Method for the Massive Culture of Cells Producing Recombinant Human Erythropoietin

Abstract

The present invention relates, in general, to a method for the massive culture of recombinant mammalian cells for the production of recombinant human erythropoietin (EPO) in culture medium containing insulin. The present invention also refers to a method of producing EPO and to the EPO thus produced.

Description

CROSS REFERENCE TO RELATED APPLICATIONS

This application is a continuation of U.S. application Ser. No. 09/830,968, § 371(c) Date: Nov. 6, 2001, which is the U.S. National Phase entry of International Application No. PCT/US99/26240, filed Nov. 8, 1999, which claims the benefit of the filing date of Argentinian Application No. P98 01 05611, filed Nov. 6, 1998, and of Argentinian Application No. P99 01 00681, filed Feb. 23, 1999.

BACKGROUND OF THE INVENTION

1. Field of the Invention

The present invention relates, in general, to a method for the massive culture of recombinant mammalian cells for the production of recombinant human erythropoietin (EPO) in a culture medium containing insulin. The present invention also refers to a method of producing EPO and to the EPO thus produced.

2. Background Information

EPO is a glycoprotein that stimulates erythroblast differentiation in the bone marrow, thus increasing the circulating blood erythrocyte count. The mean life of erythrocytes in humans is 120 days and therefore, a human being loses 1/120 erythrocytes each day. This loss must be continuously restored to maintain an adequate level of red blood cells.

The existence of EPO was first postulated by the turn of the century and was definitely proved by Reissman and Erslev early in the '50s. See Carnot, et al., C.R. Acad. Sci. (France), 143, 384-6 (1906); Carnot, et al., C.R. Acad. Sci. (France), 143, 432-5 (1906); Carnot, et al., C.R. Soc. Biol., 111, 344-6 (1906); Carnot, C.R. Soc. Biol., 111, 463-5 (1906); Reissman, Blood, 1950, 5, 372-80 (1950) and Erslev, Blood, 8, 349-57 (1953). Reissman and Erslev's experiments were promptly confirmed by other researchers. See Hodgson, et al., Blood, 9, 299-309 (1954); Gordon, et al., Proc. Soc. Exp. Biol. Med., 86, 255-8 (1954) and Borsook, et al., Blood, 9, 734-42 (1954).

The identification of the EPO production site in the organism was an issue of debate. Successive experiments led to identify the kidney as the main organ and peritubular interstitial cells as the synthesis site. See Jacobson, et al., Nature, 179, 633-4 (1957); Kuratowska, et al., Blood, 18, 527-34 (1961); Fisher, Acta Hematol., 26, 224-32 (1961); Fisher, et al., Nature, 205, 611-2 (1965); Frenkel, et al., Ann. N.Y. Acad. Sci., 149, 1, 292-3 (1968); Busuttil, et al., Proc. Soc. Exp. Biol. Med., 137, 1, 327-30 (1971); Busuttil, Acta Haematol., (Switzerland), 47, 238-42 (1972); Erslev, Blood, 44, 1, 77-85 (1974); Kazal, Ann. Clin. Lab. Sci., 5, 2, 98-109 (1975); Sherwood, et al., Endocrinology, 99, 2, 504-10 (1976); Fisher, Ann. Rev. Pharmacol. Toxicol., 28, 101-22 (1988); Jelkmann, et al., Exp. Hematol., 11, 7, 581-8 (1983); Kurtz, et al., Proc. Natl. Acad. Sci. (USA), 80, 13, 4008-11 (1983); Caro, et al., J. Lab. Clin. Med., 103, 6, 922-31 (1984); Caro, et al., Exp. Hematol., 12, 357 (1984); Schuster, et al., Blood, 70, 1, 316-8 (1986); Bondurant, et al., Mol.Cell. Biol., 6, 7, 2731-3 (1986); Bondurant, et al., Mol. Cell. Biol., 6, 7, 2731-3 (1986); Schuster, et al., Blood, 71, 2, 524-7 (1988); Koury, et al., Blood, 71, 2, 524-7 (1988); Lacombe, et al., J. Clin. Invest., 81, 2, 620-3 (1988); Koury, et al., Blood, 74, 2, 645-51 (1989).

