Expression and export of angiogenesis inhibitors as immunofusins

FIELD OF THE INVENTION

[0002] This invention relates generally to methods and compositions for making and using fusion proteins containing an angiogenesis inhibitor. More particularly, the invention relates to methods and compositions for making and using fusion proteins called immunofusins which contain an immunoglobulin Fc region and an angiogenesis inhibitor.

BACKGROUND OF THE INVENTION

[0003] Two potent angiogenesis inhibitors, angiostatin (O'Reilly et al. (1994) Cell 79:315) and endostatin (O'Reilly et al. (1997) Cell 88:277), were discovered and found to be generated naturally by primary tumors. Both proteins are specific inhibitors of endothelial cell proliferation and inhibit tumor growth by blocking angiogenesis, the formation of new blood vessels that nourish tumors. Studies have shown that these angiogenesis inhibitors are non-toxic even at very high doses and that they may suppressed the growth of metastases and primary tumors may regress to a dormant microscopic state. Both inhibitors were identified as proteolytic fragments of much larger intact molecules. Angiostatin was found to be a fragment of plasminogen, and endostatin a fragment of collagen XVIII.

[0004] These two proteins have generated great interest in the cancer area because they have been shown to suppress the growth of many different types of tumors in mice, with no obvious side effects or drug resistance. Traditional chemotherapy generally leads to acquired drug resistance caused primarily by the genetic instability of cancer cells. Rather than targeting cancer cells, therapies using angiogenesis inhibitors target the normal endothelial cells, which support the growth of the tumor. Because endothelial cells are genetically stable, it is possible that angiogenesis inhibitor therapies may result in less drug resistance. Studies indicate that drug resistance did not develop in mice exposed to prolonged anti-angiogenic therapy using endostatin. Furthermore, repeated cycles of endostatin treatment in mice resulted in prolonged tumor dormancy and no recurrence of tumors following discontinuation of therapy (Boehm et al. (1997) Nature 390:404).

[0005] Despite promising results in mice, it has not been possible to produce clinical grade soluble, active angiostatin and endostatin in commercial quantities using E. coli, baculoviral, yeast, and mammalian expression systems. Expression in E. coli yielded insoluble protein aggregates of undefined composition, which could not be injected into humans. Other production methods, such as baculovirus and mammalian expression systems, yielded very low levels of the recombinant proteins (O'Reilly et al. (1997) Cell 88:277).

[0006] The poor yields of the expression systems to date may be explained by both angiostatin and endostatin being internal fragments of much larger proteins. The truncated proteins may not fold properly in the absence of the residues that are cleaved from the precursor molecules. For example, angiostatin has 26 cysteine residues which form numerous disulfide bonds. Expression of angiostatin by itself may not provide the optimal environment for these numerous disulfide bonds to form correctly in the secretory pathway. Also, the recombinant endostatin protein produced in E. coli precipitated during dialysis, possibly due to the hydrophobicity of endostatin (O'Reilly et al. (1997) Cell 88:277).

[0007] A major hurdle with the use of angiostatin and endostatin in their present forms is that relatively large amounts of proteins have to be injected daily for weeks to months to achieve the desired clinical outcome. For example, in current mouse models, dosages of 20 mg/kg/day of endostatin are needed to demonstrate optimal efficacy (Boehm et al. (1997) Nature 390:404). Given that there is an urgent need to test endostatin and angiostatin clinically, a production method that can generate large quantities of clinical grade material is important.

[0008] One expression system that has been used to produce high level expression of fusion proteins in mammalian cells is a DNA construct encoding, a signal sequence, an immunoglobulin Fc region and a target protein. The fusion product of this construct generally is termed an “immunofusin.” Several target proteins have been expressed successfully as immunofusins which include: IL2, CD26, Tat, Rev, OSF-2, βIG-H3, IgE Receptor, PSMA, and gp120. These expression constructs are disclosed in U.S. Pat. No. 5,541,087 and U.S. Pat. No. 5,726,044, the disclosures of which are incorporated herein by reference.

[0009] A major purpose of expressing recombinant fusion proteins in mammalian cells has been to attempt to confer novel or useful properties to the hybrid molecules, e.g., proper folding, increased solubility, targeting of a cytokine or toxin in vivo, Fe receptor binding, complement fixation, protein A binding, increased circulation half-life, and increased ability to cross the blood-brain barrier. Examples of recombinant fusion proteins produced in mammalian cells include cytokine immunoconjugates (Gillies et al. (1992) Proc. Natl. Acad. Sci. USA 89:1428; Gillies et al. (1993) Bioconjugate Chemistry 4:230), immunoadhesins (Capon et al. (1989) Nature 337:525), immunotoxins (Chaudhary et al. (1989) Nature 339:394), and a nerve growth factor conjugate (Friden et al. (1993) Science 259:373). Each of the foregoing publications is incorporated herein by reference.

[0010] It is an object of the invention to provide novel DNAs which facilitate efficient production and secretion of angiogenesis inhibitors in a variety of mammalian host cells. It is another object of the invention to provide methods for treating mammals with nucleic acids encoding, or amino acid sequences defining angiogenesis inhibitor proteins, including non-native, biosynthetic, or otherwise artificial proteins such as proteins which have been created by rational design.

SUMMARY OF THE INVENTION

[0011] The present invention features methods and compositions useful in making and using fusion proteins containing an angiogenesis inhibitor protein. The fusion proteins can facilitate a high level expression of biologically active angiogenesis inhibitor proteins. The angiogenesis inhibitor proteins can then be cleaved from the fusion protein and combined with a pharmaceutically acceptable carrier prior to administration to a mammal, for example, a human. Alternatively, nucleic sequences encoding, or amino acid sequences defining the angiogenesis inhibitor containing fusion proteins can be combined with a pharmaceutically acceptable carrier and administered to the mammal.

[0012] In one aspect, the invention provides nucleic acid molecules, for example, DNA or RNA molecules, encoding a fusion protein of the invention. The nucleic acid molecule encodes a signal sequence, an immunoglobulin Fc region, and at least one target protein, also referred to herein as the angiogenesis inhibitor protein, selected from the group consisting of angiostatin, endostatin, a plasminogen fragment having angiostatin activity, a collagen XVIII fragment having endostatin activity, and combinations thereof. In a preferred embodiment, the nucleic acid molecule encodes, serially in a 5′ to 3′ direction, the signal sequence, the immunoglobulin Fc region and the target protein sequence. In another preferred embodiment, the nucleic acid molecule encodes, serially in a 5′ to 3′ direction, the signal sequence, the target sequence, and immunoglobulin Fc region.

[0013] In another preferred embodiment, the immunoglobulin Fc region comprises an immunoglobulin hinge region and preferably comprises at least one immunoglobulin constant heavy region, for example, an immunoglobulin constant heavy 2 (CH2) domain, an immunoglobulin constant heavy 3 (CH3) domain), and depending upon the type of immunoglobulin used to generate the Fc region, optionally an immunoglobulin constant heavy region 4 (CH4) domain. In a more preferred embodiment, the immunoglobulin Fc region comprises a hinge region, a CH2 domain and a CH3 domain. Under certain circumstances, the immunoglobulin Fc region preferably lacks at least the CH1, domain. Although the immunoglobulin Fc regions may be based on any immunoglobulin class, for example, IgA, IgD, IgE, IgG, and IgM, immunoglobulin Fc regions based on IgG are preferred.

[0014] In another embodiment, the nucleic acid of the invention can be incorporated in operative association into a replicable expression vector which can then be transfected into a mammalian host cell. In another preferred embodiment, the invention provides host cells harboring such nucleic acid sequences of the invention.

[0015] In another aspect, the invention provides a fusion protein comprising an immunoglobulin Fc region linked, either directly through a polypeptide bond or by means of a polypeptide linker, to a target protein selected from the group consisting of angiostatin, endostatin, a plasminogen fragment having angiostatin activity, a collagen XVIII fragment having endostatin activity, and combinations thereof. The target protein may be fused via its C-terminal end to an N-terminal end of the immunoglobulin Fc region. However, in a more preferred embodiment the target protein is fused via its N-terminal end to a C-terminal end of the immunoglobulin Fc region.

[0016] In another embodiment, the fusion protein may comprise a second target protein selected from the group consisting of angiostatin, endostatin, a plasminogen fragment having angiostatin activity, and a collagen XVIII fragment having endostatin activity. In this type of construct the first and second target proteins can be the same or different proteins. For example, in a preferred embodiment, the fusion protein comprises a first target protein of angiostatin, an immunoglobulin Fc region and a second target protein of endostatin. The first and second target proteins may be linked together, either directly or by means of a polypeptide linker. Alternatively, both target proteins may be linked, either directly or via a polypeptide linker, to the immunoglobulin Fc region. In the latter case, the first target protein is connected to an N-terminal end of the immunoglobulin Fc region and the second target protein is connected to a C-terminal end of the immunoglobulin Fc region.

