AU3883101A

AU3883101A - Subunit optimized fusion proteins

Info

Publication number: AU3883101A
Application number: AU38831/01A
Authority: AU
Inventors: Klaus Bosslet; Harry M. Meade; Dan Pollock
Original assignee: Genzyme Transgenics Corp
Current assignee: rEVO Biologics Inc
Priority date: 1999-09-17
Filing date: 2000-09-18
Publication date: 2001-04-17
Anticipated expiration: 2020-09-18
Also published as: NO20021244L; CA2384766A1; CN1379782A; KR20020039346A; WO2001019842A1; AU781462B2; EP1237900A4; BR0014524A; NO20021244D0; RU2002110116A; MXPA02002768A; HUP0202702A2; NZ517774A; WO2001019842A9; IL148549A0; EP1237900A1; JP2003509038A

Description

WO 01/19842 PCT/US00/25558 SUBUNIT OPTIMIZED FUSION PROTEINS 5 Related Applications This application claims the benefit of a previously filed Provisional Application No. 60/101,083 filed September 18, 1998, which is hereby incorporated by reference. 10 Field of the Invention The invention relates to a fusion protein having a first and a second member, wherein the second member of the fusion protein assembles into a multimer and the other member is chosen, or modified, such that it promotes assembly of the second member into a preselected or an optimal number of subunits. 15 Background of the Invention Fusion proteins can combine useful properties of distinct proteins. E.g., a fusion protein can combine the targeting property of an antibody molecule with the cytotoxic effect of a toxin. 20 Summary of the Invention In general, the invention features, a method of making a fusion protein having: a first member, e.g., a targeting moiety, e.g., an immunoglobulin subunit (e.g., an immunoglobulin heavy chain or light chain, or a fragment of either) fused to a second 25 member, e.g., an enzyme, e.g., a toxin (e.g., an enzyme or toxin subunit). The first and second members are chosen such that the fusion protein assembles into a complex having a number of subunits which optimizes activity of the multimeric form of the second member. In preferred embodiments the first member, or the fusion protein, assembles into a form having the same number of subunits as are present in an active, e.g., native, form of the 30 second member. In preferred embodiments the first member, or the fusion protein, assembles into a form having fewer subunits than are present in an active, e.g., native, form of the second member.

WO 01/19842 PCT/US00/25558 -2 In preferred embodiments, the fusion protein assembles into a complex, e.g., a di-, tri-, tetra-, or higher multi-meric complex. Preferably, the fusion protein assembles into a dimer or a tetramer. In preferred embodiments, the fusion protein assembles into a complex having 5 enzymatic activity. In a preferred embodiment, the first member is a monomer. E.g., it is a species which is normally monomeric, or which has been modified, e.g., by mutation of a site which modulates formation or maintenance of a multimer of subunits. In some embodiments the monomeric form is useful because it does not prevent formation of a multimer by the 10 second member. In another preferred embodiment, the first member is a forms a dimmer, e.g., a heterodimer or homodimer. E.g., it is a species which is normally dimeric, or which has been modified, e.g., by mutation of a site which modulates formation or maintenance of a multimerof subunits, to be dimeric. In some embodiments the dimeric form is useful 15 because it does not prevent formation of a multimer by the second member. In preferred embodiments, the fusion protein has the formula: R1-L-R2; R2-L-R1; R2-R1; or R1-R2, wherein R1 is a first member, e.g., an immunoglobulin subunit, L is a peptide linker and R2 is a second member, e.g., an enzyme subunit. Preferably, R1 and R2 20 are covalently linked, e.g., directly fused or linked via a peptide linker. In preferred embodiments, the first or the second member of the fusion protein, or both are modified by, e.g., substituting or deleting, a portion of the amino acid sequence. In a particularly preferred embodiment the fusion protein includes a first member which is an Ig superfamily member, preferably an Ig subunit, which has been modified to inhibit 25 formation of a multimeric form, e.g., a tetrameric form. Preferably the modification, which can be a change, insertion, or deletion of one or more amino acid residues, results in a subunit which does not form a multimer or which forms a lower order multimer that it normally would form, e.g., it forms a dimer rather than a tetramer. Preferably, a region which mediates formation or maintenance of a multimeric structure is 30 modified and thereby wholly or partly inactivated. E.g., a portion of an immunoglobulin WO 01/19842 PCT/US00/25558 -3 subunit, e.g., a heavy chain, e.g., the hinge region, is modified, e.g., deleted. In those embodiments where the hinge region of the immunoglobulin is modified, e.g., removed, the modified immunoglobulin is monovalent. In preferred embodiments, the modification of the first member inhibits the 5 assembly of the first member, or the fusion protein into a multimer, e.g., results in the production of a monomer, or, e.g., of a dimer, where a higher order multimer would otherwise be formed. In preferred embodiments, the first member is a targeting agent, e.g., a polypeptide having a high affinity for a target, e.g., an antibody, a ligand, or an enzyme. 10 In preferred embodiments, the first member is an immunoglobulin or a fragment thereof, e.g., an antigen binding fragment thereof. Preferably, the immunoglobulin is a monoclonal antibody, e.g., a human, murine (e.g., mouse) monoclonal antibody; or a recombinant monoclonal antibody. Preferably, the monoclonal antibody is a human antibody. In other embodiments, the monoclonal antibody is a recombinant antibody, e.g., a 15 chimeric or a humanized antibody (e.g., it has a variable region, or at least a complementarity determining region (CDR), derived from a non-human antibody (e.g., murine) with the remaining portion(s) are human in origin); or a transgenically produced human antibody (e.g., an antibody produced by a hybridoma which includes a B cell obtained from a transgenic non-human animal, e.g., a transgenic mouse, having a genome 20 comprising a human heavy chain transgene and a light chain transgene fused to an immortalized cell). In preferred embodiments, the first member is a full-length antibody (e.g., an IgG1 or IgG4 antibody) or includes only an antigen-binding portion (e.g., a Fab, F(ab') 2 , Fv or a single chain Fv fragment). 25 In preferred embodiments, the first member is an immunoglobulin subunit selected from the group consisting of a subunit of: IgG (e.g., IgG1, IgG2, IgG3, IgG4), IgM, IgAl, IgA2, IgA.sub.sec, IgD, of IgE. Preferably, the immunoglobulin subunit is an IgG isotype, e.g., IgG3.

