant DNA technology methods briefly described above, the microbial production of the family of homologous leukocyte interferons (in non-glycosylated form) as mature polypeptides, i.e. H. essentially free of corresponding pre-sequences or parts thereof. These interferons can be directly expressed, isolated and purified to the extent that they are suitable for use in the treatment of viral and malignant diseases in animals and humans.

Individual members of the interferon family, as far as they were expressed, were effective in in vitro tests and, in a first such attempt, also in an in vivo test, the latter being the test of the first microbial mature leukocyte. Interferons acted. -3-

AT 394 208 B

The term "mature leukocyte interferon" used in the context of the present application defines a microbially (e.g. bacterially) produced interferon molecule which does not carry any glycosyl groups. Mature leukocyte interferon, according to the present invention, is expressed by a translation start signal (ATG) immediately before the first amino acid codon of the natural product. The &quot; mature &quot; Polypeptide can therefore contain methionine (encoded for the ATG) as the first amino acid of its sequence, without this significantly changing its character. On the other hand, the microbial host can split off the methionine from the immediate translation product. Mature leukocyte interferon can be expressed together with a conjugated protein, which is not a common leader, whereby the conjugate can be cleaved specifically intracellularly or extracellularly (see British patent application publication number 2007676A). Finally, the mature interferon can be combined with a microbial &quot; signal peptide &quot; which transports the conjugate to the cell wall, where the signal sequence is cleaved and the mature polypeptide is secreted. Expression of mature leukocyte interferon means bacterial or other microbial production of an interferon molecule that contains neither glycosyl groups nor a presequence.

The amino acid sequences of the individual leukocyte interferons (FIGS. 4 and 9) were determined on the basis of the DNA sequences of the corresponding genes (FIGS. 3 and 8). Allelic variants which differ from individual to individual exist for all these individual leukocyte interferons. These variants can be characterized by exchange (substitution, inversion), deletion or insertion of one or more amino acids in the sequence. Such allelic variations of the leukocyte interferons A to J according to the invention are included in the respective formulas (LelF A to LelF J) and are also the subject of the present invention.

Description of the figures

FIG. 1 describes two amino acid partial sequences which are common to all interferon species which have been isolated from human leukocytes and purified to homogeneity. They are labeled T-1 and T-13. All possible mRNA sequences encoding these peptides are shown, as are the corresponding cDNA sequences. The letters A, T, G, C and U denote the nucleotides with the bases adenine, thymine, guanine, cytosine and uracil. The letter N means A, G, C or U. The polynucleotides are indicated from the 5'-end (left) to the 3'-end (right) and where they are shown double-stranded, the lower DNA (non-coding strand) is from 3'- reproduced in the 5 'direction.

FIG. 2 shows an autoradiogram which shows the hybridization of possible LeDF plasmids with • ^ P-labeled synthetic deoxyoligonucleotides.

FIG. 3 shows the nucleotide sequences (coding strands) of eight isolated gene fragments (designated A-H) which can be used for the expression of leukocyte interferons. The ATG Translations start codon and stop triplets for each LelF are underlined. Untranslated sequences follow the stop triplets. The complete gene for LelF A lacks a codon in position 200 to 202 that is present in the other LelFs. Before the leader sequence there are also untranslated sequences. The isolated fragment E lacks the full leader presequence, but it contains the complete gene for any mature LelF E. The isolated fragment G, on the other hand, lacks part of the coding sequence.

FIG. 4 gives an overview of the amino acid sequences of 8 leukocyte interferons (A to H) derived from the determined nucleotide sequences. The capital letters used to denote the individual amino acids are those recommended by the IUPAC-IUB Commission for Biochemical Nomenclature. It means: A alanine; C cysteine; D aspartic acid; E glutamic acid; F phenylalanine; G glycine; H histidine; I isoleucine; K lysine; L leucine; M methionine; N asparagine; P proline; Q glutamine; R arginine; S serine; T threonine; V valine; W tryptophan; and Y tyrosine. The numbers indicate the position of the amino acids (S refers to the signal peptide). The hyphen in the 165 amino acid sequence of LelF A at position 44 was introduced to align the sequence of LelF A with the 166 amino acids of the other leukocyte interferons. The nucleotide in position 187 (FIG. 3) was not taken into account for determining the amino acid sequence of the LelF E. The asterisks (sequence of LelF E) represent in-phase stop codons. Amino acids that are common to all LelFs (with the exception of the pseudogen LelF E) are shown in line &quot; All &quot; shown. The underlined residues are amino acids that also occur in human fibroblast interferon.

FIG. 5 shows the restriction endonuclease maps of the cloned cDNAs from 8 different leukocyte interferons (A to H). The hybrid plasmids were produced by the dC: dG tailing method (Goeddel, D.V. et al, Nature 287.411-416 [1980]). The cDNA inserts can therefore be cut out with Pstl. The lines at the ends of each cDNA insert represent the subsequent homopolymeric dC: dG tails. The position of the PvuII, EcoRI and BglII restriction sites are indicated. Shaded regions in the figure represent coding sequences for mature LelFs; the slanted-dashed regions denote coding sequences for signal peptides, while the white regions represent non-coding sequences.

FIG. 6 shows schematically the construction of a gene which encodes the direct microbial synthesis of mature LelF A. The restriction sites (Pstl, etc.) and the fragments used are shown. The abbreviation &quot; b.p. &quot; means base pairs. -4-

AT 394 208 B

Figure 7 (not to scale) schematically represents the restriction map of 2 gene fragments that were used to express mature LelF B. The designated nucleotide codons are the ends of the coding strands which result from treatment with the restriction enzyme Sau3a in the two cases shown.

Figures 8 and 9 show the DNA and amino acid sequences of the leukocyte interferons A, C, Η, I and J again. In Figure 9, the asterisk means a stop codon in the corresponding DNA sequence, while the hyphen in the sequence of LelF A connects the adjacent amino acids, phenylalanine and glycine with one another

Description of the Preferred Embodiments A. The Microorganisms Used

The following two microorganisms were used: E. coli x 1776 as described in U.S. Patent 4,190,495 and E. coli K-12 strain 294 (end A, thi &quot;, hsr &quot;, hsmjc +) as in the UK patent application Publication no. 2055382 A. Both organisms were deposited at the American Type Culture Collection under ATCC Nos. 31537 and 31446. All work was done in accordance with the guidelines of the National Institute of Health.

In addition to the two E. coli strains mentioned above, other strains, for example E. coli B or other microbe strains, can also be used, most of which have been deposited at a depository, such as the American Type Culture Collection, and in general are accessible; compare also DE-OS 2644432. Other microorganisms that can be used according to the invention are e.g. Bacilli, such as Bacillus subtilis and other Enterobacteriaceae, including Salmonella typhimurium and Serratia marcescens, which can express heterologous genes using replicable plasmids. Interferon genes can also be expressed in yeasts, for example in Saccharomyces cerevisiae, under the control of a yeast promoter.

B. Source and purification of LelF mRNA

LelF mRNA can be obtained from human leukocytes (usually from CML patients) that have been stimulated to produce interferon with Sendai virus or NDV, as described, for example, in German Offenlegungsschrift No. 2947134. A particularly preferred source used for the present work is a cell line called KG-1, which is derived from a patient with acute myelogenous leukemia. The cell line, described by Koeffler, Η. P. and Golde, DW, Science 200.1153 (1978), grows rapidly in a medium containing RPMI (Rosewell Park Memorial Institute) 1640 with 10% FCS (fetal calf serum), heat inactivated, 25 mM HEPES buffer (N-2-hydroxyethyl piperazin-N'-2-ethanesulfonic acid) and 50 μg / ml gentamycin. The cells can be frozen in the medium described above after adding 10% dimethyl sulfoxide. The KG-1 line has been deposited with the American Type Culture Collection under ATCC No. CRL 8031.

