BG60508B2 - Dna sequences encoding erythropoietin - Google Patents

Abstract

Erythropoietin is used in the diagnosis and treatment of various forms of anemia and hypoxia, as it increases the production of red blood cells. Recombinant DNA methods are used to obtain it. They yield DNA sequences encoding part or all of the human or monkey erythropoietin sequence. Plasmids or expression vectors obtained from them are used to transform or transfect prokaryotic or eukaryotic host cells. Upon expression, polypeptides exhibiting the biological properties of natural erythropoietin are produced. 31 claims

Description

Field of application

The invention generally relates to the manipulation of genetic materials and, in particular, to recombinant methods that make it possible to obtain polypeptites having part or all of the primary structural conformations and / or one or more of the biological properties of natural erythropoietins.

BACKGROUND OF THE INVENTION

A. Manipulation of genetic materials

Genetic materials can be broadly defined as chemicals that program for and guide the production of cell and virus elements and guide cell and virus responses. A long-chain polymer substance, known as deoxyribonucleic acid (DNA), constitutes the genetic material of all living cells and viruses, with the exception of some of which are programmed by ribonucleic acids (RNA). The repeating units in DNA polymers are four different nucleotides, each consisting of purine (adenine or guanine) or pyrimidine (thymine or cytosine) linked to a deoxyribose to which a phosphate group is attached. Binding of nucleotides in a linear polymeric form is accomplished by attaching the 5′-phosphate of one nucleotide to the 3′-hydroxyl group of another. Functional DNA is in the form of stable double-stranded associations of single nucleotide chains (known as deoxyribonucleotides), which are obtained by hydrogen bonding between the purine and pyridine bases, i.e., "additional" associations existing between adenine (A) thymine (T) or guanine (G) and cytosine (C). Conveniently, nucleotides are referred to by the names of their constituent purine or pyrimidine bases, and additional nucleotide associations in double stranded DNA (i.e., AT and G-C) are referred to as base pairs. Ribonucleic acid is a polynucleotide containing adenine, guanine, cytosine and uracil, not thymine associated with ribose and a phosphate group.

In short, the programming function of DNA is manifested by a process in which specific DNA nucleotide sequences (genes) are transcribed into relatively unstable RNA polymers. In turn, mRNA serves as a template for the formation of structural, regulatory and catalytic proteins from amino acids. This mRNA translation process involves small RNA strands (mRNAs) that transport and arrange individual amino acids along the mRNA strand to form polypeptides with the correct amino acid sequence. The mRNA information originating from the DNA providing the mRNA base and the orientation of each of the twenty amino acids for expression of the polypeptides is in the form of a triplet codon, a triplet forming groups of three nucleotide base sequences. In a sense, protein formation is the latest expression of programmed genetic information provided by the nucleotide sequence of a gene.

A promoter DNA sequence typically precedes a single gene in the DNA polymer and provides a site for initiation of transcription in mRNA. A regulatory DNA sequence, also typically preceding one gene in a given DNA polymer, binds proteins that determine the frequency (or rate) of transcription initiation. Collectively referred to as a promoter / regulatory or control DNA sequence, those sequences that precede certain genes (or series of genes) into a functional DNA polymer help determine whether transcription (or possibly expression) of a gene will take place. DNA sequences that follow a single gene in a given DNA polymer and signal the completion of transcription in mRNA are called transcriptional "terminator" sequences.

The focus of microbiological development over the last decade has been on the attempts to produce industrially and pharmaceutically relevant substances by using microorganisms that either do not have initially genetically encoded information included in their DNA about the desired product, or (in the case of mammalian cell culture) ) do not usually "express" a chromosomal gene at the appropriate level. In simple terms, the gene that determines the structure of the desired polypeptide product is either isolated from a "donor" organism or chemically synthesized and then injected into another microorganism, which is preferably a self-replicating unicellular microorganism such as bacteria, yeast or mammalian cells in culture. Once done, the existing mechanism of gene expression in a transformed or transfected host microbial cell acts to construct the desired product, using exogenous DNA as an mRNA transcription template, which is then translated into a continuous amino acid sequence.

This field is rich in patent and literature publications relating to recombinant DNA methodologies to isolate, synthesize, purify and soften genetic material for use in transforming selected host organisms. For example, US 4,237,224 to Cohen et al. refers to the transformation of single-celled host organisms with hybrid viral or circular plasmid DNA that includes selected exogenous DNA sequences. The processes in this patent include first obtaining a transformation vector by enzymatically cleaving viral or circular DNA to form linear DNA strands. Under foreign exogenous or heterogeneous DNA strands, typically including sequences coding for the desired product, are obtained in linear form using similar enzymes. Linear viral or plasmid DNA is incubated with foreign DNA in the presence of ligation enzymes suitable for creating the repair process, hybrid vectors are formed that include the selected exogenous DNA segment "entangled" in the viral or circular DNA plasmid.

Transformation of suitable unicellular host organisms with the hybrid vector completes the formation of multiple copies of exogenous DNA in the host cell population. In some cases, the expected result is simply an increase in foreign DNA and the collected "product" is DNA. More commonly, the purpose of transformation is to express exogenous DNA through host cells in the form of large-scale synthesis of isolated amounts of commercially significant protein or polypeptide fragments encoded by foreign DNA. See also, e.g., US 4,264,731, 4,273,875, 4,293,652 and EP 093,619, published Nov. 9, 1983.

The development of specific DNA sequences for the formation of DNA vectors is carried out by different methodologies, depending largely on how foreign the donor is to the host and the size of the polypeptide to be expressed in the host. At the risk of simplification, it can be said that three alternative principles can be used: (1) isolation of a double stranded DNA sequence from the genomic DNA of the donor; (2) the chemical preparation of a DNA sequence providing a code for the polypeptide of interest; and (3) in vitro synthesis of double stranded DNA sequencing by enzymatic reverse transcription of and RNA isolated from donor cells. The latter mentioned methods, which involve the formation of a single strand of mRNA, are commonly reported as cDNA methods.

The production of DNA sequences is often a matter of choice when the entire sequence of amino acid residues of the desired polypeptide product is known. The processes for obtaining DNA according to Joint Patent Application US 483,451 (filed April 15, 1983 and corresponding to PCT / US83 / 00605, published November 24, 1984 as WO83 / 04053) provide a very good means of obtaining such highly desirable results as : providing for the availability of other codons commonly found in geneGy1 that are highly expressed in the host cells selected for expression (eg, the production of yeast or E. coli preferred codons); avoidance of untranslated intron sequences typically contained in mammalian genomic DNA sequences and RNA transcriptions thereof) that are not completely processed by prokaryotic host cells; avoiding the expression of unwanted "leader" polypeptide sequences, usually encoded by genomic DNA and cDNA sequences, but often not completely delimited by the polypeptide of interest in bacterial or yeast host cells; providing accurate incorporation of DNA into ordinary expression vectors in relation to desired promoter / regulatory and terminator sequences; and creating the conditions for the proper construction of genes encoding for polypeptide fragments and analogs of the desired polypeptides.

When the entire sequence of amino acid residues of the desired polypeptide is not known, direct production of DNA sequences is not possible and isolation of DNA sequences encoding the polypeptide by the cDNA method is selected, regardless of the possible disadvantages regarding the ease of creating expression vectors, providing a high level of microbial expression. Among standard methods for isolating cDNA sequences of interest is the production of plasmid-derived cDNA banks, obtained by reverse transcription of mRNA into donor cells, selected to provide a high level of gene expression (e.g., hypo4-derived cDNA banks physical cells that are used to express large amounts of hormonal growth products). When a significant portion of the amino acid sequence of a polypeptide is known, labeled single stranded DNA sequences duplicating a sequence presumably present in the cDNA template can be used in DNA / DNA hybridization methods performed on cloned copies of cDNA that have been denatured form. This issue has been discussed and the conclusions reached are set forth in US 4394443, Weissman, et al., And the latest reports are for the use of long oligonucleotide hybridization probes in Wallace, et al., Nuc. Acids Rev., 6, pp. 3543-3557, (1979), Reyes, etal., P. N. A. S. (USA), 79, pp. 3270-3274 (1982) and Jeye, et al., Nuc. Acids Res., 11, pp. 2325-2335 (1983). The DNA / DNA hybridization methods used for diagnosis include US 4358535 from Falkow, et al., And light-emitting markers on single-stranded polynucleotide probes published EP 0070658 and 0070687, Davis, et al., "A Manual for Genetic Engineering, Advanced Bacterial Genetics', Cold Spring Harbor Laboratory, Cold Spring Harbor, NY (1980), p.55-58 and 176-174, refer to colony and plaque hybridization method; and the New England Nuclear Boston, Mass. brochure for "Gene Screen" membrane transfer materials for hybridization transfer, offering a guide for DNA and RNA transfer and hybridization, catalog N NEF-973.

Among the more significant recent advances in hybridization methods for screening recombinant clones is the use of labeled mixed synthetic oligonucleotide probes, each of which is a potentially complete complement to a specific

A DNA sequence in a hybridization probe comprising a hetero5 gene mixture of single stranded DNA and

RNA. These methods are recognized as particularly useful in detecting cDNA clones from sources that provide an extremely small amount of mRNA sequences for the polypeptides of interest. In short, the use of stringent hybridization conditions aimed at avoiding nonspecific binding may allow, for example, 15 autoradiographs of a specific cDNA clone when hybridizing target DNA to that single probe in the mixture, which is its complete complement. See Wallace et al., Nuc, 20 Acids Res. 9, pp. 879-897 (1981);

Suggs et al. P.N.A.S. United States, 78, pp. 6613-6617 (1981); Choo et al., Nature 299, pp. 178-180 (1982); Kurachi et al., P.N.A.S. (USA), 79, pp. 646125 6464 (1982); Ohkubo et al., P.N.A.S.

(USA) 80, pp. 2196-2200 (1983); and Kornblihtt et al., P.N.A.S. (USA) 80, pp. 3218-3222 (1983). Generally, the mixed-probe method of Wallace 30 et al., (1981) (see above) was extended by various researchers to obtain reliable results when isolating a cDNA clone using a 32-member mixed pool of 35 16 bases long (16-dimensional) oligonucleotide probes from single, variable DNA sequences together with an 11-dimensional to create a double-stranded positive conformation for the 40 personality of cDNA of interest.

See Singer-Sam et al., P.N.A.S (USA) 80, pp. 802-806 (1983).

The use of genomic DNA isolates has been encountered in at least the three 45 methods mentioned for developing specific DNA sequences for use in recombinant methods.

This relates especially to the field of recombinant methods aimed at providing microbial expression of mammalian polypeptides and is due in principle to the complexity of mammalian genomic DNA. Although reliable methods exist for the development of phage-derived genomic DNA banks of human and other mammalian origin (see, e.g., Down et al., Cell, 15, p. 11571 174 (1978), for methods for the production of human genomic banks commonly referred to as the Maniatis Library; Karn et al., PNAS (USA) 77, pp. 5172-5176 (1980), referring to a human genomic bank based on an alternative method of restriction endonuclease fragmentation; and Blattner et al., Science, 196, pp. 161-169, (1977) describing the production of a bovine genomic bank) has had relatively few successful attempts. and using hybridization methods for the isolation of genomic DNA in the absence of extensive prior knowledge of the amino acid sequences of DNA. As an example, Fiddes et al., J. Mol. and App. Genetics, 1, pp. 3-18 (1981) describe the successful isolation of a gene encoding a human pituitary hormone glycoprotein from a human pituitary hormone using a full length sample comprising a fragment of 621 base pairs from a pre-isolated cDNA sequence of an alpha subunit. As another example, Das et al., P.N.A.S (USA), 80, pp. 1531-1535 (1983) reported the isolation of human genomic clones of human HLADR using a synthetic oligonucleotide of 175 base pairs. Anderson et al., P.N.A.S. (USA), 80, pp. 6838-6842 (1983) report the isolation of a genomic clone for a bovine pancreatic trypsin new inhibitor (BPT1) using a single sample of 86 base pairs in length and constructed according to the known amino acid sequence of BPTI. The authors note the very low possibility of isolating mRNA suitable for cDNA bank synthesis due to the apparently low mRNA content of the designated parotid gland and lung tissue sources, and therefore direct the prospect of success in probing a genomic bank using a mixture of labeled probes, stating: "In general, oligodeoxynucleotide probes with mixed sequences for isolation of protein genes of unknown sequence from cDNA banks. Such probes are typically mixtures of 8-32 oligonucleotides, 14-17 nucleotides in length, providing all possible codon combinations for small portions (5-6 residues) of amino acid sequence. Under stringent hybridization conditions that limit samples with irregular base pairs, these mixtures are able to locate specific gene sequences of low to moderate complexity in bank clones. However, because of their small length and heterogeneity, mixed probes often lack the specificity required to probe sequences as complex as the mammalian genome. This makes such a method useless in isolating mammalian protein genes when the corresponding mRNA is lacking.

Thus, there is a continuing need in the art for improved methods for the rapid and efficient isolation of cDNA clones in cases where little is known about the amino acid sequence of the polypeptide encoded and where "enriched" tissue sources of RNA and RNA are not readily available available for use in constructing cDNA banks. Such improved methods can be used especially if they are applied to isolate genomic clones in mammals where there is scarce information about the amino acid sequences of the polypeptide for which a given gene encodes.

B. Erythropoietin as polypeptide of interest

Erythropoiesis, the production of red blood cells, is carried out continuously throughout the life of a person until the destruction of the cell occurs. Erythropoiesis is a well-controlled physiological process, ensuring that there are enough red blood cells in the blood to properly supply the tissues with oxygen, but not so much that the cells interfere with circulation. The formation of red blood cells occurs in the bone marrow and is carried out under the control of the hormone erythropoietin.

Erythropoietin, an acidic glycoprotein with a molecular weight of approximately 34,000 daltons, can exist in three forms: α, β and asialo. Alpha and beta forms differ slightly in carbohydrate components but have the same action, biological activity and molecular weight. The Asian form is an alpha or beta form with the final carbohydrate removed (sialic acid). Erythropoietin is found in very low plasma concentrations when the body is healthy, as tissues receive sufficient oxygen from the available erythrocytes. This normal low concentration is sufficient to stimulate the replacement of red blood cells that are normally lost due to aging.

