Recombinant human epo-fc fusion proteins with prolonged half-life and enhanced erythropoietic activity in vivo

RECOMBINANT HUMAN EPO-FC FUSION PROTEINS WITH PROLONGED HALF-LIFE AND ENHANCED ERYTHROPOIETIC ACTIVITY IN VIVO

Related Application [0001] This application is related to and claims the benefit of US Patent Application No. 11/340,661 filed 27 January 2006.

Technical Field [0002] This application relates to human erythropoietin fusion proteins.

Background WO 2004/004798 relates to methods and compositions for the administration of therapeutics comprising antibodies and antibody conjugates to central airways of the lung. WO 2005/001025 relates to chimeric monomer-dimer hybrid proteins comprising an immunoglobulin constant region.

[0003] Human erythropoietin (EPO), a member of the haematopoietic growth factor family, is synthesized mainly in the adult kidney and fetal liver in response to tissue hypoxia due to decreased blood oxygen availability [1]. The principal function of EPO is to act directly on certain red blood cell (RBC) progenitors and precursors in the bone marrow to stimulate the synthesis of hemoglobin and mature RBCs. It also controls the proliferation, differentiation, and maturation of RBCs. Recombinant EPO having the amino-acid sequence of naturally occurring EPO has been produced and approved to treat anemia associated with kidney functional failure, cancer and other pathological conditions [2]. In addition to its erythropoietic properties, recent research reports [3] indicate that EPO also acts on non-bone marrow cells such as neurons, suggesting other possible physiological/pathological functions of EPO in the central nervous system (CNS) and other organs/systems. Since EPO receptors have been found in many different organs, EPO may have multiple biological effects, such as acting as an anti-apoptotic agent. 5 [0004] Human EPO is a glycoprotein with a molecular weight of 30.4 kilo- daltons. Carbohydrates account for approximately 39% of its total mass. The EPO gene is located on chromosome 7ql1-22 and spans a 5.4kb region with five exons and four introns [4]. The precursor of EPO consists of 193 amino acids. Cleavage of the 10 leader sequence and the last amino acid Arg by post-translational modification yields the mature EPO having 165 amino acids. Glycosylation, with three N-linked sites at Asn 24, Asn38, Asn83 and one O-linked site at SerI26, plays a crucial role in the biosynthesis, tertiary structure and the in -vivo bioactivity of EPO [5]. EPO functions by binding to an erythropoietin receptor, a glycosylated and phosphorylated 15 transmembrane polypeptide with the molecular weight of 72-78 kilodaltons. This binding triggers the homod imerization of the receptors that leads to the activation of several signal transduction pathways: JAK2/STAT5 system, G-protein, calcium channel, and kinases. Two molecules of EPO protein are needed to bind simultaneously to one receptor molecule to achieve optimal receptor activation [6], 20 [0005] As the first hematopoietic growth factor approved for human therapy, recombinant human EPO (rHuEPO) has been used for the treatment of anemia resulting from chronic renai failure, cancers (primarily chemotherapy-associated anemia), autoimmune diseases, AIDS, surgery, bone marrow transplantation and 25 myelodysplastic syndromes, etc. Interestingly, recent studies have also observed that 2 rHuEPO has non-blood system functions and shows the potential of being used as a neuroprotective drug for cerebral ischemia, brain trauma, inflammatory disease and neural degenerative disorders [7].

[0006] Currently, three kinds of rHuEPO or rHuEPO analogs are commercially available, namely rHuEPO alpha, rHuEPO beta, and darbepoetin alfa [8]. These three recombinant proteins bind to the same erythropoietin receptor, but differ in structure, degree of glycosylation, receptor-binding affinity and in vivo metabolism. Since the initial introduction of rHuEPO-alpha in thel980s, clinicians quickly recognized the frequent dose/injection requirement of the drug as a significant shortcoming. The mean in vivo half-lives of rHuEPO alpha and rHuEPO beta administered intravenously or subcutaneously are only 8.5 and 17 hours respectively [9, 10]. Patients therefore need an injection schedule of daily, twice weekly or three times per week which imposes a burden on both patients and health care providers. Thus, there has been a longstanding need to develop recombinant EPO analogs having a longer in vivo half-life and/or enhanced erythropoietic activity.

[0007] Attempts have been made in the prior art to genetically change or chemically modify the structure of the native EPO protein to either slow down its in vivo metabolism or improve its therapeutic properties. For example, there appears to be a direct correlation between the amounts of sialic acid-containing carbohydrates on the EPO molecule and its in vivo metabolism and functional activity. Increasing the carbohydrate content of the EPO molecule thus results in a longer half-life and enhanced activities in vivo [11, 12]. Amgen has designed the rHuEPO analog darbepoetin alpha to include 5 N-linked carbohydrate chains, two more than rHuEPO. 3

Darbepoetin alpha is also known as Novel Erythropoiesis Stimulating Protein (NESP) and is sold under the trademark Aranesp™. Darbepoetin alpha differs from native human EPO at five positions (AIa30Asn, His32Thr, Pro87Val, Trp88Asn, Pro90Thr) which allows for the attachment of two additional N-Iinked oligosaccharides at 5 asparagines residue positions 30 and 88. Darbepoetin alpha binds to the EPO receptor in an identical manner as native EPO to induce intracellular signaling involving tyrosine phosphorylation by JAK-2 kinase and the same intracellular molecules Ras/MAP-k, P13-k and STAT-5. Due to the increased carbohydrate content, the half-life of darbepoetin alpha in both animals and humans is almost three fold-longer than 10 that of rHuEPO-alpha (25.3 hours vs 8.5 hours) [9]. Darbepoetin alpha (Aranesp™) also appears to exhibit enhanced bioactivity in comparison to naturally occurring or recombinant human EPO in vivo [13] and has been approved by FDA as a second generation rHuEPO drug; this drug only needs to be administrated once per week to achieve the identical therapeutic effects of 2-3 time injections per week of 15 rHuEPO[10, 14. 15], [0008] Other attempts to extend the half-life of EPO have focused on increasing the molecular weight of the EPO protein through chemical conjugation with polyethylene glycol (PEGylation) and the like. PEGylated- EPO has a much 20 larger molecular weight and is protected from being cleared from circulation and therefore has a longer plasma half-life [16]. However, PEGylation may alter the protein structure resulting in unanticipated changes of function and specificity of the EPO moiety. There are also reports of increasing the molecular weight of EPO by other methods, such as to link the EPO molecule to a carrier protein (human albumin), 25 or to form the homodimerization of two complete EPO molecules by using linking 4 peptides (3- to 17-amino acids) or by chemical cross-linking reagents [17,18, 19. 20],

While all these methods have achieved some success in extending the half-life and enhancing the activities of EPO, combining the EPO molecule with the Fc fragment of human immunoglobulin (IgG) in a fusion protein as described in the present 5 application achieves unique advantages.

[0009] Human immunoglobulin IgG is composed of four polypeptides linked covalently by disulfide bonds (two identical copies of light chain and heavy chain).

The proteolysis of IgG molecule by papain generates two Fab fragments and one Fc 10 fragment. The Fc fragment consists of two polypeptides linked together by disulfide bonds. Each polypeptide, from N-to C-terminal, is composed of a hinge region, a CH2 domain and a CH3 domain. The Fc fragment structure is almost the same among all subtypes of human immunoglobulin. IgG is among one of the most abundant proteins in the human blood and makes up 70 to 75 % of the total 15 immunoglobulins in human serum. The half-life of IgG in circulation is the longest among all five types of immunoglobulin and may reach 21 days.

[00010] Modem bio-engineering technology has been successfully applied to the creation of fusion proteins consisting of therapeutic protein fragments, such as 20 cytokines and soluble receptors, and the Fc fragment of human IgG [21,22. 23, 24],

These fusion proteins have a significantly longer in vivo half-life while retaining their biological and therapeutic properties. So far two fusion proteins comprising an Fc fragment have been successfully developed as biomedicines and approved by FDA for the treatment of rheumatoid arthritis and chronic plaque psoriasis [25,26], 25 5 [00011 ] It has been shown in the prior art that dimers of two EPO molecules linked either by chemical cross-linking or by a polypeptide exhibit enhanced in vivo activities and a prolonged half-life[ 17, 19]. The enhanced activity may due to the more efficient binding of the EPO dimer to one receptor, and the prolonged in vivo half-life due to the larger size of the dimer protein. However, the chemical crosslinking process is not efficient and is difficult to control. Moreover, the linkage peptide in the dimer of EPO may alter the three-dimensional structure of EPO molecule and the peptide itself may stimulate immunogenic responses in vivo. These shortcomings impair the therapeutic potential of EPO dimers, particularly since EPO replacement therapy in renal patients is life-long.

