RU2120996C1 - Dna fragments encoding human erythropoietin, recombinant dnas, cell line, recombinant o-glycosylated human erythropoietin, pharmaceutical composition - Google Patents

Abstract

FIELD: biotechnology, medicine. SUBSTANCE: genome DNA fragments from human fetal liver and the corresponding fragments to DNA providing high expression levels of erythropoietin are cloned in mammalian cells. Cell line CHO - a producer of recombinant erythropoietin (CHO EPO Cla 44.02-7) is obtained. As a result of expression of the cloned DNA fragments in mammalian cells O-glycosylated human erythropoietin is obtained at specific activity 200 000 U/mg, not less, and pharmaceutical composition using the indicated recombinant erythropoietin as an active component. EFFECT: improved method of preparing, increased specific activity of erythropoietin. 12 cl, 11 tbl, 12 dwg

Description

 The invention relates to cloned DNAs encoding human erythropoietin, the expression of these DNA and obtaining in vitro active human erythropoietin.

 Erythropoietin (hereinafter EPO) is a circulating glycoprotein that stimulates the formation of red blood cells in higher organisms (see Carnot et al., Compt. Rend, 143: 384 (1906)). EPO is sometimes also referred to as a factor stimulating erythropoiesis.

 The life span of human red blood cells is about 120 days. Thus, about 1/120 of the total number of red blood cells is destroyed daily in the reticuloendothelial system. However, a relatively constant number of red blood cells is produced daily to maintain a constant level of red blood cells (Guyton, Textbook of Medikal Physiology, pp. 56-60, W.B. Saunders Co., Philadelphia (1976)).

 Red blood cells are produced by maturation and differentiation of erythroblasts in the bone marrow, and EPO is a factor that acts on less differentiated cells and induces their differentiation into red blood cells (Guyton, above).

 EPO is a promising therapeutic agent for the clinical treatment of anemia, and in particular renal anemia. Unfortunately, the use of EPO is not yet generally accepted in practical therapy due to its low availability.

 When using EPO as a therapeutic agent, the issue of antigenicity is of particular importance. Preferably, EPO is obtained from raw materials of human origin, for example, from the blood or urine of patients suffering from aplastic anemia or similar diseases in which large amounts of EPO are released. These starting materials, however, exist in limited quantities (See, for example, White et al., Rec, Progr. Horm. Res., 16-219 (1960); Espada et al., Biochem. Med. 3: 475 ( 1970); Fischer, Phurm. Rev. 24: 459 (1972) and Gordon, Vitam, Horm. (NY) 31: 105 (1973).

 Obtaining EPO, as a rule, is carried out by concentrating and purifying the urine of patients with these diseases (see, for example, US patents NN 439780, 4303650 and 3865801, the disclosure of which is incorporated herein by reference). The limited supply of such urine is an obstacle to the practical use of EPO. The difficulty in using urine from healthy people is the low content of EPO in it compared with the urine of patients suffering from anemia. In addition, the urine of healthy people contains certain inhibitory factors, therefore, a satisfactory therapeutic effect can be obtained only with the subsequent significant purification of the EPO thus obtained.

 EPO can also be isolated from the blood plasma of sheep, and the selection of EPO from this source provides the creation of sufficiently potent and stable water-soluble drugs (see Coldwasser, Control Cellular Dif. Develop., Part A, pp. 487-494, Alan R. Liss, Inc., New York (1981), which is incorporated herein by reference). Sheep EPO, however, may have antigenic effects.

 Thus, the traditional methods of isolation and purification used in obtaining EPO from natural sources of supply are of little use for its mass production.

 Sugimoto et al. In US Pat. No. 4,377,513 describes one method for mass production of EPO, which involves in vivo propagation of human lymphoblastoid cells, such as Namalva, BAL L - 1, TAL L - 1, TAL L - 1 and JBL.

 A description of the production of EPO using genetic engineering methods has also appeared in the specialized literature. However, neither a suitable disclosure of these methods nor the chemical nature of the product has yet been published. In contrast, the present application provides a mass production method for producing proteins exhibiting the biological properties of human EPO. Using these techniques, it is also possible to obtain proteins that may be chemically different from authentic human EPO and yet exhibit similar (and in some cases improved) properties. For convenience, all such proteins exhibiting the biological properties of human EPO are hereinafter referred to as EPO.

 The present invention relates to the cloning of a DNA sequence that provides high levels of expression of human EPO. Suitable expression vectors and cell systems for the production of EPO, purification schemes of the recombinant product are also described, and its characteristics are given.

 As described in more detail below, EPO was isolated in partially purified form, subjected to further purification to homogeneity, and digested with trypsin to obtain specific fragments. These fragments were purified and sequenced. Based on these sequences, oligonucleotides were constructed and synthesized, which were then used to screen the human genomic library, the EPO gene was isolated to isolate the EPO gene.

 Based on mRNA from the liver of a human fetus (at the age of 20 weeks), a cDNA library was obtained and screened. Three clones of EPO cDNA were isolated (after screening> 750,000 recombinants). Two of them were identified as having full length. A sample of E. coli, transformed with a full-length EPO coding sequence, was deposited at ATCC, Rockville, Maryland under number 40153. These cDNAs were expressed as in monkey transformed with SV-40 virus (cell line COS-1; Gluzman, Cell 23: 175 -182 (1981)), as well as in Chinese hamster ovary cells (CHO cell line; Urlaub, G. and Chasin, LA Proc. Natl. Acad. Sci USA 77: 4216-4280 (1980)). EPO obtained in mammalian cells are biologically active in vitro and in vivo.

 Table 1 is an exon sequence of 87 base pairs of the human EPO gene.

 FIG. 1 illustrates the detection of EPO mRNA in human fetal liver mRNA.

 Table 2 presents the amino acid sequence of EPO deduced from the nucleotide sequence of clone λ-HEPOF L 13.

 Table 3 presents the nucleotide sequence of EPO cDNA in λ-HEPOF L 13 (shown schematically in FIG. 2) and the amino acid sequence derived from it.

 FIG. 3 illustrates the relative positions of DNA inserts of four independent genomic clones of human EPO.

 FIG. 4 is a map of the intron and exon structure of the human EPO gene.

 Table 4 illustrates the DNA sequence of the EPO gene schematically shown in FIG. 4.

 FIG. 5 to 9 illustrate the construction of the vector p91023 (B).

 FIG. 10 depicts the results of a comparative analysis on SDS-polyacrylamide gel of EPO obtained in COS-1 cells and native EPO.

 Table 5 illustrates the nucleotide and amino acid sequences of the insertion of the λ-HEPOF L 6 clone.

 Table 6 illustrates the nucleotide and amino acid sequences of the insertion of the λ-HEPOF L 8 clone.

 Table 7 illustrates the nucleotide and amino acid sequences of the insertion of the λ-HEPOF L 13 clone.

 FIG. 11 is a schematic representation of plasmid pRK 1-4.

 FIG. 12 is a schematic representation of the plasmid pdBPY-MMTneo (342-12).

 The present invention relates to the cloning of DNA sequences encoding EPO obtained by expression of these sequences of a recombinant product.

A. Isolation of the genomic clone of human EPO
Human EPO was purified to homogeneity from the urine of patients suffering from aplastic anemia, as described below. Complete digestion of this purified EPO with trypsin yielded fragments that were separated by reverse phase high performance liquid chromatography, recovered from the gradient fractions and subjected to microsequencing. The sequences of tryptic fragments are highlighted in tables 2 and 3 and are discussed in more detail below. Two of the amino acid sequences (Val-Asn-Phe-Tyr-Ala-Trp-Lys and Val-Tyr-Ser-Asn-Phe-Leu-Arg) were selected for the construction of oligonucleotide probes (oligonucleotide pool of 17-member with 32-fold degeneration and an oligonucleotide pool of 18-member with 128-fold degeneration for the first tryptic fragment, as well as two pools of 14-member, each with 48-fold degeneration for the second tryptic fragment). A 32-fold degenerative 17-member was used to screen the human genomic DNA library in the Ch4 A vector (22) using a modification of the in situ amplification method Woo and O'Malley [4] to obtain filters for screening.

 Arabic numbers in brackets [1 - 59] are used in the present description for references to publications that are given in numerical order at the end of the description of the present invention.

 The phage hybridizing with the 17-member was packaged, distributed into small groups and tested with 14-member and 18-member pools. The phage hybridizing with 17-membered, 18-membered and 14-membered was purified, and the fragments were subcloned into M13 vectors for sequencing using the dideoxy chain termination method described by Sanger and Conlson [23].