[0017] In another embodiment, two fusion proteins may associate, either covalently, for example, by a disulfide or peptide bond, or non-covalently, to produce a multimeric protein. In a preferred embodiment, two fusion proteins are associated covalently by means of one or more disulfide bonds through cysteine residues, preferably located within immunoglobulin hinge regions disposed within the immunoglobulin Fc regions of both chains.

[0018] In a preferred embodiment, the target protein comprises a plasminogen fragment having a molecular weight of approximately 40 kD and, optionally comprises, an amino acid sequence as set forth in SEQ ID NO: 3. In another preferred embodiment, the target protein comprises a collagen XVIII fragment having an amino acid sequence set forth in SEQ ID NO: 1. Furthermore, the target protein can be full-length angiostatin or endostatin or bioactive fragments thereof. The source of the target protein in generating certain fusion proteins will depend upon the intended use of the target protein. For example, if the target protein is to be administered to a human, the target protein preferably is of human origin.

[0019] In another aspect, the invention provides methods of producing a fusion protein comprising an immunoglobulin Fc region and a target protein selected from the group consisting of angiostatin, endostatin, a plasminogen fragment having angiostatin activity, and a collagen XVIII fragment having endostatin activity. The method comprises the steps of (a) providing a mammalian cell containing a DNA molecule encoding such a fusion protein, either with or without a signal sequence, and (b) culturing the mammalian cell to produce the fusion protein. The resulting fusion protein can then be harvested, refolded, if necessary, and purified using conventional purification techniques well known and used in the art. Assuming that the fusion protein comprises a proteolytic cleavage site disposed between the immunoglobulin Fc region and the target protein, the target can be cleaved from the fusion protein using conventional proteolytic enzymes and if necessary, purified prior to use.

[0020] In another aspect, the invention provides methods for treating mammals, for example, a human, in need of an angiogenesis inhibitor based therapy. For example, it is contemplated that the angiogenesis inhibitors of the invention may be administered to a human afflicted with a tumor. Treatment with the angiogenesis inhibitor may slow down or stop tumor growth and, under certain circumstances, may cause tumor regression. Treatment may include administering to the mammal an amount of the angiogenesis inhibitor in an amount sufficient to slow down or stop tumor growth. The angiogenesis inhibitor may be provided in the form of a fusion protein or as a nucleic acid, preferably operatively associated with an expression vector, in combination with a pharmaceutically acceptable carrier.

[0021] The foregoing and other objects, features and advantages of the present invention will be made more apparent from the detailed description, drawings, and claims that follow.

BRIEF DESCRIPTION OF THE DRAWINGS

[0022] FIGS. 1A-1F are schematic illustrations of exemplary angiogenesis inhibitor fusion proteins constructed in accordance with the invention (see Examples 10-15). The Figures depict, respectively, FIG. 1A, Fc-Kringle 1 of Angiostatin; FIG. 1B, Fc-inner Kringle 1 of Angiostatin; FIG. 1C, Fc-Endostatin-GlySer linker-inner Kringle 1 of Angiostatin; FIG. 1D, Fc-Endostatin-GlySer linker-Kringle 1 of Angiostatin; FIG. 1E, Fc-Endostatin-GlySer linker-Angiostatin; FIG. 1F, Angiostatin-Fc-Endostatin. The vertical lines represent optional disulfide bonds connecting cysteine residues (C) disposal within a hinge region of the Fc molecule.

DETAILED DESCRIPTION OF THE INVENTION

[0023] The invention provides fusion proteins, referred to herein as immunofusins, which were useful in the production of commercial quantities of clinical grade angiogenesis inhibitors. The angiogenesis inhibitors may be cleaved from the immunofusin protein constructs prior to use. However, it is contemplated that the immunofusins or nucleic acids encoding the immunofusins may be administered directly to mammals in need of treatment with an angiogenesis inhibitor.

[0024] The invention thus provides fusion proteins comprising an immunoglobulin Fc region and at least one target protein, referred to herein as an angiogenesis inhibitor. The angiogenesis inhibitor preferably is selected from the group consisting of angiostatin, endostatin, a plasminogen fragment angiostatin activity, a collagen XVIII fragment having endostatin activity. It is contemplated, however, that other polypeptides having angiogenesis inhibitor activity, now known or late discovered, may be expressed as fusion proteins of the type described herein.

[0025] Six exemplary embodiments of protein constructs embodying the invention are illustrated in the drawing as FIGS. 1A-1F. Because dimeric constructs are preferred, all are illustrated as dimers cross-linked by a pair of disulfide bonds between cysteines on adjacent subunits. In the drawings, the disulfide bridges are depicted as linking together the portions of two immunoglobulin Fc regions via an immunoglobulin hinge region, and thus are characteristic of native forms of these molecules. While constructs including the hinge region of Fc are preferred and have been shown promise as therapeutic agents, the invention contemplates that the crosslinking at other positions may be chosen as desired. Furthermore, under some circumstances, dimers or multimers useful in the practice of the invention may be produced by non-covalent association, for example, by hydrophobic interaction.

[0026] Because homodimeric constructs are important embodiments of the invention, FIG. 1 illustrates such constructs. It should be appreciated that heterodimeric structures also are useful but, as is known to those skilled in the art, often can be difficult to purify. However, viable constructs useful to inhibit angiogenesis in various mammalian species, including humans, can be constructed comprising a mixture of homodimers and heterodimers. For example, one chain of the heterodimeric structure may comprise endostatin and the another may comprise angiostatin.

[0027]
FIG. 1A illustrates a dimer construct produced in accordance with the procedure set forth in Example 10. Each monomer of the dimer comprises an immunoglobulin Fc region 1 including a hinge region, a CH2 domain and a CH3 domain. Attached directly to the C terminus of the Fc region 1 is the first Kringle region of angiostatin 2, both inner and outer rings. FIG. 1B shows a second embodiment of the invention (see Example 11) comprising the same Fc region as in FIG. 1A, this time having only the inner ring of Kringle one of angiostatin 3 attached to the C terminal end of the Fc region 1. FIGS. 1C through 1E depict various embodiments of the protein constructs of the invention, which include as a target protein plural angiogenesis inhibitors arranged in tandem and connected by a linker. In FIG. 1C, the target protein comprises full-length endostatin 4, a polypeptide linker 5, and the inner ring of Kringle one of angiostatin 3. FIG. 1D depicts a protein comprising an Fc region the same as that of FIG. 1A and a target protein comprising a full-length endostatin 4, a polypeptide linker 5, and a full Kringle one region of angiostatin (both inner and outer rings) 2. FIG. 1E differs from the construct of FIG. 1D in that the most C terminal protein domain comprises a full-length copy of angiostatin 7.

[0028] Although FIGS. 1A-1E represent Fc-X type constructs, where X is the target protein, it is contemplated that X-Fc type constructs may also be useful in the practice of the invention. Furthermore, it is contemplated the useful proteins of the invention may also be depicted by the formula X-Fc-X, wherein the Xs may represent the same or different target proteins. FIG. 1F depicts such a construct which comprises in an N- to C-terminal direction, full-length human angiostatin 7, a human immunoglobulin Fc region 6 including a hinge region, and full-length human endostatin domain 4.

[0029] The term “angiogenesis inhibitor,” as used herein, refers to any polypeptide chain that reduces or inhibits the formation of new blood vessels in a mammal. With regard to cancer therapy, the angiogenesis inhibitor reduces or inhibits the formation of new blood vessels in or on a tumor, preferably in or on a solid tumor. It is contemplated that useful angiogenesis inhibitors may be identified using a variety of assays well known and used in the art. Such assays include, for example, the bovine capillary endothelial cell proliferation assay, the chick chorioallantoic membrane (CAM) assay or the mouse corneal assay. However, the CAM assay is preferred (see, for example, O'Reilly et al. (1994) Cell 79: 315-328 and O'Reilly et al. (1997) Cell 88: 277-285, the disclosures of which are incorporated herein by reference). Briefly, embryos with intact yolks are removed from fertilized three day old white eggs and placed in a petri dish. After incubation at 37° C., 3% CO2 for three days, a methylcellulose disk containing the putative angiogenesis inhibitor is applied to the chorioallantoic membrane of an individual embryo. After incubation for about 48 hours, the chorioallantoic membranes were observed under a microscope for evidence of zones of inhibition.