[00012] The need has therefore arisen for EPO analogs that have a significantly longer half-life and enhanced erythropoietic activities in vivo but have no increased immunogenic properties.

Summary of the Invention [00013] In accordance with the invention, a recombinant fusion protein comprising a human erythropoietin peptide portion linked to an immunoglobulin peptide portion is described. The fusion protein has a prolonged half-life in vivo in comparison to naturally occurring or recombinant native human erythropoietin. In one embodiment of the invention, the protein has a half-life in vivo at least three fold higher than native human erythropoietin. The fusion protein may also exhibit enhanced erythropoietic bioactivity in comparison to native human erythropoietin. 6 [00014] In one embodiment of the invention the immunoglobulin peptide portion is an Fc fragment, such as an IgGl fragment. The Fc fragment includes CH2 and CH3 domains and a hinge region. The EPO peptide portion may be directly linked to the hinge region. Preferably the hinge region is at least 9 amino acids in 5 length. In one embodiment, the EPO peptide portion has a cysteine residue proximate the C terminal thereof and the hinge region includes a cysteine residue located nearest the EPO peptide portion. Preferably these two cysteine residues are spaced at least 12 amino acids apart. In one embodiment, the EPO peptide portion may comprise a complete EPO molecule directly linked to the immunoglobulin portion (i.e. no 10 external peptide linkers are interposed between the EPO and immunoglobulin portions).

[00015] The invention also relates to multimeric protein constructs comprising multiple units of the fusion protein of the invention. For example, two fusion proteins 15 may be assembled as a dimer, wherein the hinge regions of the proteins are joined by disulphide bonds. The dimer has the general shape of a IgG molecule and is more stable than free EPO molecules.

[00016] The invention also relates to nucleic acid and amino acid sequences 20 encoding the fusion protein and transfected cell lines and methods for producing the fusion protein. The invention further includes pharmaceutical compositions comprising the fusion protein and methods of using the fusion protein and/or the pharmaceutical compositions, for example to stimulate erythropoiesis in subjects in need of therapy. 7

Brief Description of the Dr awings [00017] In drawings which illustrate various embodiments of the invention but 5 which are not intended to be construed in a limiting manner: [00018] Figure 1A is a schematic diagram showing the general structure of the recombinant human EPO-Fc fusion protein (rHuEPO-Fc) of the invention.

[00019] Figure IB is a sequence listing showing the nucleotide sequence and 10 the deduced amino-acid (aa) sequence of rHuEPO-Fc protein. The total length of DNA is 128lbp. The 426 amino acids in the deduced protein sequence include 27 aa for the signal peptide and 399 aa for the complete rHuEPO-Fc protein. The complete rHuEPO-Fc protein consists of human EPO domain (166aa), hinge region (16aa, underlined), and CH2 and CH3 domains (217 aa) of the Fc fragment of human IgGl. 15 The calculated molecular weight of the polypeptide of the mature rHuEPO-Fc fusion protein is 44.6kDa, composed of 18.5kDa (41.4%) ofEPO fragment and 26.1kDa(58.6%) of IgGl Fc fragment. A homodimer is formed by two disulfide bonds via the two cysteine residues (boxed) within the hinge region. At residue 172 of the mature fusion protein (i.e. the 6* amino acid of hinge region) the native 20 cysteine residue has been substituted by glycine (bold).

[00020] Figure 2 is a schematic diagram showing the structure and features of the mammalian expression plasmid pCDl used for inserting the DNA sequence encoding the polypeptide of the rHuEPO-Fc fusion protein, and for transfecting 25 CHO cells that express the rHuEPO-Fc fusion protein. 8 [00021 ] Figure. 3 is a SDS-PAGE image showing the sizes of the dimeric form of pure rHuEPO-Fc protein in non-reduced condition and monomeric form of pure rHuEPO-Fc protein in reduced condition by SDS-PAGE analysis. The purified 5 rHuEPO-Fc protein from the supernatants of the cultured CHO cell-line expressing rHuEPO-FC exists mainly as the dimeric form and has a molecular weight of about 180 kDa on 8% Bis-Tris gel in non-reduced condition. In reduced condition {lOOmM dithiothreitol, DTT) to break disulfide bonds, the dimer is separated into two identical monomeric units with a molecular weight of 75 kDa. 10 [00022] Figure 4A and 4B are graphs showing the dose-dependent increase of hemoglobin (Hb) levels in normal mice treated with three times per week subcutaneous injection (s.c.) of rHuEPO-Fc or rHuEPO. Each point represents the mean Hb level of the group (6 mice). Day 0 levels represent the Hb levels before

15 treatment. A: Mice treated with rHuEPO-Fc. B: Mice treated with native rHuEPO

[00023] Figure 5A and 5B are graphs showing the dose-dependent increase of hemoglobin (Hb) levels in normal mice treated with once per week s.c. of rHuEPO-Fc or rHuEPO. Each point represents the mean Hb level of the group (6 mice). Day 0 20 levels represent the Hb levels before treatment. A: Mice treated with rHuEPO-Fc. B:

Mice treated with native rHuEPO

[00024] Figure 6A and 6B are graphs showing the increase of hemoglobin (Hb) levels in normal mice treated with intravenously injection (i.v.) of 12.5pg/kg of 25 rHuEPO-Fc or rHuEPO. Each point represents the mean Hb level of the group (6 9 mice). Day 0 levels represent the Hb levels before treatment. A: Mice with treatment once a week. B: Mice with treatment 3 times a week.

[00025] Figure 7 is a graph showing the dose-dependent increase of 5 hemoglobin (Hb) levels in 5/6 nephrectomized rats treated with once per week s.c. of rHuEPO-Fc, rHuEPO or darbepoetin-alfa (abbreviated Darbe.). Each point represents the mean Hb level of the group. Normal controls were normal rats with injection of carrier solution. Model controls were the 5/6 nephrectomized rats with injection of carrier solution. Week 0 levels represent the Hb levels before treatment. *: week(s) 10 post treatment.

[00026] Figure 8 is a graph showing the dose-dependent increase of hemoglobin (Hb) levels in 5/6 nephrectomized rats treated once every two weeks s.c. with rHuEPO-Fc, rHuEPO or darbepoetin-alfa (abbreviated Darbe.). Each point 15 represents the mean Hb level of the group. Normal controls were normal rats with injection of carrier solution. Model controls were the 5/6 nephrectomized rats with injection of carrier solution. Week 0 levels represent the Hb levels before treatment. *: week(s) post treatment. 20 [00027] Figure 9 is a graph showing the dose-dependent increase of hemoglobin (Hb) levels in 5/6 nephrectomized rats treated once every two weeks i.v. with 62.5pg/kg of rHuEPO-Fc, or darbepoetin-alfa (abbreviated Darbe.). Each point represents the mean Hb level of the group. Normal controls were normal rats with injection of carrier solution . Model controls were the 5/6 nephrectomized rats with 10 injection of carrier solution. Week 0 levels represent the Hb levels before treatment. *: week(s) post treatment.

[00028] Figure 10A - 10C show the potency comparisons of rHuEPO-Fc, 5 rHuEPO and darbepoetin-alfa for stimulating the colony formation of CFU-E and BFU-E in 5/6 nephrectomized rats treated with different doses and schedules. rHuEPO-Fc and darbepoietin-alpha (abbreviated Darbe.) treatment showed similar dose-dependent potencies For stimulating the CFU-E and BFU-E colony formation, while rHuEPO was less potent. A, s.c. once every week. B, s.c. once every 2 weeks. 10 C,, i.v. once every two weeks.

[00029] Figure 11 is a graph showing the serum levels of rHuEPO-Fc and rHuEPO after the intraveous injection of 5pg/kg of rHuEPO-Fc or rHuEPO to Rhesus monkeys (mean levels of 5 monkeys). 15 [00030] Figure 12 is a sequence listing showing the nucleotide sequence and the deduced amino-acid (aa) sequence of a wild type rHuEPO-FcC protein. The sequence particulars are the same as shown in Figure 1 except that a native, wild type cysteine residue is present at residue 172 of the mature fusion protein (i.e. the 6th 20 amino acid of the hinge region).

[00031] Figure 13 is a graph showing dose-dependent increase of hemoglobin (Hb) levels in normal mice treated with three times per week subcutaneous injection (s. c.) of rHuEPO-Fc (the mutant fusion protein of the present invention), rHuEPO- 25 FcC (the wild type fusion protein) and rHuEPO. Each point represents the mean Hb 11 level of the group(8). Normal control were normal mice with injection of carrier solution. Day 0 levels represent the Hb levels before treatment.