A smaller proportion, ranging from 10% to 15% of total EPO, is produced by the liver in adults. See Naughton, et al., J. Surg. Oncol., 12, 3, 227-42 (1979); Liu, et al., J. Surg. Oncol., 15, 2, 121-32 (1980); Dornfest, et al., Ann. Clin. Lab. Sci., 11,1, 37-46 (1981); Dinkelaar, et al., Exp. Hematol., 9, 7, 796-803 (1981); Caro, et al., Am. J. Physiol., 244, 5 (1983); Dornfest, et al., J. Lab. Clin. Med., 102, 2, 274-85 (1983); Naughton, et al., Ann. Clin. Lab. Sci., 13, 5, 432-8 (1983); Jacobs, et al., Nature, 313, 6005, 806-10 (1985); Erslev, et al., Med. Oncol. Tumor. Pharmacother., 3, 3-4, 159-64 (1986). The EPO produced is directly proportional to the extent of tissular hypoxia and its expression rises by increasing the number of the EPO producing cells.

EPO has shown great efficiency in the treatment of anemia, especially anemia derived from renal failure. See Eschbach, et al., N. England J. of Med., 316, 2, 73-78 (1987); Krane, Henry Ford Hosp. Med. J. , 31, 3, 177-181 (1983). Its therapeutical usefulness, however, has been limited due to the unavailability of a massive production method. The quantity and quality of the EPO obtained by the extractive systems known were insufficient. Recently, the use of recombinant DNA technology has made it possible to obtain large amounts of proteins. The application of these techniques to eukaryotic cells has allowed a large scale production of EPO. See patents U.S. Pat. No. 5,688,679 (to Powell), U.S. Pat. No. 5,547,933 (to Lin), U.S. Pat. No. 5,756,349 (to Lin), U.S. Pat. No. 4,703,008 (to Lin) and U.S. Pat. No. 4,677,195 (to Hewick et al.)

Notwithstanding the recent considerable literature referring to the production of EPO in mammalian cell cultures, no method has yet been devised to produce efficiently EPO in industrial scale. The existent EPO culture systems are further characterized by their low reproducibility and output quality. See U.S. Pat. No. 5,688,679 (to Powell); U.S. Pat. No. 5,547,933 (to Lin); U.S. Pat. No. 5,756,349 (to Lin); U.S. Pat. No. 4,703,008 (to Lin) and U.S. Pat. No. 4,677,195 (to Hewick et al.); Andersen, et al., Curr. Op. Biotech, 5,546-549 (1994); Butler, Ed., “Mammalian Cell Biotechnology”, (IRL Press, Oxford, England, 1991); Murakami, Ed., “Trends in Animal Cell Culture Technology”, (Kodansha Ltd., Tokyo, Japan, 1990); Freshney, Ed., “Animal Cell Culture. A Practical Approach,” Ch. 3, (IRL Press, Oxford, England, 1986); Pirt, “Principles of Microbe and Cell Cultivation”, (Blackwell Scientific Pub., London, England, 1985); Hames et als., “Transcription and Translation. A Practical Approach”, (IRL Press, Oxford, England, 1984).

BRIEF SUMMARY OF THE INVENTION

The present invention provides a method for obtaining human erythropoietin (EPO) comprising culturing mammalian cells which express recombinant human erythropoietin in a culture medium comprising insulin.

The invention specifically relates to a method for obtaining human EPO that comprises the separation of a supernatant comprising EPO and insulin from cells which express recombinant human EPO, the concentration of said supernatant and the freezing of the resulting concentrated product.