WO 01/19842 PCT/US00/25558 -4 In preferred embodiments, the first member is a monomer, e.g., a single chain antibody; or forms a dimer, e.g., a dimer of an immunoglobulin heavy chain and a light chain. In preferred embodiments, the first member is a monovalent antibody (e.g., it 5 includes one pair of heavy and light chains, or antigen binding portions thereof). In other embodiments, the first member is divalent antibody (e.g., it includes two pairs of heavy and light chains, or antigen binding portions thereof). In preferred embodiments, the first member includes an immunoglobulin heavy chain or a fragment thereof, e.g., an antigen binding fragment thereof. Preferably, the 10 immunoglobulin heavy chain or fragment thereof (e.g., an antigen binding fragment thereof) is linked, e.g., linked via a peptide linker or is directly fused, to an enzyme. Preferably, the immunoglobulin heavy chain-enzyme fusion protein is capable of assembling into a functional complex, e.g., a di-, tri-, tetra-, or multi-meric complex having enzymatic activity. The most preferred form is dimeric 15 In preferred embodiments, the first member includes an immunoglobulin heavy chain or fragment thereof (e.g., an antigen binding fragment thereof), and a light chain or a fragment thereof (e.g., an antigen binding fragment thereof). Preferably, the immunoglobulin heavy chain is linked, e.g., linked via a peptide linker or directly fused, to an enzyme. Preferably, the fused immunoglobulin heavy chain -enzyme fusion protein 20 assembles with a light chain, e.g., to produce a functional complex, e.g., a di-, tri-, tetra-, or multi-meric complex having enzymatic activity. The most preferred form is dimeric. In preferred embodiments, the first member is an immunoglobulin that interacts with (e.g., binds to) a cell surface antigen on a target cell, e.g., a cancer cell. For example, the immunoglobulin binds to a tumor cell antigen, e.g., carcinoembryonic antigen (CEA), TAG 25 72, her-2/neu, epidermal growth factor receptor, transferrin receptor, among others. In preferred embodiments, the first member localizes, e.g., increases the concentration of, a fusion protein in proximity to a target cell, e.g., a cancer cell. In preferred embodiments, the second member is a subunit of an enzyme, e.g., an enzyme having one or more subunits (e.g., catalytic subunits). Preferably, the enzyme 30 include one, preferably two, more preferably three, most preferably four subunits. A WO 01/19842 PCT/US00/25558 -5 preferred enzyme is beta-glucuronidase, e.g., a human beta-glucuronidase. The enzyme can be a homo-, or a hetero-multimer. If the enzyme is a heteromultimer, two (or more) fusion proteins are needed to form the active product. In preferred embodiments, the second member is capable of converting a precursor 5 drug, e.g., a prodrug, to a toxic drug. In preferred embodiments, the first member is an immunoglobulin G (IgG) heavy and light chains, and the second member is human beta-glucuronidase fusion protein. In preferred embodiments, the light chain of the first member has an amino acid sequence as shown in Figure IB (SEQ ID NO:2); the light chain of the first member has an 10 amino acid sequence at least 60%, 70%, 75%, more preferably at least 85%, more preferably at least 90%, more preferably at least 95%, most preferably at least 98%, 99% sequence identity or homology with an amino acid sequence from Figure lB (SEQ ID NO:2). In preferred embodiments, the light chain of the first member has an amino acid 15 sequence that is encoded by a nucleotide sequence as shown in Figure lB (SEQ ID NO:1), or Figure 2 (SEQ ID NO:37); the light chain of the first member has an amino acid sequence that is encoded by a nucleotide sequence at least 60%, 70%, 75%, more preferably at least 85%, more preferably at least 90%, more preferably at least 95%, most preferably at least 98%, 99% sequence identity or homology with a nucleotide sequence shown in Figure 20 1B (SEQ ID NOs:2, 3, or 4), or Figure 2 (SEQ ID NO:37); the light chain of the first member has an amino acid sequence that is encoded by a nucleotide sequence that is capable of hybridizing under stringent conditions to the nucleotide sequence shown in Figure lB. In preferred embodiments, the heavy chain of the first member has an amino acid 25 sequence as shown in Figure 4B (SEQ ID NO:6, 7, 8, 9, 10 and/or 11), or Figure 5 (SEQ ID NOs: 13, 14, 15 and/or 16); the heavy chain of the first member has an amino acid sequence at least 60%, 70%, 75%, more preferably at least 85%, more preferably at least 90%, more preferably at least 95%, most preferably at least 98%, 99% sequence identity or homology with an amino acid sequence from Figure 4B (SEQ ID NO: 6, 7, 8, 9, 10 and/or 11), or 30 Figure 5 (SEQ ID Nos:13, 14, 15 and/or 16).