[00032] Figure 14 is a graph showing dose-dependent increase of hemoglobin 5 (Hb) levels in normal mice treated with once per week subcutaneous injection(s. c.) of rHuEPO-Fc, rHuEPO-FcC and rHuEPO. Each point represents the mean Hb level of the group(8). Normal control were normal mice with injection of carrier solution. Day 0 levels represent the Fib levels before treatment. 10 Detailed Description of the Invention [00033] Throughout the following description specific details are set forth in order to provide a more thorough understanding of the invention. However, the invention may be practiced without these particulars. In other instances, well known 15 elements have not been shown or described in detail to avoid unnecessarily obscuring the present invention. Accordingly, the specification and drawings are to be regarded in an illustrative, rather than a restrictive sense.

[00034] This application relates to a novel fusion protein having erythropoietic 20 properties. The fusion protein, referred to herein as rHuEPO-Fc, comprises a human erythropoietin (EPO) molecule recombinantly linked to an immunoglobulin Fc fragment. As discussed further below, the fusion protein may be in the form of a dimer consisting of two identical polypeptide subunits. In the embodiment shown schematically in Figure 1A, each polypeptide subunit, from the N-terminal to C- 25 terminal, consists of the polypeptide sequence of the human EPO molecule and the 12 polypeptide sequence of the hinge region, CH2 domain and CH3 domain of the Fc fragment of human immunoglobulin IgGl. The two polypeptide subunits are connected together by disulfide bonds between the respective hinge regions to form the dimer structure. The dimer thus has the same general shape as an IgG molecule 5 and exhibits better stability than free EPO molecules as discussed in the examples below.

[00035] As will be apparent to a person skilled in the art, the hinge region of an intact immunoglobulin provides the protein sufficient flexibility for effective antigen- 10 antibody binding. Similarly, in the present invention the hinge region is included in the design of the rHuEPO-Fc fusion protein to maintain its flexibility, especially when the fusion protein is in the dimer form. As described below, this allows the normal binding of the EPO portion of the rHuEPO-Fc fusion protein to EPO receptors to activate EPO biological functions. It is believed that the dimer form of the rHuEPO- 15 FC fusion protein, by providing two EPO molecules, is capable of inducing the optimal activation of EPO receptors (for example, by facilitating receptor crosslinking).

[00036] As demonstrated in the examples set forth below, the rHuEPO-Fc 20 fusion protein has been successfully synthesized using recombinant DNA techniques. The fusion protein has been shown in mice, rat and primate studies to exhibit a prolonged in vivo half-life and enhanced erythropoietic properties in comparison to naturally occurring or recombinant native human EPO. As used in this patent application, the terms “native human erythropoietin” and “native human EPO” mean 25 EPO having an unmodified wild type structure. As will be appreciated by a person 13 skilled in the art, native human EPO may be naturally occurring or recombinantly produced (e.g. rHuEPO alpha). The term “native human EPO” does not include rHuEPO analogs, such as darbepoetin alpha where the EPO structure has been significantly modified, such as by hyperglycosylation. 5 [00037] The nucleic acid sequence of the rHuEPO-Fc fusion protein of the present invention is shown in SEQ. ID. No. 1. The corresponding deduced amino acid sequence is shown in SEQ. ID. No. 2. The complete rHuEPO-Fc fusion protein is 399 amino acids in length. As shown in Figure IB, the complete rHuEPO-Fc fusion 10 protein consists of the EPO1 domain (166 amino acids), the hinge region (16 amino acids, underlined) and the CH2 and CH3 domains (217 amino acids). A signal or leader peptide sequence consisting of 27 amino acids is also shown in Figure IB. The signal peptide is cleaved during synthesis of rHuEPO-Fc. The nucleic and amino acid sequences of rHuEPO-Fc including the signal or leader peptide are shown in SEQ. ID. 15 No. 3 and SEQ. ID. No. 4 respectively.

[00038] As shown best in Figure IB and SEQ. ID. No. 2, the EPO domain has a cysteine residue near the C-terminal thereof at amino acid number 161. The hinge region includes 2 cysteine residues, at amino acid numbers 178 and 181 which are 20 boxed in Figure IB. The hinge region cysteine residues form the disulphide bonds between the polypeptide subunits of the homodimer as discussed above. The naturally occurring hinge region of a human IgGl fragment also has a cysteine at residue number 6 of the hinge region portion (measured from the N-terminal). In the present invention, the cysteine residue 6 of the hinge portion has been substituted by a 25 non-cysteine residue. In particular, in the embodiment of Figure IB and SEQ. ID, 14

No. 2, the amino acid cysteine has been substituted by glycine (at amino acid residue 172 of rHuEPO-Fc, which corresponds to residue 6 of the hinge region). As will be apparent to a person skilled in the art, other non-cysteine residues could also be substituted for cysteine at this location to avoid formation of a disulphide bond, 5 [00039] As a result of the amino acid substitution at residue 172, the first cysteine residue of the hinge region (at residue 178) is spaced 17 amino acids from the above-described cysteine residue of the EPO domain (at residue 161). The inventors believe that the minimum spacing between the cysteine residue 161 of the 10 EPO domain and the first cysteine residue of the hinge region should be at least 12 amino acids to enable successful assembly and/or EPO receptor binding of a homodimer of rHuEPO-Fc. That is, if residue 172 is a cysteine residue, an undesirable disulphide bond may potentially be formed, such as between cysteine

residues 161 and 172. This may alter the three dimensional structure of the EPO 15 molecule, resulting in biological inactivity or reduced biological activity.

[00040] In one embodiment of the invention, the EPO domain is linked directly to the Fc fragment portion of the fusion protein. By avoiding providing an external linker peptide, the preferred three dimensional structure of the rHuEPO-Fc fusion 20 peptide is maintained and the risk of triggering an undesirable immunogenic response is minimized. The hinge region of the Fc fragment is preferably at least 9 amino acids in length and is preferably in the range of about 10-20 amino acids in length. 15

Examples [00041] The following examples will further illustrate the invention in greater detail although it will be appreciated that the invention is not limited to the specific 5 examples. 1. Construction of the recombinant plasmid pCdEpo-Fc encoding the fusion protein of of HuEPO-Fc. 10 [00042] The full length DNA molecule, which encodes the amino-acid sequence of the polypeptide of rHuEPO-Fc, was generated by overlapping PCR using the following oligo primers (QIAGEN Inc., US): EF5: 5 ’ -ccggaattcgccaccatgggggtgcacgaatgtcctgcct-3'; 15 EF3: 5’-ttttccttttgcggccgcttatttacccggagacagggagag-3’; EFL 5: 5 ’ -aggcctgcaggacaggggacagagttgagcccaaatctggtgaca-3 ’; EFL3: 5 ’ -tgtcaccagatttgggctcaactctgtcccctgtcctgcaggcct-3 ’.

The sequences of the above-noted primers are listed in SEQ. I.D. Nos. 5-8 20 respectively.

[00043] EcoR I and Not I sites were introduced in EF5 and EF3, respectively. For optimal expression of the HuEPO-Fc protein in mammalian cells, the Kozak sequence (GCCACCATGG) was also introduced in EF5. EFL5 and EFL3 are 16 complementary sequences consisting of 3’-terminal DNA sequence of Epo (23bp) and S’-terminal DNA sequence of IgGl hinge (22bp).

[00044] First, an EPO DNA fragment of 0.6 kb was amplified by PCR 5 (Platinum Taq DNA Polymerase High Fidelity) with primers EF5 and EFL3 from plasmid p9E containing the full length of human EPO cDNA, Fc fragment of 0.7 kb with primers EF3 and EFL5 from plasmid pD containing the full length of human IgGl cDNA sequence, respectively (p9E and pD are from the inventors’ own lab).

The two fragments were then purified and mixed in equal amount. Using the mix as 10 template, the full length rHuEPO-Fc DNA of 1.3 kb was amplified by primers EF5 and EF3. The purified 1.3 kb fragment was digested by EocR I and Not I (New England Biolab Inc. US) and then cloned into EcoR 1/Not I-digested mammalian expression vector pCDl (Figure 2). The resulting recombinant vector was named pCdEpo-Fc and the inserted nucleic-acid sequence encoding the amino-acid sequence 15 of the HuEPO-Fc protein was confirmed by DNA sequencing. 2. Establishment of rHuEPO-Fc expression cell line [00045] Chinese hamster ovary cell with dihydrofolate reductase(dhfr) 20 deficiency (CHO/dhff, ATCC No.CRL-9096), which has been approved by FDA for biological substance production, was used as the host cell for rHuEPO-Fc expression.