[0030] Preferred angiogenesis inhibitors useful in the practice of the invention include, for example, angiostatin (O'Reilly et al. (1994) Cell 79: 315-328, and U.S. Pat. Nos. 5,733,876; 5,837,682; and 5,885,795), and endostatin (O'Reilly et al. (1997) Cell 88: 277-285 and U.S. Pat. No. 5,854,205). As stated previously, angiostatin and endostatin are specific inhibitors of endothelial cell proliferation and are capable of inhibiting tumor growth by blocking angiogenesis, the formation of new blood vessels that nourish tumors.

[0031] Angiostatin has been identified as a proteolytic fragment of plasminogen (O'Reilly et al. (1994) Cell 79: 315-328, and U.S. Pat. Nos. 5,733,876; 5,837,682; and 5,885,795, the disclosure of which is incorporated herein by reference). Specifically, angiostatin is a 38 kDa internal fragment of plasminogen containing at least three of the Kringle regions of plasminogen. Endostatin has been identified as a proteolytic fragment of collagen XVIII (O'Reilly et al. (1997) Cell 88: 277-285, the disclosure of which is incorporated herein by reference). Specifically, endostatin is a 20 kDa C-terminal fragment of collagen XVIII. The terms “angiostatin” and “endostatin,” as used herein, refer not only to the full length proteins, but also to variants and bioactive fragments thereof, as well as to bioactive fragments of plasminogen and collagen XVIII, respectively. The term bioactive fragment, with respect to angiostatin refers to any protein fragment of plasminogen or angiostatin that has at least 30%, more preferably at least 70%, and most preferably at least 90% of the activity of full-length angiostatin as determined by the CAM assay. The term bioactive fragment, with respect to endostatin refers to any protein fragment of collagen XVIII or endostatin that has at least 30%, more preferably at least 70% and most preferably at least 90% of the activity of full length endostatin as determined by the CAM assay.

[0032] The term variants includes specifies and allelic variants, as well as other naturally occurring or non-naturally occurring variants, for example, generated by conventional genetic engineering protocols, that are at least 70% similar or 60% identical, more preferably at least 75% similar or 65% identical, and most preferably 80% similar or 70% identical to either the naturally-occurring sequences of endostatin or angiostatin disclosed herein.

[0033] To determine whether a candidate polypeptide has the requisite percentage similarity or identity to a reference polypeptide, the candidate amino acid sequence and the reference amino acid sequence are first aligned using the dynamic programming algorithm described in Smith and Waterman (1981), J. Mol. Biol. 147:195-197, in combination with the BLOSUM62 substitution matrix described in FIG. 2 of Henikoff and Henikoff (1992), “Amino acid substitution matrices from protein blocks”, Proc. Natl. Acad Sci. USA 89:10915-10919. For the present invention, an appropriate value for the gap insertion penalty is −12, and an appropriate value for the gap extension penalty is −4. Computer programs performing alignments using the algorithm of Smith-Waterman and the BLOSUM62 matrix, such as the GCG program suite (Oxford Molecular Group, Oxford, England), are commercially available and widely used by those skilled in the art.

[0034] Once the alignment between the candidate and reference sequence is made, a percent similarity score may be calculated. The individual amino acids of each sequence are compared sequentially according to their similarity to each other. If the value in the BLOSUM62 matrix corresponding to the two aligned amino acids is zero or a negative number, the pair-wise similarity score is zero; otherwise the pair-wise similarity score is 1.0. The raw similarity score is the sum of the pair-wise similarity scores of the aligned amino acids. The raw score then is normalized by dividing it by the number of amino acids in the smaller of the candidate or reference sequences. The normalized raw score is the percent similarity. Alternatively, to calculate a percent identity, the aligned amino acids of each sequence again are compared sequentially. If the amino acids are non-identical, the pair-wise identity score is zero; otherwise the pair-wise identity score is 1.0. The raw identity score is the sum of the identical aligned amino acids. The raw score is then normalized by dividing it by the number of amino acids in the smaller of the candidate or reference sequences. The normalized raw score is the percent identity. Insertions and deletions are ignored for the purposes of calculating percent similarity and identity. Accordingly, gap penalties are not used in this calculation, although they are used in the initial alignment.

[0035] The target proteins disclosed herein are expressed as fusion proteins with an Fc region of an immunoglobulin. As is known, each immunoglobulin heavy chain constant region is comprised of four or five domains. The domains are named sequentially as follows: CH1-hinge-CH2-CH3(-CH4). The DNA sequences of the heavy chain domains have cross-homology among the immunoglobulin classes, e.g., the CH2 domain of IgG is homologous to the CH2 domain of IgA and IgD, and to the CH3 domain of IgM and IgE.

[0036] As used herein, the term, “immunoglobulin Fc region” is understood to mean the carboxyl-terminal portion of an immunoglobulin chain constant region, preferably an immunoglobulin heavy chain constant region, or a portion thereof. For example, an immunoglobulin Fc region may comprise 1) a CH1 domain, a CH2 domain, and a CH3 domain, 2) a CH1, domain and a CH2 domain, 3) a CH1, domain and a CH3 domain, 4) a CH2 domain and a CH3 domain, or 5) a combination of two or more domains and an immunoglobulin hinge region. In a preferred embodiment the Fc region used in the DNA construct includes at least an immunoglobulin hinge region a CH2 domain and a CH3 domain and preferably lacks at least the CH1, domain.

[0037] The currently preferred class of immunoglobulin from which the heavy chain constant region is derived is IgG (Igγ) (γ subclasses 1, 2, 3, or 4). Other classes of immunoglobulin, IgA (Igα), IgD (Igδ), IgE (Igε) and IgM (Igμ), may be used. The choice of appropriate immunoglobulin heavy chain constant regions is discussed in detail in U.S. Pat. Nos. 5,541,087, and 5,726,044. The choice of particular immunoglobulin heavy chain constant region sequences from certain immunoglobulin classes and subclasses to achieve a particular result is considered to be within the level of skill in the art. The portion of the DNA construct encoding the immunoglobulin Fc region preferably comprises at least a portion of a hinge domain, and preferably at least a portion of a CH3 domain of Fcγ or the homologous domains in any of IgA, IgD, IgE, or IgM.

[0038] Depending on the application, constant region genes from species other than human e.g., mouse or rat may be used. The Fc region used as a fusion partner in the immunofusin DNA construct generally may be from any mammalian species. Where it is undesirable to elicit an immune response in the host cell or animal against the Fc region, the Fc region may be derived from the same species as the host cell or animal. For example, human Fc can be used when the host animal or cell is human; likewise, murine Fc can be used where the host animal or cell will be a mouse. Further, substitution or deletion of constructs of these constant regions, in which one or more amino acid residues of the constant region domains are substituted or deleted also would be useful. One example would be to introduce amino acid substitutions in the upper CH2 region to create a Fc variant with reduced affinity for Fc receptors (Cole et al. (1997) J. Immunol. 159:3613). One of ordinary skill in the art can prepare such constructs using well known molecular biology techniques.

[0039] The use of human Fcγ1 as the Fc region sequence has several advantages. For example, if the angiogenesis inhibitor Fc fusion protein is to be used as a biopharmaceutical, the Fcγ1 domain may confer the effector function activities to the fusion protein. The effector function activities include the biological activities such as complement fixation, antibody-directed cellular cytotoxicity, placental transfer, and increased serum half-life. The Fc domain also provides for detection by anti-Fc ELISA and purification through binding to Staphylococcus aureus protein A (“Protein A”). In certain applications, however, it may be desirable to delete specific effector functions from the Fc region, such as Fc receptor binding or complement fixation.

[0040] In the case of angiogenesis inhibitor immunofusins, one function of the immunoglobulin Fc fusion partner is to facilitate proper folding of the angiogenesis inhibitor protein to yield active angiogenesis inhibitor protein and to impact solubility to the active moieties, at least in the extracellular medium. Since the Fc fusion partner is hydrophilic, the angiogenesis inhibitor immunofusin readily is soluble unlike, for example, the recombinant endostatin produced in E. coli (O'Reilly (1997) Cell 88:277.)

[0041] In all of the Examples disclosed herein, high levels of production of the immunofusins were obtained. The angiogenesis inhibitor immunofusins were secreted into media at concentrations typically of about 30 to 100 μg/ml, and could be purified readily to homogeneity by Protein A chromatography. In addition, the angiogenesis inhibitor immunofusins could be cleaved and further purified using conventional purification protocols using, for example, by heparin sepharose, lysine sepharose or affinity purification.