WO 01/19842 PCT/US00/25558 -6 In preferred embodiments, the heavy chain of the first member has an amino acid sequence that is encoded by a nucleotide sequence as shown in Figure 4B (SEQ ID NO: 5), or Figure 5 (SEQ ID NO:12); the heavy chain of the first member has an amino acid sequence that is encoded by a nucleotide sequence at least 60%, 70%, 75%, more preferably 5 at least 85%, more preferably at least 90%, more preferably at least 95%, most preferably at least 98%, 99% sequence identity or homology with a nucleotide sequence shown in Figure 4B (SEQ ID NO:5), or Figure 5 (SEQ ID NO: 12); the heavy chain of the first member has an amino acid sequence that is encoded by a nucleotide sequence that is capable of hybridizing under stringent conditions to the nucleotide sequence shown in Figure 4B, or 5. 10 In a preferred embodiment, the fusion protein includes a peptide linker and the peptide linker has one or more of the following characteristics: a) it allows for the rotation of the first and the second member relative to each other; b) it is resistant to digestion by proteases; c) it does not interact with the first or the second; d) it allows the fusion protein to form a complex (e.g., a di-, tri-, tetra-, or multi-meric complex) that retains enzymatic 15 activity; and e) it promotes folding and/or assembly of the fusion protein into an active complex. In a preferred embodiment: the fusion protein includes a peptide linker and the peptide linker is 5 to 60, more preferably, 10 to 30, amino acids in length; the peptide linker is 20 amino acids in length; the peptide linker is 17 amino acids in length; each of the amino 20 acids in the peptide linker is selected from the group consisting of Gly, Ser, Asn, Thr and Ala; the peptide linker includes a Gly-Ser element. In a preferred embodiment, the fusion protein includes a peptide linker and the peptide linker includes a sequence having the formula (Ser-Gly-Gly-Gly-Gly)y wherein y is 1, 2, 3, 4, 5, 6, 7, or 8. Preferably, the peptide linker includes a sequence having the 25 formula (Ser-Gly-Gly-Gly-Gly) 3 . Preferably, the peptide linker includes a sequence having the formula ((Ser-Gly-Gly-Gly-Gly) 3 -Ser-Pro). In preferred embodiments, the fusion protein is produced recombinantly, e.g., produced in a host cell (e.g., a cultured cell), or in a transgenic animal, e.g., a transgenic mammal (e.g., a goat, a cow, or a rodent (e.g., a mouse).

WO 01/19842 PCT/US00/25558 -7 In preferred embodiments, the fusion protein is produced in a transgenic mammal (e.g., a goat, a cow, or a rodent (e.g., a mouse). Thus, the method further includes: providing a transgenic animal, which includes a transgene which provides for the expression of a fusion protein described herein; allowing the transgene to be 5 expressed; and, preferably, recovering fusion protein, from the milk of the transgenic mammal. For embodiments where the fusion protein is produced transgenically, the fusion protein can further include: a signal sequence which directs the secretion of the fusion protein, e.g., a signal 10 from a secreted protein (e.g., a signal from a protein secreted into milk; or an immunoglobulin secretory signal); and (optionally) a sequence which encodes a sufficient portion of the amino terminal coding region of a secreted protein, e.g., a protein secreted into milk, or an immunoglobulin, to promote secretion, e.g., in the milk of a transgenic mammal, of the fusion protein. 15 In preferred embodiments, the fusion protein is made in a mammary gland of the transgenic mammal, e.g., a ruminant, e.g., a goat or a cow. In preferred embodiments, the fusion protein is secreted into the milk of the transgenic mammal, e.g., a ruminant, e.g., a dairy animal, e.g., a goat or a cow. In preferred embodiments, the fusion protein is secreted into the milk of a transgenic 20 mammal at concentrations of at least about 0.1 mg/ml, 0.5 mg/ml, 1.0 mg/ml, 1.5 mg/ml, 2 mg/ml, 3 mg/ml, 5 mg/ml or higher. In preferred embodiments, the fusion protein is made under the control of a mammary gland specific promoter, e.g., a milk specific promoter, e.g., a milk serum protein or casein promoter. The milk specific promoter can be a casein promoter, beta 25 lactoglobulin promoter, whey acid protein promoter, or lactalbumin promoter. Preferably, the promoter is a goat 13 casein promoter. In preferred embodiments, the transgene encoding the fusion protein is a nucleic acid construct which includes: (a) optionally, an insulator sequence; WO 01/19842 PCT/US00/25558 -8 (b) a promoter, e.g., a mammary epithelial specific promoter, e.g., a milk protein promoter; (c) a nucleotide sequence which encodes a signal sequence which can direct the secretion of the fusion protein, e.g. a signal from a milk specific protein, or an 5 immunoglobulin; (d) optionally, a nucleotide sequence which encodes a sufficient portion of the amino terminal coding region of a secreted protein, e.g. a protein secreted into milk, or an immunoglobulin, to allow secretion, e.g., in the milk of a transgenic mammal, of the fusion protein; 10 (e) one or more nucleotide sequences which encode a fusion protein, e.g., an immunoglobulin-enzyme fusion protein as described herein; and (f) (optionally) a 3' untranslated region from a mammalian gene, e.g., a mammary epithelial specific gene, (e.g., a milk protein gene). In preferred embodiments, elements a (if present), b, c, d (if present), and f of the 15 transgene are from the same gene; the elements a (if present), b, c, d (if