[00046] The CHO-dhfr' cells were transfected with the recombinant vector pCdEpo-Fc using Lipofectamine (Gibco, Cat.No: 18292-037, USA). The supernatants 25 from the culture of selected clones were assayed by ELISA (Roche, Cat.No: 1 -693 17 417, Canada) for EPO activity. Positive clones were further screened under increasing Methotrexate( MTX) pressures. One cell line with highest rHuEPO-Fc protein expression was selected as the rHuEPO-Fc-expressing CHO cell-line, and gradually adapted to serum-free media (CD CHO Medium, Gibco, Cat.No: 10743-029, USA). 5 This rHuEPO-Fc-expressing CHO cell-line was used for the production of rHuEPO-Fc protein. 3. Purification of rHuEPO-Fc Protein 10 [00047] rHuEPO-Fc protein molecules contained in the supernatants collected from the serum-free media culturing the rHuEPO-Fc- expressing CHO cells were isolated at first by Protein A affinity chromatography (Amersham, Cat.No: 17 01, Canada). The isolated proteins were further purified by gel filtration in HiLoad 16/60 Superdex 200pg colume (Amersham, Cat.No:l 7 01, Canada). The purity 15 of the rHuEPO-Fc protein was more than 98% as determined by electrophoresis. 4. Determination of the Sizes of the pure rHuEPO-Fc Protein [00048] First, SDS-PAGE was carried out to determine the sizes of the pure 20 rHuEPO-Fc protein. As shown in Figure 3, a single band with molecular weight of about 180 kDa was seen on 8% Bis-Tris gel in the non-reduced condition, which measured the overall size of the protein with the existence of disulfide bonds. This indicated that most rHuEPO-Fc protein molecules were produced as the dimeric form, as expected from the design of the fusion protein. When SDS-PAGE analysis was 25 conducted in the reducing condition (lOOmM dithiothreitol, DTT) to break the 18 disulfide bonds, only the band with molecular weight of 75Kda was identified, consistent with the estimated molecular weight of single polypeptide chain of HuEPO-hinge region-CH2-CH3. 5 [00049] The accurate molecular weight of the pure rHuEPO-Fc fusion protein with glycosylation, determine by Mass Spectrum (MALDI-TOF-MS), was 111099 daltons (111.1 Kda). In this assay, only a single peak of protein was observed, indicating the purified rHuEPO-Fc protein was nearly 100% pure. The 15 amino acids of the N-terminal of the pure rHuEPO-Fc protein was determined by protein 10 sequence analysis as: APPRLICDSRVLERY. This was consistent with the sequence of the first 15 amino acids of the native human EPO polypeptide, and confirms that the purified rHuEPO-Fc protein does have the right and complete EPO molecule sequence as predicted by the DNA sequence encoding the amino-acid sequences of the rHuEPO-Fc fusion protein. 15 5. Enhanced erythropoietic activities of rHuEPO-Fc in normal mice [00050] In vivo experiments in mice were conducted to confirm the retaining of the erythropoietic activity of the rHuEPO-Fc protein and determine its efficacy 20 compared to rHuEPO and darbepoetin-alpha. For comparison purpose, all the doses of three EPOs used in the described animal experiments of the invention: our rHuEPO-Fc, rHuEPO (i.e. native human EPO) and darbepoetin-alpha, were the amounts of EPO molecule portion alone based on the molar basis. In respect to rHuEPO-Fc protein, the EPO portion contributes to 41.4% of the total rHuEPO-Fc 25 molecular weight as calculated by the ratio of the weight of amino acids of EPO in the 19 weight of the total amino acids of the whole rHuEPO-Fc molecule (166 aa among 399 aa). The EPO amount for rHuEPO-Fc was then decided as 41.4% of the total amount of the rHuEPO-Fc protein.

[00051 ] rHuEPO-Fc (stock concentration: 0.5mg/ml, purity of 98.6%) and native human rHuEPO (i.e.with natural human EPO structure)( 6000IU/0.5ml, manufactured by Kirin Brewery Co., Japan) were diluted in carrier solution (2.5mg/ml of human serum albumin, 5.8mg/ml of sodium citrate, 0.06mg/ml of citric acid and 5.8mg/ml of sodium chloride , pH5.5-5.6). The dose of rHuEPO in amount was calculated according to its activity/amount ration, BALB/c mice (6- to 8-week old, weighing 18-22g, equal numbers of male and female, purchased from Experiment Animal Center, AMMS, China) were grouped randomly with 6 in each group. Each group of mice was treated with one combination of one dose (0.1, 0.5, 2.5,12.5, 62.5pg/kg), one injection route( i.v. through the tail vein or s. c.) and one injection schedule (three times per week or once per week). The control group of mice was injected with the equal volume of carrier solution. The treatment lasted for 3 weeks and the total observation times were 5 weeks. Peripheral blood samples (tail vein) for measurement were taken before treatment, on the 4th day and 7th day of every week for 5 weeks. Hb was measured as the index by absorptiometry. Mean±SD was calculated from the data of each group and t test was conducted among different groups.

[00052] The administration of EPO three times per week to mice, provided that the EPOs have normal erythropoietic activity, would induce saturated stimulation of erythropoiesis. As shown in Figure 4, both groups treated with 3 times per week s.c. 20 had significant elevation of Hb levels even at the dose of 2.5pg/kg. This experiment demonstrated that rHuEPO-Fc exhibited an in vivo erythropoietic activity as effective as rHuEPO. The elevation of Hb levels in the treated group was dose-dependent. However, saturated elevation of the Hb levels was induced in mice at the dose of 5 I2.5pg/kg of rHuEPO-Fc, whereas the similar saturated elevation of the Hb levels was only achieved at the dose of 62.5pg/kg of rHuEPO. The elevation of Hb levels induced by 2.5pg/kg of rHuEPO-Fc was also greater than that by 2.5pg/kg of rHuEPO. These results suggested more potent erythropoietic stimulation by rHuEPO-

Fc than rHuEPO. 10 [00053] The erythropoietic potency of rHuEPO-Fc was further explored by reducing the injection times to once per week subcutaneously. As shown in Figure 5, the rHuEPO-Fc-treated groups showed dose-dependent elevation of Hb levels at the doses of 12.5, or 62.5pg(kg. Both doses of 12.5 and 62.5pg/kg of rHuEPO also 15 induced the elevation of Hb levels to the similar extent, which was much lower than that by 62.5pg/kg of rHuEPO-Fc. This strongly indicates that rHuEPO-Fc has enhanced erythropoietic activity in vivo. It is presumably due to either the prolonged half-life of the rHuEPO-Fc in vivo or improved EPO receptor binding/activation by the dimer EPO molecules in the rHuEPO-Fc protein, or by the combined effects of 20 both.

[00054] When the same doses (12.5pg/kg) of rHuEPO-Fc or rHuEPO were administrated intravenously either three times per week or once per week, elevation of the Hb levels was observed for all the treated groups (Figure. 6). However, i.v. 25 administration once per week of rHuEPO-Fc induced greater, more persistent 21 elevation of the Hb levels, which continued longer after the treatment was over. This data provides further support for the enhanced erythropoietic properties of the rHuEPO-Fc protein in comparison with rHuEPO having the structure of naturally occurring EPO protein. 5 6. Enhanced erythropoietic activities of rHuEPO-Fc in 5/6 nephrectomized rats [00055] Experiments in normal mice proved the enhanced erythropoietic activities of rHuEPO-Fc in vivo. To further observe the efficacy of rHuEPO-Fc in 10 stimulating erythropoiesis, pharmacodynamic studies were conducted in rats with experimental renal anemia that was made by 5/6 nephrectomy. The efficacy of rHuEPO-Fc was compared with those of rHuEPO and darbepoetin-alpha (60pg/ml, lot. No.N079, manufactured by Kirin Brewery Co., Japan). 15 [00056] Wistar rats (male and female in equal number, weighing 160-180g, purchased from Vitalriver Experiment Animal Inc., Beijing, China, Licence No. SCXK11 0008) were used in this invention to create the anemia model due to the renal functional failure by a two-step nephrectomy [27], 5/6 nephrectomy was done to rats with general anesthesia by two separate operations under sterile condition, 20 After 2/3 of the left kidney was resected, the rats were allowed to recover for twenty days. The right kidney was then resected carefully. Antibiotics were administrated to prevent infection after each operation. In total 5/6 of the kidney tissue was finally resected. The nephrectomized rats gradually developed renal function dissufficiency and anemia. The rats entered stable status of anemia 50 days after nephrectomy, and 25 were then randomly grouped (9/group) to start the administration of the EPOs. Each 22 group of rats was treated with one combination of one dose (2.5,12.5, 62.5pg/kg), one injection route (i.v. through the tail vein or s. c.) and one injection schedule (once per week or once every 2 weeks). The control group and model group of rats were injected with the equal volume of carrier solution. The treatment lasted for 4 weeks 5 and the total observation times were 6 weeks.