The KG-1 cells were designed according to the method described by Rubinstein et al. in (Proc. Natl. Acad. Sci. U.S.A. 7 £, 640-644 [1979]) with Sendai virus or NDV to produce leukocyte interferon mRNA. The cells were harvested 5 hours after induction and the mRNA was prepared according to the guanidine thiocyanate-guanidine hydrochloride method (Chirgwin et al., Biochemistry iS, 5294-5299 [1979]). RNA from non-induced cells was isolated in the same way. Oligo-deoxythymidine (dT) cellulose chromatography and sucrose gradient ultracentrifugation were used to isolate the 12S fraction of poly (A) mRNA as described by Green et al. (Arch. Biochem. Biophys. 172, 74-89 [1976]) and Okuyuma et al. (Arch. Biochem. Biophys. 188.98-104 [1989]). This mRNA had an interferon titer of 8000-10,000 units per microgram in the Xenopus laevis oocyte test (Cavalieri et al., Proc. Natl. Acad. Sci. U.S.A. 74, 3287-3291 [1977]). C. Manufacture of colony banks. containing theLELF cDNA sequences. 5 pg mRNA was used to produce double-stranded cDNA according to standard methods (Wickens et al., J. Biol. Chem. 253.2483-2495 [1978] and Goeddel et al., Nature 281.544-548 [1979]). The cDNA was size-auctioned by electrophoresis on a 6% polyacrylamide gel and 230 ng of the 500-1500 base pair material was isolated by electroelution. 100 ng of these cDNA were treated with deoxycytidine (dC) residues according to the method described by Chang et al. (Nature 275.617-624 [1978]), combined with 470 ng of the plasmid pBR322, which contains deoxyguanosine (dG) residues at the Pstl site (Bolivar et al., Gene 2.95-113 [1977]) had been extended and used to transform E. coli x 1776. About 130 tetracycline-resistant, ampicillin-sensitive transformants were obtained per ng of cDNA.

In a second similar experiment, approximately 1000 tetracycline-resistant, ampicillin-sensitive E. coli K-12 strain 294 transformants were obtained per ng of cDNA. In this case the size of the fractionated cDNA material was between 600 and 1300 base pairs. It was by electroelution for the extension with -5-

AT 394 208 B

Deoxycytidine (dC) obtained. D. Preparation of synthetic oligonucleotides and their use

Knowledge of the amino acid sequences of various tryptic fragments of human leukocyte interferons allowed the synthesis of deoxyoligonucleotides that were complementary to certain regions of the LelF mRNA. The two tryptic peptides TI and T13 were chosen because their amino acid sequences required the synthesis of only 12 and 4 undecamers, respectively, in order to capture all possible nucleotide sequences (FIG. 1). 4 sets of deoxynucleotide probes were synthesized for each sequence, containing either 3 (T-1A, B, C, D) or one (T-13A, B, C, D) oligonucleotide. The designated complementary deoxyoligonucleotides of 11 bases each were chemically synthesized according to the phosphortriester method (Crea et al., Proc. Natl. Acad. Sei. U.S.A. 25, 5765-5769 [1978]). 4 individual probes were manufactured in the T-13 series. The twelve T-1 probes were prepared in four pools of 3 probes as shown in Figure 1.

The four individual T-13 series probes and the twelve T-1 probes, made in four groups of 3 probes each, were used to start the synthesis of radiolabeled single-stranded cDNA for use as hybridization probes. The template mRNA was either 12S RNA from Sendai virus-induced KG-1 cells (8000 units of IF activity per pg) or total poly (A) mRNA from non-induced Ύ1

Leukocytes (10 units per pg). P-labeled cDNA was made from these initiators using known reaction conditions (Noyes et al., Proc. Natl. Acad. Sci. U.S.A. 76, 1770-1774 [1979]). The 60 pl reactions were carried out in 20 mM Tris-HCl (pH 8.3), 20 mM KCl, 8 mM MgCl2 and 30 mM β-mercaptoethanol, each with 1 pg of each starter (ie 12 pg in total for the Tl series, 4 pg total for the T-13 series), 2 pg induced 12S mRNA (or 10 pg non-induced poly (A) mRNA), 0.5 mM dATP, dCTP, dTTP, 200 pCi (a ^ P) dCTP (Amersham , 2-3000 Ci / mmole) and 60 units of reverse transcriptase (Bethesda Research Laboratories). The product was separated from unlabelled material by gel filtration on a 10 ml Sephadex® G-50 column, treated with 0.3N NaOH for 30 minutes at 70 ° C. to destroy the RNA and neutralized with HCl. The hybridizations were carried out as described by Kafatos et al. (Nucleic Acids Res. 7, 1541-1552 [1979]). E. Identification of clones pLl-pL30

The method of Birnboim et al., Nucleic Acids Res. 7, 1513-1523 (1979), for the rapid isolation of plasmids was used to convert 1 pg of plasmid DNA from each of the 500 individual E. coli K-12 strain 294 transformants (compare C) to win. Each DNA sample was denatured and triplicated on nitrocellulose filters using the method of Kafatos et al. (see above).

The three sets of nitrocellulose filters containing 500 plasmid samples were hybridized with a) induced cDNA strengthened with a T-1 starter set, b) T-13 started induced cDNA and c) uninduced cDNA prepared using both starter sets. Clones were considered positive if they hybridized more strongly to one or both of the induced cDNA probes than to the entire uninduced probe. 30 positive clones (pLl-pL30) were selected from the 500 for further analysis. F. Identification of clones pL-31-nL39.

Isolation of a plasmid fNr. 1041. which contained an LelF gene fragment.

E. coli x 1776 transformants were tested by the colony hybridization method of Grunstein and Hogness (Proc. Natl. Acad. Sci. USA 22, 3961-3965 [1975]) using ^ P-labeled induced mRNA as a probe (Lillenhaug et al., Biochemistry, 15, 1858-1865 [1976]). Unlabelled mRNA from non-induced cells was mixed with the probe in a ratio of 200: 1 to compete Ύ) with uninduced mRNA present in the J P-labeled preparation. The hybridization of labeled mRNA should mainly take place in colonies that contain induced sequences. Three classes of transformants were obtained: (1) 2-3% of the colonies hybridized very strongly with ^ P-mRNA, (2) 10% hybridized significantly weaker than the first class and (3) the rest gave no discernible hybridization signal.

The positive colonies (classes (1) and (2)) were checked for the presence of interferon-specific sequences by means of an assay which is based on the specific hybridization of interferon-mRNA to plasmid DNA. First, 60 of the highly positive colonies (class (1)) were individually in 100 ml M9 medium, the -6-

AT 394 208 B was supplemented with tetracycline (20 pg / ml), diaminopimelic acid (100 pg / ml), thymidine (20 pg / ml) and d-biotin (1 μg / ml). The M9 medium contains per liter: Na2HPC> 4 (6 g), KH2P04 (3 g), NaCl (0.5 g) and NH4CI (1 g). After autoclaving, 1 ml of sterile IM MgS04 and 10 ml of sterile 0.01M CaCl2 were added. In each case 10 of the 60 cultures were combined and the plasmid DNA was isolated as described by Clewell et al., Biochemistry £, 4428-440 [1970]. 10 pg of each plasmid DNA pool were cleaved with HindIII, denatured and covalently bound to DBM paper (diazobenzyloxymethyl paper). 1 pg of purified mRNA from induced cells was hybridized to each filter. Non-hybridized mRNA was removed by washing. The specifically hybridized mRNA was eluted and transferred to Xenopus laevis oocytes. In this assay, all six pools were negative. Five pools of 10 colonies each and 1 pool of 9 colonies were formed from 59 weakly positive colonies (class (2)) and tested according to the method described above. Of the six pools, one (K10) hybridized significantly more than the rest each time it was tested. To identify the specific interferon cDNA clone, plasmid DNA was prepared from each of the nine colonies pooled in pool K10 and tested individually. Two of the nine plasmids (Nos. 101 and 104) bound interferon mRNA more than the others. A special Bgl II- γ) was made from plasmid No. 104

Restriction fragment containing 260 base pairs isolated, P-labeled according to that of Taylor et al., Biochim. Biophys. Acta 442, 324-330 (1976), and used as a probe to transform 400 E. coli 294 transformants by an in situ colony screening method (Grunstein and Hogness, Proc. Naü. Sci. USA 22, 3961-3965 [1975]) to test. Nine colonies (pL31-pL39) were identified that hybridized to varying degrees with this probe.