present), and f of the transgene are from two or more genes. For example, the signal sequence, the promoter sequence and the 3' untranslated sequence can be from a mammary epithelial specific gene, e.g., a milk serum protein or casein gene (e.g., a P3 casein gene). Preferably, the signal sequence, the promoter sequence and the 3' untranslated sequence are from a goat 13 casein 20 gene. In preferred embodiments, the promoter of the transgene is a mammary epithelial specific promoter, e.g., a milk serum protein or casein promoter (e.g., a 13 casein promoter). The milk specific promoter can be a casein promoter, beta lactoglobulin promoter, whey acid protein promoter, or lactalbumin promoter. Preferably, the promoter is a goat 13 casein 25 promoter. In preferred embodiments, the signal sequence encoded by the transgene is an amino terminal sequence which directs the expression of the protein to the exterior of a cell, or into the cell membrane. For example, the signal sequence can be obtained from an immunoglobulin protein. Preferably, the signal sequence is from a protein which is secreted 30 into the milk, e.g., the milk of the transgenic animal.

WO 01/19842 PCT/US00/25558 -9 In preferred embodiments, the one or more nucleotide sequences encoding a fusion protein include one or more of: a nucleotide sequence encoding a first member, e.g., an immunoglobulin heavy chain (or an antigen binding portion thereof) operably linked to a second member, e.g., an enzyme; (optionally) a nucleotide sequence encoding an 5 immunoglobulin light chain (or an antigen binding portion thereof), or both. In one embodiment, the nucleotide sequences encoding the heavy chain fusion and the light chain are operatively linked in a single construct, e.g., a single cosmid. In another embodiment, the nucleotide sequences encoding the heavy chain fusion and the light chain are introduced into a transgenic animal in separate constructs. Preferably, when linked, the nucleotide 10 sequences are arranged in the following order: 5'-N1-3' linked to 5'-N2-3'; or 5'-N2-3' linked to 5'-N1-3' wherein N1 is a first member, e.g., an immunoglobulin heavy chain (or an antigen binding portion thereof) operably linked to a second member, e.g., an enzyme; and N2 is an immunoglobulin light chain (or an antigen binding portion thereof). The nucleotide sequences can be in any orientation with 15 respect to each other, e.g., sense/sense; reverse/reverse; sense/reverse; or reverse/sense. In preferred embodiments, the 3' untranslated region of the transgene includes a polyadenylation site, and is obtained from a mammalian gene, e.g., a mammary epithelial specific gene, e.g., a milk serum protein gene or casein gene. The 3' untranslated region can be obtained from a casein gene (e.g., a P3 casein gene), a beta lactoglobulin gene, whey 20 acid protein gene, or lactalbumin gene. Preferably, the 3' untranslated region is from a goat 3 casein gene. In preferred embodiments, the transgene, e.g., the transgene as described herein, integrates into a germ cell and/or a somatic cell of the transgenic animal. 25 In another aspect, the invention features, a method for providing a transgenically produced fusion protein, e.g., a fusion protein as described herein, in the milk, of a transgenic mammal. The method includes obtaining milk from a transgenic mammal, which includes a fusion protein encoding transgene, e.g., one which has been introduced into its germline, e.g., a nucleic acid construct as described herein, that result in the WO 01/19842 PCT/US00/25558 -10 expression of the protein-coding sequence of fusion protein in mammary gland epithelial cells, thereby secreting the fusion protein in the milk of the mammal. In preferred embodiments the transgenic mammal is selected from the group consisting of sheep, mice, pigs, cows and goats. The preferred transgenic mammal is a 5 goat. In preferred embodiments, the fusion protein is secreted into the milk of a transgenic mammal at concentrations of at least about 0.1 mg/ml, 0.5 mg/ml, 1.0 mg/ml, 1.5 mg/ml, 2 mg/ml, 3 mg/ml, 5 mg/ml or higher. In preferred embodiments, the transgene encoding the immunoglobulin-enzyme 10 fusion protein is a nucleic acid construct which includes: (a) optionally, an insulator sequence; (b) a promoter, e.g., a mammary epithelial specific promoter, e.g., a milk protein promoter; (c) a nucleotide sequence which encodes a signal sequence which can direct the 15 secretion of the fusion protein, e.g., a signal from a milk specific protein, or an immunoglobulin; (d) optionally, a nucleotide sequence which encodes a sufficient portion of the amino terminal coding region of a secreted protein, e.g., a protein secreted into milk, or an immunoglobulin, to allow secretion, e.g., in the milk of a transgenic mammal, of the non 20 secreted protein; (e) one or more nucleotide sequences which encode a fusion protein, e.g., a fusion protein as described herein; and (f) optionally, a 3' untranslated region from a mammalian gene, e.g., a mammary epithelial specific gene, (e.g., a milk protein gene). 25 In preferred embodiments, elements a (if present), b, c, d (if present), and f of the transgene are from the same gene; the elements a (if present), b, c, d (if present), and f of the transgene are from two or more genes. For example, the signal sequence, the promoter sequence and the 3' untranslated sequence can be from a mammalian gene, e.g., a mammary epitlielial specific gene, e.g., a milk serum protein or casein gene (e.g., a 3 casein gene).