The amount of erythropoietin in the bloodstream increases under conditions of hypoxia when the transport of oxygen from blood cells to the bloodstream is reduced. Hypoxia can be caused by the loss of large amounts of blood due to hemorrhage, the destruction of red blood cells upon prolonged exposure to radiation, the reduction of oxygen uptake at high altitudes, prolonged insomnia, or various forms of anemia. In response to tissues subjected to hypoxic stress, erythropoietin will increase the production of red blood cells by stimulating the conversion of primitive transitional cells into bone marrow into pro-erythroblasts, which subsequently mature, synthesize hemoglobin and release into the bloodstream. When the number of red blood cells in the bloodstream is greater than necessary to meet normal tissue needs for oxygen, erythropoietin in the circulation decreases.

See generally Testa et al., Exp. Hematol. 8, (Supp. 8), 144-152 (1980); Tong et al., J. Biol. Chemo. 256 (24), 12666-12672 (1981); Goldwasser, J. Cell. Physiolin 110 (Supp. 1), 133-135 (1982); Finch, Blood, 60 (6), 1241-1246 (1982); Sytdowski et al., Expt. Hematol. 8 (Supp. 8), 52-64 (1980); Naughton, Ann.Clin.Lab.Sci. 13 (5), 432-438 (1983); Weiss et al., Am. J. Vet. Really. 44 (10), 1839-1835 (1983); Lappin et al .. Exp. Hematol. 1 1 (7), 661-666 (1983); Baciu et al., Ann. Ν. Y. Acad Sci. 414, 66-72 (1983); Murphy et al., Acta Haematdologica Japanica 46 (7), 1380-1396 (1983); Desspyris et al., Brit. J. Haematol. 56, 295-306 (1984);

Emmannouel et al .. Am. J. Physiol. 247 (Pt 2). F 168-76 (1984).

Because erythropoietin is essential in the process of red blood cell formation, the hormone has a potential utility in both the diagnosis and treatment of blood disorders characterized by poor or impaired red blood cell production. See, e.g., Renna thur-Das et al., Blood, 63 (5), 1168-71 (1984) and Haddy, Am. Jour. Ped. Hematol / Oncol. 4, 191-196 (1982), relating to erythropoietin in the treatment of sickle cell anemia and Eschbach et al. J. Clin. Invest. 74 (2), p. 434441 (1984), which describes therapeutic regimens for uremic sheep based on an in vivo erythropoietin-rich plasma infusion response and provides a dosage of 10 EPO / kg per day for 15 to 40 days as a remedy for anemia of the type associated with chronic renal failure. See also Krane, Henry Ford. Hosp. Med. J. 31 (3), 177-181 (1983).

Recently, it has been found that the availability of erythropoietin in quantity would allow the treatment of anemia per year for 1,600,000 people in the United States alone. See, for example, Morrison, “Bioprocessing in Space - an Overview pp. 557-571 in The World Biotech Report 1984, Volume 3, USA (Online Publication,,. Y. 1984). Recent studies have laid the foundation for expanding the field of erythropoietin treatment in a variety of conditions, disorders and haematological disorders. Vedorato et al., Acta Haematol. 71, 211-213, 1984 (beta-thalassemia); Vichinsky et al., J. Pediatr. 104 (1), 15-21 (1984) (cystic fibrosis), Cotes et al., Brit. J. Obstet Gunecol. 90 (4), 304-311 (1983) (pregnancy, menstrual disorders); Haga et al., Acta Pediatr. Scand. 72. 827-831 (1983) (early anemia in preterm birth); ClausWalker et al., Arch. Phys. Med. Rehabil, 65, 370-374 (1984) (spinal cord damage); Dunn et al., Eur. J. Appl. Physiol. 52, 187-182 (1984) (space flight); Miller et al., Brit. J. Haematol. 52, 545-590 (1982) (severe bleeding); Udupa et al., J. Lab. Clin. Med. 103 (4). 574-580 (1984); Lipschitz et al., Blood. 63 (3), 502-509 (1983) (aging); Dainiak et al., Cancer 51 (6), 1 101-1106 (1983) and Schwarts et al., Otolaryngol. 109, 269272 (1983) (various neoplastic disease states accompanied by abnormal erythropoiesis).

Previous attempts to produce erythropoietin in good quantities of plasma or urine have proved relatively unsuccessful. Complex and complicated laboratory procedures are required and usually result in the collection of too small amounts of crude and unstable extracts containing erythropoietin.

U.S. Patent 3,033,753 describes one method for the partial purification of erythropoietin from sheep blood plasma that yields a small yield of a crude solid extract containing erythropoietin.

Initial attempts to isolate erythropoietin from urine yielded unstable, biologically inactive hormone preparations. US Patent No. 3,865,801 describes a method of stabilizing the biological activity of a crude substance containing erythropoietin derived from urine. The resulting crude preparation containing erythropoietin has 90% of erythropoietin activity and is stable.

Another method for the purification of human erythropoietin from the urine of patients with aplastic anemia has been described by Miyake et al., J. Biol. Chem .. Volume 252, No. 15, Aug. 10, 1975, p. 5558-5564. This seven-step method involves ion exchange chromatography, ethanol precipitation, gel filtration and adsorption chromatography and yields a pure erythropoietin preparation with 70,400 units / mg protein in 21% yields.

US Patent No. 4,397,840 to Takezawa et al. Describes methods for the preparation of a 'single erythropoietin product' from urine in healthy human subjects with weak basic ion exchange resins and suggests that low molecular weight products have no inhibitory effect on erythropoietin.

Patent Application GB 2,085,887 from Sugimoto et al., Published in May 1982, describes a method for producing hybrid human lymphoblastoid cells in yields of 3 to 420 units of erythropoietin per ml of cell suspension (culture, after multiplication of mammalian hosts, contains up to 10 ⁷ cells per milliliter). At the highest yields that can be achieved, the production rate of erythropoietin may be 40 to about 4,000 units / 10 ⁶ cells / 48 hours in vitro, followed by cell transfer from an in vivo propagation system . (See also equivalent US 4,377,513) Many proposals have been made to isolate erythropoietin from tissue sources, including neoplastic cells, but yields are quite low. See, e.g., Jelkman et al., Expl. hematol. 11 (7), 581-588 (1983); Tambourin et al., PNAS (USA), 8, 6269-6273 (1983); Katsuoka et al., Gann. 74, 534-541 (1983); Hagiwara et al., Blood, 63 (4), 828-835 (1984); Choppin et al., Blood, 64 (2), 341-347 (1984).

Other isolation techniques used to prepare purified erythropoietin include immunological procedures. A polyclonal serum-derived antibody reactive to erythropoietin is obtained by injecting an animal, preferably a rat or rabbit, with human erythropoietin. Injected human erythropoietin is recognized as a foreign antigenic substance by the animal's immune system and initiates the production of antigen antibodies. Different cells, in response to stimulation by the antigenic substance, produce and release into the bloodstream antibodies slightly different from those produced by other responsive cells. Antibody activity is maintained in the serum of the animal when its blood is extracted. Although the crude serum or antibody preparation purified as a fraction of serum immunoglobulin G may be used in human erythropoietin detection and complexation studies, these materials have one major drawback. The serum antibody, made up of all the different antibodies produced by the individual cells, is polyclonal in nature and forms complexes with the constituents of the crude extract other than erythropoietin.

Of interest in the prior art are the recent advances in the development of permanent cell cultures capable of producing a particular type of antibody that is specifically immunologically reactive with a single antigenic determinant of a selected antigen. See generally Cisholm, High Technology, Volume 3, Issue 1, 57-63 (1983). Attempts have been made to use cell fusion and hybridization methods to produce monoclonal antibodies against erythropoietin and to use these antibodies to isolate and quantify human erythropoietin. A message about the successful development of mouse-murine hybridoma cell lines secreting monoclonal antibodies against human erythropoietin appeared in abstract in Lee-Huang, Fed. Abstract N 1463. Proc. 41, 520 (1982). Another example is a detailed description of the preparation and use of a monoclonal anti-erythropoietin antibody by Weis et al., P.N.A.S. (USA) 79, 5465-5469 (1982). See also Sasaki, Biomed. Biochim. Acta, 42 (1 1/12), 202-206 (1983); Yanagawa et al., Blood, 64 (2), 357-364 (1984); Yahagawa et al., J. Biol. Chem. 259 (5), 2707-2710 (1984); U.S. Patent No. 4,465,624.

Also included in the prior art are reports of the immunological activity of synthetic peptides, which essentially duplicate the amino acid sequence of naturally occurring proteins, glycoproteins and nucleoproteins. In particular, relatively low molecular weight polypeptides have been shown to exhibit immune responses that are the same in duration and range with immune responses to physiologically relevant proteins such as viral antigens, polypeptide hormones and the like. Particularly of the immune responses of such polypeptides is the provocation of the formation of specific antibodies in immunologically active animals. See, for example, Lerner et al., Nature 294, 654-656 (1981); Walter et al., P.N.A.S. (USA) 77, 5197-5200 (1980); Lerner et al., P.N.A.S. (USA) 78.3403-3407; Walter et al., P.N.A.S. (USA) 78, 4882-4886 (1981); Wong et al., P.N.A.S. (USA) 78,

7412-7416 (1981); Green et al., Cell. 28. 477-487 (1982); Nigg et al .. P.N.A.S. (USA) 79. 5322-5326 (1982); Baron et al., Cell, 28, 395-404 (1982); Dreemsn et al., Nature 295, 158-160 (1982); Lerner, Scientific American 248, N 2, 66-74 (1983); see also Kaiser et al., Science, p. 249-255 (1984) concerning the biological and immunological activities of synthetic peptides. which have approximately secondary structures of peptide hormones but may not have such primary structural conformations. The above studies refer, of course, to amino acid sequences of proteins other than erythropoietin. a substance for which no essential information has been published on the amino acid sequence. Patent Application US 463 724, filed 4.02.83 by J. Egrie, published on 08/22/83, as EP 0116446, describes a mouse-murine hybridoma cell line (ATCC N HB 8209) that produces a highly specific monoclonal anti-erythropoietin an antibody that is also specifically immunoreactive with a polypeptide having the following sequence of amino acids:

N H ₂ - Ala-Pro-Pro-Arg-Leu-Ile-Cys-AspSer-ArgVal-Leu-Glu-Arg-Tyr-Leu-Leu-Glu-AlaLys-COOH

The polypeptide sequence is the same as that of the first twenty amino acid residues of mature human erythropoietin isolated by the method of Miyake et al., J. Biol. Chem. 252, 5558-5564 (1977), the amino acids of which were analyzed by a gas phase sequencer (Applied Biosystems, Inc.) according to the method of Hewick, M. et al., J. Biol. Chem. 256, 7990- 7997 (1981). See also Sue et al. Proc. Natl. Acad. Sci. (USA) 80, p.

3651-3655 (1983) concerning the development of polyclonal antibodies against a synthetic 26-mer based on a different amino acid sequence and Sytowski et al., J. Immunol. Methods 69, pp. 181-186 (1984).

Although polyclonal and monoclonal antibodies, as described above, create very useful products for the application of immune assays for the detection and quantification of erythropoietin and can be used in the affinity purification of erythropoietin, it is unlikely that these materials can be used to isolate large amounts of erythropoietin from mammalian sources sufficient for other analyzes, clinical trials and potentially wide-ranging therapeutic administration of the substance in the treatment of, for example, chrono no kidney disease in which damaged tissues cannot support the production of erythropoietin. Therefore, developments in this field aimed at producing high quality mammalian erythropoietin and providing large quantities of it for diagnostic and clinical use include the successful application of recombinant methods for performing large-scale microbial synthesis of the compound.

Despite considerable efforts to direct the isolation of DNA sequences encoding for human and other mammalian erythropoietin types, none of the experiments proved successful. This is due in principle to the scarcity of tissue sources, especially human tissue sources enriched in mRNAs, such that would allow the creation of a cDNA bank from which DNA sequencing encoding erythropoietin by conventional methods can be isolated. Moreover, little is known about the continuous sequence of the amino acid residues of erythropoietin. so it is not possible to create, for example, long polynucleotide probes that are reliable and capable of being used for DNA / DNA hybridization screening of cDNA, and in particular of a genomic DNA bank. The twenty amino acid sequence used to prepare the above monoclonal antibody produced by ATTC N HB 8209 does not allow the creation of the required oligonucleotide probe composed of 60 bases by the method described by Andersen et al. (look above). It has been shown that the human erythropoietin gene may appear as a single copy gene in the human genome, and in some cases the genetic material encoding human erythropoietin may account for less than 0.00005% of the total human genomic DNA, available in the genomic bank.

The most successful of the known experiments with recombinant methods for obtaining DNA sequences suitable for use in microbial expression of isolated amounts of mammalian erythropoietin do not achieve the intended purpose. As an example, Farber et al., Exp. Hematol. 11, Supp. 14, Abstract 101 (1983) reported the extraction of kidney tissue mRNA from phenylhydrazine-treated baboons and the injection of that mRNA into Xenopus laevis oscites with a rather transient result consisting in the in vitro production of a mixture of translational products, including some exhibiting the biological properties of erythropoietin. Recently, Farber et al., Blood, 62, Ν 5, vol. Ν 1, abstract 392 on page 122a (1983) reports the in vitro translation of human renal mRNA by frog oscites. The resulting translational product was found to include, in the order of 220 mU, a translational product exhibiting erythropoietin activity per microgram of injected mRNA. Although such in vitro content of exogenous mRNA encoding erythropoietin is considered to be quite low (compared to the previously found content in translation of baboon mRNA in the product sought), the result is considered to confirm the human kidney as a site for the expression of erythropoietin allowing the production of a cDNA enriched bank from a human kidney from which the desired gene can be isolated (see also Farber, Clin. Res. 31 (4), 769A, 1983).

Following patent applications US N 561 024 and 582 185, a single message for the cloning and expression of what is believed to be human erythropoietin cDNA in E. coli appears. In short, a number of cDNA clones are included in E. coli plasmids, and the products obtained with betalactamase fusion have been found to be immunoreactive with a monoclonal antibody to a theoderic "epitope" of human erythropoietin. See LeeHuang, Proc., Nat. Acad. Sci. (USA) 81, pp. 2708-2712 (1984).