 The sequence of the region of one of the clones, hybridizing with a 32-fold degenerative 17-member, is shown in table. 1. This DNA sequence contains within the open reading frame a nucleotide sequence encoding a tryptic fragment used to construct a 17-membered pool of oligonucleotides. In addition, DNA sequence analysis shows that the region hybridizing with the 17-membered is contained within 87 bp exon.

 The confirmation that these two clones (indicated here as λ-HEPO1 and λ-HEPO2) are genomic EPO clones, obtained by sequencing additional exons containing coding regions and another tryptic fragment.

 B. Isolation of EPO cDNA Clones.

 Northern- (56) analysis of human fetal (age 20 weeks) hepatic mRNA was performed using a 95-membered single-stranded oligonucleotide probe obtained from clone M13 containing the 87 bp exon region described in Table 1. 1. As shown in FIG. 1, a strong signal is detected in mRNA of the fetal liver. Accurate identification of this band as EPO mRNA was achieved using the same probe for screening λ-cDNA to the fetal liver mRNA library [25]. Several hybridizing clones were obtained (approximately 1 positive per 250,000 screened recombinants). The complete nucleotide and deduced amino acid sequences for two clones (λ-HEPOF L 13 and λ-HEPOF L 8) are presented in table. 5 and 6. EPO-coding information is contained within 594 nucleotides in the 5'-half of the cDNA, including the encoding of a highly hydrophobic 27-amino acid leader and 166-amino acid mature protein.

 Identification of the N-terminus of the mature protein was based on the N-terminal sequence of the protein isolated from the urine of people with aplastic anemia (Table 1 and Goldwasser [26], Sue and Sytkowski [27] and Yangawa [21]). It is currently unknown whether this N-terminus (Ala-Pro-Pro-Arg ---) represents the actual N-terminus for EPO in the circulatory system or whether some cleavage occurs in the kidney or urine.

 Amino acid sequences that are underlined in table. 2 and 3, denote tryptic fragments and the N-terminal region for which information from the protein sequence was obtained. The deduced amino acid sequence exactly matches the sequences of tryptic fragments that were sequenced, confirming that the isolated gene encodes a human EPO.

C. Structure and sequence of the human EPO gene
The relative positions of the DNA inserts of the four independent human genomic EPO clones are shown in FIG. 3. Hybridization analysis of these cloned DNAs using oligonucleotide probes determined the position of the EPO gene within a region of approximately 3.3 kV shown by the shaded line in FIG. 3. Complete sequencing of this region (see Example 4) and comparison with cDNA clones resulted in a map of the intron and exon structures of the EPO gene shown in FIG. 4. The EPO gene is divided into 5 exons. Part of exon 1, complete exons II, III, and IV, as well as part of exon V contain protein-coding information. The remaining portions of exons I and V represent 5 ′ and 3 ′ untranslated sequences, respectively.

6. Expression of EPO in COS cells
In order to clearly demonstrate that biologically active EPO can be expressed in the in vitro cell culture system, expression studies in COS cells were carried out [57]. The vector used for transient studies, namely p91023 (B), is described in Example 5. This vector contains the main late adenovirus promoter, the polyadenylation sequence, the origin of replication and the SV40 enhancer, and the adenovirus VA gene. The cDNA insert in λ-HEPOF L 13 (see table 6) was inserted into the p91023 (B) vector below the main late promoter of the adenovirus. This new vector is identified as pPTF-L 13.

Twenty-four hours after transfection of this construct into strain M6 of COS-1 cells (Horowitz et al., T. Mol. Appl. Genet 2: 147-149 (1983) (the cells were washed and transferred to serum-free media and after 48 hours the cells were harvested.) Then, using the quantitative radioimmunoassay for EPO (57), the release of EPO into the culture supernatant was examined. In a separate experiment, a vector containing EPO cDNA from λ-HEPOF L 13 was introduced into COS-1 cells; media were then collected, as described above EPO in media was then quantified by either of the two biol ogic in vitro assays, in particular with ³ H-thymidine and CFV-E [12, 29], and any of the two in vivo assays, namely in a hypoxic mouse and starving rat [29, 30] (see table 9 and Example 17) These results show that biologically active EPO is produced in COS-1 cells. Western blotting using a polyclonal anti-EPO antibody showed that erythropoietin produced by COS cells has a mobility in SDS-polyacrylamide gels identical to that of the native EPO obtained from human urine (example 8). Thus, the glycosylation of EPO produced in COS-1 can be similar to what occurs in the case of native EPO.

 Various vectors containing other promoters can also be used in COS cells or in cells of other mammals. Examples of these other promoters useful in the practice of the present invention include the early and late SV-40 promoters, the mouse metallothionein gene promoter, the promoter found in long terminal repeats of bird or mammalian retroviruses. Examples of other types of cells useful in the practice of the present invention are mammalian cells such as CHO (Chinese Hamster Ovary), C127 (monkey epithelium), ZT3 (mouse fibroblast), CU-1 (African green monkey kidney).

 Examples of EPO expression in CHO, C127 and ZT3 are presented in examples 10 and 11 (CHO) and 13 (C127 and ZT3).

Recombinant EPO obtained in CHO cells, as in Example 11, was purified by conventional column chromatography methods. The relative amounts of sugars present in the glycoprotein were analyzed by two independent methods (1) Reinhold, Methods in Enzymol, 50: 244-249 (Methanolysis) and (II) Takemoto, H. et al. Anal. Biochem. 145: 245 (1985) (pyryl amination, together with the independent determination of sialic acid). The results obtained by each of these methods were fully consistent with each other. Thus, several determinations were made, which gave average values, where N-acetylglucosamine for comparative purposes is given as a value of 1:
Sugar - Relative molar level
N-Acetylglucosamine - 1
Hexoses: - 1.4
Galactose - 0.9
Mannose - 0.5
N-Acetylneurominic Acid - 1
Fucose - 0.2
N-Acetylgalactosamine - 0.1
It is noteworthy that significant levels of fucose N-acetylgalactosamine were reproducibly observed when using both independent sugar analysis methods. The presence of N-acetylgalactosamine indicates the presence of O-linked protein glycosylation. The presence of O-linked glycosylation was further confirmed by SDS-PAGE assay. In particular, following the enzymatic cleavage of all N-linked carbohydrate on glycoproteins using endo-F N-glycoside peptidase, the molecular weight of the protein was further reduced after sequential digestion with neuraminidase.

 The in vitro biological activity of purified recombinant EPO was determined by C. Krystal, Exp. Hematol., 11: 649 (1983) (bioassay for proliferation of spleen cells). After repeated determinations, the specific in vitro activity of purified recombinant EPO was found to be more than 200,000 units / mg of protein. The average value was in the range of 275000-300000 units / mg protein. In addition, values higher than 300,000 were also observed. In vivo activity ratios (analysis of polycythemic mice, Kazal and Erslev, hm. Clinical Lab. Sci., Volume B, p. 91 (1975)) to in vitro activity obtained for the recombinant product were in the range of 0.7-1.3.

 It is interesting to compare the above glycoprotein profile with that of the recombinant CHO-produced EPO previously described in International Patent Application WO 85/02810 (published June 20, 1985). A similar comparative sugar analysis presented on page 65 of this application shows that the values for fucose and N-acetylgalactosamine are 0, while the ratio of hexose: N-acetylgalactosamine is 15.09: 1. The absence of N-acetylgalactosamine indicates the absence of O -linked glycosylation in the resulting glycoprotein. In contrast, the recombinant CHO-produced EPO of the present invention described above contains significant and reproducibly detectable amounts of both fucose and N-acetylgalactosamine, less than 1/10 of the relative amount of hexoses, and is characterized by the presence of O-linked glycosylation. The high specific activity of the above recombinant EPO can be directly related to the peculiarities of its glycosylation.

 Biologically active EPO obtained by expression of cloned DNA sequences for the EPO of the present invention can be used for treatment. The amount of the active component, of course, will depend on the severity of the regimen, the chosen route of administration and the specific activity of the active EPO, and will ultimately be determined by the attending physician or veterinarian. This amount of active EPO, determined in each case by the attending physician, is referred to herein as an “effective amount of EPO for treatment”. For example, in the treatment of induced hypoproliferative anemia associated with chronic renal failure, an effective daily amount of EPO in a sheep was found to be 10 units / kg for 15-40 days. See Eschbach et al., J. Clin. Invest. 73: 434 (1984).