WO 01/19842 PCT/US00/25558 -11 Preferably, the signal sequence, the promoter sequence and the 3' untranslated sequence are from a goat P3 casein gene. In preferred embodiments, the promoter of the transgene is a mammary epithelial specific promoter, e.g., a milk serum protein or casein promoter (e.g., a P3 casein promoter). 5 The milk specific promoter can be a casein promoter, beta lactoglobulin promoter, whey acid protein promoter, or lactalbumin promoter. Preferably, the promoter is a goat P3 casein promoter. In preferred embodiments, the signal sequence encoded by the transgene is an amino terminal sequence which directs the expression of the protein to the exterior of a cell, or into 10 the cell membrane. Preferably, the signal sequence is from a protein which is secreted into the milk, e.g., the milk of the transgenic animal. In preferred embodiments, the one or more nucleotide sequences encoding an fusion protein include one or more of: a nucleotide sequence encoding an immunoglobulin heavy chain (or an antigen binding portion thereof) fused to an enzyme; a nucleotide sequence 15 encoding an immunoglobulin light chain (or an antigen binding portion thereof), or both. In one embodiment, the nucleotide sequences encoding the heavy chain fusion and the light chain are operatively linked in a single construct, e.g., a single cosmid. In another embodiment, the nucleotide sequences encoding the heavy chain fusion and the light chain are introduced into a transgenic animal in separate constructs. Preferably, when linked, the 20 nucleotide sequences are arranged in the following order: 5'-N1-3' linked to 5'-N2-3'; or 5'-N2-3' linked to 5'-N1-3' wherein Nl is an immunoglobulin heavy chain (or an antigen binding portion thereof) linked to an enzyme; and N2 is an immunoglobulin light chain (or an antigen binding portion thereof). The nucleotide sequences can be in any orientation with respect to each other, e.g., sense/sense; 25 reverse/reverse; sense/reverse; or reverse/sense. In preferred embodiments, the 3' untranslated region of the transgene includes a polyadenylation site, and is obtained from a mammalian gene, e.g., a mammary epithelial specific gene, (e.g., a milk serum protein gene or casein gene). The 3' untranslated region can be obtained from a casein gene (e.g., a 3 casein gene), a beta lactoglobulin gene, whey WO 01/19842 PCT/US00/25558 -12 acid protein gene, or lactalbumin gene. Preferably, the 3' untranslated region is from a goat 3 casein gene. In preferred embodiments, the transgene, e.g., the transgene as described herein, integrates into a germ cell and/or a somatic cell of the transgenic animal. 5 In another aspect, the invention features, a transgene, e.g., a nucleic acid construct, preferably, an isolated nucleic acid construct, which includes: (a) optionally, an insulator sequence; 10 (b) a promoter, e.g., a mammary epithelial specific promoter, e.g., a milk protein promoter; (c) a nucleotide sequence which encodes a signal sequence which can direct the secretion of the fusion protein, e.g. a signal sequence from a milk specific protein, or an immunoglobulin; 15 (d) optionally, a nucleotide sequence which encodes a sufficient portion of the amino terminal coding region of a secreted protein, e.g. a protein secreted into milk, or an immunoglobulin, to allow secretion, e.g., in the milk of a transgenic mammal, of the fusion protein protein; (e) one or more nucleotide sequences which encode a fusion protein, e.g., a fusion 20 protein as described herein; and (f) optionally, a 3' untranslated region from a mammalian gene, e.g., a mammary epithelial specific gene, (e.g., a milk protein gene). In preferred embodiments, elements a (if present), b, c, d (if present), and f of the transgene are from the same gene; the elements a (if present), b, c, d (if present), and f of the 25 transgene are from two or more genes. For example, the signal sequence, the promoter sequence and the 3' untranslated sequence can be from a mammalian gene, e.g., a mammary epithelial specific gene, e.g., a milk serum protein or casein gene (e.g., a 3 casein gene). Preferably, the signal sequence, the promoter sequence and the 3' untranslated sequence are from a goat P3 casein gene. 30 In preferred embodiments, the promoter of the transgene is a mammary epithelial specific promoter, e.g., a milk serum protein or casein promoter (e.g., a 13 casein promoter).