The present invention describes a method for the massive culture of recombinant cells adequate for the industrial production of EPO. The method described in the present invention produces an unexpectedly high amount of EPO with a low concentration of contaminant proteins in the culture medium. This condition enhances the subsequent EPO purification steps and results in a high protein recovery ratio. The high yield of EPO obtained by the claimed method is achieved by the addition of insulin to the culture medium.

Other advantages of the claimed method are: 1) its reproducibility and 2) the high quality of the EPO obtained regardless of the protein production scale.

The method claimed in the present invention utilizes EPO producing cells preserved in liquid nitrogen (Master and Working Bank). The cells are thawed and multiplied through successive expansion steps at 37° C. with different culture media. The expansion is sustained until a cellular mass adequate for industrial scale production is achieved. The cells are grown in T25 (25 cm² surface area) flasks and transferred to flasks with increasing surface areas until a final roller flask surface of 850 cm² is reached. The culture medium employed for the cell expansion is then replaced by an improved culture medium containing insulin to enhance the production of EPO. After 72 hours, the EPO containing supernatant is first recovered, purified and assayed.

The obtained culture supernatant is assayed by SDS-PAGE, Western blot, glycanase treatment followed by SDS-PAGE, isoelectric focusing and a complete protein sequence analysis to verify the identity of the EPO produced. The in vivo biological activity of the EPO thus produced is determined by an ex-hypoxic polycythemic mice assay using the World Health Organization EPO standard as reference.

BRIEF DESCRIPTION OF THE DRAWINGS/FIGURES

FIG. 1 illustrates a polyacrylamide gel electrophoresis (SDS-PAGE) assay performed on an EPO sample obtained according to the method herein described. Lanes 1, 4 and 7 show molecular weight markers. On lanes 2, 3, 5 and 6 different pure EPO samples were run according to the claimed method. The purity of the obtained product, as well as its apparent molecular weight, somewhat superior to 30 kDa, is coincident with that of urinary human EPO.

FIG. 2 illustrates a “Western Blot” assay performed on an EPO sample obtained according to the method described herein. The produced EPO identity is verified by the recognition of an antibody against human EPO. On lane 1 a human EPO standard sample was run; on lane 2 molecular weight markers were run and on lanes 3 to 5, EPO samples obtained according to the method claimed herein were run.

FIG. 3 illustrates a SDS-PAGE assay performed on a pure EPO sample obtained according to the method described herein and submitted to glycanases treatment. Molecular weight markers were run on lanes 1, 4 and 8. Lanes 2 and 7 show untreated EPO samples. On lane 3 an EPO sample treated with O-glycanase was run, showing the existence of an O-glycosilation site. On lane 5 an EPO sample partially degraded with N-glycanase was run, verifying the presence of the molecular weight markers corresponding to those expected for EPO. On lane 6 an EPO sample degraded with O-glycanase and N-glycanase was run, showing the pattern expected for the fully deglycosilated protein.

FIG. 4 illustrates an electrofocusing assay performed on pure EPO samples produced according to the method herein described. EPO samples were run on lanes 2, 3 and 4; isoelectric point markers were run on lanes 1 and 5. The presence of the EPO isoforms with isoelectric points ranging from 3.0 to 4.5 is observed.

DETAILED DESCRIPTION OF THE INVENTION

The present invention provides a method for obtaining EPO comprising the culture of mammalian cells which express recombinant human erythropoietin in a substrate medium comprising insulin. The claimed invention is further characterized by the absence of fetal calf serum from the culture medium.

The EPO expressing cells in the present invention are selected from the group comprising mammalian cells. Preferably said mammalian cells are selected from the group comprising CHO, COS, BHK, Namalwa, and HeLa cells, and even more preferably said mammalian cells are CHO cells.