[0042] In addition to the high levels of expression, fusion proteins of the invention also exhibit longer serum half-lives, presumably due to their larger molecular sizes. For example, human Fc-human angiostatin has a serum half-life of 33 hours in mouse, as compared to 4-6 hours for human angiostatin (O'Reilly et al. (1996) Nature Medicine 2:689). It is believe that angiostatin with a molecular weight of 40 kD, and endostatin with a molecular weight of 20 kD, are small enough to be cleared efficiently by renal filtration. In contrast, the dimeric forms of Fc-angiostatin and dimeric Fc-endostatin are 145 kD and 100 kD, respectively, because there are two immunoglobulin Fc regions attached to either two angiostatin molecules or two endostatin molecules. Such a bivalent structure may exhibit a higher binding affinity to the angiostatin or endostatin receptor. If the angiogenesis inhibiting activity is receptor-mediated, the Fc fusion proteins are potentially more effective to suppress tumors than monovalent angiostatin or monovalent endostatin by themselves. Furthermore, if angiostatin and/or endostatin belong to a class of dimeric protein ligands, the physical constraint imposed by the Fc on angiostatin or endostatin would make the dimerization an intramolecular process, thus shifting the equilibrium in favor of the dimer and enhancing its binding to the receptor. Cysteine residues can also be introduced by standard recombinant DNA technology to the monomer at appropriate places to stabilize the dimer through covalent disulfide bond formation.

[0043] As used herein, the term “multivalent” refers to a recombinant molecule that incorporates two or more biologically active segments. The protein fragments forming the multivalent molecule may be linked through a polypeptide peptide linker which attaches the constituent parts without causing a frame shift and permits each to function independently.

[0044] As used herein, the term “bivalent” refers to a multivalent recombinant molecule having two target proteins in a fusion construct of the invention, e.g., an Fc-X molecule, where X independently is selected from angiostatin, endostatin, or a variant thereof. Since there are two X moieties fused to an immunoglobulin Fc region (which typically itself is a dimer of the heavy chain fragments including at least a portion of the hinge region and CH3 domain, and optionally the CH2 domain), the molecule is bivalent (see, e.g., FIG. 1A). If the fusion construct of the invention has the form Fc-X-X, the resulting Fc dimer molecule is tetravalent. The two proteins forming the Fc-X-X molecule may be linked through a peptide linker. A bivalent molecule can increase the apparent binding affinity between the molecule and its receptor. For instance, if one endostatin moiety of an Fc-endostatin can bind to a receptor on a cell with a certain affinity, the second endostatin moiety of the same Fc-endostatin may bind to a second receptor on the same cell with a much higher avidity (apparent affinity). This is because of the physical proximity of the second endostatin moiety to the receptor after the first endostatin moiety is already bound. In the case of an antibody binding to an antigen, the apparent affinity is increased by at least 104.

[0045] As used herein, the terms “multimer” and “multimeric” refers to the stable association of two or more polypeptide chains either covalently, for example, by means of covalent interaction, for example, by a disulfide bond or non-covalently, for example, by hydrophobic interaction. The term multimer is intended to encompass both homomultimers, wherein the polypeptides are the same, as well as heteromultimers, wherein the polypeptides are different.

[0046] As used herein, the term “dimeric” refers to a specific multimeric molecule where two protein polypeptide chains are stably associated through covalent or non-covalent interactions. It should be understood that the immunoglobulin Fc region Fc fragment itself typically is a dimer of the heavy chain fragments including at least a portion of the hinge region and CH3 domain, and optionally the CH2 domain. Many protein ligands are known to bind to their receptors as a dimer. If a protein ligand X dimerizes naturally, the X moiety in an Fc-X molecule will dimerize to a much greater extent, since the dimerization process is concentration dependent. The physical proximity of the two X moieties connected by associated immunoglobulin Fc region would make the dimerization an intramolecular process, greatly shifting the equilibrium in favor of the dimer and enhancing its binding to the receptor.

[0047] It is understood that the present invention exploits conventional recombinant DNA methodologies for generating the Fe fusion proteins useful in the practice of the invention. The Fc fusion constructs preferably are generated at the DNA level, and the resulting DNAs integrated into expression vectors, and expressed to produce the immunofusins. As used herein, the term “vector” is understood to mean any nucleic acid comprising a nucleotide sequence competent to be incorporated into a host cell and to be recombined with and integrated into the host cell genome, or to replicate autonomously as an episome. Such vectors include linear nucleic acids, plasmids, phagemids, cosmids, RNA vectors, viral vectors and the like. Non-limiting examples of a viral vector include a retrovirus, an adenovirus and an adeno-associated virus. As used herein, the term “gene expression” or “expression” of a target protein, is understood to mean the transcription of a DNA sequence, translation of the mRNA transcript, and secretion of an Fc fusion protein product.

[0048] A useful expression vector is pdCs (Lo et al. (1988) Protein Engineering 11:495, the disclosure of which is incorporated herein by reference) in which the transcription of the Fc-X gene utilizes the enhancer/promoter of the human cytomegalovirus and the SV40 polyadenylation signal. The enhancer and promoter sequence of the human cytomegalovirus used was derived from nucleotides −601 to +7 of the sequence provided in Boshart et al., 1985, Cell 41:521, the disclosure of which is incorporated herein by reference. The vector also contains the mutant dihydrofolate reductase gene as a selection marker (Simonsen and Levinson (1983) Proc. Nat. Acad. Sci. USA 80:2495, the disclosure of which is incorporated herein by reference).

[0049] An appropriate host cell can be transformed or transfected with the DNA sequence of the invention, and utilized for the expression and secretion of a target protein. Currently preferred host cells for use in the invention include immortal hybridoma cells, NS/O myeloma cells, 293 cells, Chinese hamster ovary cells, Hela cells, and COS cells.

[0050] The fusion proteins of the invention preferably are generated by conventional recombinant DNA methodologies. The fusion proteins preferably are produced by expression in a host cell of a DNA molecule encoding a signal sequence, an immunoglobulin Fc region and a target protein (also referred to herein as an angiogenesis inhibitor). Preferred constructs may encode in a 5′ to 3′ direction, the signal sequence, the immunoglobulin Fc region and the target protein. Alternatively, the constructs may encode in a 5′ to 3′ direction, the signal sequence, the target protein and the immunoglobulin Fc region.

[0051] As used herein, the term “signal sequence” is understood to mean a peptide segment which directs the secretion of the angiogenesis inhibitor immunofusin protein and is thereafter cleaved following translation in the host cell. The signal sequence of the invention is a polynucleotide, which encodes an amino acid sequence that initiates transport of a protein across the membrane of the endoplasmic reticulum. Signal sequences which will be useful in the invention include antibody light chain signal sequences, e.g., antibody 14.18 (Gillies et. al., 1989, Jour. of Immunol. Meth., 125:191-202), antibody heavy chain signal sequences, e.g., the MOPC141 antibody heavy chain signal sequence (Sakano et al., 1980, Nature 286:5774), and any other signal sequences which are known in the art (see for example, Watson, 1984, Nucleic Acids Research 12:5145). Each of these references is incorporated herein by reference.

[0052] Signal sequences have been well characterized in the art and are known typically to contain 16 to 30 amino acid residues, and may contain greater or fewer amino acid residues. A typical signal peptide consists of three regions: a basic N-terminal region, a central hydrophobic region, and a more polar C-terminal region. The central hydrophobic region contains 4 to 12 hydrophobic residues that anchor the signal peptide across the membrane lipid bilayer during transport of the nascent polypeptide. Following initiation, the signal peptide is usually cleaved within the lumen of the endoplasmic reticulum by cellular enzymes known as signal peptidases. Potential cleavage sites of the signal peptide generally follow the “(−3, −1) rule.” Thus a typical signal peptide has small, neutral amino acid residues in positions −1 and −3 and lacks proline residues in this region. The signal peptidase will cleave such a signal peptide between the −1 and +1 amino acids. Thus, the portion of the DNA encoding the signal sequence may be cleaved from the amino-terminus of the immunofusin protein during secretion. This results in the secretion of a immunofusin protein consisting of the Fc region and the target protein. A detailed discussion of signal peptide sequences is provided by von Heijne (1986) Nucleic Acids Res., 14:4683 the disclosure of which is incorporated herein by reference.

[0053] As would be apparent to one of skill in the art, the suitability of a particular signal sequence for use in the invention may require some routine experimentation. Such experimentation will include determining the ability of the signal sequence to direct the secretion of an immunofusin and also a determination of the optimal configuration, genomic or cDNA, of the sequence to be used in order to achieve efficient secretion of immunofusins. Additionally, one skilled in the art is capable of creating a synthetic signal peptide following the rules presented by von Heijne, referenced above, and testing for the efficacy of such a synthetic signal sequence by routine experimentation. A signal sequence may also be referred to as a “signal peptide,” “leader sequence,” or “leader peptide.”