WO 01/19842 PCT/US00/25558 - 13 The milk specific promoter can be a casein promoter, beta lactoglobulin promoter, whey acid protein promoter, or lactalbumin promoter. Preferably, the promoter is a goat P casein promoter. In preferred embodiments, the signal sequence encoded by the transgene is an amino 5 terminal sequence which directs the expression of the protein to the exterior of a cell, or into the cell membrane. Preferably, the signal sequence is from a milk specific protein, or an immunoglobulin. Preferably, the signal sequence directs secretion of the encoded fusion protein into the milk of a transgenic animal, e.g., a transgenic mammal. In preferred embodiments, the one or more nucleotide sequences encoding a fusion 10 protein include one or more of: a nucleotide sequence encoding an immunoglobulin heavy chain (or an antigen binding portion thereof) fused to an enzyme; a nucleotide sequence encoding an immunoglobulin light chain (or an antigen binding portion thereof), or both. In one embodiment, the nucleotide sequences encoding the heavy chain fusion and the light chain are operatively linked in a single construct, e.g., a single cosmid. In another 15 embodiment, the nucleotide sequences encoding the heavy chain fusion and the light chain are introduced into a transgenic animal in separate constructs. Preferably, when linked, the nucleotide sequences are arranged in the following order: 5'-N1-3' linked to 5'-N2-3'; or 5'-N2-3' linked to 5'-N1-3' wherein Nl is an immunoglobulin heavy chain (or an antigen binding portion thereof) linked to an enzyme; 20 and N2 is an immunoglobulin light chain (or an antigen binding portion thereof). The nucleotide sequences can be in any orientation with respect to each other, e.g., sense/sense; reverse/reverse; sense/reverse; or reverse/sense. In preferred embodiments, the 3' untranslated region of the transgene includes a polyadenylation site, and is obtained from a mammalian gene, e.g., a mammary epithelial 25 specific gene, (e.g., a milk serum protein gene or casein gene). The 3' untranslated region can be obtained from a casein gene (e.g., a P3 casein gene), a beta lactoglobulin gene, whey acid protein gene, or lactalbumin gene. Preferably, the 3' untranslated region is from a goat 13 casein gene. 30 WO 01/19842 PCT/US00/25558 -14 In another aspect, the invention features a nucleic acid molecule encoding a fusion protein, e.g., a fusion protein as described herein. In preferred embodiments, nucleic acid has a nucleotide sequence as shown in Figure lB (SEQ ID NO:l), Figure 2 (SEQ ID NO:37), Figure 4B (SEQ ID NO:5), or Figure 5 5 (SEQ ID NO:12); the nucleic acid has a nucleotide sequence at least 60%, 70%, 75%, more preferably at least 85%, more preferably at least 90%, more preferably at least 95%, most preferably at least 98%, 99% sequence identity or homology with a nucleotide sequence shown in Figure lB (SEQ ID NO:l), Figure 2 (SEQ ID NO:37), Figure 4B (SEQ ID NO:5), or Figure 5 (SEQ ID NO:12); the nucleic acid has a nucleotide sequence that is 10 capable of hybridizing under stringent conditions to the nucleotide sequence shown in Figure I B, Figure 2, Figure 4B, or Figure 5. In a preferred embodiment, the nucleic acid has a nucleotide sequence which encodes an amino acid sequence as shown in Figure lA (SEQ ID NOs:2, 3, 4), Figure 4B (SEQ ID NO:6, 7, 8, 9, 10, 11), or Figure 5 (SEQ ID NO:13, 14, 15, 16); the nucleic acid 15 has a nucleotide sequence which encodes an amino acid sequence which is at least 60%, 70%, 75%, more preferably at least 85%, more preferably at least 90%, more preferably at least 95%, most preferably at least 98%, 99% sequence identity or homology with an amino acid sequence from Figure lA (SEQ ID NO:2, 3, 4), Figure 4B (SEQ ID NO:6, 7, 8, 9, 10, 11), or Figure 5 (SEQ ID NO:13, 14, 15, 16). 20 In another aspect, the invention features a host cell, e.g., an isolated host cell (e.g., a cultured cell), which includes a nucleic acid of the invention (e.g., a nucleic acid, or a transgene, e.g., a nucleic acid construct, as described herein). 25 In another aspect, the invention features, a fusion protein described herein, or a purified preparation thereof WO 01/19842 PCT/US00/25558 - 15 In another aspect, the invention features, a pharmaceutical or nutraceutical composition having a therapeutically effective amount of a fusion protein, e.g., a fusion protein as described herein, and a pharmaceutically acceptable carrier. In a preferred embodiment, the composition includes milk. 5 In another aspect, the invention features, a transgenic animal which includes a transgene that encodes a fusion protein, e.g., a transgene which encodes a fusion protein described herein. 