Preferred recombinant host cells comprise a vector which comprises a nucleotide sequence encoding the EPO polypeptide consisting of the amino acid sequence in SEQ ID NO: 1, a viral promoter and a viral terminator. Preferred host cell vectors confer resistance to both methotrexate and neomycin-derived antibiotics. Preferably, the EPO nucleic acid molecule comprises the nucleic acid molecule described in Lin, “DNA Sequences Encoding Erythropoietin,” U.S. Pat. No. 4,703,008. Preferably, the viral promoter is an SV40 early promoter.

The insulin in the culture medium in the above and below-described methods can be present in amounts from about 0.1 mg per liter of culture medium to about 40 mg per liter of culture media, and even more preferably from about 0.5 mg per liter of culture media to about 20 mg per liter of culture media.

In a further embodiment, the invention provides a method for obtaining human erythropoietin comprising the culture of mammalian cells which express recombinant human erythropoietin in culture media which comprises fetal calf serum-free media.

The invention also provides a method for obtaining EPO by separating the supernatant comprising EPO and insulin from mammalian cells which express recombinant human erythropoietin, concentrating said supernatant and freezing the concentrated product.

The invention provides the above and below-described methods wherein said concentration step concentrates the supernatant from about 25 to 200 fold, preferably about 50 to 150 fold, and most preferably 100 fold.

The invention provides the above and below-described methods wherein said concentration step comprises using a tangential filtration system through membranes with a molecular weight cut-off of about 3,000 Daltons.

The invention also provides a method for obtaining human erythropoietin by separating the supernatant comprising EPO and insulin from mammalian cells which express recombinant human erythropoietin, concentrating said supernatant, freezing the concentrated product and sterile filtering said concentrated product. Preferably said sterile filtering is through membranes with a pore diameter ranging from about 0.1 to about 0.2 μm, and most preferably about 0.2 μm.

The EPO protein can be further purified from recombinant cell cultures by well-known methods including ammonium sulfate or ethanol precipitation, acid extraction, anion or cation exchange chromatography, phosphocellulose chromatography, hydrophobic interaction chromatography, affinity chromatography, hydroxylapatite chromatography and lectin chromatography.

A preferred method of further purifying the EPO comprises treating cell culture supernatants comprising EPO by a combination of the following steps: (a) differential precipitation, (b) hydrophobic interaction chromatography, (c) diafiltration, (d) anionic exchange chromatography, (e) cationic exchange chromatography and (f) molecular exclusion chromatography. Preferably, said steps are performed in the following order: (a), (b), (c), (d), (e), and (f).

A preferred method of using the EPO produced by the culturing of the present invention comprises lyophilization into a form suitable for injection into humans for the treatment of diseases. Specifically, the preferred lyophilization procedure comprises placing the EPO into a pharmaceutical composition, loading the first EPO composition into a container, wherein said container is at a temperature equal to or less than −30° C.; incubating said EPO composition at a temperature equal to or less than −30° C. under atmospheric pressure for a time equal to or greater than 4 hours; exposing said composition at a pressure of equal to or less than 30 absolute microns for a time equal to or greater than one hour; and raising the temperature equal to or less than 3° C. per hour until reaching at least 25° C., while keeping pressure values equal to or less than 5 absolute microns.

A preferred pharmaceutical composition for lyophilization comprises EPO, sugar, salts and human albumin. An especially preferred composition for lyophilization comprises EPO, mannitol, NaCl, NaH₂PO₄, Na₂HPO₄ and human albumin.

The invention also provides a method for obtaining human erythropoietin by separating the supernatant comprising EPO and insulin from mammalian cells which express recombinant human erythropoietin, adding media comprising insulin to separated cells and culturing said cells.

The present invention is described in further detail in the following non-limiting examples.