[0054] The fusion of the signal sequence and the immunoglobulin Fc region is sometimes referred to herein as secretion cassette. An exemplary secretion cassette useful in the practice of the invention is a polynucleotide encoding, in a 5′ to 3′ direction, a signal sequence of an immunoglobulin light chain gene and an Fcγ1 region of the human immunoglobulin γ1 gene. The Fcγ1 region of the immunoglobulin Fcγ1 gene preferably includes at least a portion of the hinge domain and at least a portion of the CH3 domain, or alternatively at least portions of the hinge domain, CH2 domain and CH3 domain. The DNA encoding the secretion cassette can be in its genomic configuration or its cDNA configuration.

[0055] In another embodiment, the DNA sequence encodes a proteolytic cleavage site interposed between the secretion cassette and the angiogenesis inhibitor protein. A cleavage site provides for the proteolytic cleavage of the encoded fusion protein thus separating the Fc domain from the angiogenesis inhibitor protein. As used herein, “proteolytic cleavage site” is understood to mean amino acid sequences which are preferentially cleaved by a proteolytic enzyme or other proteolytic cleavage agents. Useful proteolytic cleavage sites include amino acids sequences which are recognized by proteolytic enzymes such as trypsin, plasmin or enterokinase K. Many cleavage site/cleavage agent pairs are known. See, for example, U.S. Pat. No. 5,726,044, the disclosure of which is incorporated herein by reference. Where the target protein sequence is a precursor molecule to angiostatin, endostatin, or an active variant thereof, the desired protein product may be produced by cleavage with the endogenous proteolytic enzyme, such as elastin or plasmin or urokinase.

[0056] The present invention also encompasses fusion proteins containing different combinations of recombinant angiostatin and endostatin, or fragments thereof, which can be made in large quantities. Despite the demonstrated efficacy in suppressing tumor growth, the mechanism of how angiostatin and endostatin block angiogenesis is not completely known. Angiostatin has several Kringle structures and endostatin has different structural motifs, each of which may be solely responsible for or assist in binding of the proteins to endothelial cells and exerting an anti-angiogenic effect. Accordingly, this invention includes target proteins which are bioactive fragments of angiostatin, such as Kringle 1, Kringle 2, Kringle 3, and combinations thereof, and endostatin which exhibit physiologically similar behavior to naturally occurring full-length angiostatin and endostatin.

[0057] Another embodiment of the present invention provides for bifunctional hybrid constructs of angiogenesis inhibitors. As used herein, a bifunctional hybrid molecule or construct means a protein produced by combining two protein subunits, where the two subunits can be derived from different proteins. Each protein subunit has its own independent function so that in the hybrid molecule, the functions of the two subunits may be additive or synergistic. Such functional hybrid proteins would allow the synergistic effect of angiostatin and endostatin to be explored in animal models. A preferred bifunctional hybrid may comprise at least two different angiogenesis inhibitors linked in tandem, either directly or by means of a polypeptide linker. For example, in a preferred embodiment, the target sequence encodes at least a portion of angiostatin linked in frame with at least a portion of endostatin and both the angiostatin and endostatin domains exhibit anti angiogenesis activity or angiogenesis inhibition. The two units may be linked by a polypeptide linker.

[0058] As used herein the term “polypeptide linker is understood to mean an peptide sequence that can link two proteins together or a protein and an Fc region. The polypeptide linker preferably comprises a plurality of amino acids such as glycine and/or serine. Preferably, the polypeptide linker comprises a series of glycine and serine peptides about 10-15 residues in length. See, for example, U.S. Pat. No. 5,258,698, the disclosure of which is incorporated herein by reference. It is contemplated however, that the optimal linker length and amino acid composition may be determined by routine experimentation.

[0059] It is found that when different parts of the angiostatin are expressed as Fc fusion molecules, high levels of expression are obtained, presumably because the Fc portion acts as a carrier, helping the polypeptide at the C-terminus to fold correctly. In addition, the Fc region can be glycosylated and highly charged at physiological pH, thus the Fc region can help to solubilize hydrophobic proteins.

[0060] The present invention also provides methods for the production of angiostatin and endostatin of non-human species as Fc fusion proteins. Non-human angiogenesis inhibitor fusion proteins are useful for preclinical studies of angiogenesis inhibitors because efficacy and toxicity studies of a protein drug must be performed in animal model systems before testing in humans. A human protein may not work in a mouse model because the protein may elicit an immune response, and/or exhibit different pharmacokinentics skewing the test results. Therefore, the equivalent mouse protein is the best surrogate for the human protein for testing in a mouse model.

[0061] The standard Lewis lung carcinoma model in mice (O'Reilly et al. (1997) Cell 88:277) was used to compare soluble huFc-huAngiostatin, huFc-huEndostatin, muFc-muAngiostatin, muFc-muEndostatin with the insoluble proteins produced in an E. Coli expression system. The soluble Fc fusion proteins were more efficacious in suppressing tumor growth in the Lewis lung model than the corresponding proteins produced in E. coli. Furthermore, laboratory mice are inbred and their tumors are induced and not spontaneous. Therefore, efficacy in a mouse model may not correlate to probable efficacy against human tumors. Preclinical studies in dogs will provide more precise information about the efficacy of these angiogenesis inhibitors on spontaneous tumors because there are numerous naturally occurring, spontaneous canine tumors. The methods of producing murine (mu) Fc-mu angiostatin, muFc-mu endostatin, and canine (ca) Fc-ca angiostatin, caFc-ca endostatin of the present invention will facilitate preclinical studies of angiogenesis inhibitors in both murine and canine systems.

[0062] The present invention provides methods of treating a condition mediated by angiogenesis by administering the DNA, RNA or proteins of the invention. Conditions mediated by angiogenesis include, for example: solid tumors; blood born tumors, tumor metastasis, benign tumors including hemangiomas, acoustic neuromas, neurofibromas, trachomas, and pyrogenic granulomas; rheumatoid arthritis; psoriasis; ocular angiogenic diseases (diabetic retinopathy, retinopathy of prematurity, macular degeneration, corneal graft rejection, neovascular glaucoma) retrolental fibroplasia, rubeosis, Osler-Webber Syndrome; myocardial angiogenesis; plaque neovascularization; telangiectasia; hemophiliac joints' angiofibroma; and wound granulation; and excessive or abnormal stimulation of endothelial cells, intestinal adhesions, artherosclerosis, sclerodermal and hypertrophic scars, i.e., keloids.

[0063] The DNA constructs disclosed herein can be useful in gene therapy procedures in which the endostatin or angiostatin gene is delivered into a cell by one of various means e.g., native DNA associated with a promoter or DNA within a viral vector. Once inside a cell, the angiostatin and/or endostatin gene or gene fragment is expressed and the protein is produced in vivo to carry out its normal biological function. The DNA construct of the present invention results in high levels of expression of the fusion protein. The fusion proteins of the present invention may also be useful in treating conditions mediated by angiogenesis and may have greater clinical efficacy than native angiogenesis inhibitors and other recombinant angiogenesis inhibitors because the angiogenesis inhibitor immunofusins of the present invention have a longer serum half-life than the other recombinant angiogenesis inhibitors or native angiogenesis inhibitors alone. The bivalent and dimeric forms of the present invention should have higher binding affinity due to the bivalent and dimeric structure. The bifunctional hybrid molecules of the present invention may have a higher clinical efficacy due to possible synergistic effects of two different angiogenesis inhibitors connected by the fused Fc region or a flexible polypeptide linker.

[0064] The compositions of the present invention may be provided to an animal by any suitable means, directly (e.g., locally, as by injection, implantation or topical administration to a tissue locus) or systemically (e.g., parenterally or orally). Where the composition is to be provided parenterally, such as by intravenous, subcutaneous, ophthalmic, intraperitoneal, intramuscular, buccal, rectal, vaginal, intraorbital, intracerebral, intracranial, intraspinal, intraventricular, intrathecal, intracisternal, intracapsular, intranasal or by aerosol administration, the composition preferably comprises part of an aqueous or physiologically compatible fluid suspension or solution. Thus, the carrier or vehicle is physiologically acceptable so that in addition to delivery of the desired composition to the patient, it does not otherwise adversely affect the patient's electrolyte and/or volume balance. The fluid medium for the agent thus can comprise normal physiologic saline (e.g., 9.85% aqueous NaCl, 0.15 M, pH 7-7.4).

[0065] Preferred dosages of the immunofusins per administration are within the range of 50 ng/m2 to 1 g/m2, more preferably 5 μg/m2 to 200 mg/m2, and most preferably 0.1 mg/m2 to 50 mg/m2. Preferred dosages of nucleic acids encoding the immunofusins per administration are within the range of 1 μg/m2 to 100 mg/m2, more preferably 20 μg/m2 to 10 mg/m2, and most preferably 400 μg/m2 to 4 mg/m2. It is contemplated, however, that the optimal modes of administration, and dosages may be determined by routine experimentation well within the level of skill in the art.