SUMMARY OF THE INVENTION

The present invention relates to novel purified and isolated polypeptide products having part or all of the primary structural conformation (i.e., a continuous sequence of amino acid residues) and one or more biological properties (e.g., immunological properties and in vivo and in vitro biological activity of natural erythropoietins, including allelic variants thereof, These polypeptides are also uniquely characterized to be the product of expression in a prokaryotic or eukaryotic host (e.g. p bacteria, yeast and mammalian cells in culture) of an exogenous DNA sequence obtained by genomic or cDNA cloning or by gene synthesis Products of microbial expression in vertebrate cells (eg mammals and birds) may also be characterized by that they do not associate with human proteins or other impurities that are sometimes associated with erythropoietin in its natural environment in mammalian cells or in extracellular fluids such as plasma or urine. The products of typical yeast (eg Saccharomyces cerevisiae) or prokaryotic (eg E. coli) host cells do not bind to any mammalian proteins. Depending on the host used, the polypeptides of the invention may be glycosylated with carbohydrates derived from mammals or other eukaryotic cells or may not be glycosylated. The polypeptides of the invention may also include an initial methionine amino acid residue (at position -1).

The novel glycoprotein products of the invention include those having a primary structural conformation sufficiently similar to natural (e.g., human) erythropoietin to provide the manifestation of one or more of its biological properties and having an average carbohydrate content other than natural (e.g. human) erythropoietin.

The vertebrate cells (eg COS-Ι and CHO) included in the present invention are the first available cells that can reproduce in vitro continuously and which, when cultured, are capable of producing in the culture medium in excess of 100 units (preferably in excess of 500 units and especially in excess of 1000 to 5000 units) erythropoietin per 10 ⁶ cells for 48 hours as determined by radioimmunoassay.

The present invention also relates to synthetic polypeptides that completely or partially repeat the continuous sequences of the amino acid residues of erythropoietin, which are described herein for the first time. These sequences, in fact having the primary, secondary or tertiary structural and conformational characteristics of natural erythropoietin, may exhibit biological activity and / or immunological properties in common with the natural product so that they can be used as biologically active or immunological substitutes for erythropoietin in therapeutic and immunological methods. Accordingly, monoclonal and polyclonal antibodies obtained by known methods are provided which are immunoreactive with such polypeptides and, preferably, are also immunoreactive with native erythropoietin.

The objects of the present invention are cloned DNA sequences of monkey and human origin and polypeptide sequences respectively derived from them that represent the primary structural conformation of erythropoietins of May Mun and human origin.

The present invention also provides novel biologically functional viral and circular plasmid DNA vectors including DNA sequences according to the invention and microbial (eg bacterial, yeast or mammalian cells) host microorganisms stably transformed or transfected with such vectors. Accordingly, the invention relates to novel methods for the preparation of useful polypeptides consisting in the cultivation of such transformed or transfected microbial hosts under conditions facilitating large-scale expression of exogenous vector-generated DNA sequences and isolation of the desired polypeptides from the culture medium , cell lysates or fractions of cell membranes.

The isolation and purification of the microbially produced polypeptides of the invention may be by conventional means, for example preparative chromatographic separation and immunological separation involving the production of a monoclonal and / or polyclonal antibody.

Thus, by elucidating the sequence of amino acid residues of erythropoietin, the present invention provides the complete or partial production of DNA sequences encoding erythropoietin, as well as the inclusion of codons preferred for expression by selected non-mammalian enzyme-derived host receptors and additional start, end, or intermediate DNA sequences that facilitate the construction of expression vectors. Accordingly, the present invention relates to the production (and development by site-specific mutagenesis of cDNA and genomic DNA) of DNA sequences encoding microbial expression of polypeptide analogs or erythropoietin derivatives that differ from their native forms by type or the arrangement of one or more amino acid residues (e.g., deletion analogues containing not all EPO-typical residues and / or substitutions analogues, wherein one or more of the EPO residues are replaced by others and / or analogs obtained by extension, wherein one or more amino acid residues are added to the terminal or middle chain of the polypeptide); with each of the analogues exhibiting one or all of the properties of natural erythroietin.

The novel DNA sequences according to the invention include all sequences that can provide expression in prokaryotic or eukaryotic host cells of polypeptide products having at least part of the primary structural conformation and one or more of the biological properties of erythropoietin that can be seen in (a) DNA sequences shown in Tables V and VI herein or their complementary strands; (c) DNA sequences that hybridize (under hybridization conditions as described herein or stricter) to the DNA sequences described in (a) or fragments thereof; and (c) DNA sequences that, because of genetic code degeneration, hybridize to the DNA sequences described in (a) and (c) above. Part of the (c) described are genomic DNA sequences encoding allelic variant forms of monkey or human erythropoietin and / or encoding other mammalian erythropoietin species. Part of (c) is DNA sequences encoding for EPO, EPO fragments, and EPO analogs, which may include codons facilitating the translation of information RNA into invertebrate hosts.

The present invention also includes a class of polypeptides for which encode portions of DNA complementary to the upstream human genomic DNA sequence of Table VI, i.e., complementary inverted protein, as described by Tramontano et al., Nucleic Acids Research, 12, p. 5049-5059 (1984).

Also included within the scope of the invention are pharmaceutical compositions containing effective amounts of the polypeptide products of the invention together with suitable diluents, additives and / or carriers for use in erythropoietin therapy, especially in the treatment of anemias, and in particular such anemias, which lead to chronic kidney failure.

The polypeptide products of the invention can be "labeled" by covalent binding to specific readily detectable substances - markers (eg, isotope labeling with ¹²⁵ l) to provide reagents used for the detection and quantification of erythropoietin in solid tissue and in liquid samples, such as such as blood and urine. DNA products according to the invention can also be labeled with detectable markers (isotope and non-isotope markers such as biotin) and used in DNA hybridization methods to determine the position of the erythropoietin gene and / or the position of any similar gene family in human, monkey and other mammalian species chromosome map.

They can also be used to detect defects in the erythropoietin gene at the DNA level and as gene markers to identify neighboring genes and their defects.

As described below in detail, the present invention also provides significant improvements in methods for detecting specific single-stranded polynucleotides of unknown sequence in a heterogeneous cell or viral sample comprising coupled single-stranded polynucleotides, where:

(a) preparing a mixture of labeled single-stranded polynucleotide probes with uniformly varying base sequences, each of said probes being potentially specifically complementary to one sequence of bases characteristic of the polynucleotides to be detected;

(c) the sample is fixed to a solid substrate;

(c) the substrate to which the sample is fixed is treated to reduce the possibility of other binding of polynucleotides to it other than by hybridization to polynucleotides in said sample;

(d) the treated substrate with the fixed sample sample is temporarily brought into contact with said mixture of labeled samples under conditions facilitating hybridization only between fully complementary polynucleotides, and (e) the specific polynucleotide is detected by monitoring for its hybridization reaction and a fully complementary probe in said mixture of labeled probes, which is indicated by the higher density of the labeled material on the substrate at the site of the specific polynucleotide compared to the background n atnost the labeled material, which is a result of nonspecific binding of labeled probes to the substrate.

The method is particularly effective in cases requiring the use of 64, 128, 256, 512 and 1024 or more mixed polynucleotide probes with a length of 17 to 20 bases of DNA / DNA, RNA / RNA or DNA / RNA hybridizations.

As described herein, the methods mentioned above are used to identify cDNA clones encoding erythropoietins originating from monkeys from a single bank derived from anemic monkey kidney cell mRNA. In particular, a mixture of 128 uniformly varying 20-dimensional probes based on amino acid sequence information obtained from sequencing human erythropoietin fractions is used for colony hybridization to detect seven positive erythropoietin cDNA clones between 200,000 colonies. Even the use of improved methods according to the invention has led to the rapid isolation of three positive clones by screening 1,500,000 phage plaques constituting the human genomic bank. The process is completed using the described mixture of 128 20-dimensional probes together with a second set of 128 17-dimensional probes based on amino acid analysis of various continuous sequences of human erythropoietin.

The methods described represent the first known case of using15 unrelated mixed oligonucleotide probes in DNA / DNA hybridization methods to isolate genomic clones of mammals and the first known case of using a mixture of more than 32 oligonucleotide probes in isolating cDNA clones.

A number of aspects and advantages of the invention will become apparent to those skilled in the art upon consideration of the following detailed description that illustrates the currently preferred applications of the invention.

According to the present invention, DNA sequences encoding part or all of the human and monkey erythropoietin polypeptide sequence (hereinafter referred to as EPO) are isolated and characterized. Further, monkeys of human and human origin are subject to eukaryotic and prokaryotic expression, providing isolated amounts of polypeptides exhibiting the biological (e.g. immunological) properties of natural EPO, as well as the in vivo and in vitro biological activities of EPO.

Monkey-derived DNA is isolated from an mRNA construct constructed from mRNA derived from the kidney tissue of a monkey with chemically-induced anemia and whose serum has been found immunologically to have a high EPO content compared to normal monkey serum. Isolation of desired cDNA clones containing DNA encoding for EPO is accompanied by DNA / DNA colony hybridization using a pool of 128 mixed, isotopically labeled, 20-dimensional oligonucleotide probes, and involves rapid screening of 200,000 colonies. The preparation of the oligonucleotide probes is based on the amino acid sequence information obtained by enzymatic fragmentation and sequencing of a small sample of human EPO.

Human DNA is isolated from a human genomic DNA bank. The isolation of clones containing the coding for EPO DNA was performed by DNA / DNA hybridization of plaques using the above pool of 128 mixed isotope-labeled 17-mer probes, whose sequences were based on amino acid sequence information derived from a different enzyme fragment. human EPO.

Positive colonies and plaques were verified by dideoxy sequencing of clonal DNA using one subset of 16 pool sequences of 20-dimensional probes, and selected clones were subjected to nucleotide sequence analysis to determine the primary structural conformation of EPO polypeptides encoded by them. The particular polypeptide sequences exhibit a high degree of homology to each other and to a partial sequence obtained by amino acid analysis of human EPO fragments.

One selected positive monkey cDNA clone and one selected positive human genomic clone are included in a DNA vector that is amplified in E. coli and used to transfect a culture of mammalian cells. Cultivation of transfected host cells results in an EPO content of the current culture fluid up to 3000 mU per ml of culture fluid.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a graphical representation of a radioimmunoassay of products16 according to the invention;

Figures 2 to 4 show the construction of vectors according to the invention and

Figures 5 to 21c DNA and polypeptide sequences according to the invention.

The following examples are provided to illustrate the invention and relate specifically to the processes preceding the identification of EPO coding for monkey cDNA clones and human genomic clones, the processes leading to such identification and sequencing, the development of expression systems and immunoassay. in these systems.

In particular, Example 1 relates to the amino acid sequencing of human EPO fragments and the preparation of mixtures of "isotope" labeled probes based on the results of this sequencing. Example 2 relates generally to methods for identifying positive monkey cDNA clones and thus provides information relating to animal treatment and pre-radioimmunoassay (RIA) for analysis of animal sera. Example 3 relates to the preparation of a cDNA bank, hybridization screening of colonies and verification of positive clones, sequencing of positive cDNA clones, and providing information on the primary structural conformation (amino acid sequence) of monkey EPO. Example 4 relates to processes involved in identifying positive human genomic clones and thus providing genomic bank source information, plaque hybridization methods, and verification of positive clones. Example 5 relates to DNA sequencing of a positive genomic clone and obtaining information about the human EPO polypeptide amino acid sequence, including comparison with the sequence information of monkey EPO. Example 6 relates to methods for constructing a vector comprising DNA encoding for EPO derived from a positive monkey cDNA clone, using the vector to transfect COS-1 cells and culturing the transfected cells. Example 7 relates to methods for constructing a vector comprising EPO-encoding DNAs derived from a positive human genomic clone, using the vector to transfect COS-1 cells and culturing the transfected cells. Example 8 describes an immunoassay of the present fluids of the culture media used for culturing the cells transfected according to Examples 6 and 7. Example 9 describes the in vivo and in vitro biological activity of EPO obtained by microbial expression according to Examples 6 and 7.

Example 10 relates to the creation of mammalian EPO and human genomic DNA cDNA mammalian host expression systems, including Chinese hamster ovary cells ("WMO") and to the immunological and biological activities of these expression systems, as well as the characterization of such products. Example 11 relates to the preparation of genes encoding human EPO and EPO analogs, which genes include a number of preferred columns for expression in E. coli and yeast host cells and to expression systems based thereon. Example 12 deals with profiles of the immunological and biological activity of the expression products of the system of Example 11.

Example 1

A. Amino acid sequencing of 5 fragments of human EPO

Human EPO was isolated from urine and subjected to triple digestion, resulting in the production and isolation of 17 individual fragments in amounts of approximately 100-150 pmol.

Table 1

Fragment N Result of sequence analysis T4a Т4Ь T9 T13 T16 T18 T21 T25 T26a Т26Ь T27 SCI T31 TKZ T35 T38 A-P-P-R G-K-L-K A-L-G-A-Q-K V-L-E-R A-V-S-G-L-R L-F-R K-L-F-R Y-L-L-E-A-K L-I-C-D-S-R L-Y-T-G-E-A-C-R E-A-I-S-P-P-O-A-A-M-A-A-P-L-R E-A-E-X-I-T-T-G-X-A-E-H-X-S-L-N-E-X-I-T-V-P V-Y-S-N-F-L-R S-L-T-T-L-L-R V-N-F-Y-A-W-K G-Q-A-L-L-V-X-S-S-Q-P-W-E-P-L-Q-L-H-V-D-E

The fragments are numbered arbitrarily and their amino acid sequences analyzed by microsequencing analysis using a gas-phase sequencer (Applied Biosystems) to obtain the sequence information used in Table 1, using one-letter codes, and X denotes a residue that is not uniquely defined.