EXAMPLES

Example 1

Culture

The recombinant human EPO is produced in a mammalian cell line (CHO) transfected with human erythropoietin genomic DNA. The cell banks (“Master” and “Working Bank”) containing the EPO producing cells were preserved in liquid nitrogen according to commonly used laboratory methods. See Hames et al., “Transcription and Translation. A Practical Approach”, (IRL Press, Oxford, England, 1984), incorporated herein as reference. Four “seeds” from the EPO working bank were thawed and added separately to four T 25 flasks containing 10 ml of culture medium no. 1 each. See Table 1. For each T 25 flask the following protocol was applied:

The EPO producing cells were cultured for 24 hours at 37° C. The culture medium was then removed and 10 ml of medium no. 2 were added to each T 25 flask. See Table 1. The cell cultures were preserved at the same conditions for an additional 24 hour period. The resulting EPO producing cell cultures were then subject to a series of expansion stages as described in the following examples.

Example 2

Expansion 1

Each cell culture from Example 1 was removed from the T 25 flasks by trypsin treatment according to commonly used laboratory methods. See Hames et al., supra. Afterwards, the cell cultures were grown according to the following protocol: to each one of five T 25 flasks, 20% of each grown cell culture and 10 ml of culture medium no. 2 were added. See Table 1. The EPO producing cell cultures were incubated further for 48 hours at 37° C.

Example 3

Expansion 2

The five cell cultures from the previous example were removed from the T 25 flasks by trypsin treatment according to commonly used laboratory methods. See Hames et al., supra. Each cell culture was transferred to a T 150 flask. Subsequently, 75 ml of culture medium no. 1 were added to each T 150 flask. See Table 1. The EPO producing cell cultures were then incubated for 72 hours at 37° C.

Example 4

Expansion 3

The cell cultures from the previous example were removed from the T 150 flasks by trypsin treatment according to commonly used laboratory methods. See Hames et al., supra. 10% of the cell culture of each T 150 flask was transferred to a new T 150 flask. 75 ml of culture medium no. 1 were added to each new T 150 flask. See Table 1. The EPO producing cell cultures were then incubated for 72 hours at 37° C.

Example 5

Expansion 5

The cell cultures from the previous example were removed from the T 150 flasks by trypsin treatment according to commonly used laboratory methods. See Hames et al., supra. The cell cultures from each T 150 flask were transferred to a “roller” flask of 850 cm² of inner surface. 200 ml of culture medium no. 1 were added to each roller flask. See Table 1. The EPO producing cell cultures were then incubated for 72 hours at 37° C.

Example 6

Expansion 6

The cell cultures from the previous example were removed from the roller flasks by trypsin treatment according to commonly used laboratory methods. See Hames et al., supra. 1/15^th of each cell culture was transferred to a new roller flask. In total 3,000 roller flasks containing the EPO producing cell culture were prepared and grown according to the following protocol:

200 ml of culture medium no. 1 were added to each roller flask. The EPO producing cell cultures were then incubated for 72 hours at 37° C. The roller rotation speed was 11 rounds per hour.

The formation of a cell monolayer was verified by an inverted microscopic analysis. The culture medium of each roller flask was discarded and the EPO producing cell cultures were rinsed with 300 ml Hank's solution per roller. See Table 1. 200 ml of culture medium no. 3 was then added to each roller flask.

Example 7

Harvest

The culture supernatant of each cell culture from the preceding example was harvested every 48 hours, under strict sterile conditions, and replaced with 200 ml of fresh culture medium no. 3. This procedure was repeated 5 times for each roller flask.

The harvested culture supernatant was concentrated a hundred-fold with a tangential filtration system utilizing 3,000 D cut off Amicon S10Y3 membranes. The concentrated material was filtered under sterile conditions and stored at −20° C.

A cellular density between 180,000 cells per cm² and 800,000 cell per cm² was achieved. Cell viability ranged between 95% and 98% throughout the whole procedure.

The harvested supernatant was 2,900 liters. The material concentrated according to example 7 yielded 29.5 liters.