[0066] The invention is illustrated further by the following non-limiting examples.

EXAMPLES

Example 1

Expression of huFc-huEndostatin

[0067] Human endostatin was expressed as a human Fc-human endostatin (huFc-huEndo) fusion protein according to the teachings of Lo et al. (1998) Protein Engineering 11:495. Fc refers to the Fc fragment of the human immunoglobulin gamma (DNA sequence set forth in SEQ ID NO: 1; amino acid sequence set forth in SEQ ID NO: 2). (Polymerase chain reactions PCR) was used to adapt the endostatin cDNA (SEQ ID NO: 3; whose amino acid sequence is disclosed in SEQ ID NO: 4), for expression in an Fc-Endo fusion protein. The forward primer was either 5′-CC CCG GGT AAA CAC AGC CAC CGC GAC TTC C (SEQ ID NO: 5; encoded amino acids disclosed in SEQ ID NO: 6) or 5′-C AAG CTT CAC AGC CAC CGC GAC TTC C (SEQ ID NO: 7; encoded amino acids disclosed in SEQ ID NO: 8), where the XmaI site or the HindIII site was followed by sequence encoding the N-terminus of endostatin. The primer with the XmaI site adapted the endostatin cDNA for ligation to the XmaI site at the end of the CH3 domain of the IgGFc region. The primer with the HindIII site adapted the endostatin cDNA for ligation to the HindIII site of the pdCs-Fc(D4K) vector, which contains the enterokinase recognition site Asp4-Lys (LaVallie et al. (1993) J. Biol. Chem. 268:23311-23317) at the junction of the fusion protein. The reverse primer was 5′-C CTC GAG CTA CTT GGA GGC AGT CAT G (SEQ ID NO: 9), which was designed to put a translation STOP codon (anticodon, CTA) immediately after the C-terminus of endostatin, and this was followed by an XhoI site. The PCR products were cloned and sequenced, and the XmaI-XhoI fragment was ligated to the resulting XmaI and XhoI digested pdCs-Fc vector. Similarly, the HindIII-XhoI fragment encoding endostatin was ligated into appropriately digested pdCs-huFc(D4K) vector. Stable clones expressing Fc-endo or Fc(D4K)-endostatin were obtained by electroporation of NS/0 cells followed by selection in growth medium containing 100 nM methotrexate. Protein expression level was assayed by anti-human Fc ELISA (Example 3) and confirmed by SDS-PAGE, which showed a protein product of ˜52 kD. The best producing clones were subcloned by limiting dilutions.

Example 2

Cell Culture and Transfection

[0068] For transient transfection, the plasmid was introduced into human kidney 293 cells by co-precipitation of plasmid DNA with calcium phosphate (Sambrook et al. (1989) Molecular Cloning—A Laboratory Manual, Cold Spring Harbor, N.Y.) or by lipofection using LipofectAMINE Plus (Life Technologies, Gaithersburg, Md.) according to supplier's protocol.

[0069] In order to obtain stably transfected clones, plasmid DNA was introduced into the mouse myeloma NS/0 cells by electroporation. NS/0 cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. About 5×106 cells were washed once with PBS and resuspended in 0.5 ml PBS. Ten μg of linearized plasmid DNA then was incubated with the cells in a Gene Pulser Cuvette (0.4 cm electrode gap, BioRad, Hercules, Calif.) on ice for 10 min. Electroporation was performed using a Gene Pulser (BioRad, Hercules, Calif.) with settings at 0.25 V and 500 μF. Cells were allowed to recover for 10 min. on ice, after which they were resuspended in growth medium and then plated onto two 96 well plates. Stably transfected clones were selected by growth in the presence of 100 nM methotrexate (MTX), which was introduced two days post-transfection. The cells were fed every 3 days for three more times, and MTX-resistant clones appeared in 2 to 3 weeks. Supernatants from clones were assayed by anti-Fc ELISA to identify high producers. High producing clones were isolated and propagated in growth medium containing 100 nM MTX.

Example 3

ELISA Procedures

[0070] Three different ELISAs were used to determine the concentrations of protein products in the supernatants of MTX-resistant clones and other test samples. The anti-human Fc (huFc) ELISA was used to measure the amount of human Fc-containing proteins. The anti-murine Fc (muFc) and anti-canine Fc (caFc) antibodies were used in ELISAs to measure the amount of murine Fc-and canineFc-containing proteins, respectively. The procedure for the anti-huFc ELISA is described in detail herein below.

[0071] A. Coating Plates

[0072] ELISA plates were coated with AffiniPure Goat anti-Human IgG (H+L) (Jackson ImmunoResearch Laboratories, West Grove, Pa.) at 5 μg/ml in PBS, and 100 μl/well in 96-well plates (Nunc-Immuno plate MaxiSorp™, Nalge Nunc International, Rochester, N.Y. ). Coated plates were covered and incubated at 4° C. overnight. Plates then were washed 4 times with 0.05% Tween 20 in PBS, and blocked with 1% BSA/1% Goat Serum in PBS, 200 μl/well. After incubation with the blocking buffer at 37° C. for 2 hours, the plates were washed 4 times with 0.05% Tween in PBS and tapped dry on paper towels.

[0073] B. Incubation with Test Samples and Secondary Antibody

[0074] Test samples were diluted to the proper concentrations in a sample buffer, containing 1% BSA/1% Goat Serum/0.05% Tween in PBS. A standard curve was prepared with a chimeric antibody (with a human Fc), the concentration of which was known. To prepare a standard curve, serial dilutions were made in the sample buffer to give a standard curve ranging from 125 ng/ml to 3.9 ng/ml. The diluted samples and standards were added to the plate, 100 μl/well and the plate was then incubated at 37° C. for 2 hr. After incubation, the plate was washed 8 times with 0.05% Tween in PBS. To each well was then added 100 μl of secondary antibody, the horse radish peroxidase (HRP)-conjugated anti-human IgG (Jackson ImmunoResearch Laboratories, Inc. West Grove, Pa), diluted about 1:120,000 in sample buffer. The exact dilution of the secondary antibody had to be determined for each lot of the HRP-conjugated Anti-Human IgG. After incubation at 37° C. for 2 hr, the plate was washed 8 times with 0.05% Tween in PBS.

[0075] C. Development

[0076] A substrate solution was prepared by dissolving 30 mg (1 tablet) of o-phenylenediamine dihydrochloride (OPD) into 15 ml of 0.025 M citric acid/0.05 M Na2HPO4 buffer, pH 5, containing 0.03% of freshly added H2O2. The substrate solution was added to the plate at 100 μl/well. The color was allowed to develop for 30 min. at room temperature in the dark. The developing time can be subject to change, depending on lot to lot variability of the coated plates, the secondary antibody, etc. The reaction was stopped by adding 4N H2SO4, 100 μ/well. The plate was read by a plate reader, which was set at both 490 and 650 nm, and programmed to subtract the background OD at 650 nm from the OD at 490 nm.

[0077] The procedure for the anti-muFc ELISA was similar, except that ELISA plate was coated with AffiniPure Goat anti-murine IgG (H+L) (Jackson ImmunoResearch, West Grove, Pa.) at 5 μg/ml in PBS, and 100 μl/well; and the secondary antibody was horse radish peroxidase-conjugated goat anti-muIgG, Fcγ (Jackson ImmunoResearch West Grove, Pa.), used at 1 in 5000 dilution. Similarly, for the anti-caFc ELISA, the ELISA plate was coated with AffiniPure Rabbit anti-dog IgG, Fe Fragment specific (Jackson ImmunoResearch, West Grove, Pa.) at 5 μg/ml in PBS, and 100 μl/well; and the secondary antibody was horse radish peroxidase-conjugated AffiniPure rabbit anti-dog IgG, Fc fragment specific (Jackson ImmunoResearch, West Grove, Pa.), used at 1 in 5000 dilution.

Example 4

Expression of huFc-huAngiostatin

[0078] Human angiostatin (DNA sequence set forth in SEQ ID NO: 10; amino acid sequence set forth in SEQ ID NO: 11) was expressed as a human Fc-human angiostatin (huFc-huAngio) fusion protein essentially as described in Example 1. PCR was used to adapt the angiostatin cDNA (SEQ ID NO: 3), for expression in the pdCs-huFc or pdCs-huFc(D4K) vectors. The respective forward primers were 5′-CC CCG GGT AAG AAA GTG TAT CTC TCA GAG (SEQ ID NO 12; encoded amino acids disclosed in SEQ ID NO: 13), and 5′-C CCC AAG CTT AAA GTG TAT CTC TCA GAG (SEQ ID NO: 14; encoded amino acids disclosed in SEQ ID NO: 15), where the XmaI site or the HindIII site was followed by sequence encoding the N-terminus of angiostatin. The reverse primer was 5′-CCC CTC GAG CTA CGC TTC TGT TCC TGA GCA (SEQ ID NO: 16), which was designed to put a translation STOP codon (anticodon, CTA) immediately after the C-terminus of angiostatin, and this was followed by an XhoI site. The PCR products were cloned and sequenced, and the resulting XmaI-XhoI fragment and the HindIII-XhoI fragment encoding angiostatin were ligated to the pdCs-huFc and the pdCs-huFc(D4K) vectors, respectively. Stable NS/0 clones expressing huFc-huAngio and huFc(D4K)-huAngio were selected and assayed as described in Examples 2 and 3.