 Active recombinant EPO can be administered by any route appropriate to the conditions used. Injection of EPO into the bloodstream of a mammal is preferred. The specialist, however, it is clear that the preferred route of administration varies depending on the selected mode.

 Active EPO can be administered as a pure (or substantially pure) compound, but it is preferable to have it in the form of a pharmaceutical composition or preparation.

 The compositions of the present invention for both veterinary and human use contain an active recombinant EPO, as described above, together with one or more pharmaceutically acceptable carriers and, if necessary, other therapeutic agents. The carrier (s) may be "acceptable" in the sense of being compatible with other components of the composition and harmless to its recipient. Preferably, the composition does not include an oxidizing agent and other substances. Compositions can be obtained by any of the methods well known in the pharmaceutical field. All methods include the step of binding the active component to a carrier that comprises one or more accessory components. Basically, formulations are prepared by uniformly and thoroughly mixing the active component with liquid carriers, or finely divided solid carriers, or both immediately, after which, if necessary, the product is formed.

 Compositions suitable for parenteral administration include sterile aqueous solutions of the active ingredient with solutions, preferably isotonic. Such compositions can be prepared by dissolving the solid active component in water to obtain an aqueous solution. Giving sterility to the specified solution can be achieved in disposable or multi-dose containers, for example, sealed ampoules or tubes.

 For a better understanding of the present invention, the actual scope of which is set forth in the attached claims, the following examples are given. It is clear that in the proposed methods, modifications can be made within the essence of the present invention.

Examples
Example 1: Isolation of the genomic clone of EPO.

EPO is purified from the urine of patients suffering from aplastic anemia, mainly as previously described (Miyake et al., J. Biol. Chem., 252: 5558 (1977)), except that the phenolic treatment is replaced by heat treatment at a temperature of 80 ^o C for 5 minutes to inactivate neuraminidase. The final purification step is fractionation on a C-4 Vydac column with high performance liquid chromatography (Separation Group) using an acetonitrile gradient from 0 to 95% with 0.1% trifluoroacetic acid (TFA) for 100 minutes. The position of the EPO in the gradient is determined by gel electrophoresis and analysis of the N-terminal sequence [21, 26, 27] of the main peaks. EPO elute at 53% acetonitrile. The fractions containing EPO were evaporated to 100 μl, adjusted to pH 7.0 with ammonium bicarbonate, digested completely with 2% TRSC-treated trypsin (Worthington) for 18 hours at 37 ^° C. Tryptic digest was then subjected to reverse phase HPLC as described above. Optical density is determined at 280 and 214 nm. Well-separated peaks are evaporated almost dry and directly analyzed for the N-terminal amino acid sequence [59] using a Model 480F gas phase sequencer Applied Biocyctems. The obtained sequences are given in table. 2 and 3. As described above, two of these tryptic fragments are selected for the synthesis of oligonucleotide probes. From the sequence Val-Asn-Phe-Tyr-Ala-Trp-Lys (amino acids 46 - 52 in Tables 2 and 3), a 17-membered member with 32-fold degeneracy is obtained.

и 18-членник с 128-кратной вырожденностью

Из последовательности Val-Tyr-Ser-Asn-Phe-Leu-Arg (аминокислоты 144 - 150 в табл. 2 и 3) получают два пула 14-меров, каждый - с 32-кратной вырожденностью.

and 18-member with 128-fold degeneracy

From the sequence Val-Tyr-Ser-Asn-Phe-Leu-Arg (amino acids 144 - 150 in Tables 2 and 3), two pools of 14 measures are obtained, each with 32-fold degeneracy.

которые отличаются в первом положении лейцинового кодона. Олигонуклеотиды помечают на 5'-конце ³²P, используя полинуклеотидкиназу (New England Biolabs) и гамма ³²P-ATP (New England Nuclear). Удельная активность олигонуклеотидов варьирует между 1000 и 3000 дюйм³/ммол олигонуклеотида. Геномную ДНК-библиотеку человека в бактериофаге лямбда (Lawn и др., [22]) подвергают скринингу с использованием модификации методики амплификации in situ, первоначально описанной Woo и др. [47]. Приблизительно 3,5 • 10⁵ фагов высевают с плотностью 6000 фагов на 150 мм чашку Петри (среды NZCYM) и инкубируют при температуре 37^oC до тех пор, пока пятна не станут видимыми, однако маленькими (приблизительно 0,5 мм). После охлаждения при температуре 4^oC в течение 1 ч дупликатные реплики картин пятен переносят на найлоновые мембраны (New England Nuclear) и инкубируют в течение ночи при температуре 37^oC на чашках со свежими средами NZCYM. Затем денатурируют и нейтрализуют флотированием каждого в течение 10 мин на тонкой пленке 0,5H NaOH - 1 M NaCl и 0,5 M Tris (pH 8) - 1M NaCl, соответственно. Вслед за вакуумной сушкой при температуре 80^oC в течение 2 часов фильтраты промывают в 5 x SSC, 0,5% SDS в течение 1 часа, и клеточные остатки на поверхности фильтров удаляют путем осторожного соскоба сухой тканью. Это соскабливание снижает фоновое связывание зонда с фильтрами. Фильтры затем промывают водой и предварительно гибридизируют от 4 до 8 ч при температуре 48^oC в 3М растворе тетраметиламмонийхлорида, 10 мМ NaPO₄ (pH 6,8), 5 x D-Enhardt's, 0,5% SDS и 10 мМ ЭДТА. 17-членник, меченый ³²P, затем прибавляют при концентрации 0,1 пмол/мл и осуществляют гибридизацию при температуре 48^oC в течение 72 ч. Затем фильтры последовательно промывают в 2 • SSC (0,3 M NaCl - 0,03 M цитрат натрия, pH 7) при комнатной температуре, в течение 1 ч в 3 M TMAC1 - 10 мМ NaPO₄ (pH 6,8) при комнатной температуре и от 5 до 15 мин - при температуре гибридизации. Приблизительно 120 сильных дупликатных сигналов обнаруживают после 2-дневной авторадиографии. Положительные сигналы отбирают, группируют в пулы по 8, пересеивают и вновь подвергают скринингу в трипликате с использованием 1/2 пула 14-членников на каждом из двух фильтров и 17-мера - на третьем фильтре. Условия и 17-членник для высевания и гибридизации такие же, как описано выше, за исключением того, что гибридизацию для 14-членника осуществляют при температуре 37^oC. Вслед за авторадиографией 17-членный зонд удаляют с фильтра в 50%-ном формамиде в течение 20 минут при комнатной температуре, а фильтр повторно гибридизируют при температуре 52^oC с 18-членным зондом. Два независимых фага гибридизируются со всеми тремя зондами. ДНК одного из этих фагов (обозначен здесь λ-HEPO1) переваривают полностью с помощью San 3A и субклонируют в М13 для анализа ДНК-последовательности с использованием дидезокси-метода обрыва цепи, предложенного Sanger и Conlson, [23]. Нуклеотидная последовательность и выведенная аминокислотная последовательность открытой рамки считывания, кодирующей триптический фрагмент ЭПО (подчеркнутая область), описаны здесь. Интронные последовательности даны в строчных буквах, экзонные последовательности (87 нуклеотидов) даны в прописных буквах. Последовательности, которые согласуются с сайтами акцептора сращивания (a) и донора (d), подчеркнуты (см. табл. 4).