C. Preparation and preparation of mixtures of oligonucleotide probes The amino acid sequences listed in Table I are considered in the context of genetic code degeneration to determine whether methods using mixed probes are applied to DNA / DNA hybridization processes on cDNA and / or GE30 Nominal DNA Bank. This analysis shows that in the ΝT35 fragment there is a series of 7 amino acid residues (Vai Asn-Phe-Tyr-AlaTrp-Lys) that can be uniquely identified as coding for one of the 128 possible DNA sequences spanning 20 basic couples. Therefore, a first set of 128 20-mer40 oligonucleotides was synthesized by standard phosphoamidite methods (See Beaucage et al., Tetrahedron Letters, 22, pp. 1859-1862 (1981)) on a solid support according to the sequence shown in Table 2 below.

Table 2

The rest Vai Asn Phe Tyr Ala Trp Lys 3 ' CAA TTG AAG ATG CGA ACC TT 5 ' Τ A A A T Mr Mr C C

Subsequent analysis shows that a pool of 128 can be prepared in a fragment of T38 having a series of 6 mixed 17-dimensional oligonucleotide amino acid residues (Gin-Pro-probes listed in Table 3 below. Trp-Glu-Pro- Leu), based on Table 3

The rest Gin Pro Trp Glu Pro Leu 3 ' GTT GGA ACC CTT GGA GA 5 ' C T C T A Mr Mr C C

Oligonucleotide probes were labeled at the 5 'end with gamma - ³² P-ATP 75008000 ci / mmol (ICN) using T ₄ polynucleotide kinase (NEN).

Example 2

A. Methods for Monkey Treatment Female Cynomolgus monkeys

Macaca fascicularias (2, 5 - 3 kg, 1.5 to 2 years of age) are treated subcutaneously with a solution of phenylhydrazine hydrochloride with a pH of 7.0 at doses of

12.5 mg / kg on days 1, 3 and 5. The hematocrit value is monitored before each injection. On the seventh day or when the hematocrit value fell below 25% of baseline, serum and kidney were taken after administration of ketamine hydrochloride at 25 mg / kg. The collected materials were immediately frozen in liquid nitrogen and stored at -70 ° C.

B. RIA for EPO (radioimmunoassay - RIA)

Radioimmunoassay assays are performed to quantify EPO in samples as follows:

Standard erythropoietin or unknown sample was incubated with 30 antisera for 2 h at 37 ° C. After two hours of incubation, the tubes were cooled on ice, ¹²⁵ 1-labeled erythropoietin was added, and the tubes were incubated at 0 ° C for at least another 15 hours. Each tube contained a 500 μ смес incubation mixture consisting of 50 μΐ diluted immune serum, 10,000 ²⁵ 1-erythropoietin, 5 μΐ trazilol and 0-250 μΐ standard EPO or unknown sample in PBS (phosphate buffered saline) containing 0.1% BSA to supplement the remaining volume. The antiserum used is from the second rabbit blood test immunized with 1% pure human urine erythropoietin. The final dilution of antiserum an19 study was adjusted so that the amount of bound antibody ^{| 25} 1-EPO does not exceed 1020 ° _o of the total. In fact, this corresponds to a final dilution of the antisera from 1:50 000 to 1: 100 000.

The antibody-bound ¹²⁵ 1-erythropoietin is precipitated by the addition of 150 μΐ Staph A. After 40 min incubation, the samples are centrifuged and the precipitate is washed twice with 0.75 ml 10 mM Tris hydrochloric acid at pH

8.2. containing 0.15M sodium chloride, 2 mM EDTA and 0.05% Triton X100. The washed sludge was examined with a gamma counter to determine the percentage of bound ¹²⁵ l of erythropoietin. Samples associated with preimmune sera are subtracted from the final value to correct the number of nonspecific precipitates. The content of erythropoietin in unknown samples was determined by comparison with the standard curve.

The method described above is applicable to the study of monkey serum prepared as described in Part A above as well as untreated monkey serum. Normal serum samples have been found to contain approximately 36 mU / ml, whereas treated monkey serum contains from 1000 to 1700 mU / ml.

Example 3

A. Obtaining a Monkey cDNA Bank

Information RNA was isolated from the kidneys of normal and anemic monkeys by the method of guanidine thiocyanate according to Chirgwin et al., Biochemistry, 18, p. 5294 (1979) and poly (A) ⁺ mRNA was purified by two passes by oligo column chromatography ( dT) cellulose, as described on page 197198 by Maniatis et al., Molecular Cloning, a Laboratory Manual, NY, (1982). The cDNA bank was constructed according to a modification of the general method of Okayama et al., Mol. and Cell Bid 2. pp. 161-170 (1982). The key features of the preferred methods now are: (1) pUC8 is used as a single vector cut with PstI and then affixed with oligos-dT with a length of 60-80 bases; (2) Hindi digestion is used to remove the oligo dT tail from one end of the vector; (3) The synthesis of the first strand and oligo dG end is carried out according to the method described; (4) BamHI digestion is used to remove oligo dG from the end of the vector, (5) Replacement of the RNA strand with DNA is performed in the presence of two linkers (GATCTAAAGACCGTCCCCCCCCC and ACGGTCTTTA) with a triple molar excess of oligo dG terminal vector.

B. Methods for hybridization of monkey cDNA bank screening colonies.

Transformed E. coli were spread at a density of 9000 colonies per 10x10 cm dish on solid medium containing 50 pg / ml ampicillin. Gene Screen Filters (New England Nuclear Catalog N EF-972) pre-wetted on BHI-CAM plates (as brain-infusion 37 g / Ι, casamic acid 2 g / Ι and agar 15 g / Ι containing 500 pg / ml chloramphenicol) and used to separate colonies from the medium. Colonies were cultured in the same medium for 12 hours or longer to multiply the number of plasmid copies. Multiplied colonies were treated by placing a series of filters on 2 pieces of Whatman 3 MM filter paper saturated with one of the following solutions:

(1) 50 mM glucose - 25 mM trisolic acid (pH 8) -10 mM EDTA (pH 8.0) for 5 min;

(2) 0.5 M sodium hydroxide for 10 min; and (3) 1.0 M Tris-hydrochloric acid (pH 7.5) for 3 min.

The filters were then air-dried in vacuo at 80 ° C for 2 h.

The dried filters are digested with proteinase K by treatment with a solution containing 50 pg / ml of protease enzyme in buffer K (0.1 M Trisolic Acid (pH 8.0) - 0.15 M Sodium Chloride - 10 mM EDTA (pH 8.2) - 0.2% SDS). In particular, 5 ml of the solution is added to each filter and grinding is carried out at 55 ° C for 30 min, after which the solution is removed.

The filters were then treated with 4 ml of hybridization buffer (5xSSPE - 0.5% SDS - 100 pg / ml SS E. coli DNA - 5xBFP). The pre-hybridization treatment is carried out at 55 ° C, usually for 4 hours or longer, after which the pre-hybridization buffer is removed.

The hybridization is as follows. To each filter was added 3 ml of hybridization buffer (5xSSPE - 0.5% SDS - 100pg / ml yeast TRNA) containing 0.025 pmol from each of the 128 sample sequences of Table 2 (total mixture was referred to as EPV mixture) and the filters are kept at 48 ° C for 20 h. This temperature is 2 ° C lower than the lowest of the calculated dissociation temperatures (Td) determined for each sample.

After hybridization, the filters were washed three times for 10 min on a shaker with 6xSSC -0.1% SDS at room temperature and washed two to three times with 6xSS C-1% SDS at hybridization temperature (48 ° C).

The autograph of the filters revealed seven positive branches among the 200,000 colonies screened.

The initial sequence analysis of one of the putative positive monkey cDNA clones (designated clone 83) was performed to confirm the results by modifying the method of Wallace et al., Gene, pp. 21-26 (1981). Briefly, plasmid DNA from monkey cDNA clone 83 is linearized by digestion with EcoRi and denatured by heating in a boiling water bath. The nucleotide sequence was determined by the dideoxy method of Sanger et al., P.N.A.S. (USA) 74, pp. 5463-5467 (1977). A secondary set of EPV sample mixtures consisting of 16 sequences was used as a primer for sequencing reactions.

C. Monkey EPO cDNA sequencing

Analysis of the nucleotide sequence of clone 83 was performed by the method of Messing, Methods in Enzymology, 101, pp. 20-78 (1983). Table 4 shows a preliminary analysis of the restriction maps of EcoRI / HindIII, a cloned fragment of clone 83 consisting of approximately 1600 base pairs. The approximate locations of the restriction endonuclease enzyme sites are indicated as the bases are counted from the 3 'end to the EcoRi site at the 5' end of the fragment. Nucleotide sequencing is performed by sequencing individual restriction fragments with the binding of overlapping fragments. For example, sequence overlap information was obtained by nucleotide analysis in a restriction fragment designated CH3 (Sau3A at ~ 111 / SmaI at ~ 324) and by sequencing in the reverse order of one fragment designated C73 (Alul at ~ 424 / BstEII at = 203).

Table 4

A restriction enzyme recognition site Approximate location EcoRl 1 Sau3A 111 Smal 180 BstEIl 203 Smal 324 Kpnl 371 Rsal 372 Alul 424 PstI 426 Alul 430 Hpal 466 Alul 546 PstI 601 PvuII 604 Alul 605 Alul 782 Alul 788 Rsal 792 PstI 807 Alul 841 Alul 927 Ncol 946 Sau3A 1014 Alul 1072 Alul 1115 Alul 1223 PstI 1301 Rsal 1343 Alul 1384 Hindlll 1449 Alul 1450 Hindlll 1585

Sequencing of 1342 base pairs (in a region spanning the Sau 3A site at 3 'to the EcoR1 site and the HindIII site) and analysis of all possible reading frames provide the DNA and amino acid sequence information shown in FIG. 5. In the figure, the putative starting amino acid gap45 from the amino terminus of the mature EPO chain (which is confirmed by correlation of the previously mentioned sequence analysis of the twenty amino acid residues) is indicated as +1. The pressure from the methionine-defining ATG codon (designated -27) 'up' from the starting amino-terminal alanine residue as determined for the first remaining amino acid sequence of the mature protein is indicative of the possibility of EPO expression initially in the cytoplasm in a precursor form comprising 27 amino acids, which is excised prior to entry of mature EPO into the bloodstream. Potential glycolysis sites in the polypeptide are indicated by an asterisk. The molecular weight of the translated region is 21 117 daltons, and the molecular weight of the 165 residues constituting the mature monkey EPO is 18 236 daltons.

The polypeptide sequence of FIG. 5 can be readily assayed for the presence of highly hydrophilic regions and / or secondary conformational characteristics indicative of potentially highly immunogenic sites by, for example, the methods of Hopp et al., P.N.A.S. (USA) 78, pp. 3824-3828 (1981) and Kyte et al., J. Mol. Biol. 157, p. 105132 (1982) and / or Chou et al., Biochem. 13, pp. 222-245 (1974) and Adrances in Enzymology, 47, pp. 45-47 (1978). Computer analysis by the method of Hoop et al. Is performed using a program referred to as PEP Reference Section 6.7 provided by Intelligenetics, Inc. 124 University Avenue Polo Alto, California.

Example 4

A. Human Genomic Bank

A Ch4A phage genomic bank of human embryonic liver prepared according to the method of Lawn et al., Cell (see above) was obtained and maintained for the purpose of hybridization plaque assay.

C. Plaque hybridization methods used for human genomic bank screening

Phage particles were lysed and DNA fixed to filters (50,000 filter plates) according to the Woo method, Methods in Enzymology, 68, pp. 389-395 (1979) except using Gene Screen Plus filters (New England Nuclear Catalog N NEF-972) and NZYAM plates (NaCI, 5 g; MgCI ₂ , 6H ₂ O, 2 g; NZ-amine A, 10 g; yeast extract 3 g; casamic acids 2g; maltose 2 g; and arap 15 g / l ).

The air-dried filters were washed for 1 h at 80 ° C and then ground with Protease K as described in Example 3, Part B. Pre-hybridization was performed with 1 mole of sodium chloride - 1% SDS buffer at 55 ° C for 4 h or more, after which the buffer is removed. Hybridization and post-hybridization washing were performed as described in Example 3. part B. Mixtures of one of the 128 20-dimensional samples referred to as EPV and the other of the 128 17-dimensional samples of Table 3 (labeled EPQ mixtures) were used. Hybridization was performed at 48 ° C using the EPV mixture. Hybridization of the EPQ mixture was carried out at 46 ° C - 4 ° below the calculated Td for the constituents of the mixture. Removal of the hybridized re-hybridization sample was performed by boiling with 1 x SSC -0.1% SDS for 2 min. The autoradiography of the filters showed three positive clones (reactive with both sample mixtures) between 1,500,000 screened phage plates. Confirmation of positive clones as EPO encoders was performed by DNA sequencing and electron micrographs of the mono-cDNA formulated heteroduplexes of Example 3. This procedure also reveals numerous introns in the genomic DNA sequence.

Example 5

The nucleotide sequence of one of the positive clones (designated λλ ΗΕΙ) was analyzed and the results obtained are shown in FIG. 6.

In FIG. 6, the original continuous DNA sequence represents an upstream chain of 620 bases from one untranslated sequence immediately preceding one translated portion of the human EPO gene. In particular, the sequence appears to constitute the 5 'end of the gene, which results in a translated strand of DNA encoding the first four amino acids (-27 to -24) of one leading sequence (sequencing). The four pairs of bases in the sequence preceding that encoding the leading of the leading are not yet uniquely defined and are therefore marked with an "X". Following is an intron of about 639 base pairs (439 base pairs are sequenced and the remaining 200 base properties are designated "I.S."), and immediately preceding a glutamine codon, which is designated as residue-23 of the translated polypeptide. The next exon sequence was found to code for the amino acid residues of one alanine residue (designated as + 1 residue from the mature human EPO amino acid sequence) to the codon for threonine at position +26, followed by a second intron consisting of 256 bases as specifically indicated. Following this intron, there is one exon sequence for amino acid residues 27 to 55 and then a third intron containing 612 base pairs. The subsequent exon encodes for residues 56 to 115 of human EPO and thereafter begins a fourth intron with 134 bases as described. There is one exon after the fourth intron. coding for residues NN 116 to 166 and one stop codon (TGA). Finally, Table 6 discloses a sequence of 568 base pairs that suggests that there is an untranslated 3 'region of the human EPO gene, two base pairs to which ("X") are not yet uniquely sequenced.