[00057] All doses (2.5, 12.5, 62.5pg/kg) of rHuEPO-Fc, administered subcutaneously once per week, induced dose-dependent elevation of the Hb levels comparing to the model control group that did not receive EPO treatment. Both 12.5 10 and 62,5pg/kg of rHuEPO or darbepoetin, administered subcutaneously once per week also induced elevation of Hb levels. The increased levels of Hb in both groups treated with 12.5 or 62.5 pg/kg of rHuEPO-Fc were significantly higher than those in groups treated with 12.5 or 62.5pg/kg of rHuEPO respectively. The Hb levels in 62.5pg/kg of rHuEPO-Fc -treated groups were also slightly higher than that in 15 62.5pg/kg of darbepoetin-treated group. After stopping treatment, the decrease of Hb levels in 62.5pg/kg of rHuEPO-Fc-treated group was much slower and the Hb levels remained higher than those of both normal control and model control groups until the end of observation (two weeks after treatment), indicating a stronger and/or a prolonged erythopoietic stimulation (summarized in Figure 7). 20 [00058] For the treatment of subcutaneous injection once every two weeks, only 12.5 or 62.5pg/kg of the three EPOs were administered (Figure 8). 12.5pg/kg of rHuEPO barely increased Hb levels compared to the model control group, and the weak erythropoietic response in the 62.5pg/kg of rHuEPO treated group failed to 25 bring the Hb levels to normal in comparison with the normal control group. 23

Treatments of either rHuEPO-Fc or darbepoetin at the doses of 12.5 or 62.5pg/kg induced significant elevation of Hb levels that was higher than that of the normal control group, indicating the effective correction of anemia status by both rHuEPO-Fc and darbepoetin. No significant differences were observed between same doses of rHuEPO-Fc and darbepoetin in terms of efficacy. The high dose of 62.5pg/kg resulted in the persistent increase of erythropoiesis until the termination of the observation (two weeks post treatment). This further suggested that rHuEPO-Fc and darbepoetin exhibit the property of long-lasting stimulation of erythropoiesis in vivo, which in turn could be transferred to the reduction of administration frequencies to patients clinically.

[00059] While darbepoetin has been approved for clinical application with less-frequent injections to increase the patient compliance and reduce the work burden of health care providers, these experimental data strongly indicate that rHuEPO-Fc disclosed in the current invention has at least the similar potential benefits. As discussed above, darbepoetin, as a mutant analog of the human EPO molecule containing additional sugar compounds (increased glycosylation), may have an increased risk of inducing irnmunogenesis in vivo due to the altered three dimensional structures. Only long-term observation of patients undergoing treatment with darbepoetin will give a decisive answer to the immunogenic risks of darbepoetin. In contrast, rHuEPO-Fc, without the modification of the EPO molecule portion, has a carbohydrate content identical or closely similar to that of native human EPO. The amounts of sialic acids in the inventors’ pure rHuEPO-Fc protein were around 10.0 mmol/mmol EPO, consistent with the reported parameters of rHuEPO. The Fc part of rHuEPO-Fc, with no external amino acid(s)/linking peptide, represents the general 24 structure of human IgGl, and theoretically would not lead to an immunogenic response. If approved clinically, rHuEPO-Fc may provide a better choice for patients than currently available rHuEPO and EPO analogs, especially those who need long term administration. 5 [00060] Once injected intravenously once every two weeks, both rHuEPO-Fc and darbepoetin (62.5pg/kg) were able to induce identical increases of Hb levels in the rats with renal anemia far above the normal Hb levels in the normal control rats (Figure 9). This further demonstrates the persistent stimulation of erythropoiesis by 10 rHuEPO-Fc, as darbepoetin’s efficacy has been clinically proven.

[00061] Data derived, from cell culturing experiments of bone marrow cells collected from the 5/6 nephrectomized rats after treatments (once per week or per two weeks, s.c. or i.v.) showed that rHuEPO-Fc, rHuEPO and darbepoetin all stimulated 15 the formation of CFU-E and BFU-E. The potencies of rHuEPO-Fc and darbepoetin were similar and stronger than that of rHuEPO (Figure 10).

[00062] Blood urinonitrogen (BUN) and Crea levels were similar in the treated groups and model control group. The levels of serum Fe in all the treated groups 20 w'ere higher that that of the model control group. Pathological examinations observed the increase distribution of red blood cell (RBC)-related cells in bone marrow and spleen of all EPO-treated rats. 25 7. Pharmacokinetic studies of rHuEPO-Fc in Rhesus monkeys [00063] As discussed above, the inventors have designed rHuEPO-Fc in such way that the EPO portion of the fusion protein retains the functional properties of 5 natural EPO, such as stimulating erythropoiesis, and the Fc fragment of human IgG 1 allows the stable existence of the fusion protein in circulation, thus extending its half-life in vivo. The above animal studies have demonstrated the erythropoietic activities of rHuEPO-Fc are enhanced in comparison with rHuEPO. The inventors have also conducted pharmacokinetic studies to determine the in vivo half-life of rHuEPO-Fc in 10 comparison to that of rHuEPO. Primates were used to generate data as they are biologically very similar to human beings.

[00064] Study design was based on literature reports and the experiments were conducted according to the general guidelines of pharmacokinetics. Two groups of 15 Rhesus monkeys with 5 monkeys in each group (3-5kg, purchased from the

Experiment Animal Center, AMMS, China) were injected intravenously with 5 pg/kg of rHuEPO-Fc or rHuEPO, respectively. Blood samples were taken before and at 0.017, 0.167, 0.5, 1, 2, 4, 8, 12, 24,48, 96, 168, 240 h after injection. Sera were collected by centrifugation and the serum rHuEPO-Fc or rHuEPO levels were 20 determined by using human erythropoietin enzyme-linked immunosorbent assay (ELISA) kits (purchased from R&D Systems, Minneapolis, MN). The average half-life (tl/2) of rHuEPO-Fc and rHuEPO injected intravenously was 35.24+/-5.15 h and 8.72+/-1.69 h respectively (summarized in Figure 11). 26 [00065] To observe the bioavailability of rHuEPO-Fc, 5ug/kg of rHuEPO-Fc was injected subcutaneously to 5 Rhesus monkeys. Blood samples were taken before and 1, 2, 5, 8,10, 12,15, 24, 48, 72, 96, 168,240 h after the injection, and the serum levels of rHuEPO-Fc were determined by the R&D kits. The bioavailability index 5 was calculated as 35.7]+/- 5.37% with the subcutaneous injection. This is identical to the reported bioavailability figures of darbepoetin-alpha (Aranesp™) in patients with chronic renal failure [9, 15] .

[00066] This data demonstrates that rHuEPO-Fc has a significantly prolonged 10 half-life in primates, and the in vivo half-life of rHuEPO-Fc is at least four fold longer than that of rHuEPO manufactured by Kirin Beer Brewing Co. of Japan. The prolonged half-life in vivo likely contributes to the enhanced erythropoietic activity of rHuEPO-Fc. 15 8. Immunogenicity of rHuEPO-Fc in Macaca fascicularis [00067] As indicated above, attention was given in the design of rHuEPO-Fc fusion protein to intentionally avoid or minimize the changes of the immunogenic properties of the rHuEPO-Fc fusion protein. The inventors avoided including/adding 20 any external amino acid(s) or linking peptide sequences in the fusion protein. The invented HuEPO-Fc fusion protein of the embodiment of Figure IB only contains the polypeptide sequences of the natural EPO protein and the Fc fragment (hinge region, CH2, CH3) of human IgGl, and would theoretically not induce an immunogenic response and the production of antibodies against rHuEPO-Fc protein. As will be 27 appreciated by a person skilled in the art, other embodiments having alternative structures are also encompassed by the present invention.

[00068] The following primate studies were conducted to observe the 5 immunogenicity of rHuEPC>-Fc protein. Ten crab-eating macaque (Macaca fascicularis)(maie/kma\e=5/5, ~ 5 years old, average weight of male 4.0±0.3kg, female is 2.9±0.4kg, purchased from Laboratory Animal Center, AMMS, China) were injected subcutaneously with 5pg/kg of purified rHuEPO-Fc 3 times per week for 4 weeks, and two were injected with equal volume of carrier solution as the control 10 animals. Sera were collected once a week for 5 weeks (1 week post-treatment) and tested for the specific antibodies against rHuEPO-Fc by ELISA using the purified rHuEPO-Fc (5pg/ml) as the coating antigen. In addition, RBC count and Hb levels in the peripheral blood were also determined within the experimental period. The resultant data shows that, while the stimulated erythropoiesis enhancement in the 15 rHuEPO-Fc-treated macaques was observed (the mean RBC numbers increased from 4.74xl09/ml to 6,67xl09/ml and the mean Hb levels from 12.2g/dl to 13.7g/dl), rHuEPO-Fc failed to induce detectable specific antibodies against the fusion protein.