The following table summarizes the results of each harvest stage:

HARVEST	EPO-RIA (g)	TOTAL PROTEINS (g)
1	26.55	51.33
2	38.05	95.87
3	37.76	100.0
4	32.45	134.8
5	33.04	161.7
TOTAL	167.85	543.7

Example 8

EPO Assays

The EPO obtained in the previous example was further purified and assayed for identity and biological activity according to the following protocol.

In a denaturing SDS-PAGE gel EPO was identified as a wide band of molecular weight as expected for EPO. See FIG. 1. The band was recognized by monoclonal and polyclonal antibodies raised against human EPO in a Western blot assay as expected for EPO. See FIG. 2. The treatment with glycanases proved the existence of the glycosidic chains in the extent and size expected for EPO. See FIG. 3. The EPO produced was shown to be composed of a series of species with isoelectric points ranging from 3.0 to 4.5 as expected for EPO. See FIG. 4.

The complete amino acid sequence of the isolated protein, purified from the culture supernatant of transfected cell lines showed total homology with natural human erythropoietin whose 165 amino acid sequence is as follows (SEQ ID NO: 1):

NH2-	Ala	Pro	Pro	Arg	Leu	Ile	Cys	Asp
	Ser	Arg	Val	Leu	Glu	Arg	Tyr	Leu
	Leu	Glu	Ala	Lys	Glu	Ala	Glu
	Ile	Thr	Thr	Gly	Cys	Ala	Glu	His
	Cys	Ser	Leu	Asn	Glu		Ile	Thr
	Val	Pro	Asp	Thr	Lys	Val	Asn	Phe
	Tyr	Ala	Trp	Lys	Arg	Met	Glu	Val
	Gly	Gln	Gln	Ala	Val	Glu	Val	Trp
	Gln	Gly	Leu	Ala	Leu	Leu	Ser	Glu
	Ala	Val	Leu	Arg	Gly	Gln	Ala	Leu
	Leu	Val		Ser	Ser	Gln	Pro	Trp
	Glu	Pro	Leu	Gln	Leu	His	Val	Asp
	Lys	Ala	Val	Ser	Gly	Leu	Arg	Ser
	Leu	Thr	Thr	Leu	Leu	Arg	Ala	Leu
	Gly	Ala	Gln	Lys	Glu	Ala	Ile	Ser
	Pro	Pro	Asp	Ala	Ala		Ala	Ala
	Pro	Leu	Arg	Thr	Ile	Thr	Ala	Asp
	Thr	Phe	Arg	Lys	Leu	Phe	Arg	Val
	Tyr	Ser	Asn	Phe	Leu	Arg	Gly	Lys
	Leu	Lys	Leu	Tyr	Thr	Gly	Glu	Ala
	Cys	Arg	Thr	Gly	Asp-COOH

The presence of the four glycosilation sites on the 165 amino acid chain, as well as the complex carbohydrate structure, and in particular, the sialic acid terminal residues, which characterizes EPO were verified. These results were further supported by a biological activity assay of the produced protein by an ex-hypoxic polycythemic mice test which showed complete concordance with the international EPO standard.

After purification, 30% of the EPO obtained was recovered. The surprisingly high protein yield ratio is attributed to the initial low level of impurities resulting from the claimed invention. The low level of impurities of the culture supernatant is due to the substitution of fetal calf serum by insulin. The insulin substitution prevents the occurrence of metabolites of animal origin in the culture supernatant.

Under the traditional culture methods, which utilize 10% fetal calf serum, the EPO recovery is less than 1%. The recovery ratio of the claimed method is thirty times higher than this.

All publications mentioned hereinabove are hereby incorporated in their entirety by reference.

While the foregoing invention has been described in some detail for purposes of clarity and understanding, it will be appreciated by one skilled in the art from a reading of this disclosure that various changes in form and detail can be made without departing from the true scope of the invention and appended claims.