10 Preferred transgenic animals include: mammals; birds; reptiles; marsupials; and amphibians. Suitable mammals include: ruminants; ungulates; domesticated mammals; and dairy animals. Particularly preferred animals include: mice, goats, sheep, camels, rabbits, cows, pigs, horses, oxen, and llamas. Suitable birds include chickens, geese, and turkeys. Where the transgenic protein is secreted into the milk of a transgenic animal, the animal 15 should be able to produce at least 1, and more preferably at least 10, or 100, liters of milk per year. Preferably, the transgenic animal is a ruminant, e.g., a goat, cow or sheep. Most preferably, the transgenic animal is a goat. In preferred embodiments, the transgenic mammals has germ cells and somatic cells containing a transgene that encodes a fusion protein, e.g., a transgene which encodes a 20 fusion protein described herein. In preferred embodiments, the fusion protein expressed in the transgenic animal is under the control of a mammary gland specific promoter, e.g., a milk specific promoter, e.g., a milk serum protein or casein promoter. The milk specific promoter can be a casein promoter, beta lactoglobulin promoter, whey acid protein promoter, or lactalbumin 25 promoter. Preferably, the promoter is a goat P casein promoter. In preferred embodiments, the transgenic animal is a mammal, and the fusion protein is secreted into the milk of the transgenic animal at concentrations of at least about 0.1 mg/ml, 0.5 mg/ml, 1.0 mg/ml, 1.5 mg/ml, 2 mg/ml, 3 mg/ml, 5 mg/ml or higher. 30 WO 01/19842 PCT/US00/25558 -16 In another aspect, the invention features, a method of making a transgenic organism which has a fusion protein transgene. The method includes providing or forming in a cell of an organism, a fusion protein, e.g., a transgene which encodes a fusion protein described herein; and allowing the cell, or a descendent of the cell, to give rise to a transgenic 5 organism. In a preferred embodiment, the transgenic organism is a transgenic plant or animal. Preferred transgenic animals include: mammals; birds; reptiles; marsupials; and amphibians. Suitable mammals include: ruminants; ungulates; domesticated mammals; and dairy animals. Particularly preferred animals include: mice, goats, sheep, camels, rabbits, 10 cows, pigs, horses, oxen, and llamas. Suitable birds include chickens, geese, and turkeys. Where the transgenic protein is secreted into the milk of a transgenic animal, the animal should be able to produce at least 1, and more preferably at least 10, or 100, liters of milk per year. In preferred embodiments, the fusion protein is under the control of a mammary 15 gland specific promoter, e.g., a milk specific promoter, e.g., a milk serum protein or casein promoter. The milk specific promoter can be a casein promoter, beta lactoglobulin promoter, whey acid protein promoter, or lactalbumin promoter. Preferably, the promoter is a goat 13 casein promoter. In preferred embodiments, the organism is a mammal, and the fusion protein is 20 secreted into the milk of the transgenic animal at concentrations of at least about 0.1 mg/ml, 0.5 mg/ml, 1.0 mg/ml, 1.5 mg/ml, 2 mg/ml, 3 mg/ml, 5 mg/ml or higher. In another aspect, the invention features, a method of selectively killing an aberrant 25 or diseased cell which expresses on its surface a target antigen, e.g., a cancer cell expressing a cell suface antigen. The method includes: contacting said aberrant or diseased cell with an effective amount of a fusion protein, e.g., a fusion protein described herein, wherein either the first or the second member of the fusion protein recognizes said target antigen, such that selective killing of the 30 cell occurs.

WO 01/19842 PCT/US00/25558 -17 The subject method can be used on cells in culture, e.g. in vitro or ex vivo (e.g., cultures comprising cancer cells). For example, cells can be cultured in vitro in culture medium and the contacting step can be effected by adding the fusion protein of the invention to the culture medium. Alternatively, the method can be performed on cells (e.g., 5 cancer cells) present in a subject, e.g., as part of an in vivo (e.g., therapeutic or prophylactic) protocol. In another aspect, the invention features, a method of selectively killing an aberrant 10 or diseased cell which expresses on its surface a target antigen, e.g., a cancer cell expressing a cell suface antigen. The method includes: introducing into said aberrant or diseased cell a nucleic acid encoding a fusion protein, e.g., a fusion protein described herein, wherein either the first or the second member of the fusion protein recognizes said target antigen, such that selective killing of the 15 cell occurs. The subject method can be used on cells in culture, e.g. in vitro or ex vivo (e.g., cultures comprising cancer cells). For example, cells can be cultured in vitro in culture medium and the nucleic acids of the invention can be introduced to the culture medium. Alternatively, the method can be performed on cells (e.g., cancer cells) present in a subject, 20 e.g., as part of an in vivo (e.g., therapeutic or prophylactic) gene therapy protocol. In another aspect, the invention provides, a method of treating in a subject, a disorder characterized by aberrant growth or activity of a cell which expresses on its surface 25 a target antigen, e.g., a cancer cell expressing a target antigen. The method includes administering to the subject an effective amount of a fusion protein, or a nucleic acid encoding a fusion protein (e.g., a fusion protein described herein), wherein either the first or the second member of the fusion protein recognizes said target antigen. In a preferred embodiment, the disease is characterized by aberrant growth or 30 activity of a cell, e.g., cancer cell, an immune cell.