These results indicate that rHuEPO-Fc fusion protein does not cause immunogenicity in primates. 20 28 9. Acute toxicity studies of rHuEPO-Fc in normal mice [00069] To assess the safety of rHuEPO-Fc fusion protein, acute toxic studies were conducted in animals. 5 [00070] Two groups of BALB/c mice (n=20, equal numbers of male and female, 5-6 weeks old, the average weight of female is 15.8±0.4g, male is 15.9±0.6g, purchased from Chinese Academy of Medicine, China) were injected intravenously once with excessive amount of purified rHuEPO-Fc (male=I3.3mg/kg, 10 female=l 3.2mg/kg) or equal volume of the carrier solution via their tail veins respectively. In addition to observing the instant reaction following injection, general behavior and status, activities, eating and defecation patterns and changes were monitored and recorded daily for 14 days. All mice were also weighed at day 7 and day 14. At day 15 post-injection, the anatomic examination of the main organs of the 15 mice were conducted. Pathologic examination would be conducted if any unusual changes or suspicious changes of the organs were observed.

[00071 ] All mice in the 2 groups had no obvious instant reaction following injection. Within the period of 14 days, no obvious changes of behavior, activities, 20 eating and defecation patterns were observed. Moreover, the weight of the mice in both groups increased steadily during the testing period, and no apparent differences were found between the 2 groups on day 7 or day 14 post injection. No abnormal or pathologic changes were detected in the tissues of brain, lung, heart, liver and kidney. These results indicate that administration of excessive amount of rHuEPO-Fc, far 29 more than required for exhibiting the normal erythropoiesis function, is safe and had no apparent toxic effects. 10. Comparison of wild type and mutated EPO fusion proteins [00072] Investigations were also conducted to compare wild type and mutated versions of EPO proteins. As described above, in one embodiment the invention includes a single amino acid mutation at amino acid residue 172 (C172G). For comparison purposes, a wild type fusion protein was also prepared having a cysteine amino acid at residue 172 (Figure 12). The wild type fusion protein was prepared in the same manner as Exampl es 1-3 above. With respect to the construction of the recombinant plasmid, the following oligo primers (QIAGEN Inc., US) were used (the altered amino acids in EFL5w and EFL3w in comparison to the primers of Example 1 are bolded): EF5: 5 ’-ccggaattcgccaccatgggggtgcacgaatgtcctgcct-3'; EF3: 5’-ttttccttttgcggccgcttatttacccggagacagggagag-3 EFL5w: 5’-aggcctgcaggacaggggacagagttgagcccaaatcttgtgaca-3 ’; EFL3w: 5’-tgtcacaagatttgggctcaactctgtcccctgtcctgcaggcct-3

The sequences of primers EFL5w and EFL3w are listed in SEQ. I.D. Nos. 9-10 respectively.

[00073] In vivo experiments in mice were conducted to compare the erythropoietic activity of the wild type fusion protein (herein referred to as rHuEPO- 30

FcC) with the mutated fusion protein (i.e. the rHuEPO-Fc protein of the present invention described above) and with recombinant human EPO (rHuEPO). For comparison purpose, all the doses of the three proteins used in this example, namely rHuEPO-Fc, rHuEPO-FcC and rHuEPO, were the amounts of the EPO molecule 5 portion alone on a molar basis. In respect to the rHuEPO-Fc and rHuEPO-FcC proteins, the EPO portion contributes to 41.4% of the total molecular weight as calculated by the ratio of the weight of amino acids of EPO to the weight of the total amino acids of the whole rHuEPO-Fc and rHuEPO-FcC molecules (i.e. 166 aa among 399 aa). 10

[00074] rHuEPO-Fc (stock concentration: 300gg/mI,), rHuEPO-FcC(stock concentration: 90pg/ml) and rHuEPO with the natural human EPO structure( 6000IU/0.5ml. manufactured by Kirin Brewery Co., Japan) were diluted in carrier solution (2.5mg/ml of human serum albumin, 5.8mg/ml of sodium citrate, 15 0.06mg/ml of citric acid and 5.8mg/ml of sodium chloride , pH5.5-5.6). The dose of rHuEPO in amount was calculated according to its activity/amount ratio. BALB/c mice (9- to 10-week old, weighing 18-22g, equal numbers of male and female, purchased from Experiment Animal Center, AMMS, China) were grouped randomly with 8 in each group. Each group of mice was treated with one combination of one 20 dose ( 2.5,12.5, 62.5pg/kg), one injection route(s. c.) and one injection schedule (three times per week or once per week). The control group of mice was injected with the equal volume of carrier solution. The treatment lasted for 26 days. Peripheral blood samples (tail vein) for measurement were taken before treatment, on the 2nd, 6th, 9th, 13th, 16th, 19th, 22nd and 26th days of treatment. Hb was measured as the index by 31 absorptiometry. Mean±SD was calculated from the data of each group and t test was conducted among different groups.

[00075] As shown in Figure 13, administration of all three EPO proteins at 5 intervals of three times per week stimulated erythropoiesis. At either the dose of 2.5 pg/kg or 12.5 pg/kg, rHuEPO-Fc induced higher elevation of Hb levels than rHuEPO. The highest elevation of Hb levels was achieved by the 12.5 pg/kg dose of rHuEPO-Fc. Both the 2.5 pg/kg and 12.5 pg/kg doses of rHuEPO-FcC induced much weaker erythropoiesis than equivalent doses of rHuEPO and rHuEPO-Fc as indicated by the 10 significant lower elevation of Hb levels in the rHuEPO-FcC-treated groups. In fact, 12.5 pg/kg of rHuEPO-FcC induced lower elevation of Hb levels than 2.5 pg/kg of rHuEPO. These results suggest that rHuEPO-FcC has impaired erythropoietic activity in vivo in comparison to rHuEPO having the natural EPO molecular sequence. By contrast, the rHuEPO-Fc fusion protein of the present invention exhibited more 1 5 potent erythropietic functions. The administration of the three EPO proteins at intervals of three times per week largely excluded the impact of differences in the half-life of the proteins.

[00076] The erythropoietic potency of rHuEPO-Fc and rHuEPO-FcC was 20 further evaluated by reducing the injection times to once per week subcutaneously.

As shown in Figure 14, the rHuEPO-Fc-treated groups showed higher elevation of Hb levels than rHuEPO-treated ones at the doses of 12.5 pg/kg or 62.5 pg/kg. In contrast, rHuEPO-FcC induced much weaker elevation of Hb levels than that induced by rHuEPO. For example, 12.5 pgkg of rHuEPO induced higher elevation of Hb levels 25 than that induced by 62.5 pg/kg of rHuEPO-FcC at most time points. This further 32 indicates that by reducing the administration times to include the effects of half-life, rHuEPO-FcC exhibits much weaker erythropoietic functions in vivo in comparison to rHuEPO having the natural EPO molecular sequence and in comparison to the rHuEPO-Fc fusion protein of the present invention. 5 [00077] In summary, these results demonstrate that rHuEPO-FcC, formed by the fusion of natural molecular sequences of both human EPO and human Fc fragment (hinge, CH2 and CEO), exhibits much weaker erythropoietic functions in vivo in comparison to the rHuEPO having the natural EPO molecular sequence. In 10 particular, the erythropoietic activities of the rHuEPO-FcC fusion protein are less than

1/5 of those of natural EPO molecule. This indicates that the fusion between EPO molecule and the natural sequence of human Fc fragment impairs the functional properties of the EPO molecule. By the single amino acid replacement at the first cysteine residue in the hinge region of the Fc fragment, the rHuEPO-Fc fusion protein 15 of the present invention comprising the natural EPO molecule sequence and the mutant Fc fragment shows more potent erythropoietic functions in vivo compared to the natural EPO molecule. This data suggests that the first cysteine residue in the hinge region of the wild type Fc fragment somehow interferes with the EPO molecule, likely by causing structural changes to the EPO molecule, and this in turn impairs the 20 functional properties of the EPO molecule in stimulating erythropoiesis.