Example 5

Expression of muFc-mu-Endostatin

[0079] Murine endostatin (DNA sequence set forth in SEQ ID NO: 17; amino acid sequence set forth in SEQ ID NO: 18) and murine Fe (DNA sequence set forth in SEQ ID NO: 19; encoded amino acids set forth in SEQ ID NO: 20) were expressed as a murine F-murine endostatin (muFc-muEndo) fusion protein essentially as described in Example 1. PCR was used to adapt the endostatin cDNA (SEQ ID NO: 4), for expression in the pdCs-muFc(D4K) vector. The forward primer was 5′-C CCC AAG CTT CAT ACT CAT CAG GAC TTT C (SEQ ID NO: 21; encoded amino acids disclosed in SEQ ID NO: 22), where the HindIII site was followed by sequence encoding the N-terminus of endostatin. The reverse primer was 5′-CCC CTC GAG CTA TTT GGA GAA AGA GGT C (SEQ ID NO: 23), which was designed to put a translation STOP codon (anticodon, CTA) immediately after the C-terminus of endostatin, and this was followed by an XhoI site. The PCR product was cloned and sequenced, and the resulting HindIII-XhoI fragment encoding endostatin was ligated into the pdCs-muFc(D4K) vector. Stable NS/0 clones expressing muFc(D4K)-muEndo were selected and assayed (anti-muFc ELISA) as described in Examples 2 and 3.

Example 6

Expression of muFc-muAngiostatin

[0080] Murine angiostatin (DNA sequence set forth in-SEQ ID NO: 24; amino acid sequence set forth in SEQ ID NO: 25) was expressed as a murine Fc-murine angiostatin (muFc-muAngio) fusion protein essentially as described in Example 1. PCR was used to adapt the angiostatin cDNA (SEQ ID NO: 6) for expression in the pdCs-Fc(D4K) vector. The forward primer was 5′-C CCC AAG CTT GTG TAT CTG TCA GAA TGT AAG CCC TCC TGT CTC TGA GCA (SEQ ID NO: 26; encoded amino acids disclosed in SEQ ID NO: 27), where the HindIII site was followed by sequence encoding the N-terminus of angiostatin. The reverse primer was 5′-CCC CTC GAG CTA CCC TCC TGT CTC TGA GCA (SEQ ID NO: 28), which was designed to put a translation STOP codon (anticodon, CTA) immediately after the C-terminus of angiostatin, and this was followed by an XhoI site (CTCGAG). The PCR product was cloned and sequenced, and the HindIII-XhoI fragment encoding angiostatin was ligated to the pdCs-muFc(D4K) vector. Stable NS/0 clones expressing muFc(D4K)-muAngio were selected and assayed (anti-muFc ELISA) as described in Examples 2 and 3.

Example 7

Expression of Canine Fc (caFc)

[0081] Canine peripheral blood monocytic cells (PBMCs) isolated from dog's blood were used to prepare mRNA. After synthesis of the first strand cDNA with reverse transcriptase and oligo(dT), PCR was performed to amplify the canine Fc (Kazuhiko et al., (1992) JP 1992040894-A1) using the forward primer 5′-CC TTA AGC GAA AAT GGA AGA GTT CCT CGC (SEQ ID NO: 29; encoded amino acids disclosed in SEQ ID NO: 30), in which an AfIII site was introduced immediately upstream of the sequence encoding the hinge region of the canine Fc, and the reverse primer 5′-C CTC GAG TCA TTT ACC CGG GGA ATG GGA GAG GGA TTT CTG (SEQ ID NO: 31), in which an XhoI site was introduced after the translation STOP codon (anticodon, TCA) of the canine Fc. The reverse primer also introduced a silent mutation to create a XmaI restriction site, which facilitates the construction of the pdCs-caFc(D4K) vector through a linker-adaptor and ligation to DNA constructs encoding canine endostatin or angiostatin. Similar to the construction of pdCs-huFc, which was described in detail in Lo et al. (Lo et al., Protein Engineering (1998) 11:495), the expression vector for the pdCs-caFc was constructed as follows. The AfIII-XhoI fragment encoding the canine Fc was ligated to the XbaI-AfIII fragment encoding the light chain signal peptide and the XbaI-XhoI digested pdCs vector. The resulting pdCs-caFc expression vector then was used to transfect 293 cells. About 3 days post-transfection, the supernatant was purified by Protein A chromatography. Expression of dog Fc (DNA sequence set forth in SEQ ID NO: 32; amino acid sequence set forth in SEQ ID NO: 33) was confirmed by SDS-PAGE followed by Western blot analysis using a peroxidase-conjugated Rabbit anti-Dog IgG, Fc fragment specific (Jackson ImmunoResearch, West Grove, Pa.).

Example 8

Expression of caFc-caEndostatin

[0082] The coding sequence for canine endostatin (DNA sequence set forth in SEQ ID NO: 34; amino acid sequence set forth in SEQ ID NO: 35) was adapted to a HindIII-XhoI fragment for expression as a Fc fusion protein, essentially as described in Example 5. At the 3′ end, a STOP codon was introduced, for example, by PCR, immediately after the codon encoding the C-terminal lysine residue, and this was followed by the NotI restriction site. At the 5′ end, however, there was a DraIII restriction site convenient for reconstruction. An oligonucleotide duplex consisting of a HindIII and a DraIII sticky ends was chemically synthesized and used to ligate to the DraIII-XhoI restriction fragment which encodes the rest of the canine endostatin cDNA. The duplex used is shown below:

[0083] HindIII

[0084] 5′-AGCTT CAC ACC CAC CAG GAC TTC CAG CCG GTG CTG CAC CTG (SEQ ID NO: 36)

[0085] A GTG TGG GTG GTC CTG AAG GTC GGC CAC GAC GTG-5′ (SEQ ID NO: 38) DraIII

[0086] The first CAC in the duplex encodes the N-terminal histidine residue of the canine endostatin. The HindIII-XhoI fragment encoding the full-length canine endostatin thus could be ligated to the HindIII-XhoI digested pdCs-caFc vector (see Example 7) for expression. Stable NS/0 clones expressing caFc-caEndo were selected and assayed by anti-caFc ELISA, as described in Examples 2 and 3. The protein product was analyzed on SDS-PAGE and confirmed by Western blot analysis.

Example 9

Expression of caFc-caAngiostatin

[0087] The cDNA encoding the full length canine angiostatin (DNA sequence set forth in SEQ ID NO: 39; amino acid sequence set forth in SEQ ID NO: 40) was adapted for expression as a caFc fusion protein essentially as in the aforementioned examples. Briefly, at the 3′ end, a STOP codon was introduced, for example, by PCR, immediately after the codon encoding the C-terminal lysine residue and this was followed by a NotI restriction site instead of an XhoI site, since there was an internal XhoI restriction site in the cDNA of the canine angiostatin. At the 5′ end, a HindIII site was introduced in-frame immediately upstream of the N-terminus of angiostatin. The HindIII-NotI fragment encoding the full length canine angiostatin then was ligated to the HindIII-NotI digested pdCs-caFc vector (where the NotI site was introduced at the XhoI site through linker ligation) for expression. Stable NS/0 clones expressing caFc-caAngio were selected and assayed by anti-caFc ELISA, as described in Examples 2 and 3. The protein product was analyzed on SDS-PAGE and confirmed by Western blot analysis.

Example 10

Expression of muFc-K1 of muAngio

[0088] Angiostatin comprises the first four of the five Kringle domains of plasminogen. To determine if any one or several Kringle domains are responsible for the observed anti-angiogenic activity of angiostatin, it is possible to produce single Kringle domains by themselves or combination thereof for testing. To demonstrate the utility of Fc as a fusion protein partner, the expression of the first Kringle domain of murine angiostatin (K1) was achieved in the following way. The first Kringle domain ends at Glu-87 of murine angiostatin (SEQ ID NO: 25). There was a convenient NsiI restriction site in the cDNA at this position so that after digestion by NsiI, the four-base 3′-overhang was removed by T4 polymerase to create a blunt end. A translation STOP codon was introduced immediately downstream of the GAA encoding Glu-87 via ligation to the palindromic linker TGA CTC GAG TCA (SEQ ID NO: 41), where the STOP codon TGA was followed by an XhoI site. The HindIII-XhoI fragment encoding this truncated angiostatin, i.e., first Kringle only, then was ligated into the pdCs-muFc(D4K) vector for expression. High levels of expression were obtained in both transient and stable expression, as analyzed by anti-muFc ELISA and SDS-PAGE.