In addition, the labeled 260 base pair fragment was used to test 4000 E. coli 294 transformants in the same way. 50 colonies were identified that hybridized to different degrees with this probe. One contained the LelF G fragment, one contained the LelF H fragment and one contained a fragment called LelF Hl (apparently an allele of LelF H). The hybrid plasmids obtained were called &quot; LelF G &quot;, &quot; pLelF H &quot;, etc., G. Isolation and sequencing of a first complete LelF gene.

Plasmid DNA was prepared from all 39 potential LelF cDNA clones and tested again with the same 260 base pair DNA probe using the hybridization procedure of Kafatos et al. (see above). Three plasmids (pL4, pL31, pL34) gave very strong hybridization signals, four (pL13, pL30, pL32, pL36) hybridized moderately, while three (pL6, pL8, pL14) hybridized only weakly with the probe.

The 39 potential LelF cDNA recombinant plasmids were also tested using the ^ P-labeled synthetic undecamers (individual T-1 starter pools or individual T-13 starters) as hybridization probes. The hybridization conditions were selected so that perfect base pairing was necessary to obtain detectable hybridization signals (Wallace et al., Nucleic Acids Res. £, 3543-3557 [1979]). In this way, plasmid DNA was prepared from the 39 clones by a standard method (Clewell et al., See above) and purified by means of Biorad Agarose A-50 column chromatography. Samples of 3 pg of each preparation were linearized by treatment with EcoRI, denatured with alkali and spotted on 2 nitrocellulose filters (1.5 pg / stain) as described by Kafatos et al. (See above). The individual synthetic deoxyoligonucleotide starters and starter pools were phosphorylated with (γ ^ Ρ) ΑΤΡ as follows: 50 pmol oligonucleotide and 100 pmol γ ^ Ρ ATP (New England Nuclear, 2500 Ci / mmol) were washed with 30 μΐ 50 mmol Tris-HCl, 10 mmol of MgCl2 and 15 mmol of β-mercaptoethanol were added. 2 units of T4 polynucleotide kinase were added and after 30 minutes at 37 ° C. the ^ P-labeled starters were purified by chromatography on 10 ml Sephadex G-50 columns. The hybridizations were under

Use of 10® cpm starter T-13C or 3 x 10® cpm starter pool T-1C carried out at 15 ° C for 14 hours in 6 x SSC [1 x SSC = 0.15 M NaCl, 0.015 M sodium citrate, pH 7 , 2], 10 x Denhardt's solution [0.2% bovine serum albumin, 0.2% polyvinyl pyrolidone, 0.2% Ficoll] as described by Wallace et al. (see above). The filters were washed in 6 x SSC for 5 minutes at 0 ° C., dried and exposed to X-ray film. The results are shown in Figure 2 for - ^ P starter pool T-13C and starter T-1C.

The plasmid DNA from clone 104 provided remarkable hybridization with starter pool T-1C and starter T-13C, while hybridization with the other undecamers was not found. As shown in Figure 2, several of the 39 potential LelF plasmids (pL2,4,13,17,20,30, 31, 34) hybridized with these two probes. Restriction analysis showed, however, that only 1 of the plasmids, pL31, also contained a BglII fragment containing 260 base pairs. PstI treatment of pL31 showed that the size of the cDNA segment was about 1000 base pairs.

The entire Pstl section of pL31 was carried out both according to the chemical method of Maxam-Gilbert (Methods Enzymol. 65,499-560 [1980]) and according to the dideoxy chain breaking method (Smith, Methods Enzymol. 65,560-580 [1980]), after the Sau3a fragments have been subcloned into an M13 vector -7-

AT 394 208 B. The DNA sequence is shown in Figure 3 (&quot; A &quot;). From this sequence, the entire amino acid sequence of the LelF A was predicted, including the pre- or signal peptide. The first ATG Translations start codon is located 60 nucleotides from the 5 'end of the sequence and is followed by a TGA stop triplet after 188 codons. There are 342 untranslated nucleotides at the 3 'end, which are eventually followed by a poly (A) sequence. The putative signal peptide (probably involved in the secretion of mature LelF from leukocytes) is 23 amino acids long. The mature LelF A, which consists of 165 amino acids, has a calculated molecular weight of 19.390. It can be seen from FIG. 4 that the tryptic peptides T-1 and T-13 correspond to amino acids 145-149 and 57-61 of LelF A. The actual DNA sequences found in these two regions are represented by the starter pool Tl-C and the starter T13-C (Figure 8). H. Direct expression of mature leukocyte interferon A (LelF Al I. General

The method by which the mature LelF A was directly expressed is a variant of that previously used for human growth hormone (Goeddel et al., Nature 281, 544-548 [1979]) in that it is a combination of synthetic (N- terminal) and complementary DNA.

As shown in Figure 6, there is a Sau3a restriction endonuclease fragment between codons 1 and 3 of LelF A. Two synthetic deoxyolinucleotides were created that contain an ATG translation start codon, restore the codon for amino acid 1 (cysteine) and create an EcoRI cohesive end. These oligomers were attached to a Sau3a-AvaII fragment of pL31, which consisted of 34 base pairs. The resulting 45 base pair-containing product was attached to 2 additional DNA fragments, so that an artificial-natural hybrid gene of 865 base pairs was formed, which encoded LelF A and which is limited by EcoRI and PstI restriction sites. This gene was inserted between the EcoRI and PstI restriction sites of the plasmid pBR322 and gave the plasmid pLelF Al. 2. Construction of the Trvptophan control element [containing the E. coli trp promoter. Operator and the trp-

Leader ribosome binding site without the ATG sequence for the translation hefrinnV

The plasmid pGMI carries the E. coli tryptophan operon with the deletion ALE1413 (Miozzari et al.,

J. Bacteriology 133.1457-1466 [1978]) and therefore expresses a fusion protein with the first 6 amino acids of the trp leader and about the last third of the trp E polypeptide (hereinafter referred to as LE ') and the complete trp D polypeptide, under the control of the trp promoter-operator system. The plasmid (20 pg) was cleaved at 5 sites with the restriction enzyme PvuII. The gene fragments were then combined with EcoRI connection sequences (consisting of a self-complementary polynucleotide sequence pCATGAATTCATG) in order to create an EcoRI cleavage site for later cloning into a plasmid with such an EcoRI cleavage site. The 20 pg of the DNA fragments obtained from pGMI were treated with 10 units of T4 DNA ligase in the presence of 200 pMol of the 5'-phosphorylated synthetic oligonucleotide pCATGAATTCATG in 20 pl of T4 DNA ligase buffer (20 mmol Tris, pH 7.6, 0.5 mmol ATP, 10 mmol MgC ^, 5 mmol Dithiothreit) treated at 4 ° C overnight. The solution was then heated to 70 ° C for 10 minutes. The connecting sequences were digested with EcoRI and the fragments, which now contained EcoRI ends, were separated by means of 5% polyacrylamide gel electrophoresis (PAGE). The three largest fragments were isolated from the gel. For this purpose, staining was first carried out with ethidium bromide, the localized fragments were identified under UV light and the regions concerned were cut out of the gel. Each gel fragment was placed in a dialysis bag with 300 microliters of 0.1 x TBE and subjected to electrophoresis at 100 volts for 1 hour in 0.1 x TBE buffer (TBE buffer contains: 10.8 g of Tris base, 5.5 g boric acid, 0.09 g Na2EDTA in 1 liter H2O). The aqueous solution was collected, extracted with phenol, extracted with chloroform and made 0.2M in sodium chloride. The DNA was obtained in aqueous solution after ethanol precipitation. The gene containing the trp promoter operator with EcoRI cohesive ends was identified by the method described below, which consists in inserting fragments into a tetracycline-sensitive plasmid which become tetracycline-resistant by incorporating a promoter operator.