[00078] As will be apparent to those skilled in the art in the light of the foregoing disclosure, many alterations and modifications are possible in the practice of this invention without departing from the spirit or scope thereof. 25 33

References 1. Cohen J, et al. Erythropoietin and its receptor: signaling and clinical manifestations. IMAJ 4, pp 1072-1076(2002) 2. Blackwell K, et al. rHuEPO and improved treatment outcomes: potential modes of action. The Oncologist 9(suppl 5), pp41-47(2004). 3. Lappin TR, et al. EPO’s alter ego: erythropoietion has multiple actions. Stem Cells 20, pp485-492(2002). 4. Maiese K, et al. New avenues of exploration for erythropoietin. JAMA 293(1), pp90-95(2005). 5. Fisher JW. Erythropoietin: physiologic and pharmacologic aspects. Proc. Soc. Exp. Biol. Med. 216, pp358-369(1997). 6. Jelkmann W. Molecular biology of erythropoietin. Internal Med. 43(8), pp649-659(2004). 7. Ng T, et al. Recombinant erythropoietin in clinical practice. Postgrad. Med. J. 79, pp367-376(2005). 8. Weiss G, et al. Anemia of chronic disease. N. Engl. J. Med. 352(10), pplOl l-1023(2005). 9. Macdougall IC. An overview of the efficacy and safety of novel erythropoiesis stimulating proteinfNESP). Nephrol, Dial. Transplant. 16(suppl 3), pp 14-21(2001). 10. Joy MS. Darbepoetin-alfa: a novel erythropoiesis-stimulating protein. Ann. Pharmacother. 36, ppl 183-1192(2002). 11. Ellitt S, et al. Enhancement of therapeutic protein in vivo activities through glycoengineering. Nature Biotechnology 21, pp414-421(2003). 12. Elliott S, et al. Control of rHuEPO biological activity: the role of carbohydrate. Experimental Hematology 32, ppl 146-1155(2004). 13. Egrie JC, et al. Darbepoetin-alfa has a longer circulating half-life and greater in vivo potency than recombinant human erythropoietin. Exp. Hematol. 31, pp290-299(2003). 14. Egrie JC, et al. Development and characterization of novel eiythropoiesis stimulating proteinfNESP). British J. Caner. 84(suppl 1), pp3-10(2001). 15. Macdougall IC, et al. Pharmacokinetics of novel erythropoiesis stimulating protein compared with epoetin alfa in dialysis patients. J. Am. Soc. Nephrol. 10, pp2392-2395(1999). 34 16. Jolling K, et al. Population pharmacokinetic analysis of peglated human erythropoietin in rats. J. Pharm. Sci. 93(12), pp3027-3038(2004) 17. Dalle B, et al. Dimeric erythropoietin fusion protein with enhanced erythropoietic activity in vtro and in vivo. Blood 97(12), pp3776-3782(2001). 18. Kochendoerfer GG, et al. Design and chemical synthesis of a homogeneous polymer-modified erythropoiesis protein. Science 299 pp884-887(2003). 19. Sytkowski AJ, et al. Human erythropoietin dimmers with markedly enhanced in vivo activity. Proc. Natl. Acad. Sci. USA 95, ppl 184-1188(1998). 20. Sytkowski AJ, et al.An erythropoietin fusion protein comprised of identical repeating domains exhibits enhanced biological proterties. J. Biol. Chem. 274(35), pp24773-24778(1999). 21. Jones TD, et al. The development of a modified human IFN-a2b linked to the Fc portion of human IgGl as a novel potential therapeutic for the treatment of hepatitis C virus infection. J. Interferon. Cytokine. Res. 24, pp560-5 72(2004). 22. Lo KM, et al. High level expression and secretion of Fc-X fusion proteins in mammalian cells. Protein engineering 11(6), pp495-500(1998), 23. Mohler KM, et al. Soluble tumor necrosis factor(TNF) receptors are effective therapeutic agents in lethal endotoxemia and function simultaneously as both TNF earners and TNF antabonists. J. Immunol. 151(3), ppl548-1561(1993). 24. Way JC, et al. Improvement of Fc-eiythropoietin structure and pharmacokinetics by modification at a disulfide bond. Protein Engineering, Design & Selection 18(3), ppl 11-118(2005). 25. Goldenberg MM. Etanercept, a novel drug for the treatment of patients with severe, active rheumatoid arthritis. Clin. Ther. 21(1), pp75-87(1999) 26. Wong VK, et al. The use of alefacept in the treatment of psoriasis. Skin Therapy Lett. 8(6), ppl-2(2003) 27. Chanutin A, et al. Experimental renal insufficiency produced by partial nephrectomy. Arch. Intern, Med. 49,pp767-787 (1932). 35

N240 0002 - sequence nstmg.ST25.txt SEQUENCE LISTING <110> Novagenetics, inc.

<120> RECOMBINANT HUMAN EPO-FC FUSION PROTEINS WITH PROLONGED HALF-LIFE AND ENHANCED ERYTHROPOIETIC ACTIVITY IN VIVO <130> N24O 0002 <150> US 11/340,661 <151> 2006 27 <160> 10 <170> Patentln version 3.3 <210> 1 <211> 1200

<212> DNA <213> Artificial <220> <223> DNA encoding rHuEPO-Fc fusion protein <400> 1 gccccaccac gcctcatctg tgacagccga gaggccgaga atatcacgac gggctgtgct gtcccagaca ccaaagttaa tttctatgcc gtagaagtct ggcagggcct ggccctgctg ttggtcaact cttcccagcc gtgggagccc ggccttcgca gcctcaccac tctgcttcgg cctccagatg cggcctcagc tgctccactc ctcttccgag tctactccaa tttcctccgg tgcaggacag gggacagagt tgagcccaaa tgcccagcac ctgaactcct ggggggaccg gacaccctca tgatctcccg gacccctgag gaagaccctg aggtcaagtt caactggtac acaaagccgc gggaggagca gtacaacagc ctgcaccagg actggctgaa tggcaaggag ccagccccca tcgagaaaac catctccaaa tacaccctgc ccccatcccg ggatgagctg gtcaaaggct tctatcccag cgacatcgcc aacaactaca agaccacgcc tcccgtgctg aagctcaccg tggacaagag caggtggcag catgaggctc tgcacaacca ctacacgcag <210> 2 <211> 399 gtcctggaga ggtacctctt ggaggccaag 60 gaacactgca gcttgaatga gaatatcact 120 tggaagagga tggaggtcgg gcagcaggcc 180 tcggaagctg tcctgcgggg ccaggccctg 240 ctgcagctgc atgtggataa agccgtcagt 300 gctctgcgag cccagaagga agccatctcc 360 cgaacaatca ctgctgacac tttccgcaaa 420 ggaaagctga agctgtacac aggggaggcc 480 tctggtgaca aaactagtac atgcccaccg 540 tcagtcttcc tcttcccccc aaaacccaag 600 gtcacatgcg tggtggtgga cgtgagccac 660 gtggacggcg tggaggtgca taatgccaag 720 acgtaccgtg tggtcagcgt cctcaccgtc 780 tacaagtgca aggtctccaa caaagccctc 840 gccaaagggc agccccgaga accacaggtg 900 accaagaacc aggtcagcct gacctgcctg 960 gtggagtggg agagcaatgg gcagccggag 1020 gactccgacg gccccttctt cctctacagc 1080 caggggaacg tcttctcatg ctccgtgatg 1140 aagagcctct ccctgtctcc gggtaaataa 1200 36 N240 0002 - sequence listing.ST25.txt <212> PRT <213> Artificial <22O> <223> amino acid sequence of rHuEPO-Fc fusion protein <4oo> ; 1 Ala pro pro Arg Leu lie cys Asp Ser Arg Val Leu Glu Arg Tyr Leu 1 5 10 15 Leu Glu Ala cys Glu Ala Glu Asn lie Thr Thr Gly cys Ala Glu HIS 20 25 30 cys Ser Leu Asn Glu Asn lie Thr Val Pro Asp Thr Lys Val Asn Phe 35 40 45 Tyr Ala Trp Lys Arg Met Glu Val Gly Gin Gin Ala Val Glu val Trp 50 55 60 Gin Gly Leu Al a Leu Leu Ser Gl u Ala Val Leu Arg Gly Gin Ala Leu 65 70 75 80 Leu Val Asn Ser Ser Gin pro Trp Glu Pro Leu Gin Leu Hi s val Asp 85 90 95 Lys Ala Val ser Gly Leu Arg Ser Leu Thr Thr Leu Leu Arg Ala Leu 100 105 110 Arg Ala Gin Lys Glu Ala lie Ser Pro Pro ASP Ala Ala ser Ala Ala 115 120 125 pro Leu Arg Thr lie Thr Ala Asp Thr Phe Arg Lys Leu Phe Arg val 130 135 140 Tyr Ser Asn Phe Leu Arg Gly Lys Leu Lys Leu Tyr Thr Gly Glu Ala 145 150 155 160 Cys Arg Thr Gly Asp Arg Val Glu Pro Lys Ser Gly Asp Lys Thr Ser 165 170 175 Thr cys Pro Pro cys pro Ala pro Glu Leu Leu Gly Gly pro ser val 180 185 190 Phe Leu Phe Pro Pro Lys Pro Lys Asp Thr Leu Met lie Ser Arg Thr 195 200 205 pro Glu val Thr Cys val val Val Asp val Ser Hi s Glu Asp Pro Glu 210 215 220 val Lys Phe Asn Trp Tyr val Asp Gly val Glu Val Hi s Asn Al a Lys 225 230 235 240 37