Thus in FIG. 6 identifies the primary structural conformation (amino acid sequence) of mature human EPO, comprising 166 specific amino acid residues (established molecular weight = 18 399). The figure also shows the DNA sequence encoding for a lead sequence of 27 residues together with the 5 'and 3' DNA sequences that is relevant for the promoter / operator functions of the human gene operon. The sites for potential glycolysis of the mature human EPO polypeptide are indicated in the asterisk figure. It is worth noting that the specific amino acid sequence of FIG. 6 is probably that of a naturally occurring allelic form of human erythropoietin. This is confirmed by the results of long-term sequencing experiments on urine-derived human erythropoietin isolates, and found that a significant number of erythropoietin molecules have methionine at residue 126 rather than serine, as shown in the figure.

Figure 9 shows the degree of homology between the human and monkey EPO polypeptide sequences. In the upper unbroken row of the figure, the single translated letters represent the derived translated polypeptide sequences of human EPO beginning with residue -27, and the lower unbroken row shows the deduced polypeptide sequence of monkey EPO beginning with designated residue -27. Asterisks are used to indicate sequence homology. It should be noted that the deduced human and monkey EPO sequences reveal a single "additional" lysine (K) residue at position 116 of the human EPO. The notes to FIG. 6 indicate that this residue is on the verge of putative mRNA binding in the genomic sequence. The presence of a lysine residue in the human polypeptide sequence was then confirmed by sequencing a human cDNA clone cDNA clone isolated from COS-1 cells transformed with the human genomic DNA of Example 7 below.

Example 6

The selected expression system for the initial microbial synthesis assays of isolated amounts of EPO polypeptide material encoded by monkey cDNA obtained by the method of Example 3 is one comprising mammalian host cells (e.g., COS-1 cells, ATCC N CRL-1650). Cells were transfected with a single vector capable of autonomous replication in E. coli host thanks to DNA obtained from pBR322 and in mammalian hosts (via the presence of SV40 derived from DNA virus).

In particular, one expression vector is constructed as follows: Plasmid clone 83 obtained in Example 3 was amplified in E. coli and approximately 1.4 square monkey DNA encoding EPO was isolated by EcoRI and HindIII digestion. Approximately 40 square HindIII / Sall fragment from pBR322 was isolated. An approximately 30 bp EcoRI / Sall linker fragment was obtained from M13mpl0 RF DNA (P and L laboratories). This linker includes, in series, an EcoRI adhesive end followed by SstI, Smal, BamHMI and Xbal recognition sites and a Sall adhesive end. The above three fragments are ligated to produce approximately 5.4 square intermediate plasmid ("pERS"), where EPO DNA is attached on one side by a "bank" of useful restriction endonuclease recognition sites. The pERS was then digested with HindIII and Sall to obtain EPO DNA and EcoRI to a Sall (M13mpl0) linker. The 1.4 kb fragment was ligated with approximately 4.0 kv BamHl / Sall on pBR322 and another M13mpl0 Hindlll / BamHl RF linker fragment also from approximately 30 pp. The M13 linker fragment was characterized by a Hindlll adhesive end, followed by PstI, Sall, Xbal recognition sites and BamHI sticky end. The ligation product is again a useful intermediate plasmid (" pBRERO ") comprising EPO DNA attached from both sides by banks to a restriction site.

The selected vector for expression of EPO DNA in COS-1 cells ("pDSVL1") is designed to allow selection and autonomous replication in E. coli. These features are provided by the replication source and ampicillin gene-resistant cDNA sequences in the region extending from nucleotides 2448 to 4362 of pBR322. This sequence is modified structurally by the addition of a single linker axis25 that binds HindIII directly adjacent to nucleotide 2448 before incorporation into the vector. Among the other useful properties selected for the vector is its ability to autonomously replicate in COS-1 cells and the presence of a viral promoter sequence acting in mammalian cells. These properties are provided by the origin of the DNA replication sequence and the "late gene" viral promoter / DNA sequence present in a 342 bp sequence spanning nucleotides numbers 5171 to 270 of the SV40 genome. The vector has one unique restriction site (BamHI) next to it a viral promoter sequence provided by the use of a commercially available linker sequence (Collaborative Research). The vector also includes one sequence of 237 base pairs (derived from nucleotide numbers 2553 to 2770 of SV40) containing the viral mRNA polyadenylation signal of the "late gene" (commonly referred to as a transcription terminator). This fragment is positioned in the vector in the right direction just opposite the "late gene" viral promoter through the only BamHI site. There is also another mammalian gene in the vector in a position inaccessible to potential transcription of a gene included on a single BamHI site between the viral promoter and the terminator sequence. The mammalian gene is a murine dihydrofolate reductazone (DMFR) mini-gene of approximately 2,500 bp isolated from plasmid pMG-1 as in Gasser et al., P.N.A.S. (USA) 79, pp. 6522-6526 (1982). Again the major operational components of plasmid pDSVLI comprise nucleotides 2448 to 4362 of pBR322 together with nucleotides 5171 to

270 (342 bp) and 2553 to 2770 (237 bp) on SV40 DNA.

Following the method described by Maniatis et al., See above, the coding EPO DNA was isolated from plasmid pBR-EPO as a BamHI fragment and ligated into plasmid pDSVLI cut with BamHI. Restriction enzyme analysis was used to confirm the insertion of the EPO gene in the right direction in two of the cloned vectors obtained (pair vectors H and L). See FIG. 2, showing plasmid pDSVLI-MkE. Vectors with EPO genes in the wrong direction (vectors F, X, and G) are retained for use as negative controls in transfection experiments to determine the extent of EPO expression in hosts transformed with vectors containing EPO DNA in the right direction.

The vectors H, L, F, X and G are combined with DNA carrier (liver and spleen DNA) and used to transfect pairs of 60 mm plates by microprecipitation with calcium phosphate. Pairs of 60mm tiles were also transfected with carrier DNA as a negative control with "false" transformation. After five days, all nutrient media were tested for the presence of polypeptides having the immunological properties of natural EPO.

Example 7

A. Initial EPO Expression System Including COS-1 Cells The system selected for initial microbial synthesis assays of isolated amounts of a human EPO polypeptide encoded by a human genomic DNA EPO clone also includes expression in mammalian host cells (e.g., COS-1 cells, ATTC N CRL-1650). The human EPO gene is cloned primarily into a "shuttle vector capable of autonomous replication both in E. coli hosts ((due to the presence of DNA obtained from pBR322) and in the mammalian cell line COS-1 (due to the presence of SV40 Then the shuttle vector containing the EPO gene is transfected into COS-1 cells, the EPO polypeptide material is produced in the transfected cells and secreted into the culture media.

In more detail, the expression vector is constructed as follows: DNA isolated from a lambda clone XhEl containing the human genomic EPO gene is digested with BamH1 and HindIII restriction endonuclease and a 5.6kV DNA fragment is known to be known. that it contains the entire EPO gene. This fragment was mixed and ligated with the bacterial plasmid pUC8 (Bethesda Research Laboratories, Inc.), which was similarly digested, creating the intermediate plasmid "pUC8-HuE", providing a convenient source of this restriction fragment.

The vector selected for the expression of EPO DNA in COS-1 cells (pSV4SEt) was constructed earlier. The plasmid pSV4SEt contains DNA sequences allowing selection and autonomous replication in E. coli. These features are provided by the source of replication and DNA sequences of the ampicillin-resistant gene located in the region extending from nucleotide 2448 to 4362 of the bacterial plasmid pBR322. This sequence is structurally modified by the addition of a linker providing the HindiII recognition site directly to nucleotide 2448. The plasmid pSV4SEtV is also capable of autonomous replication in COS-1 cells. This property is due to a 342 bp fragment containing the SV40 viral replication source (nucleotide numbers 5171 to 270). This fragment is modified by adding a linker providing an EcoR1 recognition site adjacent to nucleotide 270 and a linker providing a Sall recognition site adjacent to nucleotide 5171. In this vector there is also a 1061 bp fragment from SV40 (nucleotide numbers 1711 to 2772 plus one linker that creates a Sall recognition site near nucleotide number 2772). In this fragment there is a unique BamH1 recognition sequence. In general, the plasmid pSV4SEt contains unique BamH1 and HindIII recognition sites allowing insertion of the human EPO gene, sequences allowing replication and selection in E. coli. and sequences allowing replication in COS-1 cells.

In order to incorporate the EPO gene into pSV4SEt, the plasmid pUC8-HuE was digested with the restriction endonucleases BamH1 and HindIII and 5.6k EPO encoding DNA was isolated. pSV4SEt was also digested with BamH1 and HindIII and the 2513 bp major fragment was isolated (retaining all necessary functions). These fragments were mixed and ligated to produce the final vector "pSVg NiEPO". (see Fig. 3). This vector is propagated in E. coli and a DNA vector is isolated. Restriction enzyme analysis is used to confirm the insertion of the EPO gene.

Plasmid pSVgHuEPO DNA is used to express human EPO polypeptide material in COS1 cells. In fact, pSVgHuEPO DNA is combined with carrier DNA and transfected into triple 60 mm dishes with COS-1 cells. As a control, carrying DNA alone, it is also transfected into COS-1 cells. Samples were taken from the cell culture medium five and seven days later and tested for the presence of polypeptides having immunological properties such as natural EPO.

B. A second EPO expression system comprising COS-1 cells

To provide improved production of human EPO polypeptide material encoded by the human genomic DNA EPO clone in COS-1 cells (ATTC N N CRL-1650) another system is used.

In the preceding system, the expression of EPO in COS-1 cells is carried out using its own promoter, which is in the 5.6k BamH1 to HindIII restriction fragment. In the following construction, the EPO gene is altered so that its expression is accomplished by the SV40 late promoter.

The cloned 5.6kB BamH1 to HindIII genomic human EPO restriction fragment was modified as follows: The plasmid pUC8HuE, as described above, was cleaved with the restriction endonuclease BamH1 and BstEII. BstEII cleaves the 5.6k EPO gene at a position that is at 44 base pairs 5 'from the starting ATG code for the polypeptide and at approximately 680 base pairs 3' from the HindIII restriction site. A fragment of approximately 4900 base pairs was isolated. A synthetic linker DNA fragment having Sall and BstEII adhesive ends and an internal BamH1 recognition site is synthesized and purified. The two fragments were mixed and ligated with plasmid pBR322, which was cut with Sall and BamH1 to produce the intermediate plasmid pBRgHE. The genomic human EPO gene can therefore be isolated as a single BamH1 digested fragment of 4900 base pairs carrying the entire structural gene with a single ATG located at 44 base pairs 3 'from the BamH1 site adjacent to the amino-terminal coding region. This fragment was isolated and included as a BamH1 fragment in the cleaved BamH1 expression vector plasmid pDSVLI (described in Example 6). The resulting plasmid, pSVLgHu EPO, as shown in FIG. 4 was used to express EPO polypeptide material from COS1 cells as described in Examples 6 and 7A.

Example 8

The nutrient media of the six transfected COS-1 cultures of Example 6 were analyzed by radioimmunoassay according to the method described previously in Example 2, Part B. Each sample was analyzed at aliquot levels of 250, 125, 50 and 25 μΐ. The present fluids obtained by culturing falsely transfected or transfected vectors in the wrong direction of the EPO gene gene are clearly negative for EPO immunoreactivity. For each sample of the two supernatants obtained from culturing COS-1 cells transfected with vectors (H and L) with EPO DNA in the right direction, the percentage inhibition of ¹²⁵ 1-EPO binding to antibody was 72 to 88%. which puts all the values obtained at the top of the standard curve. The exact concentration of EPO in the culture supernatant cannot be determined. A fairly constant value of 300 mU / ml was found, determined from the values of the largest aliquot (250 μΙ) calculated.

A representative sample of the culture fluid according to example 6 and the culture fluids of the fifth and seventh days according to example 7A were tested by RIA to compare the activity of recombinant monkey and human EPO products with that of the natural human EPO accepted as standard, and the results are presented in graphical form in FIG. 1. In summary, the results are expected to show that the recombinant monkey EPO remarkably blocks an antibody against the human EPO antibody, although it is not able to completely inhibit binding under the experimental conditions. However, the maximum inhibition rate for recombinant human EPO is close to that for standard human EPO. The parallelity of the dose response curves shows the immunological identity of the sequences (epitopes). The previous results of monkey EPO analysis in culture fluids were re-counted at higher dilution and values ranged from 2.91 to 3.12 U / ml. The established levels of human EPO production were 392 mU / ml respectively for the five-day culture sample and 567 mU / ml for the seven-day culture sample. The calculated levels of production of monkey EPO with the described expression system of Example 7B were of the same order or higher.

Example 9

The culture fluids obtained according to Examples 6 and 7 were subjected to in vitro testing for EPO activity by the method of Goldwasser et al., Endocrinology 97, 2, pp. 315-323 (1975). The determined values for monkey EPO in the tested culture fluids ranged from 3.2 to 4.3 U / ml. Cultural fluid with human EPO also showed in vitro in vitro EPO activity that could be neutralized by an anti-EPO antibody. Culture fluids from recombinant monkey EPO according to Example 6 were also subjected to in vivo biological activity assays according to the general methods of Cotes et al., Nature, 191, pp. 1065-1067 (1961) and Hammond et al., Anna !. N.Y. Acad. Sci., 149, pp. 516-527 (1968), and the levels of activity found are in the range of 0.94 to 1.24 U / ml.

Example 10

In the foregoing examples, recombinant monkey or human EPO material is produced from vectors used to transfect COS-1 cells. These vectors replicate in COS-1 cells due to the presence of SV40 T antigen in the cells and an SV40 source of replication on the vectors. Although these vectors produce usable amounts of EPO in COS-1 cells, expression is only transient (7 to 14 days) due to eventual loss of the vector. In addition, only a small percentage of COS-1 cells are productively transfected with vectors. This example describes an expression system using Chinese hamster ovary (CMO) DHFR cells and an selectable marker, DHFR. (For examples of similar expression systems, see US 4 399 216 and EP NN 117 058, 117 059 and 117 060, all published on 08/29/84.)