which differ in the first position of the leucine codon. Oligonucleotides are labeled at the 5'-end of ³² P using polynucleotide kinase (New England Biolabs) and gamma ³² P-ATP (New England Nuclear). The specific activity of oligonucleotides varies between 1000 and 3000 inch ³ / mmol of oligonucleotide. The human genomic DNA library in the bacteriophage lambda (Lawn et al., [22]) is screened using a modification of the in situ amplification technique originally described by Woo et al. [47]. About 3.5 x 10 ⁵ phages are seeded at a density of 6000 phages per 150 mm Petri dish (NZCYM media) and incubated at 37 ^° C until the spots become visible, but small (approximately 0.5 mm). After cooling at 4 ^° C for 1 h, duplicate replicas of stain patterns were transferred onto nylon membranes (New England Nuclear) and incubated overnight at 37 ^° C in plates with fresh NZCYM media. Then denatured and neutralized by flotation of each for 10 min on a thin film of 0.5 N NaOH - 1 M NaCl and 0.5 M Tris (pH 8) - 1 M NaCl, respectively. Following vacuum drying at 80 ^° C. for 2 hours, the filtrates were washed in 5 x SSC, 0.5% SDS for 1 hour, and cell debris on the surface of the filters was removed by careful scraping with a dry cloth. This scraping reduces the background binding of the probe to the filters. The filters are then washed with water and pre-hybridized for 4 to 8 hours at 48 ^° C. in a 3M solution of tetramethylammonium chloride, 10 mM NaPO ₄ (pH 6.8), 5 x D-Enhardt's, 0.5% SDS and 10 mM EDTA. The 17-member labeled with ³² P is then added at a concentration of 0.1 pmol / ml and hybridized at 48 ^° C. for 72 hours. Then the filters are washed successively with 2 • SSC (0.3 M NaCl - 0.03 M sodium citrate, pH 7) at room temperature, for 1 h in 3 M TMAC1 - 10 mM NaPO ₄ (pH 6.8) at room temperature and from 5 to 15 minutes at a hybridization temperature. About 120 strong duplicate signals are detected after 2-day autoradiography. Positive signals are selected, grouped into 8 pools, re-screened and screened again in triplicate using 1/2 pool of 14-member on each of the two filters and 17-measure on the third filter. The conditions and 17-membered for sowing and hybridization are the same as described above, except that hybridization for the 14-membered is carried out at a temperature of 37 ^o C. Following autoradiography, the 17-membered probe is removed from the filter in 50% formamide in for 20 minutes at room temperature, and the filter is re-hybridized at a temperature of 52 ^o C with an 18-membered probe. Two independent phages hybridize with all three probes. The DNA of one of these phages (designated here as λ-HEPO1) is completely digested with San 3A and subcloned into M13 for DNA sequence analysis using the dideoxy chain termination method proposed by Sanger and Conlson [23]. The nucleotide sequence and the deduced amino acid sequence of an open reading frame encoding a tryptic fragment of EPO (underlined region) are described herein. Intron sequences are given in lowercase letters, exon sequences (87 nucleotides) are given in capital letters. Sequences that are consistent with the sites of the splice acceptor (a) and the donor (d) are underlined (see table 4).

 Example 2: Northern — analysis of human fetal liver mRNA.

5 μg of human fetal liver mRNA (20-week) and adult liver mRNA are electrophoresed on a 0.8% agarose formaldehyde gel and transferred onto nitrocellulose using the method of Derman et al., Cell, 23: 731 (1981). From M13 containing the insert shown in table 1, then get a single-stranded probe. The primer is a 20-member originating from the same tryptic fragment as the original 17-member probe. The probe was prepared as described by Anderson et al., PNAS, [49], except that following Smal cleavage (which produces the desired 95-nucleotide probe containing 74 nucleotides of the coding sequence), the small fragment was purified from M13 by column chromatography with Sepharose C14B and 0.1 N NaOH — 0.2 M NaCl. The filter is hybridized with a probe to approximately 5 • 10 ⁶ counts per minute for 12 hours at 68 ^° C, washed in 2 x SSC at 68 ^° C and exposed to a reinforcing screen for 6 days. A single-stranded marker mRNA of 1200 nucleotides (indicated by an arrow) moves along an adjacent "path" (Fig. 1).

 Example 3: cDNA from a fetal liver.

A probe identical to that described in Example 2 was prepared and used to screen a fetal liver cDNA library obtained in the λ-Ch21A vector (Toole et al., Nature, [25]) (1984) using standard spot screening techniques (Berten, Davis, Science, [53] (1978)). Three independent positive clones (designated here are λ-HEPOF L6, 1350 base pairs; λ-HEPOF L8, 700 bp and λ-HEPOF L13, 1400 bp) were isolated after screening 1 × 10 ⁶ spots. The complete inserts of the λ-HEPOF L13 and λ-HEPOF L6 clones were completely sequenced after subcloning into M13 (Tables 7 and 5, respectively). The λ-HEPOF L8 was partially sequenced; the remaining regions were assumed identical to the other two clones (Table 6). 5'- and 3'-untranslated sequences are represented by lowercase letters. The coding region is in capital letters.

 In relation to the table. 2 and 3, it should be mentioned that the deduced amino acid sequence shown below the nucleotide sequence is numbered starting at 1 for the first amino acid of the mature protein. The leader peptide is in capital letters. The cysteine residues in the mature protein are additionally indicated by SH, and potential N-linked glycosylation sites are indicated by an asterisk. Amino acids that are underlined mark those residues that are identified directly by sequencing the N-terminus or by sequencing tryptic fragments of EPO as described in Example 1. Partial underlining indicates residues in the amino acid sequence that could not be uniquely determined. The cDNA clones λ-HEPOF L6, λ-HEPOF L8 and λ-HEPOF L13 are deposited and are available from the American Type Culture Collection, Rockville, Maryland, with accession numbers ATCC 40156, ATCC 40152 and ATCC 40153, respectively.

 Example 4: EPO gene structure.

 The relative sizes and positions of four independent genomic clones (λ-HEPO1,2,3 and 6) from the Hae-111 / AluI library are illustrated by overlapping lines in FIG. 3. The thickened line indicates the position of the EPO gene. The scale (in Kb) and positions of known restriction endonuclease cleavage sites are given. The region containing the EPO gene is completely sequenced for both strands using directed exonuclease 111-generated series of deletions through this region. A schematic representation of five exons encoding EPO mRNA is shown in FIG. 4. The exact 5'-boundary of exon 1 is currently unknown. Protein-coding exon regions are shaded. The complete nucleotide sequence of exons is shown in Table 4. The known limits of each exon are outlined by solid vertical lines. The genomic clones λ-HEPO1, λ-HEPO2, λ-HEPO3 and λ-HEPO6 are deposited and are available from the American Type Culture Collection, Rockville, Maryland, under the accession numbers ATS 40154, ATCC 40155, ATCC 40150 and ATCC 40151, respectively.

 Example 5: Construction of the vector p91023 (B).

 The vector for transformation is pAdD26sVpA (3), described by Kaufman et al., Mol. Cell Biol., 2: 1304 (1982). The construction of this vector is shown in FIG. 5. This plasmid contains the mouse dihydrofolate reductase cDNA gene (DFHP), which is under transcriptional control of the main late promoter of adenovirus 2 (Ad2). The 5'-splicing site is indicated in adenoviral DNA, and the 3'-splicing site derived from the immunoglobulin gene is present between the main late promoter of adenovirus 2 and the DFHP coding sequence. The SV40 early polyadenylation site is present downstream of the DFHP coding sequence. The prokaryotic region of pAdD26-SVpA (3) originates from psVod / Mellon et al., Cell, 27: 279 (1981) and does not contain pBR322 sequences known for inhibiting replication in mammalian cells (Lusky et al., Nature, 293 : 79 (1981).

 As shown in FIG. 5, pAdD26sVpA (3) is first transferred to the plasmid pAdD26sVpA (3) (d) by deletion of one of the two Pst 1 sites. This is done by partial digestion of Pst 1 using enzyme deficiency so that a subpopulation of linearized plasmids is obtained in which there is only one site Pst 1 is cleaved, followed by processing of the Klenov fragments, crosslinking for recycling, and screening for division of the Pst 1 site located at the 3'-position with respect to the SV40 polyadenylation sequence.

In the plasmid pAdD26SVpA (3) (d), the three-part adenovirus leader and virus-associated genes (VA genes) are inserted. pAdD26SVpA (3) (d) cleave Pvull to create a linear molecule. Then pJAW 43 / zatn et al., Cell, 16: 851 (1979)) digested Xhol, treated with a Klenov fragment, digested PvuIl, and a 140 bp fragment. secreted by acrylamide gel electrophoresis [6] in Tris-borate buffer (Maniatis et al., supra). Fragment 140 bp then crosslinked with the PvuII-digested plasmid pAdD26SVpA (3) (d) (Fig. 6). The crosslinking product is used to transform E. coli with resistance to tetracycline; colonies were screened using the Grunstein-Hoghess technique. Using a ³² P labeled probe hybridizing with a 140 bp fragment, positive colonies are obtained from which DNA is isolated to test whether the reconstructed PvuII site is the 5 ′ or 3 ′ end of the inserted 140 bp DNA. The correct orientation of the PvuII site is on the 5'-side of the 140 bp insert. This plasmid is designated pTPL in FIG. 6.