Example 11

Expression of muFc-innerK1 of muAngio

[0089] A Kringle domain consists of multiple loops, including an outer loop and an inner loop. In the first Kringle of murine angiostatin, the inner loop is defined by Cys 55 and Cys 79, which together form a disulfide bond at the base of the loop. The Cys-67 of the inner loop forms another disulfide bond with a Cys residue of the outer loop to give the Kringle structure. To test if the inner loop has any anti-angiogenic activity, it was expressed as a muFc-inner K1 (Kringle 1) as follows. With a DNA fragment encoding the first Kringle as template, a mutagenic primer having the sequence 5′GGG CCT TGG AGC TAC ACT ACA (SEQ ID NO: 42; encoded amino acids disclosed in SEQ ID NO: 43) was used to mutagenize TGC (Cys-67) to AGC (Ser), by PCR. This ensures that the Cys-67 does not form a disulfide bond when the inner loop of Kringle 1 is expressed without the outer loop. An upstream primer having the sequence 5′GCGGATCCAAGCTT AGT ACA CAT CCC AAT GAG GG (SEQ ID NO: 44; encoded amino acids disclosed in SEQ ID NO: 45) was used to introduce a HindIII site in frame immediately 5′ to the codon for Ser-43 (AGT). A BamHI site was also introduced immediately upstream of the HindIII site. The BamHI site is useful for ligating to the BamHI site at the end of the flexible Gly-Ser linker shown in Example 12 below. Thus a HindIII-XhoI DNA fragment encoding Ser-43 through Glu-87 of murine angiostatin was ligated to the pdCs-muFc(D4K) vector for expression. High levels of expression of muFc-innerK1 were obtained in both transient and stable expression, as analyzed by anti-muFc ELISA and SDS-PAGE.

Example 12

Expression of muFc-muEndo-GlySer Linker-InnerK1 of muAngio

[0090] The hybrid molecule muFc-muEndo-innerK1 comprises muFc-muEndo joined by a polypeptide linker containing glycine and serine residues, to the inner loop of the first Kringle of murine angiostatin. The DNA construct was assembled as follows.

[0091] There is a BspHI site at the 3′ end of the murine endostatin cDNA. To introduce a flexible linker of glycine and serine residues at the C-terminus of murine endostatin, a 540-bp HindIII-BspHI fragment encoding endostatin was ligated to an overlapping oligonucleotide duplex formed by the oligonucleotides disclosed in SEQ ID NO: 46 and SEQ ID NO: 48. The amino acid linker encoded by SEQ ID NO: 46 is disclosed in SEQ ID NO: 47.

[0092] The HindIII-BamHI fragment encoding murine endostatin and the Gly-Ser linker w a s subcloned into a standard cloning vector. The BamHI site was then used to introduce the BamHI-XhoI fragment encoding the innerK1 in Example 11. The resulting HindIII-XhoI fragment encoding muEndo- GlySer linker-innerK1, was ligated to the pdCs-muFc(D4K) vector for expression. High levels of expression of muFc-muEndo-GlySer linker-innerK1 were obtained in both transient and stable expression, as analyzed by anti-muFc ELISA and SDS-PAGE.

Example 13

Expression of muFc-muEndo-GlySer Linker-K1 of muAngio

[0093] The hybrid molecule muFc-muEndo-K1 comprises muFc-muEndo joined by a polypeptide linker containing glycine and serine residues, to the first Kringle of murine angiostatin. The DNA construct was assembled as follows.

[0094] The BamHI end of the HindIII-BamHI fragment encoding the muEndo-GlySer linker (Example 12) was ligated to the HindIII-XhoI fragment encoding the Kringle 1 of murine angiostatin (Example 10) via the following adaptor:

1BamHI5′ GA TCC TCA GGC C(SEQ ID NO:49) G AGT CCG GTCGA(SEQ ID NO:50) HindIII

[0095] The adaptor has a HindIII′ sticky end, which upon ligation, would not regenerate the HindIII site. Thus, the resulting HindIII-XhoI fragment, which encodes the muEndo-GlySer linker-Kringle 1, was ligated to the pdCs-muFc(D4K) vector for expression. High levels of expression of muFc-muEndo-GlySer linker-K1 were obtained in both transient and stable expression, as analyzed by anti-muFc ELISA and SDS-PAGE.

Example 14

Expression of muFc-muEndo-GlySer Linker-muAngio

[0096] The hybrid molecule muFc-muEndo-GlySer linker-muAngio comprises muFc-muEndo joined by a polypeptide linker containing glycine and serine residues, to murine angiostatin. The DNA construct was assembled essentially as follows. The BamHI end of the HindIII-BamHI fragment encoding the muEndo-GlySer linker (Example 12) was ligated to the HindIII-XhoI fragment encoding murine angiostatin via the adaptor described in Example 13. The resulting HindIII-XhoI fragment, which encodes the muEndo-GlySer linker-muAngio, was ligated to the pdCs-muFc(D4K) vector for expression. High levels of expression of muFc-muEndo-GlySer linker-muAngio were obtained in both transient and stable expression, as analyzed by anti-muFc ELISA and SDS-PAGE.

Example 15

Expression of huAngio-huFc-huEndo

[0097] The hybrid molecule huAngio-huFc-huEndo comprises human angiostatin joined by a peptide bond to huFc-huEndo. The DNA construct was assembled as follows. A HindIII-XhoI fragment which encodes human angiostatin without a STOP codon was first generated by PCR, so that the codon for the last amino acid residue of angiostatin was followed immediately by CTCGAG of the XhoI site. The HindIII at the 5′ end was ligated to an XbaI-AfIII fragment of the light chain signal peptide (Lo et al., Protein Engineering (1998) 11:495) via a AfIII-HindIII′ adaptor:

2 AfIII5′ TTA AGC GGC C(SEQ ID NO:51) CG CGG GTCGA(SEQ ID NO:52) HindIII′

[0098] The HindIII′ sticky end of the adaptor, upon ligation, would not regenerate a HindIII site. At the 3′ end, the XhoI site was ligated to the AfIII site of the AfIII-XhoI fragment encoding the huFc-hu-Endo via the following XhoI′-AfIII adaptor:

3 XhoI′5′ TC GAC TCC GGC(SEQ ID NO:53) G AGG CCG AATT(SEQ ID NO:54) AfIII

[0099] The XhoI sticky end of the adaptor, upon ligation, would not regenerate a XhoI site. The resulting XbaI-XhoI fragment encoding the signal peptide-human angiostatin-huFc-human endostatin was cloned into the pdCs vector for expression. High levels of expression of were obtained in both transient and stable expression, as analyzed by anti-muFc ELISA and SDS-PAGE.

Example 16

[0100] Pharmacokinetics

[0101] In one set of pharmacokinetic studies, C57/BL6 mice with implanted Lewis lung tumors at 100-200 mm3 were injected in the tail vein with 720 μg huFc-huAngio per mouse. The size of the tumors and the dosage of huFc-huAngio used in this study were chosen to simulate the actual treatment protocol described by O'Reilly (O'Reilly et al., (1996) Nature Medicine 2:689). Blood was harvested by retro-orbital bleeding at ½, 1, 2, 4, 8, 24, and 48 hr. post injection. The blood samples were analyzed by anti-huFc ELISA followed by Western analysis. HuFc-huAngio was found to have a circulating half-life of about 32 hr. in mouse and Western analysis showed that over 90% of the hu-Fc-huAngio remained as an intact molecule in circulation.

[0102] The pharmacokinetic studies was also repeated in Swiss mice without tumors at a dosage of 200 μg/mouse. In this case huFc-huAngio was found to have a circulating half-life of about 33 hr.

[0103] Equivalents

[0104] The invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof The foregoing embodiments are therefore to be considered in all respects illustrative rather than limiting on the invention described herein. Scope of the invention is thus indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.

	Number	Date	Country
Parent	09383315	Aug 1999	US
Child	10292418	Nov 2002	US

Expression and export of angiogenesis inhibitors as immunofusins

Information

Publication Number

Date Filed

Date Published

Inventors

CPC

US Classifications

International Classifications

Abstract

Description

Claims

RELATED APPLICATIONS

Provisional Applications (1)

Continuations (1)