The plasmid pBRHI (Rodriguez et al., Nucleic Acids Res. 6, 3267-3287 [1979]) expresses ampicillin resistance and contains the gene for tetracycline resistance, but, because there is no associated promoter, does not express this resistance. The plasmid is therefore sensitive to tetracycline . The plasmid can be made resistant to tetracycline by inserting a promoter-operator system at the EcoRI site. pBRHI was treated with EcoRI and the enzyme removed by phenol extraction and chloroform extraction. The DNA molecule obtained after ethanol precipitation in water was combined in separate reaction mixtures with each of the three DNA fragments obtained above and linked with T4 DNA ligase as already described. The DNA of the reaction mixture was used to transform competent E. coli K-12 strain 294 organisms by standard methods (Hershfield et al., Proc. Natl. Acad. Sci. USA 71.3455-3459 [1974]). The bacteria were analyzed on LB (Luria -Bertani) plates brought the 20 pg / ml -8-

AT 394 208 B

Ampicillin and 5 μ ^ πιΐ tetracycline contained. Various tretacycline-resistant colonies were selected, the plasmid DNA isolated and the presence of the desired fragment confirmed by restriction enzyme analyzes. The resulting plasmid was called pBRHtrp

An EcoRI and BamHI digestion product of the genome of the hepatitis B virus was obtained by known methods and inserted into the EcoRI and BamHI cleavage sites of the plasmid pGH6 (Goeddel et al., Nature 281.544 [1979]), whereby the plasmid pHS32 was formed. The plasmid pHS32 was digested with Xbal, extracted with phenol, extracted with chloroform and subjected to ethanol precipitation. It was then mixed with 1 μE of E. coli DNA polymerase I, Klenow fragment (Boehringer-Mannheim), in 30 μE of polymerase buffer (50 mM potassium phosphate, pH 7.4, 7 mM MgC ^, 1 mM β-mercaptoethanol), containing 0.1 mM dTTP and 0.1 mM dCTP, treated for 30 minutes at 0 ° C and 2 hours at 37 ° C. This treatment changed the Xbal cleavage site as follows 5 'CTAGA— 5' CTAGA- 3 'T— 3 ’TCT-

This linear residue of the plasmid pHS32 (after phenol and chloroform extraction and ethanol precipitation in water) was cleaved with EcoRI. The large plasmid fragment was separated from the small EcoRI-Xbal fragment by PAGE and isolated after electroelution. This DNA fragment of pHS32 (0.2 µg) was bound to the EcoRI-Taql fragment of tryptophan operon (about 0.01 µg) from pBRHtrp under conditions similar to those described above. The following happens: -T ~ AG £ CTAGA— + TCT-- -TCTAGA- ~ AG £ TCT ”with the unusual base pairing T-C. A portion of the reaction mixture was transformed into E. coli 294 cells, heat treated and placed on LB plates containing ampicillin. 24 colonies were selected, grown in 3 ml LB medium and the plasmid isolated. Six of these were found to have regenerated the Xbal cleavage sites in E. coli by catalyzed DNA repair and replication in the following manner: —TCTAGA— —TCTAGA— - &gt; —AG £ TCT— —AGATCT—

These plasmids could also be cleaved with EcoRI and Hpal and gave the expected restriction fragments. A plasmid, designated pTrpl4, was used for the expression of heterologous peptides as described below.

The plasmid pHGH 107 (Goeddel et al., Nature 281, 544, [1979]) contains a gene for human growth hormone, consisting of 23 synthetically produced amino acid codons and 153 amino acid codons, which was obtained from cDNA via reverse transcription of the mRNA. This gene, although it does not contain the pre-sequence human growth hormone codon, has an ATG translation start codon. The gene was isolated from 10 μg pHGH 107 after treatment with EcoRI and E. coli DNA polymerase I Klenow fragment, as well as dTTP and dATP, as described above. After phenol and chloroform extraction and ethanol precipitation, the plasmid was treated with BamHI.

The fragment containing the HGH gene was isolated by PAGE and electroelution. The DNA fragment obtained also contains the first 350 nucleotides of the structural gene conferring tetracycline resistance, but lacks the tetracycline promoter-operator system, so that when it is subsequently cloned into an expression plasmid, those plasmids which contain this section pass through the recovered tetracycline resistance can be identified. Because the EcoRI end of the fragment was filled in according to the Klenow polymerase I method, the fragment has a blunt end and a cohesive end, which ensures correct orientation when it is later inserted into an expression plasmid.

Next, the plasmid pTrpl4 was prepared to receive the HGH gene-containing fragment prepared above. pTrpl4 was treated with Xbal and the resulting cohesive ends became -9-

AT 394 208 B by the Klenow polymerase I method using dATP, dTTP, dGTP and dCTP. After phenol and chloroform extraction and ethanol precipitation, the DNA obtained was treated with BamHI and the resulting large plasmid fragment was eluted by PAGE and electroelution. This fragment had a blunt and a cohesive end and allowed recombination in the correct direction with the fragment containing the HGH gene, the production of which was described above.

The fragment containing the HGH gene and the pTrpl4 AXba-BamHI fragment were combined and linked together under conditions similar to those described above. By connecting the filled Xbal and EcoRI ends, the Xbal and EcoRI restriction sites were restored:

Xbal padded EcoRI padded HGH gene beginning —TCTAG AATTCTATG— —TCTAGAATTCTATG— + - &gt; —AGATC TTAAGATAC— —AGATCTTAAGATAC—

Xbal EcoRI

This construction also restored the tetracycline resistance gene. Because the plasmid pHGH 107 expresses tetracycline resistance from a promoter upstream of the HGH gene (Lac promoter), the plasmid constructed above, designated pHGH 207, allows expression of the tetracycline resistance gene under the control of the tryptophan promoter operator . The resulting mixture was transformed into E. coli 294 and the colonies were selected on LB plates containing 5 pg / ml tetracycline

The plasmid pHGH 207 was treated with EcoRI and a fragment containing 300 base pairs was obtained by PAGE and electroelution, which contained the trp promoter, operator and the trp leader ribosome binding site, but which had an ATG sequence for the start of the Translation was missing. This DNA fragment was cloned into the EcoRI site of pLelF A. Expression plasmids which contain the modified trp regulon (E. coli operon, which lacks the attenuator sequence so that expression is increased) can be grown to predetermined concentrations in a medium which contains enough tryptophan to produce the promoter. Suppress operator system. If the tryptophan is then withdrawn and the system is unblocked, the desired product is expressed

In the present case and as shown in FIG. 6, 250 pg of the plasmid pL31 were treated with Pstl and the section containing 1000 base pairs was isolated by gel electrophoresis on a 6% polyacrylamide gel. About 40 pg of the section from the gel were obtained by electroelution and divided into 3 equal parts for further treatment: a) A 16 pg sample of the fragment was partially digested with 40 units of BglII for 45 minutes at 37 ° C. and that The reaction mixture was purified on a 6% polyacrylamide gel. About 2 pg of the desired 670 base pair-containing fragment was obtained, b) Another sample (8 pg) of the 1000 base pair-containing PstI fragment was treated with Avall and Bglll. 1 pg of the 150 base pair-containing fragment shown in FIG. 6 was obtained after gel electrophoresis, c) 16 pg of the 1000 base pair-containing fragment were treated with Sau3a and Avall. After electrophoresis on 10% polyacrylamide gel, about 0.25 pg (10 pmol) of the Receive 34 base pair-containing fragment. The two indicated deoxyolinucleotides, 5'-dAATTCATGTGT (fragment 1) and 5'-dGATCACACATG (fragment 2) were synthesized according to the phosphotriester method. Fragment 2 was phosphorylated as follows: 200 pl (about 40 pM) of (γ ^ ρ) ATP (Amersham, 5000 Ci / mM) were brought to dryness and in 30 pl 60 mM Tris-HCl (pH 8), 10 mM MgC ^, 15 mM β-mercaptoethanol, 100 pM of the DNA fragment and 2 units of T4 polynucleotide kinase resuspended. After 15 minutes at 37 ° C, 1 μl of 10 mM ATP was added and the reaction was continued for a further 15 minutes. The mixture was then heated to 70 ° C. for 15 minutes, 100 pM of the 5'-OH fragment 1 and 10 pM of the 34 base pair-containing Sau3a-AvaII fragment were added. The compound was dissolved at 4 ° C for 5 hours in 50 ml of 20 mM Tris-HCl (pH 7.5), 10 mM MgC ^, 10 mM dithiothreitol, 0.5 mM ATP and 10 units of T4 DNA.