 The SV40 virus DNA fragment containing the SV40 enhancer sequence is obtained by digesting the SV40 Ava 11 DNA, blunting the ends with a Klenov fragment, stitching the Xho 1 linkers with the fragments from Xho 1 digestion and isolating the largest fragment (D) by gel electrophoresis. This fragment is then crosslinked with pTPL digested with Xho 1 to obtain plasmid pCYSVL 2-TPL. The orientation of the SV40 fragment D in pCYSVL 2-TPL is such that the late promoter of the SV40 virus is in the same orientation as the main late promoter of the adenovirus.

 To introduce the adenovirus-related genes (VA genes) into pCYSVL 2-TPL, the plasmid pBR322 is first constructed which contains a fragment of type 2 adenovirus Hind 111. Type 2 adenovirus DNA is digested with Hind 111 and fragment B is isolated by gel electrophoresis. This fragment is inserted into the plasmid pBR322, which was previously digested with Hind 111. After transformation of E. coli, ampicillin-resistant recombinants are screened for insertion of the fragment in Hind 111; the orientation of the insert is determined by digestion with a restriction enzyme. pBR322 - Ad Hind 111 B contains the B fragment Hind 111 of type 2 adenovirus in the orientation shown in FIG. 7.

 As shown in FIG. 8, VA genes are obtained from the plasmid pBR322 - Ad Hind 111 B by digestion of Hpa 1, addition of EcoRI linkers and digestion of EcoRI followed by restoration of the 1.4 kb fragment. A fragment having the sticky ends of EcoRI is then sutured into the EcoRI site of pPTL, previously digested by EcoRI. After transformation of E. coli HB101 and selection of transformants for tetracycline resistance, the colonies are screened by filter hybridization with DNA specific for the VA genes. DNA is obtained from positively hybridizing clones and characterized by restriction endonuclease digestion. The resulting plasmid is designated p91023.

 As shown in FIG. 9, two EcoR1 sites in plasmid p91023 are cleaved by complete cleavage of p91023 with EcoRI and restoration of two DNA fragments (about 7 kb and about 1.3 kb). The last fragment contains VA genes. The ends of both fragments are filled using a Klenov fragment, and the two fragments are then crosslinked to one another. Plasmid p91023 (A), which contains VA genes and is similar to plasmid p91023, but which has lost the interstitial fragment for two EcoRI sites, is identified by screening Gruns tein - Hogness of VA gene fragments, as well as traditional restriction analysis.

 One Pst 1 site in plasmid p91023 (A) is cleaved and replaced with the EcoRI site. p91023 (A) is completely digested with Pst 1 and treated with a Klenov fragment. EcoRI linkers are crosslinked with the blunt site Psr1 of plasmid p91023 (A). The linear plasmid p91023 (A) with EcoRI linkers attached to the blunt site Pst1 is separated from the non-crosslinked linkers and EcoRI is completely digested, and then crosslinked. Plasmid p91023 (B), as shown in FIG. 9 are restored and identified as having a structure similar to plasmid p91023 (A), but containing the EcoRI site instead of the previous Pst1 site. Plasmid p91023 (B) has been deposited and is available from the American Type Culture Collection, Rockville, Maryland, accession number ATCC 39754.

 Example 6: Expression of EPO cDNA in CO5 cells.

The cDNA inserts from λ-EPOF L6 and λ-EPOF L13 are inserted into plasmid p91023 (B), forming pRTF L6 and pRTF L13, respectively. 8 μg of each of the purified DNA is then used to transfect 5 • 10 ⁶ COS cells using the DEAE-dextran method (below). After 12 hours, the cells were washed and treated with Chloroquine (0.1 mM) for 2 hours, washed again and incubated for 24 hours in media (10 ml) containing 10% fetal calf serum. The media are replaced with new ones (4 ml each), free of serum, and collected after 48 hours.

 EPO is quantified by radioimmunoassay as described by Sherwood and Goldwasser [54]. Antibody delivered by Dr. Judith Sherwood. An iodinated isotopic indicator is obtained from a homogeneous EPO as described in Example 1. The sensitivity of the sample is approximately 1 ng / ml. The results are shown below in table 8.

 pPTF L13 is deposited and is available from the American Type Culture Collection, Rockville, Maryland, with accession number ATCC 39990.

 Example 7: Expression of EPO cDNA in COS cells.

 cDNA from λ-HEPOF L13 is transferred to plasmid p91023 (B), transformed into COS-1 cells and harvested as described above (Example 6), except that the chloroquine treatment is omitted.

The biological activity of EPO is measured in vitro using either a colony-forming sample with mouse fetal liver cells as a source of CFV-E, or a ³ H-thymidine uptake method using spleen cells from mice injected with phenylhydrazine. The sensitivity of these assays is approximately 25 mU / ml. In vivo biological activity is measured using either the hypoxic mouse method or the starving rat method. The sensitivity of these assays is approximately 100 mU / ml. With a simulated conditioned medium, activity was not detected in any case. The results of the study of EPO expressed by the clone EPOF L13 are shown below in table. 9. The above activities are expressed in units / ml, and using a quantitatively determined EPO as a trading standard (Toyobo, 1 ps.).

 Example 8: Analysis of EPO from COS cells in SDS PAA gel.

180 ng of EPO obtained from the medium of COS cells transfected with cDNA (from λ-HEPOF L13) as part of vector 91023 (B) (above), is subjected to electrophoresis in 10% PAA gel in the presence of SDS (according to Laewlli) and transferred to nitrocellulose paper (Towbin et al., Proc. Natl. Acad. Sci. USA, 76: 4350 (1979)). The filter is probed with an anti-EPO antibody as described in table. 8, washed and re-probed with protein A ¹²⁵ J. The filter is autoradiographed for two days. The native homogeneous EPO described in Example 1 is also subjected to electrophoresis either before (lane B) or after iodination (lane C) (see FIG. 10). Used markers include methionine labeled with ³⁵ S, serum albumin (68000 d) and egg albumin (45000 d).

 Example 9: Construction of pRKI-4.

 From plasmid PSV2DHFR / Subramani et al., Mol Cell. Biol. 1: 854-864 (1981)) isolate a BamHI-Pvule fragment containing the early SV40 promoter adjacent to the mouse dihydrofolate reductase gene (DHFR), SV40 enhancer, small antigenic intron t, and SV40 polyadenylation sequence (fragment A). Other fragments are obtained from the vector p91023 (A) (above) as follows: p91023 (A) is digested with Pst 1 in a single Pst1 site near the adenovirus promoter to linearize the plasmid and either crosslinked with Pst1 - EcoRI synthetic converters and cyclized (creating Pst 1 - EcoR1 sites - Pst 1 at the initial site of Pst1 p91023 (B ')), either treated with a large fragment of DNA polymerase 1 to destroy the Pst1 sites and crosslinked with a synthetic EcoRI linker and recycled (creating an EcoRI site at the initial site of Pst1 p91023 (B). Each of the two obtained plasmids p91023 (B) and p91023 (B ') digest Xba and E coRI to produce two fragments (F and G) By combining fragment G from plasmid p91023 (B) and fragment F from plasmid p91023 (B '), as well as fragment G from plasmid p91023 (B) and fragment F from plasmid p91023 (B) ') create two new plasmids that contain either the EcoRI site - Pst1 or the Pst1 site - EcoRI. A vector where the Pst1 site is close to the main late adenovirus promoter is named p91023 (C).

 Vector p91023 (C) is completely digested with Xho1, and the resulting linearized sticky-ended DNA is blunted by filling the ends with a large fragment of E. coli DNA polymerase 1. A 340 bp Hind 111 - EcoRI fragment containing the SV40 enhancer, prepared as described below, is sewn onto this DNA.