Thr Lys Pro N240 0002 - sequen Arg Glu Glu Gin Tyr Asn ser ce listi Thr Tyr ng.ST2 5. txt Val 255 Ser Arg Val 245 250 Val Leu Thr val Leu His Gin ASp Trp Leu Asn Gly Lys Glu Tyr Lys 260 265 270 cys Lys Val Ser Asn Lys Ala Leu pro Ala pro lie Glu Lys Thr lie 275 280 285 Ser Lys Ala Lys Gly Gin Pro Arg Glu Pro Gin val Tyr Thr Leu Pro 290 295 300 Pro Ser Arg Asp Glu Leu Thr Lys Asn Gin val Ser Leu Thr cys Leu 305 310 315 320 Val Lys Gly Phe Tyr pro ser Asp lie Ala val Glu Trp Glu Ser Asn 325 330 335 Gly Gin Pro Glu Asn Asn Tyr Lys Thr Thr Pro Pro Val Leu Asp Ser 340 345 350 Asp Gly Pro Phe phe Leu Tyr Ser Lys Leu Thr val Asp Lys Ser Arg 355 360 365 Trp Gin Gin Gly Asn val Phe Ser cys Ser Val Met His Glu Al a Leu 370 375 380 His Asn His Tyr Thr Gl n Lys Ser Leu Ser Leu ser Pro Gly Lys 385 390 395 <210> 3 <211> 1281 <212> DNA <213> Artificial <220> <223> DNA encoding rHuEPO-Fc fusion protein with 27 amino acid signal peptide <400> 3 atgggggtgc acgaatgtcc tgcctggctg tggcttctcc tgtccctgct gtcgctccct 60 ctgggcctcc cagtcctggg cgccccacca cgcctcatct gtgacagccg agtcctggag 120 aggtacctct tggaggccaa ggaggccgag aatatcacga cgggctgtgc tgaacactgc 180 agcttgaatg agaatatcac tgtcccagac accaaagtta atttctatgc ctggaagagg 240 atggaggtcg ggcagcaggc cgtagaagtc tggcagggcc tggccctgct gtcggaagct 300 gtcctgcggg gccaggccct gttggtcaac tcttcccagc cgtgggagcc cctgcagctg 360 catgtggata aagccgtcag tggccttcgc agcctcacca ctctgcttcg ggctctgcga 420 gcccagaagg aagccatctc ccctccagat gcggcctcag ctgctccact ccgaacaatc 480 actgctgaca ctttccgcaa actcttccga gtctactcca atttcctccg gggaaagctg 540 38 N240 0002 - sequence listing.ST25.txt aagctgtaca caggggaggc ctgcaggaca ggggacagag ttgagcccaa atctggtgac 600 aaaactagta catgcccacc gtgcccagca cctgaactcc tggggggacc gtcagtcttc 660 ctcttccccc caaaacccaa ggacaccctc atgatctccc ggacccctga ggtcacatgc 720 gtggtggtgg acgtgagcca cgaagaccct gaggtcaagt tcaactggta cgtggacggc 780 gtggaggtgc ataatgccaa gacaaagccg cgggaggagc agtacaacag cacgtaccgt 840 gtggtcagcg tcctcaccgt cctgcaccag gactggctga atggcaagga gtacaagtgc 900 aaggtctcca acaaagccct cccagccccc atcgagaaaa ccatctccaa agccaaaggg 960 cagccccgag aaccacaggt gtacaccctg cccccatccc gggatgagct gaccaagaac 1020 caggtcagcc tgacctgcct ggtcaaaggc ttctatccca gcgacatcgc cgtggagtgg 1080 gagagcaatg ggcagccgga gaacaactac aagaccacgc ctcccgtgct ggactccgac 1140 ggccccttct tcctctacag caagctcacc gtggacaaga gcaggtggca gcaggggaac 1200 gtcttctcat gctccgtgat gcatgaggct ctgcacaacc actacacgca tccctgtctc cgggtaaata a <210> 4 <211> 426 <212> PRT <213> Artificial <220> gaagagcctc 1260 1281 <223> amino acid sequence of thuepo-fc fusion protein acid signal peptide <400> 4 with 27 amino

Met 1 Gly val Hi s Glu 5 cys pro Ala Trp Leu Trp Leu Leu Leu 10 Ser 15 Leu Leu Ser Leu Pro Leu Gly Leu Pro Val Leu Gly Ala Pro Pro Arg Leu 20 25 30 lie Cys Asp Ser Arg Val Leu Glu Arg Tyr Leu Leu Glu Ala Lys Glu 35 40 45 Ala GlU Asn lie Thr Thr Gly cys Ala Glu His cys Ser Leu Asn Glu 50 55 60 Asn lie Thr val pro Asp Thr Lys val Asn Phe Tyr Ala Trp Lys Arg 65 70 75 80 Met Glu val Gly Gin Gin Ala Val Glu val Trp Gin Gly Leu Ala Leu 85 90 95 Leu Ser Glu Al a val Leu Arg Gly Gin Ala Leu Leu val Asn Ser Ser 100 105 110 Gin Pro Trp Glu Pro Leu Gin Leu His Val Asp Lys Ala val Ser Gly 115 120 125 39 N240 0002 - sequence listing.ST25.txt

Leu Arg 130 Ser Leu Thr Thr Leu Leu Arg Ala 135 Leu Arg 140 Ala Gin Lys Gl u Ala lie Ser Pro Pro Asp Ala Ala Ser Ala Ala Pro Leu Arg Thr lie 145 150 155 160 Thr Ala Asp Thr Phe Arg Lys Leu Phe Arg val Tyr Ser Asn Phe Leu 165 170 175 Arg Gly Lys Leu Lys Leu Tyr Thr Gly Glu Ala cys Arg Thr Gly Asp 180 185 190 Arg val Glu Pro Lys Ser Gly Asp Lys Thr Ser Thr cys Pro Pro cys 195 200 205 Pro Ala Pro Glu Leu Leu Gly Gly Pro Ser val Phe Leu Phe Pro Pro 210 215 220 Lys Pro Lys Asp Thr Leu Met lie Ser Arg Thr pro Glu Val Thr cys 225 230 235 240 Val Val Val Asp Val Ser Hi s Glu Asp pro Glu val Lys Phe Asn Trp 245 250 255 Tyr val Asp Gly val Glu Val His Asn Ala Lys Thr Lys Pro Arg Glu 260 265 270 Glu Gl n Tyr Asn Ser Thr Tyr Arg val val Ser Val Leu Thr Val Leu 275 280 285 His Gin Asp Trp Leu Asn Gly Lys Glu Tyr Lys cys Lys val Ser Asn 290 295 300 Lys Ala Leu Pro Ala Pro lie Glu Lys Thr lie Ser Lys Ala Lys Gly 305 310 315 320 Gin Pro Arg Glu Pro Gin val Tyr Thr Leu Pro Pro Ser Arg Asp Glu 325 330 335 Leu Thr Lys Asn Gin Val Ser Leu Thr cys Leu val Lys Gly Phe Tyr 340 345 350 Pro Ser Asp lie Al a Val Glu Trp Glu Ser Asn Gly Gin Pro Glu Asn 355 360 365 Asn Tyr Lys Thr Thr Pro Pro Val Leu Asp Ser Asp Gly pro Phe Phe 370 375 380 Leu Tyr ser Lys Leu Thr Val Asp Lys Ser Arg Trp Gin Gin Gly Asn 385 390 395 400 40 N240 0002 - sequence listing.ST25.txt

Val Phe Ser Cys Ser val Met His Glu Ala Leu His Asn His Tyr Thr 405 410 415

Gin Lys Ser Leu ser Leu Ser Pro Gly Lys 420 425 <210> 5 <211> 40

<212> DNA <213> Artificial <220> <223> oligonucleotide primer <400> 5 ccggaattcg ccaccatggg ggtgcacgaa tgtcctgcct 40 <210> 6 <211> 42

<212> DNA <213> Artificial <220> <223> oligonucleotide primer <400> 6 ttttcctttt gcggccgctt atttacccgg agacagggag ag 42 <210> 7 <211> 45

<212> DNA <213> Artificial <220> <223> oligonucleotide primer <400> 7 aggcctgcag gacaggggac agagttgagc ccaaatctgg tgaca 45 <210> 8 <211> 45

<212> DNA <213> Artificial <220> <223> oligonucleotide primer <400> 8 tgtcaccaga tttgggctca actctgtccc ctgtcctgca ggcct 45 <210> 9 <211> 45

<212> DNA <213> Artificial <22O> <223> oligonucleotide primer <400> 9 aggcctgcag gacaggggac agagttgagc ccaaatcttg tgaca 45 41 N240 0002 - sequence listing.ST25.txt <210> 10 <211> 45

<212> DNA <213> Artificial <220> <223> oligonucleotide primer <400> 10 tgtcacaaga tttgggctca actctgtccc ctgtcctgca ggcct 45 42