Ligase produced. The mixture was subjected to electrophoresis on a 6% polyacrylamide gel and the product containing 45 base pairs was obtained by electroelution. About 30 ng (1 pM) of the 45 base pair-containing product was combined with 0.5 pg (5 pM) of the 150 base pair-containing Avall-Bglll fragment and 1 pg (2 pM) of the 670 base-pair Bglll - Pstl fragment united. Binding was made at 20 ° C for 16 hours using 20 units of T4 DNA ligase. The ligase was then inactivated by heating to 65 ° C for 10 minutes and the mixture was treated with EcoRI and Pstl to cleave polymers of the gene. The mixture was then purified by PAGE (6%). About 20 ng (0.04 pM) of the 865 base pair containing product was isolated. Half of it (10 ng) was inserted between the EcoRI and PstI sites of pBR322 (0.3 pg). The transformation of E. coli 294 -10-

AT 394 208 B supplied 70 tetracycline-resistant, ampicillin-sensitive transformants.

The plasmid DNA isolated from 18 of these transformants was treated with EcoRI and Pstl. 16 of the 18 plasmids had an EcoRI - PstI fragment that was 865 base pairs long. 1 pg of one of these plasmids (pLelF Al) was treated with EcoRI and bound to the 300 base pair EcoRI fragment (0.1 pg) containing the E. coli trp promoter and the trp leader ribosome binding site. The transformants containing the ϋρ promoter were identified using a trp probe according to the Grunstein-Hogness colony screening method. An asymmetrically located Xbal cleavage site in the tip fragment allowed the determination of recombinants in which the trp promoter was oriented in the direction of the LelF A gene.

I. In vitro and in vivo activity of LelF A

Extracts for the IF assay were prepared as follows: 1 ml cultures were grown in L broth containing 5 pg / ml tetracycline to an A ^ g value of approximately 1.0, then with 25 ml M9 medium, containing 5 pg / ml tetracycline diluted. When the A ^ g reached 1.0, 10 ml samples were centrifuged and the cell cake suspended in 1 ml of 15% sucrose, 50 ml-Tris-HCl (pH 8.0) and 50 mM EDTA. 1 mg lysozyme was added and after 5 minutes storage at 0 ° C the cells were destroyed by ultrasound. Centrifugation was carried out for 10 minutes (15,000 rpm) and the interferon activity in the supernatant was determined by comparison with LelF standard after the CPE inhibition test. To determine the number of EF molecules per cell, a specific LelF activity of 4 x 10 ^ units per mg was used

As shown in Table 1, clone pLelF A trp 25, in which the trp promoter is oriented in the correct direction, provides high activity (up to 2.5 x 10 ^ units per liter). As shown in Table 2, the interferon produced by E. coli K-12 strain 294 / pLeIF A trp 25 behaves like authentic human LeBF; it is stable at pH 2 and is neutralized by rabbit anti-human leukocyte antibodies. The interferon has an apparent molecular weight of approximately 20,000.

Detection of interferon in vivo activity requires the presence of macrophages and natural killer (NK) cells, and these cells appear to be stimulated. It was therefore possible that the interferon produced by E. coli 294 / pLeIF A 25, which showed antiviral activity in the cell culture assay, is inactive in infected animals. Furthermore, it is entirely possible that the in vivo antiviral activity of the bacterially produced, non-glycosylated LelF A differs from that of the glycosylated interferon which is obtained from human leukocytes. The biological activities of bacterially produced LelF A (2% pure) and of LelF (8% pure) isolated from leukocytes were therefore compared using the lethal encephalomyocarditis (EMC) in squirrel monkeys (Table 3).

Table 1

Interferon activity in extracts from E. coli E. coli K-12 strain 294 cell density IF activity LelF molecules transformed by (cells / ml) E / ml culture per cell pLelF A trp 25 Ίλ X t - 'o 00 36,000 9,000 pLelF A trp 25 1.8 X 109 250,000 12,000 -11- AT 394 208 B Table 2

Comparison of the activity of extracts from E. coli 294 / pLeIF A 25 with standard LelF *)

Interferon activity (U / ml) untreated pH 2 rabbit anti-human leukocyte antibody 294 / pLeIF A trp 25 extract 500 500 &lt; 10 LelF standard 500 500 &lt; The extract of E. coli 294 / pLeIF A trp 25 containing 250,000 U / ml, which is described in Table 1, was diluted 500-fold with minimal medium so that it had a specific activity of 500 U / ml . A leukocyte interferon standard (Wadley Institute) previously titrated against NIH leukocyte interferon standard was also diluted to a final concentration of 500 U / ml. Aliquots (1 ml each) were brought to pH 2 with 1N HCl, 52 hours Incubated at 4 ° C, neutralized by adding NaOH and the IF activity determined according to the standard CPE inhibition test. Aliquots (25 μΐ) of the 500 U / ml samples (untreated) were incubated with 25 μΐ rabbit anti-human leukocyte interferon for 60 minutes at 37 ° C., centrifuged at 12,000 × g for 5 minutes and the supernatant was tested.

Table 3

Antiviral effect of various LelF preparations against EMC virus infections in squirrel monkeys

Treatment survivor serum PFU / ml day 2 day 3 day 4 control 0/3 1 ° 1 3 x 104 io5l (bacterial proteins) 0 &gt; 3 0 &quot; 104 1,200} 3.4 x 104 oj 0 ° J Bacterial 3/3 0 0 0 LelF 0 0 0 0 0 0 LelF standard 3/3 0 0 0 0 0 0

All monkeys were male (mean weight 713 g) and had no EMC virus antibodies prior to infection. The monkeys were infected intramuscularly with 100 x LD ^ q EMC virus (determined in mice). The control animals died 134, 158 and 164 hours after infection. The interferon

Treatment with 106 units was intravenous at -4, +2.23.29.48.72.168 and 240 hours after infection. The bacterial leukocyte interferon was a column chromatography- -12-

AT 394 208 B

Fraction of a lysate from E. coli 294 / pLeIF A 25 with a specific activity of 7.4 x 10® U / mg protein. The bacterial proteins used in the controls were equivalent to the column fraction of the E. coli 294 / pBR322 lysate at twice the total protein concentration. The leukocyte interferon standard was Sendai virus-induced interferon from normal human cells, which was chromatographically purified to a specific activity of 32 x 10® U / mg protein.

The KontroMere lost their balance, showed progressive lethargy, limp paralysis of the hind limbs and watery eyes, beginning about 8 hours before death. The monkeys treated with interferon did not show any of these abnormalities: they remained active throughout and did not develop viremia. The one monkey in the control group who did not develop viremia within 4 days died last (164 hours after infection), but showed high virus titers in the heart and brain after death. The interferon-treated monkeys did not develop antibodies against EMC viruses, which was found 14 and 21 days after infection. The results thus show that the antiviral effect of the LelF preparations on the infected animals can only be attributed to the action of the interferon, since the contaminating proteins of the bacterial and of the leukocyte-derived interferon are completely different. In addition, the results indicate that glycosylation is not necessary for in vivo antiviral activity of LelF A. J. Isolation of cDNA for further leukocyte interferon DNA from the plasmid, which contains DNA for the complete LelF A, was cut out with Pstl, Ύ) electrophoretically isolated and labeled with J P. The resulting radiolabelled DNA was used as a probe for the screening of additional E. coli 294 transformants obtained in the same manner as described in Part C [by the in situ colony test method of Grunstein and Hogness ( see above)]. The colonies hybridizing with the probe were isolated. Plasmid DNA from these colonies, as well as from the ten hybridizing colonies mentioned in Part G above, were excised by treatment with PstI and characterized by three different methods. Initially, these PstI fragments were characterized by the degradation products that were generated by treatment with the restriction endonucleases BglII, PvuII and EcoRI. The analysis allowed the classification of at least eight different types (LelF A, LelF B, LelF C, LelF D, LelF E, LelF F, LelF G and LelF H), shown in Figure 5, as an overview of the location of the different restriction sites relative to the previously known pre-sequence and the coding sequence of LelF A. One type of these, LelF D, appears to be identical to the LelF described by Nagata et al., Nature 284.316-320 (1980).