 The Hind 111 fragment is Pvull from the SV 40 virus, which contains the beginning of replication, as well as an enhancer inserted into the plasmid with lac / Little et al., Mol.Biol. Med., 1: 473-488 (1983)), which was previously digested with BamHI, filled the sticky ends with a large fragment of DNA polymerase 1 and digested with Hind 111 DNA. In the resulting plasmid (c SVHPlac), the BamH1 site was regenerated by crosslinking with blunt end of Pvull. An EcoRI-Hind 111 fragment is obtained from the plasmid with SVHPlac and ligated to the EcoRI-Hind 111 pSV Od / Mellon et al. Fragment above), which contains the plasmid replication base. The resulting plasmid is designated pSVHPOd. An EcoRI - Hind 111 fragment of 340 bp is then obtained. plasmids pSVHPOd containing ori SV40 / enhancer are blunted at both ends using a large fragment of DNA polymerase 1 and crosslinked with the Xhol-digested, blunt-ended vector p91023 (C) described above. The resulting plasmid (p91023 (C) / Xho / plus EcoRI / Hind 111) in which the orientation of the Hind 111-EcoRI fragment is such that the BamH1 site within the fragment is adjacent to the VA gene, called pES105. Plasmid pES105 is digested with BamH1 and Pvull, as well as separately with Pvull, and the BamH1 - Pvull fragment containing the main late adenovirus promoter (fragment B) and the Pvull fragment containing the plasmid gene (tetracycline resistance / resistance) and other sequences (fragment are isolated) C) Fragments A, B and C are crosslinked and the resulting plasmid shown in FIG. 11 are isolated and designated pRKI-4. Plasmid pRKI-4 was deposited with the American Type Culture Collection Rockville, Maryland, where it is available under accession number ATCC 39940.

 Example 10: Expression of EPO cDNA in CHO Cells - Method 1.

 The DNA (20 μg) of the pRTF L 13 plasmid described above (Example 6) was digested with the restriction endonuclease Cla 1 and crosslinked with the Cla 1-digested DNA of the plasmid pAdD265SVp (A) 1 (2 μg), which contains the intact gene of dihydrofolate reductase (DHFR), driven by the main late promoter of adenovirus (Kaufman and Sharp, Mol.and Cell Biol., 2: 1304-1319 (1982)). This cross-linked DNA is used to transfect DHFR-negative CHO cells (DUKX-B11, Chasin L. A. and UrlaubG. (1980) PNAS 77: 4216-4220). After two days of cultivation, cells that included at least one DHFR gene were selected in nucleotide-free alpha media and replenished with 10% dialyzed fetal bovine serum. After cultivation in selective media, for two weeks, the colonies are removed, combined into groups of 10-100 colonies, re-plated and grown to confluence in alpha-media that do not contain nucleotides. Media from pools grown under methotrexate selection are tested for containing EPO by radioimmunoassay (RIA). Pools that show the presence of EPO are grown in the presence of methotrexate (0.02 μM). The Cla strain 4.4.02 EPO subcloned from the positive pool releases 460 ng-ml EPO into the medium containing 0.02 μM MTX (Table 10). Cla 44.02-7 EPO is a selected cell line for EPO production, which is deposited by the American Type Culture Colection under the accession number ATCC CRL 8695. For pools that are negative with RIA, methotrexate-resistant colonies obtained from their equivalent cultures growing in the presence of methotrexate (0.02 μM), is again tested in pools for the presence of EPO by RIA. Those cultures that are not positive subclone and are cultured at even higher concentrations of methotrexate.

 Stepwise methotrexate (MTX) selection is achieved by repeated cell culture cycles in the presence of increasing concentrations of methotrexate and selection for surviving microorganisms. At each cycle, EPO is detected in the culture supernatant using RIA and in vitro biological activity analysis. The levels of methotrexate used in each step amplification are 0.02 μM, 0.1 μM and 0.5 μM. as shown in the table. 10, after the first round of selection at 0.02 μM MTX, significant levels of EPO are released into the culture media.

 Example 11: Expression of EPO cDNA in CHO cells — Method II.

 Clone λ-HEPOF L 13 DNA is digested with EcoRI and a small fragment of RI. containing the EPO gene, subcloned into the EcoRI site of plasmid pRKI-4 (see example 10). This recombinant DNA (pRKF L 13) is then used to transfect DHFR-negative CHO cells directly (without digestion), and selection and amplification are carried out as described in Example 10 above.

 RKF L 13 DNA is also inserted into CHO cells by fusion of protoplasts and microinjection. Plasmid pRKF L 13 has been deposited and is available from the American Type Culture Collection Rockvill, Maryland, accession number ATCC 39989.

 A preferred clone of a single colony is deposited and is available from the American Type Culture Collection Rockville, Maryland, under accession number TCC CRL 8695.

 Example 12: Expression of a Genomic EPO Clone in COS-1 Cells

 The vector used for expression of the genomic clone of EPO is pSVOd / Mellon et al. (Above). DNA from pSVOd is completely digested with Hind 111 and blunted using a DNA polymerase fragment 1. The genomic λ-HERO3 clone is completely digested with EcoRI and Hind 111, and a 4.0 kb fragment containing the EPO gene is isolated and blunted as described above. The nucleotide sequence of this fragment from the Hind 111 site to the region above the polyadenylation signal is shown in FIG. 4 and tab. 4. A fragment of the EPO gene is inserted into the plasmid fragment of psvod, and correctly constructed recombinants in both orientations are isolated and checked. Plasmid pCZ2-1 has the EPO gene in orientation “a” (ie, with the 5'-end of the EPO gene close to the beginning of SV 40), and plasmid pCZ1-3 contains the gene in the opposite orientation (orientation “b”).

 Plasmids pCZ1-3 and pCZ2-1 are transfected into COS-1 cells, as described in example 7, the media are collected and analyzed for the presence of immunologically reactive EPO. Approximately 31 ng / ml EPO was detected in the settled cultures after administration of pCZ2-1 and 16-31 ng / ml after administration of pCZ1-3.

 Coding fragments of the genomic clones HEPO1, HEPO2 and HEPO6 can be inserted into COS cells for expression in a similar manner.

 Example 13: Expression in C127 and ZTZ Cells.

a. Construction of pBPYEPO
A plasmid containing the cDNA sequence of EPO under the transcriptional control of the mouse metallothionein promoter and linked to the complete DNA of the bovine papilloma virus is obtained as follows.

pEPO49f
Plasmid pSP6 / 5 purchased by Promeda Biotec. This plasmid is completely digested with EcoRI. 1340 bp EcoRI fragment from λ-HEPOF L 13 is inserted into it using DNA ligase. The resulting plasmid, in which the 5 ′ end of the EPO gene is closest to the SP6 promoter (as determined by digestion of BglI and Hind 111), is designated pEPO49f. In this orientation, the BamH1 site in the pSP6 / 5 polylinker is directly adjacent to the 5 ′ end of the EPO gene.

pMMTneo BRY
The plasmid pdBPV-MMTneo (342-12) (Law et al., Mol. And Cell Biol., 3: 2110-2115 (1983), shown in Fig. 12, is digested until BamH1 is ready to obtain two fragments: a large fragment of size ≈ 8 kb containing the ASR genome and a smaller fragment of ≈ 6.5 kb containing pML2 start of replication and ampicillin resistance, metallothionein promoter, neomycin resistance and SV40 polyadenylation signal. The digested DNA is recycled with DNA ligase, and plasmids that contain only 6.8 kb fragment identified by restriction endonucleic digestion EcoRI and BamH1, one such plasmid is designated pMMTneo BPY.

pEPO15a
pMMTneo BPY completely digested BglII. pEPO49f is digested until ready to BamH1 and BglII and receive in a gel of 700 BP a fragment containing the whole EPO coding region. BglII-digested pMMTneo BPY and a 700 bp BamH1 / BglII fragment crosslinked, and the resulting plasmids are identified by colony hybridization using an oligonucleotide d / GGTCATCTGTCCCCTGTCC) probe that is specific for the EPO gene. Of the plasmids that are positive hybridization data, one (pEPO15a) that has EPO cDNA in orientation, such that the 5 ′ end of EPO cDNA is adjacent to the metallothionein promoter, is identified by digestion of EcoR1 and Kpn1.

pBPY-EPO
Plasmid pEPO15a was digested with BamH1 for linearization. Plasmid pdBPY-MMTneo (342-12) was also completely digested with BamH1 to give two fragments of size 6.5 and 8 kb. A fragment that contains the whole genome of bovine papilloma virus is isolated in a gel. Cleaved pEPO15a and a 8 kb fragment are cross-linked to the pBPY-EPO plasmid, which is identified by colony hybridization using the oligonucleotide probe d / P-CCACACCCGGTACACA-OH) specific for the BPY genome. Digestion of pBPY-EPO DNA using Hind 111 indicates that the transcription direction of the BPY genome is the same as the transcription direction of the metallothionein promoter (similar to pdBPY-MMTneo / 342-12 /). Plasmid pdBPY-MMTneo / 342-12 / is available from the American Tire Culture Collection, Rockville, Maryland, under accession number ATCC 37224.

b. Expression
Method 1
PBPY-EPO DNA is prepared and approximately 25 μg is used to transfect 1 x 10 ⁶ cells of C127 / Lowy et al., J.of Virol., 26: 291-298 (1978) using standard calcium phosphate precipitation techniques (Crahm et al. ., Virology, 52: 456-467 (1973)). Five hours after transfection, the medium is removed, the cells are shocked with glycerol, washed and fresh α-medium containing 10% fetal bovine serum is added. After forty-eight hours, the cells are trypsinized, diluted 1: 10 with DME medium containing 500 μg / ml G418 (Sonthern et al., Mol. Appl. Genet., 1: 327-341 (1982)), and grown for two three weeks. G418-resistant colonies are then transferred individually to microtiter wells and grown in the presence of G418. The cells are then washed, fresh media containing 10% fetal bovine serum are added, which are harvested after 24 hours. Conditioned media test and detect the presence of EPO by radioimmunoassay and in vitro biological assay.