TABLE NUMBER 1
Culture Medium no. 1
Basal Culture Medium + 10% Fetal Calf Serum
ISCOVE'S DNEM	8.85	g/l	Tryptophan	27	mg/l
HAM F12	5.35	g/l	Asparagine	40	mg/l
NaHCO3	2.10	g/l	Serine	80	mg/l
Glucose	1.30	g/l	Ethanolamine	3_g/l
Lactose	0.20	g/l	Glutamine	1.90	g/l
Galactose	0.20	g/l	Fetal Calf Serum	100	ml/l
Sodium Pyruvate	0.11	g/l
Culture Medium no. 2
Basal Culture Medium + 10% Fetal Calf Serum + Geneticin 0.5 mg/ml
ISCOVE'S	8.85	g/l	Tryptophan	27	mg/l
DNEM
HAM F12	5.35	g/l	Asparagine	40	mg/l
NaHCO3	2.10	g/l	Serine	80	mg/l
Glucose	1.30	g/l	Ethanolamine	3	mg/l
Lactose	0.20	g/l	Glutamine	0.20	g/l
Sodium Pyruvate	0.11	g/l	Fetal Calf Serum	100	ml/l
Glutamine	1.90	g/l	Geneticin	500	mg/l
Culture Medium no. 3
Basal Culture Medium + Insulin
ISCOVE'S DNEM	8.85	g/l	Tryptophan	27	mg/l
HAM F12	5.35	g/l	Asparagine	40	mg/l
NaHCO3	2.10	g/l	Serine	80	mg/l
Glucose	1.30	g/l	Ethanolamine	3	mg/l
Lactose	0.20	g/l	Glutamine	0.20	mg/l
Galactose	0.20	g/l	Sodium Pyruvate	0.11	g/l
Glutamine	1.90	g/l	Insulin	10	mg/l
HANK's Solution
CaCl2•2H20	185	mg/l	Glucose	1.0	g/l
MgSO4•7H2O	140	mg/l	NaHCO3	350	mg/l
KCl	400	mg/l	KH2PO4	60	mg/l
NaCl	8.0	g/l	Na2HPO4	47.8	mg/l

Claims

1-12. (canceled)

13. A method for preparing a lyophilized recombinant human erythropoietin composition comprising:

(i) providing a composition comprising: (a) recombinant human erythropoietin; (b) mannitol; (c) two or more salts selected from the group consisting of: NaCl, NaH2PO4, and Na2HPO4, and mixtures thereof; and (d) human albumin;

(ii) incubating the composition at a temperature equal to or less than −30° C. and at atmospheric pressure;

(iii) exposing the composition to a pressure equal to or less than 30 microns; and

(iv) raising the temperature of the composition at a rate equal to or less than 3° C. per hour until reaching a temperature of at least 25° C. while maintaining a pressure equal to or less than 5 microns.

14. The method of claim 13, wherein the composition comprises:

(a) recombinant human erythropoietin;

(b) mannitol;

(c) NaCl, NaH2PO4, and Na2HPO4; and

(d) human albumin.

15. The method of claim 13, wherein the composition of (ii) is incubated for a time equal to or greater than 4 hours.

16. The method of claim 13, wherein the composition of (iii) is exposed for a time equal to or greater than 1 hour.

17. The method of claim 13, wherein the composition of (ii) is incubated for a time equal to or greater than 4 hours, and wherein the composition of (iii) is exposed for a time equal to or greater than 1 hour.

18. A pharmaceutical composition made by the method of claim 13.

19. A pharmaceutical composition for lyophilization comprising:

(a) recombinant human erythropoietin;

(b) mannitol;

(c) two or more salts selected from the group consisting of: NaCl, NaH2PO4, and Na2HPO4, and mixtures thereof; and

(d) human albumin.

20. The pharmaceutical composition of claim 19, wherein the composition comprises NaCl, NaH2PO4, and Na2HPO4.

21. A pharmaceutical composition for lyophilization comprising:

(a) recombinant human erythropoietin;

(b) mannitol;

(c) NaCl, NaH2PO4, and Na2HPO4; and

(d) human albumin.