CHO DHFR cells (DuX-Bll) CHO K1 cells, Urlaub et al., Proc. Nat. Acad. Sci. (US) Volume 77, 4461 (1980) do not contain the dihydrofolate reductase (DHFR) enzyme due to mutations in structural genes and therefore require nutrient media containing glycine, hypoxanthine and thymidine. Plasmids pDSVL-MkE (Example 6) or pDSVL-gHuEPO (Example 7B) were transfected with carrier DNA in CHO DHFR cells cultured in media containing hypoxanthine, thymidine and glycine in 60 mm dishes. Plasmid pSVgHuEPO (Example 7A) was mixed with plasmid pMG2 containing a murine dihydrofolate reductase gene cloned into the bacterial plasmid vector pBR322 (according to Gasser see above). The plasmid mixture and the carrier DNA were transfected into CHO DHFR cells. (Cells that acquire one plasmid usually acquire a second plasmid as well). After three days, the cells were distributed by trypsin treatment in several 100 mm dishes in an environment lacking hypoxanthin and thymidine. Only those cells that are stably transformed with the DHFR gene and therefore the EPO gene can survive in these environments. After 7 to 21 days, the colonies of the surviving cells become visible. These transformant colonies after dispersion by trypsinization can be continuously propagated in media lacking hypoxanthin and tryptidine, creating new cell strains (e.g. CHO pDSVL-MkEPO, CHO-pSVgHuEPO, CHO-pDSVL-gHuEPO). The culture fluids from the above cell strains are tested in the RIA for the presence of recombinant monkey or human EPO. CHO-pDSVLMkEPO strain media contain EPOs with immunological properties similar to those obtained with COS-1 cells transfected with plasmid pDSVL-MkEPO. A representative 65-hour culture fluid contains monkey EPO of 0.6U / ml. Culture fluids from CHO pSVgHuEPO and CHO pDSVL-gHuEPO contain recombinant human EPO with immunological properties similar to those reported in COS-1 cells transfected with pSVgHuEPO or pDSVL-gHuEPO. One representative 3-day CHO culture fluid pSVgHuEPO contains 2.99 U / ml human EPO, and a 5-day CHO culture sample pDSVL-gHuEPO contains 18.2 U / ml human EPO measured by R1A.

The amount of EPO produced by the cell strains described can be increased by gene multiplication, yielding new cell strains with greater productivity. The enzyme dehydropholate reductase (DHFR), a product for which the DHFR gene encodes, can be inhibited by the drug methotrexate (MTX). In particular, cells propagated in environments lacking hypoxanthin and thymidine are inhibited or killed by MTX. Under suitable conditions (eg minimal concentration of MTX), cells resistant to MTX can be obtained and suitable for growth in the presence of MTX. These cells have been found to be resistant to MTX due to an increase in the number of their DHFR genes, which is reflected in increased production of the DHFR enzyme. The surviving cells can then be treated with increasing concentrations of MTX, resulting in cell strains containing a greater number of DHFR genes. "Passenger genes" (eg EPO) that fell into the expression vector together with the DHFR gene or transformed with the DHFR gene also increased their copy gene genius, as found.

As an example of application of this multiplication system, cell strain CMO pDSVL-MkE is subjected to increasing concentrations of MTX (0 nmol, 30 nmol and 100 nmol). Representative samples of the 65-hour culture medium from each multiplication step were examined by RIA and found to contain 0.60, respectively; 2.45 and

6.10 U / ml EPO. Cell CHO pMSVL-gHuEPO sham was subjected to a series of treatments with increasing MTX con30 concentrations of 30 nmol, 50 nmol, 100 nmol, 200 nmol, Igmol and 5 gmol MTX, respectively. Representative sample from the culture medium of a 3-day culture; pa of the 100 nmol MTX step contained human EPO 3098 ± 129 V / ml as judged by RIA. Representative samples of the 48-hour culture medium from steps 100 nmol and 1 gmol MTX contain human 1 EPO of 466 and 1352 U / ml, respectively, as assessed by R1A (the result is averaged over a three-fold study). In these processes, 1 x 10 ⁶ cells are spread in 5 ml of medium in 60 mm dishes. 1 24 h later, the media were removed and replaced with 5 ml serum-free medium (high glucose DMEM supplemented with 0.1 mmol non-essential amino acids and L-glutamine). Wait for EPO to accumulate for 48 h in serum-free medium. The media were collected for RIA assay and cells were trypsinized and enumerated. The mean R1A values of 467 2 U / ml and 1352 U / ml for cells cultured at 100 nmol and Igmol MTX, respectively, provide actual yields of 2335 u / plate and 6750 u / plate. The average cell count of blue-3 to 1.94 x 10 ⁶ and 3.12 x 10 ^{6, respectively} . The effective production rates under these culture conditions are 1264 and 2167 and / 10 ⁶ cells / 48 h, respectively.

The cells in the cultures described are 3 genetically heterogeneous populations. Standardized screening methods are used to isolate genetically heterogeneous clones with the highest production capacity. This is described in Chapter 4, Part 2, "Highlights of the Characteristics of Cell Lines Used for Organic Products" (1984), Bureau of Biological Research, Center for Medicines and Biological Products, US, Food and Drug Administration medicines.

The productivity of the CHO cell lines producing EPO described above can be improved by suitable cell culture methods. The multiplication of mammalian cells in culture usually requires the presence of serum in the culture media. One method of producing erythropoietin from CHO cells in serum-free media greatly facilitates the purification of erythropoietin from the culture medium. The method described below relates to the economical production of large quantities of erythropoietin sufficient for production in serum-free media.

Strain CHO pDSVL-gHK EPO cells cultured under standard cell culture conditions are used for propagation in round-bottom flasks for cell culture development. Cells were propagated as suspended cell lines in fertilized flasks in media containing a mixture of high glucose DMEM and F12 Ham containing 5% 50 50 L-glutamine, penicillin and streptomycin serum, 0.05 mmol 0 non-essential amino acids and appropriate concentration of methotrexate. The suspension cell culture allows the CHO-producing cells of EPO to be increased to large 5 volumes. CHO cells cultured in suspension are used for seedlings in round-bottomed flasks at an initial seeding density of 1.5 x 10 ⁷ viable cells for a round-bottomed flask of 850 0 ml in 200 ml of medium. The cells were cultured until confluence as a glued cell line over a three-day period. The traces used in this cultivation phase are the same as those used for suspension cultivation. At the end of the three-day culture period, serum-containing media were removed and replaced with 100 ml serum-free media; supplemented with a 50:50 mixture of high glucose DMEM and F12 of Ham supplemented with 0.05 mmol non-essential amino acids and L-glutamine. The round-bottomed flasks were returned to the incubator for a period of 1 to 3 hours, after which the medium was again removed and replaced with 100 ml of fresh serum-free medium. Incubation for 1 to 3 hours of serum free media reduces the concentration of contaminating serum proteins. The round-bottomed flasks were returned to the incubator for seven days, during which serum free erythropoietin was accumulated in serum-free media. At the end of the seven-day production phase, the conditioned medium is removed and replaced with fresh serum-free medium for a second production cycle. An example of the effect of this production system is a representative sample of a seven-day serum-free medium containing human erythropoietin in an amount of 3892 + 409 U / ml, as assessed by RIA. Based on a certain cellular density of 0.9 to 1.8 x 10 ⁵ cells / cm ³ and that each 850 cm ³ flask contains from 0.75 to 1.5 x 10 ⁸ cells, the rate of EPO production in 7 days 100 ml culture is 750 to 1450 U / 10 ⁶ cells / 48 hours.

Cell fluids from the CMO pDSVL-MkEPO cell strain obtained in 10 nmol MTX were subjected to RIA in vitro and in vivo assays for EPO activity. Conditioned media samples contained 41.2 ± 1.4 U / ml MCPO as determined by RIA; 41,2 ± 0,064 U / ml, measured by in vitro biological activity assay and

42.5 ± 5 U / ml, measured by in vivo biological activity. The amino acid sequencing of polypeptide products indicates the presence of

EPO products, usually the species having three residues of the leader sequence. adjacent to the putative aminocranium alanine. Whether this is the result of abnormal membrane separation of the polypeptide in CHO cells or reflects a difference in the amino terminus of monkey EPO compared to human EPO is still unknown.

Cell fluids from CHO cell strain pDSVL-gHuEPO were subjected to the three studies described. A 5.5-day liquid sample contains recombinant human EPO in the medium in an amount of 18.2 U / ml according to RIA analysis, 15.8 ± 4.6 U / ml according to in vitro assay and 16.8 ± 3.0 U / ml found in the in vivo study.

Culture fluid from CHO pDSVL-gHuEPO cells amplified in the described phase with the addition of 100 nmol MTX was subjected to the same three assays. A three-day sample contained recombinant human EPO in the amount of 3089 ± 129 U / ml according to the RIA, 2589 ± 71,5U / ml found in vitro and 2040 ± 160 U / ml in vivo. The amino acid sequencing of the products obtained shows one amino terminus corresponding to the one shown in FIG. 6.

Conditioned culture media from CHO cells transfected with plasmid pDSVL-MkEPO at 10 nmol MTX were pooled and MTX dialyzed over several days, resulting in media with an EPO activity of 221 ± 5.0 U / ml (EPO-CCM). To determine in vivo the effect of EPO-CCM on the hematocrit levels of normal Bam / C mice, the following experiment was performed. Conditioned cell media from untransfected CHO cells (CCM) and EPO-CCM were calibrated with PBS (phosphate buffered saline). CCMs were used in the control group (3 mice) and two doses of EPO-CCMs - 4 units per injection and 44 units per injection - in the experimental groups (2 mice / group). Seven mice were injected intraperitoneally 3 times a week for 5 weeks. After the eighth injection, mean hematocrit values were determined, which for the control group were 50.4%, for the 4U group 55.1% and for the group 44 Ό - 67.9%.

Mammalian cell expression products can be readily restored to substantially pure nutrient form by HPLC (C ₄ ) with an ethanol gradient, preferably at pH 7.

A preliminary attempt is made to characterize recombinant glycoprotein products from conditioned medium from human EPO gene expression in COS-1 and CHO cells by comparing with human urine EPO isolate by Western blot analysis and SDS-PAGE. These studies indicate that the CHO-produced EPO material has a low molecular weight higher than the COS-1 expression product, which in turn is slightly larger than the human urine extract used as a comparison. All products are somewhat heterogeneous. Treatment with a neuroamidase enzyme for sialic acid removal results in the production of COS-1 and CHO recombinant products of approximately the same molecular weight, which are both larger than the resultant asial human urine extract. The endoglucosidase F enzyme (EC 3.2.1) is used to treat recombinant CHO product and urine extract (to completely remove carbohydrates from both), resulting in substantially homogeneous products that have substantially identical molecular weight.

Purified human urine EPO and recombinant CHO-produced cells, EPO according to the invention are subjected to carbohydrate analysis by the method of Ledeen et al., Methods in Enzymology, 83 (Part E), 139-191 (1982), which is modified by use of the hydrolysis method of Nesser et al., Anal. Biochem., 142, 5867 (1984). The experimentally determined amounts of carbohydrates in the composition (expressed as molar amounts of carbohydrates in the product) for urine isolate are as follows: hexoses 1.73, N-acetylglucosamine-1, N-acetylneuraminic acid - 0.93, fucose-O and N-acetylgalactosamine - Oh. The corresponding values for the recombinant product (obtained from CHO pDSVL-gHuEPO from a 3-day culture medium containing MTX) are as follows: hexose - 15.09, N-acetylglucosamine-1, N-acetylneuraminic acid-0, fucose-0 and N -acetylgalactosamine-O. These results are consistent with those obtained with Western blot and SDSPAGE analyzes described above.

Glucoproteins prepared according to the present invention are products with a primary structural conformation sufficiently consistent with that of natural erythropoietin to allow one or more of the biological properties of the same and medium carbohydrate composition to differ from that of natural erythropoietin.

Example 11

This example relates to the general preparation by joining nucleotide bases of two structural genes encoding the human EPO sequence of FIG. 6 and including respectively "preferred" codons for expression in E. coli and yeast (S. cerevisiae) cells. The preparation of genes encoding for analogues of human EPO is also described. In short, the method used is the same as that described in the aforementioned statement by Alton et al., (WO83 / 04053). The genes are intended to initially assemble oligonucleotides into constituent duplexes, which in turn are joined in three separate parts. These portions are intended for multiplication and, upon removal from the multiplication system, can be joined in series or by multiple ligation of fragments into a suitable expression vector.

Figures 10 to 15 and 7 show the construction and binding of the resulting gene encoding a human EPO translation product that does not contain any leading sequence but includes one initial methionine residue at position -1. In addition, the gene includes an essential part of the E. coli preferred codons and is therefore referred to as the "ESERO" gene.

In particular, FIG. 10 shows oligonucleotides used to produce Section 1 of the ESEPO gene encoding amino acid residues of human polypeptides. The oligonucleotides are bound in duplexes (1 and 2, 3 and 4, etc.), which are then ligated to give ESEPO section 1, shown in FIG. 11. It should be noted that the resulting section includes the respective EcoRI and BamHI adhesive ends, with an Xbal restriction enzyme recognition site under the EcoRI adhesive end and a ΚρηΙ recognition site above the BamHI adhesive end. Section 1 could easily be multiplied using the M13 phage vector used to confirm the sequence of the section. Some difficulties have been reported in isolating the section, such as the Xbal / Kpnl fragment from RFAHK formed in E.coli, due to methylation of the bases forming the ΚρηΙ site in the host. Therefore, single-stranded phage DNA is isolated and converted to two-stranded form in vitro by primer extension, and the desired two-stranded fragment is then easily isolated.

Sections 2 and 3 of the ESEPO gene (Figs. 13 and 15) are similarly constructed from the oligonucleotides shown in Figs. 12 and 14. Each section was amplified into the M13 vector used to confirm the sequence and isolated from phage DNA. As can be seen from FIG. 13, ESERO Section 2 is prepared with EcoRI and BamHI adhesive ends and can be isolated as a KpnI / Bglll fragment. Similarly, ESERO Section 3 is prepared with BamHI and Sall adhesive ends and can be isolated from phage RFDHK as a Bglll / Sall fragment. The three sections thus obtained can easily be joined in a continuous DNA sequence (FIG.