Furthermore, various DNAs with the Cleveland et al. (Cell 2Ü, 95-105 [1980]) described the hybridization selection test for their ability to selectively remove LelF-mRNA from RNA containing poly-A from KG-1 cells. In this test, LelF A, B, C and F were positive. Finally, the latter PstI fragments were inserted into plasmids, E. coli 294 was transformed with these plasmids and the fragments were expressed. The expressed products, which were thought to be preinterferons, were all positive in the CPE assay for interferon activity, with only the LelF F fragment showing minor activity. Incidentally, all LeEF types were sequenced. K. Direct expression of a second mature leukocyte interferon (LelF Bl

The sequence of the isolated DNA fragment containing the mature LelF B gene shows that the first fourteen nucleotides for LelF A and B are identical. Thus, a fragment from pLelF A25 containing the trp promoter operator, the ribosome binding site and the beginning of the LelF A (= B) gene was isolated and combined with the rest of the B sequence in an expression plasmid. The restriction maps for the Pst fragments of pL4 (a plasmid containing the LelF B gene from FIG. 5) and pLelF A25 are shown in FIGS. 7a and 7b.

In order to obtain the approximately 950 (= base pairs) long Sau3a-PstI fragment from the sequence shown in FIG. 7a, several steps were necessary because of the occurrence of further Sau3a restriction sites located in this area. The procedure was as follows: 1. The following fragments were isolated: a) 110 b.p. from Sau3a-EcoRI; b) 132 b.p. from EcoRI-Xba; c) 700 b.p. from Xba-Pst. 2. Fragments (la) and (lb) were joined together and treated with Xba and Bgin to prevent self-polymerization over the Sau3a and Xba ends (the corresponding Sau3a cleavage site was within a Bglll cleavage site; Bglll cuts left) a Sau3a cohesive end). A 242 b.p. containing fragment isolated -13- 5 5 10

I J

J 35 40

AT 394 208 B 3. The product of (2) was combined with (lc) and treated with Pstl and Bglll, again to prevent self-polymerization. An approximately 950 b.p. containing fragment, Sau3a-Pstl (see Figure 7a) was isolated. This fragment contained that part of the LelF B gene which had no equivalent in the LelF A. 4. An approximately 300 b.p. Containing fragment (HindIII-Sau3a) containing the trp promoter operator, the ribosome binding site, the ATG start signal and the cysteine codon of LelF A was isolated from pLelF A25. 5. An approximately 3600 b.p. containing fragment (Pstl-Hindlll) was isolated from pBR322. It contained the replicon encoding tetracycline resistance but not ampicillin resistance. 6. The fragments obtained according to 3, 4 and 5 were joined together and E. coli K-12 strain 294 was transformed with the resulting plasmid.

The transformants were subjected to a mini test (Bimboim et al., Nucleic Acids Res. 2.1513-1523 [1979]) and plasmid samples were treated with EcoRI. Degradation provided the following three fragments: 1) a 15 EcoRI-EcoRI trp promoter fragment; 2) the inner EcoRI-EcoRI fragment of pL4 and 3) the fragment from the translation start signal to the EcoRI site of pL4.

In the CPE assay, bacterial extracts from clones obtained in the manner described above show activities of 10 × 10 ^ units of interferon per liter at A ^ q = 1. A representative clone that was produced in this way is E. coli 294 / pLeIF B trp 7. 20 L. Direct expression of further mature leukocyte interferons (LelF CDF Η. I and Jj

Additional full length gene fragments from other LelF types can be tailored and incorporated into vectors for expression in analogy to the manner described for LelF A. Complete sequencing in a conventional manner can determine whether a restriction site is sufficiently close to the first amino acid codon of the mature interferon so that the method described in Part H can be used for the expression of mature LelF A, i. H. the elimination of the presequence by a restriction cut and replacement of the amino acid codons lost together with the presequence by appending a synthetic DNA fragment. If this is not the case, the following method can be used. Here, the fragment encoding the pre-sequence is separated precisely at the point where the codon for the first amino acid of the mature polypeptide begins, namely by 1. converting the double-stranded DNA into single-stranded DNA in the vicinity of this point; 2. the single-stranded region formed in the first step is hybridized with a complementary starter sequence of a single-stranded DNA, the 5 'end of which is opposite the nucleotide which borders on the intended cleavage site; 3. That part of the second strand lying in the 3 'direction of the starter which was eliminated in (1) is restored by reaction with DNA polymerase in the presence of deoxynucelotide triphosphates containing adenine, thymine, guanine and cytosine, and 4. the remaining single-stranded DNA sequence, which extends beyond the intended cleavage site, is degraded.

A short synthetic DNA sequence that ends at the 3 'end of the coding strand with the translation start signal ATG can then be linked, e.g. B. over blunt ends, with the trimmed gene for the mature interferons. The genes can be inserted into a plasmid and placed under the control of a promoter 45 and its associated ribosome binding site.

In a similar manner to that described in Section K above, gene fragments encoding LelF C and LelF B were constructed for direct expression in bacteria. In the strategy that led to the expression of these further leukocyte interferons, the fragment containing approximately 300 base pairs (Hindm-Sau3a), the trp promoter operator, the ribosome binding site, the ATG start signal and the 50 cysteine was used in each case Contained codon of LelF A from pLelF A25. This fragment was combined with the fragments of other interferons which encode the corresponding amino acid sequences which follow the common cysteine residue. Each plasmid obtained was used to transform E. coli K-12 strain 294. The following steps were necessary for the individual genes: 55

LelFC

Isolation of the following fragments from pLelF C: (a) a 35 base pair fragment Sau3a-Sau96; 60 (b) a Sau96-Pstl fragment containing more than 900 base pairs; -14-

AT 394 208 B (c) isolation of a fragment (HindIII-Sau3a) containing about 300 base pairs from pLelF A-25 as described in part K (4) above; (φ Isolation of the approximately 3600 base pair-containing fragment from section K (5) above.

Construction: (1) connection of (a) with (c). Cleavage with BglII, Hindill and isolation of the product containing about 335 base pairs. (2) Connection of the fragments obtained under (1), (b) and (d) and transformation of E. coli with the resulting plasmid.

A representative clone obtained in this way is E. coli K-12 strain 294 / pLeIF C trp 35.

LelFD

The following were isolated from pLelF D: a) a fragment containing 35 base pairs (Sau3a-AvaII); b) a fragment containing 150 base pairs (Avall-Bglll) and c) an approximately 700 base pair-containing fragment (Bglll-Pstl).

The following was isolated from pLelF A25: d) a fragment containing 300 base pairs (HindIII-Sau3a).

The following was isolated from pBR322: e) an approximately 3600 base pair-containing fragment (Hindül-Pstl).

Construction (1) The association of (a) with (b), cleavage with BglII and purification provides a 185 base pair product (2) The association of (1) with (d), cleavage with Hindill and BglII and purification provides product containing about 500 base pairs (3) Fragments (2), (c) and (e) were joined together and E. coli was transformed with the plasmid obtained

A representative clone obtained in this way is E. coli K-12 strain 294 / pLeIF D trp 11.