Method II
C127 or ZT3 cells cotransfect 25 μg of pBPY-EPO and 2 mg of psvrueo (Souther et al., Supra) as described in Method I. This represents an approximately 10-fold molar excess of pBPY-EPO. Following transfection, a procedure similar to that described in method I is used.

Method III
C127 cells transfect 30 μg pBPY-EPO as described in method I. Following transfection and dilution (1:10), fresh media are changed every three days. After approximately 2 weeks, foci of BPY-transformed cells appear. Separate foci are collected in 1 cm of a microtiter plate nest, grown to a sublayer monolayer, and analyzed for the presence of EPO using an immunoassay and a study of biological activity.

C. In vitro determination of the activity of recombinant EPO expressed in C127 and 3T3 cells
EPO expressed in C127 and 3T3 cells by method II was tested for biological activity. The biological activity of purified in vitro recombinant EPO was analyzed by G. Krystal, Exp. Hematol. 11: 649 (1983) (bioanalysis of spleen cell proliferation) with protein determination based on amino acid data. According to numerous definitions, the specific in vitro activity of recombinant EPO purified from C127 cells was 267,000 units / mg protein. The specific in vitro activity of recombinant EPO purified from 3T3 cells was 244,000 u / mg protein.

 In both cases, significant levels of N-acetylgalactosamine were determined by two independent methods (Reinhold, Methods in Enzymol. 50: 244-249 (Methanolysis) and Take moto H. et al., Anal. Biochem. 145: 245 - (1985)), which indicates O-linked protein glycosylation.

 Example 14: Purification of EPO.

 Conditioned media after culturing recombinant vector transformed COS cells containing approximately 200 μg / L EPO are concentrated using ultrafiltration membranes for a molecular weight of 10,000, such as, for example, Millipoze Pellican with a membrane of 0.465 square meters, and then diafiltered against 10 mM buffer solution of sodium phosphate at pH 7. Concentrated and diafiltered conditioned media contain approximately 2.5 mg of EPO in 380 mg of total protein. The EPO solution was further concentrated to 186 ml, and the precipitated proteins were recovered by centrifugation at 110,000 x g for 30 minutes.

The pH of the supernatant that contains EPO is adjusted to 5.5 with 50% acetic acid, left at 4 ^° C for 30 minutes, and the precipitate formed is separated by centrifugation at 13,000 xg for 30 minutes.

Carboxymethyl Sepharose Chromatography
A centrifugation supernatant containing approximately 200 μg EPO (24 mg total protein) was applied to a column packed with CM-Sepharose (20 ml), equilibrated with 10 mM sodium acetate, pH 5.5, washed with 40 ml of the same buffer solution. EPO associated with CM-Sepharose was eluted with a 100 ml NaCl gradient (0-1M) in a 10 mM sodium phosphate solution, pH 5.5. EPO-containing fractions (total 50 μg) were combined and concentrated to 2 ml using an Americon YM10 ultrafiltration membrane.

Reverse phase HPLC
Concentrated CM-Sepharose fractions containing EPO were further purified by reverse phase HPLC using a Vydac C-4 column. EPO is fed to a column balanced with 10% solvent B (solvent A represents 0.1% CF ₃ CO ₂ H in water; solvent B represents 0.1% CP ₃ CO ₂ H in CF ₃ CN), with a low flow rate 1 ml / min. The column is washed with 10% B for 10 minutes, and the EPO is eluted with a linear gradient of B (10 - 70% for 60 minutes). Fractions containing EPO are combined. Get ≈ 40 μg EPO per 120 μg of total protein; the solution is lyophilized. Lyophilized EPO was redissolved in 0.1 M Tris-HCl at pH 7.5 containing 0.15 M NaCl and re-subjected to reverse phase HPLC. Fractions containing EPO are pooled and analyzed by SDS-polyacrylamide (10%) gel electrophoresis (Lamelli UK, Nature). The combined fractions of EPO contain 15.5 μg of EPO in 25 μg of total protein. In a similar manner, the EPO obtained using other types of mammalian cells according to the present invention is purified.

 Example 15: Preparation of a pharmaceutical composition.

In a sterile, 100-liter V2A double-walled boiler equipped with a stirrer, excipients are dissolved:
Urea - 700.0 g
Sodium Chloride - 70.0 g
Tween 20 (Tween 20) - 7.0 g
Monosubstituted sodium phosphate • 1 H ₂ O - 38.4 g
Disubstituted sodium phosphate • 2 H ₂ O - 350.0 g
Calcium Chloride • 2 H ₂ O - 8.4 g
Glycine - 105.0 g
L-Leucine - 140.0 g
L-Isoleucine - 140.0 g
L-Threonine - 35.0 g
L-Glutamic acid - 35.0 g
L-Phenylalanine - 70.0 g
Water for injection purposes - 70.0 g
A portion of 214.3 ml of erythropoietin feed with a titer of 140,000 units / ml is added to 30 l of this excipient solution and the final volume is 35 l and thoroughly mixed. The remainder of the excipient solution is washed with a filter system. The solution of the initial mixture is sterile filtered on a membrane filter with a pore size of 0.2 μm. The sterile filtered solution is poured into 1-ml vials under aseptic conditions and freeze-dried in a lyophilization apparatus in compliance with the following criteria:
Freezing time: 12-14 hours at -40 ^o C.

Basic drying: temperature of the solution +10 ^o C, pressure 10 ^-1 mbar, time 48 - 60 hours.

Additional drying: solution temperature +20 ^o C, pressure 10 ^-3 mbar, time 4 to 6 hours.

 In this way, a voluminous, open-pore, dry substance for injection is obtained, which remains for at least 2 years in the refrigerator and 1 year at room temperature and dissolves for several hours in 2 ml of water for injection purposes without clouding and free from particles.

 For a composition thus prepared, a therapeutic effect is indicated.

Sources of information
1. Jacobson, LO, Goldwasser, E. Fried, W., and Pizak, LF, Trans. Assoc. Am. Physicians TO: 305-317 (1957).

 2. Krantz, S. B. and Jacobson, L. O. Chicago; University of Chicago Press 1970, pp. 29-31.

 3. Hammond, D and Winnick, S. Ann. N.Y. Acad Scl. 230: 219-227 (1974).

 4. Sherwood, J. B. and Goldwasser, E., Endocrinology 103: 866-870 (1978).

 5. Fried, W. Blood 40: 671-677 (1972).

 6. Fisher, J. J. Lab. and clin. Med. 93: 695-699 (1979).

 7. Naughton, B. A., Kaplan, S. M., Roy, M., Burdowski, A. J., Gordon, A. S., and Piliero, S. J. Science 196: 301-302.

 8. Lucarelli, G. P., Howard, D., and Stohlman, F., Jr. J. Clin. Invest 43: 2195-2203 (1964).

 9. Zanjani, E. D., Poster, J., Burlington, H., Mann, L. I., and Wasserman, L. R. J. Lab. Clin. Med. 89: 640-644 (1977).

 10. Krantz, S. B., Gallien-Lartigue, O., and Goldwasser, E. J. Biol Chem. 238: 4085-4090 (1963).

 11. Dunn, C. D., Jarvis, J. H. and Greenman, J. M. Exp. Hematol. 3: 65-78 (1975).

 12. Krystal, G. Exp. Hematol. 11: 649-660 (1983).

 13. Iscove, N.N. and Guilbert, L.J., M.J. Murphy, Jr. (Ed.) New York: Sprinqer-Verlaq, pp. 3-7 (1978).

 14. Goldwasser, E., ICN UCLA Symposium, Control of Cellular Division and Development, A. R. Liss, Inc., pp. 487 - 494 (1981).