7) encoding the entire human EPO polypeptide with one amino acid methionine codon (ATG) to initiate E. coli translation. It should also be noted that above the baseline ATG, there is a series of base pairs essentially repeating the sequence of the ribosomal binding site of the high expression OMPf gene of E. coli.

Any suitable expression vector can be used to carry ESERO. The specific vector selected for expression of the ESEPO gene is the "temperature-sensitive" plasmid pCFM536 - a plasmid derivative pCFM414 (ATCC 40076) described in US patent application hereto

636 727, filed 6.08.84 (EP application no. 136 480) by Charles F. Morris.

More specifically, pCFM536 was digested with Xbal and HindIII; the large fragment was isolated and used in two-step ligation with the ESEPO gene. Sections 1 (Xbal / Kpnl), 2 (Kpnl / BglH) and 3 (Bglll / Sall) are pre-assembled in the correct order in MI3 and the EPO gene is isolated from them as a single Xbal / Hindlll fragment. This fragment includes a portion of the M13tr9 phage polylinker spanning Sall to HindIII sites. Expression control in the resultant expression plasmid, p536, is accomplished by an XP _L promoter, which itself is controlled by the C _lg37 repressor gene (as in E. coli strain K 1 2AHtrp).

The resulting ESEPO gene 5 can be variously modified to encode erythropoietin analogs such as [Asn ² , des-Pro ² to lie ⁶ ] hEPO and [His7] hEPO, as described below.

A [Asn ² , des-Pro ² to lie ⁶ ] hEPO

Plasmid 536 carrying the resulting 10 ESERO gene of FIG. 7 like Xbal to Hindlll inclusion, grind with Hindlll and Xhol. The last endonuclease cuts the ESEPO gene at the only recognition site of 6 base pairs, 15 extending from the last codon base encoding Asp ⁸ to the second Arg ¹⁰ codon base. An Xbal / Xhol linker sequence was obtained in the following sequence:

Xbal + 1 2 7 8 9 Met Ala Asn Cys Asp Xhol 5-CTAG ATG GCT AAT TGC GAC-3 ' 3 ' -TAC CGA TTA ACG CTG AGCT-5 '

The Xbal / Xhol linker and the sequence of the Xhol / Hindlll fragment of the ESEPO gene are included in the large fragment obtained from Xbal and Hindlll digestion of plasmid pCFM526, derived from plasmid pCFM414 (ATCC 40076) prepared according to the Charles F. Morris method already described producing a plasmid-generated DNA sequence encoding E. coli expression of the Met ' ¹ form of the desired analogue.

B. [His ⁷ ] hEPO

Plasmid 536 was digested with HindIII and Xhol as in part A above. An Xbal / Xhol linker is obtained in the following sequence:

Xbal +1 23 4567 8 9 Xhol

Met Ala Pro Pro Arg Leu lie His Asp 5'-CTAG ATG GCT CCG CCA CGT CTG ATC CAT GAC-3 ' 3 ' -TAC CGA GGC GGT GCA GAC TAG GTA CTG AGCT-5 '

Linker and Xhol / Hindlll fragmented included in pCFM526 for preparing minutes of Esser sequence is generated from a plasmid DNA sequence coding for E. coli expression of Met ^'1 form the desired analog.

The resulting gene construct ("SCEPO") comprising codons favoring yeast is as described in the following FIGS. 16 to 21 and 8. As in the case of the ESEPO gene, the overall construction involves the formation of three groups of oligonucleotides (Figs. 16, and 20), which are duplexed and joined in sections (Figs. 17, and 21). It should be noted that the synthesis is facilitated in part by the use of the same suboptimal codons in both SCEPO and ESEPO constructs, for example oligonucleotides 7-12 of section 1 of both genes are identical to oligonucleotides 1-6 of section 2 in every gene.

The merged SCEPO sections were sequenced in M13 and sections 1, 2 and 3 were isolated from phage as HindIII / Kpnl, KpnI / BgLII and BglH / Sall fragments.

The now preferred expression system for SCEPO genius products is a secretion system based on the S. cerevisiae secreted factor as described in US 487 753, filed 08/22/83 by Grant A. Bitter, published 11/10/84 as an EP application 0 123 294. In short, the system includes constructions in which the DNA encoding the yeast α-factor gene leader sequence is located 5 'immediately adjacent to the coding region of the exogenous gene to be expressed. As a result, the translated gene product includes a leader or signal sequence that is "processed" by an endogenous yeast enzyme during the secretion of the rest of the product. Because the construct uses the codon to initiate the translation of the α-factor (ATG), it is not necessary to provide such a codon at position -I of the SCEPO gene. As can be seen from FIG. 8, the alanine (+1) coding sequence is preceded by a linker sequence allowing direct insertion into a plasmid including DNA for the first 80 residues of the leader factor after the α-factor promoter. A preferred construct for expression of the SCEPO gene involves four times ligation. including the SCEPO section fragments described above and the large fragment of HindIII / Sall digestion of plasmid pAC3. From the plasmid pAC3 / SCEPO obtained, the α-factor promoter and the leader sequence and the SCEPO gene were isolated by digestion with BamHI and ligated into BamHI-digested plasmid pUE to form the expression plasmid pUE / SCEPO.

Example 12

This example relates to the expression of recombinant products of the obtained ESEPO and SCEPO genes in the expression system of Example 11.

Using an expression system requiring E. coli host cells, the plasmid p536 of Example 11 was transformed into AM7 E. coli cells pre-transformed with a suitable plasmid, pMW1, carrying the _CgJ7 gene. Cell cultures in LB broth (ampicillin 50 pg / ml and kanamycin 5 pg / ml, preferably with 10 mmol magnesium sulfate) were cultured at 28 ° C and at cell growth in the culture to an optical density of OD _60n = 0, l. EPO expression was induced by raising the temperature to 42 ° C. Cells cultured to about 40 OD provide an EPO production (as determined by gel) of about 5 mg / OD liter.

Cells were harvested, lysed, disrupted with French Press (10,000 psi = psi) and treated with lysozyme and NP-40 detergent. The precipitate obtained by centrifugation at 24,000 xg was dissolved in guanidine hydrochloric acid and further purified one-step by C ₄ (vydac) reverse phase HPLC (EtOH) 0-80%, 50 mmol NH ₄ Ac, pH 4,5). Protein sequencing shows that the product is more than 95% pure and the resulting products have two different amino ends, APPR ... and PPR ... in a relative quantitative ratio of about 3: 1. This recent observation of hEPO and [des Ala'JhEPO products indicates that the "processing" of the amino terminus in host cells serves to remove the methionine terminal and, in some cases, the initial alanine. The activity of the isolates found in the radioimmunoassay is 150,000 to 160,000 U / mg; in the in vitro assay, the activity was from 30,000 to 62,000 U / mg; and in the in vivo assay the activity ranged from about 120 to 720 U / mg. (For each study, the standard activity of human urine isolate is 70,000 U / mg.) The dose response curve for the recombinant product in the in vivo study differs markedly from that of the standard human urine EPO.

The plasmids of EPO analogs obtained in Parts A and B of Example 11 were transformed into pMW1-transformed AM7 E. coli cells and the cells were cultured as described. Purified isolates were tested both by RIA and in vitro assays. The values obtained for R1A and in vitro studies for [Asn ² , des Pro ² to lie ⁶ ] hEPO expression products are approximately 11,000 U / mg and 6,000 U / mg protein, respectively, while the values obtained for [His ⁷ ] hEPO are approximately 41,000 U / mg and 14,000 U / mg protein, respectively, indicating that the analogues have activity from one quarter to one tenth of that of the parent expression product.

An expression system intended for use in S. cerevisiae host cells, the plasmid pUE / SCEPO is transformed into two different strains. YSDP4 (genotype arep4-3 trpl) and RK81 (genotype aapep4-3trpl). Transformed YSDP4 hosts were cultured in SD medium (Methods in Yeast Genetics Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, p. 62 (1983) containing 0.5% casemic acids at pH 6.5 at 30 ° C. Medium collected when cells cultured to 36 OD contain EPO products in amounts of about 244 U / ml (97 gg / ODI according to RIA) Transformed RK81 cells cultured to 6.5 OD or 60 OD give media with EPO concentrations of about 80- 90 U / ml (34pg / ODl according to RIA) Preliminary analyzes show significant heterogeneity of products produced by the expression system, possibly due to variations in glucose the expression of expression proteins and the relatively high content of mannose in the bound carbohydrate.

Plasmids PAC3 and PUE in HB101 E. coli cells were deposited on 09/27/84 in the American Culture Type Collection, 12301 Parklawn Drive, Rockville, Md., Respectively ATCC 39881 and ATCC 39882. Plasmids pCFM526 in AM7 cells, pCFM 536 in JM103 cells and pMW1 in JM103 cells were also deposited on 21.11.84 under ATCC numbers 33932, 33934 and 33933, respectively. Strains of Saccharomyces cerevisiae YSPD4 and RK81 were deposited on 21.11.84 under ATCC numbers 20734 and 20733, respectively.

As will be apparent from the examples described above, the present invention provides a large number of extremely valuable products and processes that find application in many different directions.

The polypeptides of the invention are obviously useful materials, whether they are obtained by microbial expression or are synthetic products, with the primary, secondary and tertiary structural conformation being first elucidated in the present invention.

As indicated, the recombinant and synthetic products of the present invention exhibit, in varying degrees, the biological activity of in vitro EPO isolates from natural sources and can therefore be used as substitutes for EPO isolates in nutrient media used for culturing erythropoietin cells. Similarly, given the extent to which the polypeptide products of the invention share in vivo the activity of natural EPO isolates, they are obviously suitable for use in the treatment of mammalian erythropoietin, including humans, as they exert some or all of the effects, characteristic in vivo of EPO, eg stimulation of reticulocyte response, development of ferrokinetic effects (such as effects of iron turnover and effects during passage of bone marrow), changes in erythrocyte mass, stimulate is the synthesis of hemoglobin C (see Eschbach et al. above) and as outlined in Example 10, raising the level of hematocrit values in mammals. Included in the group of people who can be treated with the products of the invention are patients in total need of blood transfusion, including trauma patients, those undergoing surgery, kidney patients, including those on dialysis, and patients with various diseases affecting the composition of the blood, such as hemophilia, sickle cell anemia, physiological anemia, and the like. Minimizing the need for transfusion through the administration of EPO therapy may be expected to lead to reduced transmission of infectious agents. The products of the invention, by virtue of being obtained by recombinant methods, are expected to be free from pyrogens, natural inhibitors and the like and therefore exhibit increased overall efficacy when used in therapeutic methods compared to naturally occurring products. Erythropoietin therapy with products obtained according to the present invention is also expected to increase oxygen transfer capacity in individuals with established hypoxic conditions due to specific environmental conditions and possibly to beneficial cardiovascular effects.

One preferred method of administering the polypeptide products of the invention is parenterally (e.g., intravenously, intramuscularly, subcutaneously, or intraperitoneally), the compositions being administered typically include therapeutically effective amounts of the product in combination with acceptable diluents, carriers and / or additives. Preliminary pharmacokinetic studies have shown a longer half-life of in vivo in monkey EPO products in intramuscular than in intravenous administration. Effective doses are expected to vary greatly depending on the disease being treated, but therapeutic doses are set, with the current expected active substance being 0.1 (about 7U) to 100 (about 7000 U) µg / kg body weight. For the pharmaceutical ingredients of the invention, standard diluents such as human serum albumin and standard carriers such as saline are used.

Excipients suitable for use in the compositions of the invention include compounds known to exhibit a stimulating effect on erythropoiesis such as testosterone, insulin-like growth factor stimulators, prostaglandins, serotonin, cyclic NMR, prolactin, and triactin and prolactin and agents commonly used in the treatment of aplastic anemias, such as methanol, stanozolol, and nandrolone (see, e.g., Resegotti et al., Panminerva Medica, 23, 243-248 (1981); McGonigle et al., Kidney Int., 25 (2). 437444 (1984); Pavlovic-Kantera etal., Expt. Hematol., 8 Supp, 8), 283-291 (1980) and Kurtz, FEBS Letters, 14a (1), 105-108 (1982). Also known as additives are known as erythropoietin or asialo-EPO enhancers or synergists, such as adrenergic agonists, thyroid hormone, androgens, and BPA [see Dunn "Gurrent Concepts in Erythropoiesis", John Wiley and Sons (1983) ; Weiland et al., Blut, 44 (3), 173-175 (1982); Kalmanti, Kidney Int., 22, 383391 (1982); Shahidi, New. Eng. J. Med., 289, 72-80 (1973); Fisher et al., Steroids, 30 (6), 833-845 (1977); Urabe et al., J. Exp. Med. 149, 1314-1325 (1979) and Billat et al., Expt. Hematol., 10 (1), 133-140 (1982)], as well as the group of compounds referred to as "liver erythropoietic factors", see Naughton et al., Acta Haemat., 69, 171179 (1983)] and "erythrotropins" [ as described by Congote et al., in abstract 364, Proceedings 7th Internat. Congress of Endocrinology (Quebec, 1984); Cingote, Biochem. Biophys. Really. Comm, 115 (2), 447-483 (1983) and Congote, Anal. Biochem., 140, 428-433 (1984)] and erythrogenin [as described by Rothman et al., J. Surg. Oncol. 20, 105108 (1982)]. Screening to determine the erythropoietin responses of exhypoxic polycythemic mice pretreated with 5α-dihydrotestosterone or nandrolone and then erythropoietin according to the invention yielded the same results.

The polypeptides of the invention are widely used in diagnostics, which involves their use as labeled and unlabeled forms in various immunoassay methods, including RIA, ELISA and the like, as well as various in vitro and in vivo activity studies. See, e.g., Dunn et al., Expt. Hematol., 11 (7), 590-600 (1983); Gibson et al., Pathology 16, 155156 (1984); Krystal, Expt. Hematol. 11 (7), 649-660 (1983); Saito et al., Jap. J. Med. 23 (1) 16-21 (1984); Nathan et al., New Engl. J. Med., 308 (9), 520-522 (1983) and various sources pertaining to research in the field.