LelFF

The homology of amino acids 1-13 between LelF B and LelF F can be used in the construction of a complete LelF F gene and the direct expression of the corresponding interferon. A fragment (A) containing the trp promoter with suitably constructed ends is obtained from pHGH207, which is described above, by treatment with Pst I and Xba I and subsequent isolation of a fragment containing approximately 1050 base pairs. A second fragment (B) is available in the larger of the two fragments created by treating plasmid pHky 10 with Pstl and Bglll. Fragment A contains about half the gene encoding ampicillin resistance, while fragment B contains the rest of this gene and the complete gene for tetracycline resistance except for the associated promoter. Fragments A and B are combined using T4 ligase and the product obtained is combined with Xbal and Bglll treated to exclude dimerizations. This creates fragment (c), which contains the trp promoter operator and the genes for tetracycline and ampicillin resistance

A fragment (d) of about 580 base pairs is obtained by treating pLelF F with Avall and Bglll. This fragment contains the codons for amino acids 14-166 of LelF F.

A fragment (s) (49 base pairs) is obtained by treating pLelF B with Xbal and Avall. This fragment encodes amino acids 1-13 of LelF F.

Fragments (c), (d) and (e) are linked together in the presence of T4 ligase. The cohesive ends of the corresponding fragments are such that the resulting plasmid is formed in the correct manner, the gene for tetracycline resistance being brought under the control of the trp promoter operator, together with the gene for mature LelF F, so that the bacteria transformed with it can be selected on the basis of their tetracycline resistance. A representative clone obtained in this way is E. coli K-12 strain 294 pLelF F trp 1. -15-

AT 394 208 B

LelFH

The gene suitable for the expression of mature LelF H can be produced as follows: 1. Plasmid pLelF H is subjected to the treatment with Haell and Rsal and an 816 base pair fragment is isolated which does not cover the range from amino acid 10 of the signal peptide to 3 ' coding region encodes 2. The fragment is denatured and subjected to repair synthesis with the Klenow fragment of DNA polymerase I (Klenow et al., Proc. Natl. Acad. Sci. USA £ 5., 168 [1970]), in which the synthetic deoxyribooligonucleotide starter 5'-dATG TGT AAT CTG TCT is used. 3. The product obtained is cleaved with Sau3a and a fragment containing 452 base pairs, which encodes amino acids 1-150, is isolated. 4. Treatment of pLelF H with Sau3a and Pstl and isolation of a 500 base pair-containing fragment leads to a gene which encodes amino acids 150 to the end of the sequence. 5. The fragments isolated in (3) and (4) are linked to the following fragment

166 Asp Stop GATTGA

Pstl, 1st

With Cys ATG TGT ..

Sau3a encoding the 166 amino acids of LelF H 6. pLelF A trp 25 is first treated with Xbal, then with DNA polymerase I and finally with Pst I. The resulting large fragment can be isolated and linked to the product of (5) to a plasmid that can be used to transform E. coli K-12 strain 294 or another bacterium, which is then able to express mature LelF H .

LelFI

The gene library of the human genome in the phage R Charon 4A constructed by Lawn et al., (Cell 15, 1157 [1978]) was investigated for leukocyte interferon genes, the method described by Lawn et al. (see above) and Maniatis et al. (Cell 15,687 [1978]) were used. A radioactive LelF probe derived from the cDNA of the LelF A clone was used to test approximately 5000,000 plaques. Six LelF clones were obtained in this screening. After repeated screening and plaque cleaning, one of these clones, HLeIF2, was selected for further analysis.

Using the method described above, other probes can be used to isolate additional pLeEF clones from the human genome. These in turn can be used to produce further leukocyte interferon proteins according to the present invention. 1. The 2000 base pair EcoRI fragment of clone HLeIF2 was cloned into pBR325 at the EcoRI site. The resulting plasmid LelF E was digested with EcoRI and the fragment containing 2000 base pairs isolated. The deoxyoligonucleotide dAATTCTGCAG (an EcoRI-Pstl converter) was bound to the 2000 base pair-containing EcoRI fragment and the resulting fragment was cleaved with PstI to give a 2000 base pair-containing fragment with PstI ends. This was digested with Sau96 and a fragment of 1100 base pairs was isolated, which had a Pstl and a Sau96 end. 2. The plasmid pLelF C trp 35 was treated with Pstl and Xbal. The large fragment was isolated. 3. The small Xbal-Pstl fragment from pLelF C trp 35 was treated with Xbal and Sau96. An Xbal-Sau96 fragment containing 40 base pairs was isolated. 4. The fragments isolated under (1), (2) and (3) were combined to form the plasmid pLelF I trp 1.

LelF J 1. The plasmid pLelF J contains a HindIII fragment of the human genome consisting of 3800 bases, which comprises the LelF J gene sequence. A 760 base pair-containing Ddel-Rsal fragment was isolated from this plasmid 2. The plasmid pLelF B trp 7 was cleaved with Hindlll and Ddel and a 340 base pair-containing Hindlll-Ddel fragment was isolated -16-

AT 394 208 B 3. The plasmid pBR322 was cleaved with Pstl, the ends were cleaved off by incubation with DNA polymerase I (Klenow fragment) and finally treated with HindIII. The large, approximately 3600 base pair fragment was isolated 4. The fragments from (1), (2) and (3) were linked together to form the plasmid pLelF J trp 1. M. Purification

The leukocyte interferon content of bacterial extracts can be increased by the following measures in the order given: 1. Polyethyleneimine precipitation, in which most of the cellular proteins, including the interferon, remain in the supernatant. 2. Ammonium sulfate fractionation, the interferon being saturated of 55% with ammonium sulfate fails from the solution 3. Suspension of the ammonium sulfate cake in 0.06 M potassium phosphate, 10 mM Tris-HCl (pH 7.2) and dialysis against 25 mM Tris-HCl (pH 7.9), the interferon activity 4. Chromatography of the supernatant, adjusted to a pH of 8.5, remains in the solution on a DEAE cellulose column (elution with a linear gradient from 0 to 0.2 M NaCl in 25 mM Tris-HCl, pH 8, 5). 5. Adsorption on cibachrome blue agarose or hydroxylapatite and elution with 1.5 M KCl or 0.2 M phosphate solution (optional). 6. Fractionation on a Sephadex® G-75 column. 7. Cation exchange chromatography on CM cellulose in 25 mM ammonium acetate at pH 5.0 with an ammonium acetate gradient (up to 0.2 M ammonium acetate).

According to the above-mentioned process, a material with a purity of over 95% is obtained.

The material can also be purified by size exclusion chromatography, reverse phase HPLC (RP-8) or affinity chromatography on immobilized antiinterferon antibodies.

Furthermore, the material obtained according to paragraph 4 above can be applied to a column of monoclonal antibodies, which is produced as described by Milstein in Scientific American 243.66 (1980) and with 0.2 M acetic acid, 0.1% Triton and 0 , 15 M NaCl can be eluted.

In another preferred method, the interferon obtained according to the invention can be purified as follows: 1. Frozen cell material is mechanically comminuted. The partially thawed cells are in four times the volume of buffer A, containing 0.1 M Tris (pH 7.5-8.0), 10% (w / v) sucrose, 0.2 M NaCl, 5 mM EDTA, 0 , 1 mM PMSF and 10-100 mM MgC ^, suspended. The suspension is then passed through a homogenizer twice at 4 ° C. under a pressure of about 400 and less than 70 atmospheres and finally cooled in an ice bath. 2. Polyethyleneimine is slowly added to the homogenate to a concentration of about 0.35%. The mixture is left to stand and the solids are separated by centrifugation and filtration. In this step the temperature is kept below 10 ° C. The supernatant (or the filtrate) is concentrated by ultrafiltration to about a tenth of the original volume. If necessary, the solution is filtered again through a microporous membrane. 3. The clear solution is applied directly to a column of monoclonal antibodies at a flow rate of 5-8 cm / hour (e.g. 25-40 ml per hour with a column diameter of 2.6 cm). It is washed with about ten times the column volume of a solution containing 25 mM Tris-HCl (pH 7.5-8.5), 0.5 M NaCl and 0.2% of a detergent such as Triton X-100. It is then rinsed with about ten times the column volume of a solution which contains 0.15 M NaCl and 0.1% of a detergent, such as Triton® X-100. The column is then eluted with 0.2 M acetic acid containing 0.1% of a detergent such as Triton® X-100. The protein-containing fractions are pooled and adjusted to a pH of about 4 with 1N NaOH or 1.0 M Tris base , 5 brought. 4. This fraction is then placed on a cation exchanger, for example Whatmann CM52 cellulose, which has previously been equilibrated with a suitable buffer, for example ammonium acetate, pH 4.5 (50 mM). It was then washed with the buffer until the UV absorption in the eluate was constant. The column was then eluted with 25 mM ammonium acetate / 0.12 M sodium chloride and the eluate was worked up in the usual manner