 15. Cline, M. J. and Golde, D. W. Nature 277: 177-181 (1979).

 16. Metcalf, D., Johnson, G. R., and Burgess, A. W. Blood 55: 138- (1980).

 17. Krane, N. Henry Ford Hasp. Med. J. 31: 177-181 (1983).

 18. Eschbach, J., Madenovic, J., Garcia, J., Wahl, P., and Adamson, J. J. Clin. Invest. 74: 434-441 (1984).

 19. Anagnostou, A., Barone, J., Vedo, A., and Fried, W. Br. J. Hematol 37: 85-91 (1977).

 20. Miyake, T., Kung, C., and Goldwasser, E. J. Biol. Chem. 252: 5558-5564 (1977).

 21. Yanagawa, S., Hirade, K., Ohnota, H., Sasaki, R., Chiba, H., Veda, M., and Goto, M. J. Biol. Chem. 259: 2707-2710 (1984).

 22. Lawn, R. M., Fritsch, E. F., Parker, R. C., Blake, G., and Maniatis, T. Cell 15: 1157- (1978).

 23. Sanger, F., Nicklen, S., and Coulson, A. R. Proc. Nat'l. Acad Sci., U.S.A. 74: 5463 - (1977).

 24. Zanjanc, E. D., Ascensao, J. L., McGlave, P. B., Banisadre, M., and Ash, R. C. J. Clin. Invest. 67: 1183- (1981).

 25. Toole, J.J., Knopf, J.L., Wozney, J.M., Sultzman, L.A. Buecker, JL, Pittman, DD, Kaufman, RJ, Brown, E., Shoemaker, C., Orr, EC, Amphlett, GW, Foster, WB, Coe, ML, Knutson, GJ, Fass, DN, and Hewick, RM Nature in Press.

 26. Goldwasser, E. Blood Suppl. 1, 58, xlii (abstr) (1981).

 27. Sue, J. M. and Sytkowdki, A. J. Proc. Nat'l Acad. Sci U.S.A. 80: 3651-3655 (1983).

 28. Bersch, N. and Golde, D.W., In Vitro Aspects of Erythropoiesis, M. J. Murphy (Ed.) New York: Springer-VerIag (1978).

 29. Cotes, P. M. and Bangham, D. R. Nature 191: 1065 - (1961).

 30. Goldwasser, E. and Gross, M. Methods in Enzymol 37: 109-121 (1975).

 31. Nabeshima, Y. -i, Fujii-Kuriyama, Y., Muramatsu, M., and Ogata, K. Nature 308: 333-338 (1984).

 32. Young, R. A., Hagencuhle, O. and Schibler, U. Cell 23: 451-558 (1981).

 33. Medford, R. M., Nguyen, H. T., Destree, A. T., Summers, E. and Nadal-Ginard, B. Cell 38: 409-421 (1984).

 34. Ziff, E. B. Nature 287: 491-499 (1980).

 35. Early, P. Cell 20: 313-319 (1980).

 36. Sytkowski, A. Bio. Biop. Res. Comm. 96: 143-149 (1980).

 37. Murphy, M. and Miyake, T. Acta. Haematol. Jpn. 46: 1380-1396 (1983).

 38. Wagh, P. V. and Bahl, O. P. CRC Critical Reviews in Biochemistry 307-377 (1981).

 39. Wang, F. F., Kung, C. K. -H. and Goldwasser, E. Fed. Proc. Fed. Am. Soc. Exp. Biol. 42: 1872 (abstr) (1983).

 40. Lowy, P., Keighley, G. and Borsook, H. Nature 185: 102-103 (1960).

 41. VanLenten, L. and Ashwell, G. J. Blol. Chem. 247: 4633-4640 (1972).

 42. Lee-Huang, S. Proc. Nat'l Acad. Sci. U.S.A. 81: 2708-2712 (1984).

 43. Fyhrquist, F., Rosenlof, K., Gronhagen-Riska, C., Hortling, L. and Tikkanen, I. Nature 308: 649-562 (1984).

 44. Ohkubo, H., Kageyama, R., Vjihara, M., Hirose, T., Inayama, S., and Nakanishi, S. Proc. Nat'l Acad. Sci. U.S.A. 80: 2196-2200 (1983).

 45. Suggs, S.V., Wallace, R. B., Hirose, T., Kawashima, E. H. and Itakura, K. Proc. Nat'l. Acad Sci. U.S.A. - 78: 6613-6617 (1981).

 46. Woo, S. L. C., Dugaiczyk, A., Tsai, M. -J., Lai, E. C., Catterall, J. F. and O'Malley, B. W. Proc. Nat'l. Acad Sci. U.S.A. 75: 3688- (1978).

 47. Melchior, W. B. and Von Hippel, P. H. Proc. Nat'l Acad. Soc.U.S.A. 70: 298-302 (1973).

 48. Orosz, J. M. and Wetmis, J. G. Biopolymers 16: 1183-1199 (1977).

 49. Anderson, S and Kingston, I. B. Proc. Nat'l Acad. Sci.- U.S.A. 80: 6836-6842 (1983).

 50. Ullrich, A., Coussens, L., Hayflick, JS Dull, TJ, Gray, A., Tam, AW, Lee, J., Yarden, Y., Libermann, TA, Schlessinger, J., Downward, J ., Mayes, ELV, Whittle, H., Waterfield, MD and Seeburg, P. H. Nature 309: 418-425 (1984).

 51. Fisher, J. Proc. Soc. Exptl. Biol. and Med. 173: 289-305 (1983).

 52. Kozak, M. Nuc. Acid Res. 12: 857-872 (1984).

 53. Benton, W.D. and Davis, R.W. Science 196: 180-182 (1977).

 54. Sherwood, J.B. and Goldwasser, E. Blood 54: 885-893 (1979).

 55. Derman, E., Krauter, K., Walling, L., Weinberger, C., Ray, M., and Darnell, J. T., Cell 23: 731- (1981).

 56. Gluzman, Y., Cell 23: 175-182 (1981).

 57. Hewick, R. M., Hunkapiller, M. E., Hood, L. E., and Dreyer, W. J. J. Biol. Chem. 256: 7990-7997 (1981).

 58. Towbin, H., Stachelin, T., and Gordon, J., Proc. Nat'l Acad. Sci. 76: 4380- (1979).

 59. Carnott, P., DeFlandre, C. C. R. R. Acad. Sci. Paris 143: 432- (1960).

Claims (12)

 1. A DNA fragment encoding human erythropoietin (EPO) and having a sequence from ATG encoding the initial Met (position - 27), according to AGA encoding the terminal Arg (position 166) in the polynucleotide sequence described in the description.  2. A fragment of genomic DNA encoding human EPO and having a sequence from ATG encoding the initial Met (position 617), according to the AGA encoding the terminal Arg (position 2761) in the polynucleotide sequence described.  3. Recombinant DNA λ = HEPOFL 13 (ATCC 40153) containing the DNA sequence according to claim 1.  4. Recombinant λ-HEPO 1 DNA (ATCC 40154) containing the genomic DNA sequence of claim 2.  5. Recombinant λ-HEPO 2 DNA (ATCC 40155) containing the genomic DNA sequence of claim 2.  6. Recombinant λ-HEPO 3 DNA (ATCC 40150) containing the genomic DNA sequence of claim 2.  7. Recombinant λ-HEPO 6 DNA (ATCC 40151) containing the genomic DNA sequence of claim 2.  8. Cell line CHO EPO Cla 44.02-7 (ATCC CRL 8695) - producer of recombinant EPO.  9. Recombinant human O-glycosylated EPO obtained by culturing mammalian cells, encoded by DNA, according to claim 1 or 2 and having a specific activity of at least 200,000 units / mg.  10. EPO according to claim 9, characterized in that it is obtained by culturing CHO, 3T3 or C 127 cells.  11. EPO according to claim 10, characterized in that it is obtained by culturing the cell line CHO EPO Cla 44.02-7 (ATCC CRL 8695).  12. A pharmaceutical composition for treating EPO-deficient conditions, comprising an effective amount of erythropoietin as an active principle and pharmaceutically acceptable carriers, characterized in that the active principle of the composition is a recombinant human O-glycosylated EPO according to claim 9.

Images