The polypeptides of the invention, including the synthetic peptides representing the EPO residue sequences, first described herein, are also very useful pure materials for the preparation of polyclonal antibodies and monoclonal antibody banks specific for various continuous and interrupted epitopes of EPO. As an example, it can be noted that in the preliminary analysis of the amino acid sequences of FIG. 6 in hydropathy according to Nore et al., P.N.A.S. (US) 78, pp. 3824-3828 (1981) and secondary structures according to Chou et al., Ann. Rev. Biochem. 47, p. 251. (1979) found that synthetic peptides with continuous sequences covering residues from positions 41 to 57, including from 116 to 128 and from 144 to 166, provoke strong antigenic reactions and produce useful monoclonal and polyclonal antibodies , exhibiting immune responses to both synthetic polypeptides and whole protein. Such antibodies are expected to be useful for the detection and affinity purification of EPO and related products.

The following three synthetic peptides were obtained:

(1) hEPO 41-57, V-P-D-T-K-V-N-FY-A-W-K-R-M-E-V-G;

(2) hEPO 116-128, K-E-A-I-S-P-P-DA-A-S-A-A;

(3) hEPO 144-166, V-Y-S-N-F-L-R-GK-L-K-L-Y-T-G-E-A-C-R-T-G-D-R.

Immunization preparatory studies with the described polypeptides showed a relatively low positive response to hEPO 41-57, a minor response to hEPO 116128, and a strong positive response to hEPO 144-166, accounted for by the ability of rabbit serum antibodies to immunorecipitate ^{| 25} 1-labeled human from urine. Preliminary in vivo studies did not show significant activity of the three peptides, either alone or in combination.

Although the determined amino acid residues of mammalian EPOs obtained in the examples are essentially the same as the primary structural conformation of mature EPOs, it should be noted that the sequence of 165 amino acid residues of monkey EPOs in Table V and 166 the human EPO residue of FIG. 6 does not exhaust the group of useful polypeptides prepared according to the invention. Also included within the scope of the invention are all the various natural allelic forms of EPO for which previous studies on biologically active animal polypeptides, such as human γ-interferon, have been found to be likely to exist. For comparison, human immune interferons having an arginine residue at position 140 according to EP 0 077 670 and those having glutamine at position 140 according to Gray et al., Nature 295, p. 503508 (1982) may be mentioned. Both species are characterized as having sequences of mature human γ-interferon. The allelic forms of mature EPO polypeptides may differ from each other, as well as the sequences shown in FIG. 5 and 6, with respect to the length of the sequence and / or the deletions, substitutions, insertions or additions of amino acids in the sequence made, with subsequent potential differences in glycolysis ability. As indicated earlier, a putative allelic form of human EPO includes one methionine residue at position 126. Accepted. that there are also natural allelic forms encoding EPO genomic DNA and cDNA sequences that encode for said types of allelic polypeptides or simply use different codons to designate the same polypeptides.

In addition to the natural allelic forms of mature EPO, the present invention also includes other EPO products such as polypeptide analogues of EPO and fragments of mature EPO. According to the method of the described application of Alton et al. (WO 83/04053), genes encoding microbial expression of polypeptides with a primary conformation different from that described in mature EPO in identity can be easily designed and prepared. or the arrangement of one or more residues due to substitutions, final and intermediate additions and deletions. In contrast, changes in cDNA and genomic EPO genes can be readily accomplished by known site-directed mutagenesis methods and used to create analogs and derivatives of EPO. Such EPO products exhibit at least one of the biological properties of EPOs, and others may differ. As an example, the products obtained according to the invention EPO include those that are pre-truncated by deletions, for example [8 Asn ² , des -Pro ² to lie ⁶ ] hEPO, [des -Thr ¹⁶³ to Arg ¹⁶⁶ ] hEPO and "Δ27-55Ι1ΕΡΟ "The latter having residues encoding a completely codon removed or those more resistant to hydrolysis (and therefore may have more pronounced or longer lasting effects than natural EPOs); or those that have been altered for deletion at one or more potential glycolysis sites (which may result in higher activity in yeast-produced products); or those which in one or more cysteine residues are removed or substituted, for example, by histidine or serine residues (such as the [His ⁷ ] hEPO analogue) and are potentially easier to isolate in active form from microbial systems; or those in which one or more tyrosine residues are substituted with phenylalanine (such as [Phe ¹⁵ ] hEPO, [Phe ⁴⁹ ] hEPO and [Phe ¹⁴⁴⁵ ] hEPO ^analogues and can bind more or less readily to EPO receptors on target cells. Also included are polypeptide fragments that repeat only part of a continuous amino acid sequence or secondary conformations in mature EPO, which fragments may exhibit one aspect of EPO activity (eg, receptor binding) and others not (eg, erythropoietin activity). in this oh related are those potential EPO fragments that are included in the human genomic DNA sequence of Table VI, i.e., "fragments" of the entire continuous EPO sequence, which are encoded by intron sequences and which may constitute specific domains of biological activity. It is worth noting that the lack of in vivo activity in one or more of the EPO products of the invention does not completely preclude their therapeutic use (see Weiland et al., Above) or their use in other areas such as p in EPO assays or EPO antagonism. Erythropoietin antagonists can be quite useful in the treatment of polycysthemia or cases of overproduction of EPO [see, e.g., Adamson, Hosp. Practice, 18 (12), 49-57 (1983) and Hellmann et al., Clin. Lab. Haemat., 5, 335-342 (1983).

According to another aspect of the present invention, the cloned DNA sequences described herein that encode human and monkey EPO polypeptides are very valuable because of the information on the amino acid sequence of animal erythropoietin that they provide, as it has been lacking so far, despite many analytic studies of the isolate of the isolate natural products. DNA sequences are also valuable as products and can be used to perform large-scale microbial synthesis of erythropoietin by various recombinant methods. In other words, the DNA sequences according to the invention are used to produce new and useful viral and circular plasmid DNA vectors, new and useful transformed and transfected microbial prokaryotic and eukaryotic host cells (including bacterial and yeast cells and mammalian cells grown in culture) and novel and useful methods for culturing such host microbial cells capable of expressing EPO and EPO products. The DNA sequences of the invention are suitable for use as labeled indicators in the isolation of cDNAs encoding EPO and related proteins and the genomic DNA sequence of other mammalian and non-mammalian species specifically described herein. The extent to which the DNA sequences according to the invention can be used in various other protein synthesis methods (eg insect cells) or in the genetic therapy of humans and other mammals cannot yet be evaluated. The DNA sequences of the invention can be used in the development of transgenic mammalian species to serve as eukaryotic hosts in the production of erythropoietin and erythropoietin products in sufficient quantity. See generally Palmiter et al., Science 222 (4625),

809-814 (1983).

Therefore, the specific applications according to the exemplary embodiments do not limit the scope of the present invention, and one skilled in the art can find numerous applications. As one example, while the DNA sequences described in the exemplary embodiments include cDNA and genomic DNA sequences, because the application provides the necessary amino acid sequence information to obtain the DNA sequence, the invention also includes DNA sequences that can be constructed based on knowledge of amino acids DNA sequences. They can code for EPO (as in Example 12) and for EPO fragments, as well as for polypeptide EPO analogs (called "EPO products"), which may exhibit one or more of the biological properties of natural erythropoietin without any other ( or to manifest them to varying degrees).

The DNA sequences of the present invention therefore include all DNA sequences providing expression in a prokaryotic or eukaryotic host cell of a polypeptide product having at least part of the primary structural conformation and one or more of the biological properties of erythropoietin, i.e.: (a) DNA sequences shown in FIG. 5 and 6; (c) DNA sequences that hybridize to the DNA sequences referred to in (a) or fragments thereof; and (c) DNA sequences which, due to the degeneration of the genetic code, hybridize to the DNA sequences defined in (a) and (c). It should be noted in this sense that, for example, existing sequences of allelic monkey and human EPO genes, as well as gene sequences of other mammalian species, are expected to hybridize to the sequences of FIG. 5 and 6, or fragments thereof. In addition to those with degenerate genetic codes, the SCEPO and ESEPO genes and the cDNA or genomic DNA sequences obtained or mutagenic encoding for different EPO fragments and analogs also hybridize to said DNA sequence. Such hybridizations may be carried out under the hybridization conditions described for the initial isolation of DNA encoding monkey and human EPO, or under more stringent conditions, if lateral hybridization is desired.

Similarly, although the foregoing examples illustrate the invention with respect to microbial expression of EPO products in the context of expression in animal DNA cells included in a hybrid vector of bacterial plasmid and viral genomic sources, a wide variety of invention is included within the scope of the invention expression systems. These include expression systems including vectors of homogeneous origin applied to cultures of bacterial, yeast and mammalian cells, as well as expression systems that do not include vectors (such as transfecting cells with calcium phosphate). In this respect, it should be understood that the expression of, for example, monkey DNA in a monkey cell culture and in a human host cell culture, in fact, are cases of exogenous DNA expression. since there is no source in the genome of the host cell for EPO DNA for which a high level of expression is sought. The expression systems of the invention are further related to methods resulting in the cytoplasmic formation of EPO products and the accumulation of glycosylated and non-glycosylated EPO products in the cytoplasm of host cells or in their membranes (eg accumulation in periplasmic space) or in food space , as described, or other unusual systems such as the P. aeruginosa expression system (described by Gray et al., Biotechnology, 2, pp. 161-165 (1984).

Improved hybridization methods according to the invention, although only the application for DNA / DNA hybridization screening is described, are equally applicable to RNA / RNA or RNA / DNA screening. Mixed-indicator methods such as those described generally represent improvements in hybridization methods that allow for faster and more reliable isolation of polynucleotides. These many distinct improvements include improved colony transfer and maintenance methods; the use of nylon-based filters such as "GeneScreen" and "GeneScreen Plus" that allow re-samples with the same filters and re-use of the filter, applying new protease treatments (compared, for example, with Taub et al., Annal. Biochem., 125 , pp. 222-230 (1982)], the use of very low individual concentrations (in the order of 0.025 pmol) of a large number of mixed indicators (eg more than 32 in number) and the implementation of hybridization and post-hybridization43 stages in a strict temperature regime with values approaching the lowest (for example, differs These are improvements combined to produce results that could not be expected. This is clearly illustrated by the 10 ° C and, preferably, 2 ° C, 5 dissociation temperatures calculated for some of the mixed indicators used. the fact that methods using mixed indicators, including four times as many indicators as previously described, were successfully used for cDNA screening of RNAs in 15 relatively small amounts, have been successfully applied to isolate a gene with a unique screening sequence of 1,500,000 phages plaques in a genomic bank. This fact refutes the 20 published opinion of Anderson et al., See above that mixed-indicator screening methods are ... "inapplicable in isolating mammalian protein genes when the corresponding RNAs are not available".

Claims (31)

Patent Claims 1. A purified and isolated DNA sequence encoding erythropoietin selected from the group consisting of 30 consisting of: (a) The DNA sequences described in FIG. 5 and 6 or their complementary chains; (c) DNA sequences that, under stringent conditions, hybridize to the DNA sequences defined in (a). 2. A purified and isolated DNA sequence comprising essentially 40 DNA sequences encoding human erythropoietin. 3. A purified and isolated DNA sequence consisting essentially of a DNA sequence encoding a monkey erythropoietin. 4. A prokaryotic or eukaryotic host cell transformed or transfected with a DNA sequence according to claims 1.2 and 3 for the expression of erythropoietin in the host cell. A biologically active circular plasmid or viral DNA vector comprising a DNA sequence according to claims 1, 2 or 3. A prokaryotic or eukaryotic host cell stably transformed or transfected with a DNA vector according to claim 6. 7. A purified and isolated DNA sequence consisting essentially of a DNA sequence encoding a polypeptide having an amino acid sequence close enough to that of erythropoietin to provide the biological property of bone marrow cells to enhance the production of reticulocytes and red blood cells and increase hemoglobin synthesis and iron uptake. The cDNA sequence of claim 7. The DNA sequence of claim 8 encoding monkey erythropoietin. The DNA sequence of claim 9, comprising the protein coding region of FIG. 5. The genomic DNA sequence of claim 7. The DNA sequence of claim 11 encoding human erythropoietin. The DNA sequence of claim 12, comprising the protein coding region shown in FIG. 6. The DNA sequence of claim 7, comprising one or more codons preferred for expression in E. coli cells. The DNA sequence of claim 14 encoding for the expression of human erythropoietin. The DNA sequence of claim 15, comprising the protein coding region of FIG. 7. The DNA sequence of claim 7, comprising one or more codons preferred for expression in yeast cells. 18. The DNA sequence of claim 17 encoding for the expression of human erythropoietin. The DNA sequence of claim 18, comprising the protein coding region shown in FIG. 8. The DNA sequence of claim 7, covalently linked to a detectable marker substance. The DNA sequence of claim 20, wherein the detectable substance is isotopically labeled. The single stranded DNA sequence of claim 20. A prokaryotic or eukaryotic host cell transformed or transfected with a DNA sequence according to claims 7, 8 or 11 in a manner that provides expression of the polypeptide in the host cell. A transformed or transfected host cell according to claim 23. which is capable of glycolysis of the polypeptide. 5 A mammalian transformed or transfected host cell according to claim 24. A transformed or transfected COS cell according to claim 25. 10 A transformed or transfected CHO cell according to claim 25. 28. Biologically active circular plasmid or viral DNA vector, 15 comprising the DNA sequence of claim 7. 29. A prokaryotic or eukaryotic host cell stably transfected or transformed with The DNA vector of claim 28. 30. The DNA sequence of claim 7, encoding for [Phe ¹⁵ ] hEPO, [Phe ⁴⁹ ] hEPO, [Phe ¹⁴⁵ ] hEPO, [His ⁷ ] hEPO, [Asn ² -des25 Pro ² to lle ⁶ ] hEPO, [ des Thr ¹⁶³ to Arg ¹⁶⁶ ] hEPO or [A27-55] hEPO. 31. A purified and isolated DNA sequence as shown in FIG. 5 or 6 or complementary ve- 30 Riga of such sequence. Attachment: 21 figures