The monoclonal antibodies used in the preferred embodiment can be prepared according to the method described by Staehelin et al., Proc. Natl. Acad. Be. U.S.A. 2S, 1848-52 (1981). The purification and binding of the monoclonal antibodies to Affigel-10 is described below. -17-

AT 394 208 B

Obtaining and Purifying Monoclonal Antibodies from Ascites Fluid Five female Balb / c mice were each inoculated with 5-10 x 10 ^ hybridoma cells from the middle of the logithmic growth phase. Additional mice (10 or more) were in each case intraperitoneally inoculated with 5 x 10 ^ viable cells, which were obtained from the liquid produced by inoculated mice.

The ascites fluid from each mouse was collected. The liquid was centrifuged at 500-1000 x g for 15 minutes and then at 18000 rpm for 90 minutes. The supernatant was frozen and stored at -20 ° C. After thawing, further solid material was removed by centrifugation at 35,000 revolutions per minute for 90 minutes. The individual batches obtained from each mouse were tested for specific antibody activity after a solid-phase antibody binding test (Staehelin et al., See above) and, if possible, summarized.

The protein concentrations of the solutions were approximately determined, assuming that 1 mg of protein in an 1 cm cell gave an absorption of 1.2 at 280 nm. Ascites fluids with a high antibody content contained 30-35 mg protein / ml. This corresponds to approximately 4-7 mg specific antibody / ml. The liquid was diluted with PBS (0.01 M sodium phosphate, pH 7.3.0.15 M NaCl) to a protein concentration of 10-12 mg / ml.

90 ml of ammonium sulfate solution saturated at room temperature were slowly added to each 100 ml of the diluted solutions at 0 ° C. with vigorous stirring. The suspension was kept in ice for 40-70 minutes, then centrifuged for 15 minutes at 4 ° C at 10,000 rpm. The supernatant was decanted. The protein cake was dissolved in 0.02 M Tris-HCl (pH 7.9) / 0.04 M NaCl (buffer I) and the protein solution dialyzed against 100 volumes of buffer I at room temperature for 16-18 hours. The dialyzed solution was centrifuged at 15,000 revolutions per minute for 10 minutes. About 30-35% of the original total protein content of the ascites fluid was recovered by measuring the absorbance at 280 nm.

A solution containing 30-40 mg protein / ml was then placed on a DEAE cellulose column equilibrated with buffer I, using at least a column volume of 100 ml for every 1 g protein. The antibody was eluted from the column with a linear NaCl gradient (0.04 M-0.5 M) in 0.02 M Tris-HCl (pH 7.9). The fractions with an NaCl content of 0.06-0.1 M were pooled, concentrated and mixed with ammonium sulfate solution saturated at room temperature. It was centrifuged, the protein cake was dissolved in 0.2 M NaHCOj (pH about 8.0) / 0.3 M NaCl (buffer II) and dialyzed against this buffer at room temperature. The dialyzed solution was centrifuged at 20,000 x g for 15 minutes and the protein content was adjusted to 20-25 mg / ml with buffer II

Manufacture of immunoadsorbents

Affigel-10 (BioRad Laboratories, Richmond, California) was washed on a sintered glass filter three times with ice cold isopropanol and three times with ice cold distilled water. The gel, which still contained about 50% water, was transferred to plastic tubes and sedimented by briefly centrifuging. The supernatant was evaporated. The packed gel was mixed with an equal volume of purified antibody solution and rotated end-to-end at 4 ° C for 5 hours. After the reaction, the gel was centrifuged and washed twice with buffer III (0.1 M NaHCOj / O. IS M NaCl) to remove unbound antibodies. Protein determination showed that more than 90% of the body had been bound to the gel.

In order to saturate free, reactive binding sites, the gel was mixed with an equal volume of 0.1 M ethanolamine-HCl (pH 8) and rotated end-to-end for 60 minutes at room temperature. Finally, the gel was washed with PBS and stored at 4 ° C in PBS in the presence of 0.02% (w / v) sodium azide. N. Parenteral application

LelF can be administered parenterally to individuals who need antitumor or antiviral treatment and to those who have immunosuppressive conditions. The dosage of the interferons according to the invention can be based on that of the material currently in clinical investigation, which was obtained from human leukocytes, and can e.g. B. about 1-10 x 10 ^ units per day, in the case of material with a purity that is greater than 1%, up to about 5 x 10 ^ units per day.

A preparation suitable for parenteral administration can be prepared, for example, by dissolving 3 mg of bacterially produced LelF with a specific activity of about 2x10 ^ units per mg in 25 ml of 5 N human serum albumin, passing the solution through a bacteriological filter and filtering it Fill the solution aseptically into 100 ampoules, each containing 6 x 10 ^ units of pure interferon. The ampoules are preferably stored in the cold (-20 ° C). -18-

Claims (16)

AT 394 208 B Pharmaceutical preparations containing compounds according to the invention can be prepared by the generally known methods of galenics using the customary auxiliary and carrier materials as described in the relevant literature. 1. A method for producing a replicable vector which in a microorganism causes the expression of a polypeptide with the amino acid sequence of a mature human leukocyte interferon, characterized in that a first DNA sequence coding for this polypeptide is operable with a second DNA sequence, which contains the codons necessary for expression. 2. The method according to claim 1, characterized in that the second DNA sequence causes expression in a bacterium. 3. The method according to claim 2, characterized in that the bacterium is E. coli. 4. The method according to any one of claims 1 to 3, characterized in that the polypeptide has the amino acid sequence of an amino terminal extended by methionine mature human leukocyte interferon. 5. The method according to any one of claims 1 to 3, characterized in that the polypeptide has the amino acid sequence of an amino terminal extended by a cleavable conjugate or a microbial signal protein mature human leukocyte interferon. 6. The method according to any one of claims 1 to 3, characterized in that the polypeptide contains the amino acid partial sequence Cys-Ala-Trp-Glu-Val-Val-Arg-Ala-Glu-Ile-Met-Arg-Ser. 7. The method according to any one of claims 1 to 3, characterized in that the polypeptide which in Figure 4 in line &quot; All &quot; contains specified amino acids at the specified positions in the amino acid sequence. 8. The method according to any one of claims 1 to 3, characterized in that the polypeptide is human leukocyte interferon (LelF) A, B, C, D, F, Η, I or J. 9. The method according to any one of claims 1 to 3, characterized in that human leukocyte interferon A (LelF A) is expressed. 10. The method according to any one of claims 1 to 3, characterized in that human leukocyte interferon B (LelF B) is expressed. 11. The method according to any one of claims 1 to 3, characterized in that human leukocyte interferon C (LelF C) is expressed. 12. The method according to any one of claims 1 to 3, characterized in that human leukocyte interferon D (LelF D) is expressed. 13. The method according to any one of claims 1 to 3, characterized in that human leukocyte interferon F (LelF F) is expressed. 14. The method according to any one of claims 1 to 3, characterized in that human leukocyte interferon H (LelF H) is expressed. 15. The method according to any one of claims 1 to 3, characterized in that human leukocyte interferon I (LelF I) is expressed. -19- 5 AT 394 208 B 16. The method according to any one of claims 1 to 3, characterized in that human leukocyte interferon J (LelF J) is expressed. Add 10 sheets of drawings -20-