WO 01/19842 PCT/US00/25558 - 18 In yet another aspect, the present invention provides a method for detecting in vitro or in vivo the presence of target antigen in a sample, e.g., for diagnosing a disease. The method comprises (i) contacting a sample or a control sample under conditions that allow interaction of a labelled fusion protein, e.g.. a fusion protein as described herein, and (ii) 5 detecting formation of a complex. A statistically significant change in the formation of the complex between the fusion protein antibody and the target antigen with respect to a control sample is indicative the presence of target antigen in the sample. In preferred embodiments, the second member is an enzyme, e.g., horseradish peroxidase. 10 The invention features fusion proteins in which the ability of a first member of the fusion to from a multimer is chosen so as to optimize a characteristic, e.g., activity or solubility, of the second member. The terms peptides, proteins, and polypeptides are used interchangeably herein. 15 A purified preparation, substantially pure preparation of a polypeptide, or an isolated polypeptide as used herein, means a polypeptide that has been separated from at least one other protein, lipid, or nucleic acid with which it occurs in the cell or organism which expresses it, e.g., from a protein, lipid, or nucleic acid in a transgenic animal or in a fluid, e.g., milk, or other substance, e.g., an egg, produced by a transgenic animal. The 20 polypeptide is preferably separated from substances, e.g., antibodies or gel matrix, e.g., polyacrylamide, which are used to purify it. The polypeptide preferably constitutes at least 10, 20, 50 70, 80 or 95% dry weight of the purified preparation. Preferably, the preparation contains: sufficient polypeptide to allow protein sequencing; at least 1, 10, or 100 ptg of the polypeptide; at least 1, 10, or 100 mg of the polypeptide. 25 A substantially pure nucleic acid, is a nucleic acid which is one or both of: not immediately contiguous with either one or both of the sequences, e.g., coding sequences, with which it is immediately contiguous (i.e., one at the 5' end and one at the 3' end) in the naturally-occurring genome of the organism from which the nucleic acid is derived; or which is substantially free of a nucleic acid sequence with which it occurs in the organism 30 from which the nucleic acid is derived. The term includes, for example, a recombinant WO 01/19842 PCT/US00/25558 - 19 DNA which is incorporated into a vector, e.g., into an autonomously replicating plasmid or virus, or into the genomic DNA of a prokaryote or eukaryote, or which exists as a separate molecule (e.g., a cDNA or a genomic DNA fragment produced by PCR or restriction endonuclease treatment) independent of other DNA sequences. Substantially pure DNA 5 also includes a recombinant DNA which is part of a hybrid gene encoding additional fusion protein sequence. Homology, or sequence identity, as used herein, refers to the sequence similarity between two polypeptide molecules or between two nucleic acid molecules. When a position in the first sequence is occupied by the same amino acid residue or nucleotide as 10 the corresponding position in the second sequence, then the molecules are homologous at that position (i.e., as used herein amino acid or nucleic acid "homology" is equivalent to amino acid or nucleic acid "identity"). The percent homology between the two sequences is a function of the number of identical positions shared by the sequences (i.e., % homology = # of identical positions/total # of positions x 100). 15 For example, if 6 of 10, of the positions in two sequences are matched or homologous then the two sequences are 60% homologous or have 60% sequence identity. By way of example, the DNA sequences ATTGCC and TATGGC share 50% homology or sequence identity. Generally, a comparison is made when two sequences are aligned to give maximum homology or sequence identity. 20 The comparison of sequences and determination of percent homology between two sequences can be accomplished using a mathematical algorithm. A preferred, non-limiting example of a mathematical algorithm utilized for the comparison of sequences is the algorithm of Karlin and Altschul (1990) Proc. Natl. Acad Sci. USA 87:2264-68, modified as in Karlin and Altschul (1993) Proc. Natl. Acad Sci. USA 90:5873-77. Such an 25 algorithm is incorporated into the NBLAST and XBLAST programs (version 2.0) of Altschul, et al. (1990) J. Mol. Biol. 215:403-10. BLAST nucleotide searches can be performed with the NBLAST program, score = 100, wordlength = 12 to obtain nucleotide sequences homologous to ITALY nucleic acid molecules of the invention. BLAST protein searches can be performed with the XBLAST program, score = 50, wordlength = 3 to obtain 30 amino acid sequences homologous to ITALY protein molecules of the invention. To obtain WO 01/19842 PCT/US00/25558 - 20 gapped alignments for comparison purposes, Gapped BLAST can be utilized as described in Altschul et al., (1997) Nucleic Acids Res. 25(17):3389-3402. When utilizing BLAST and Gapped BLAST programs, the default parameters of the respective programs (e.g., XBLAST and NBLAST) can be used. See http://www.ncbi.nlm.nih.gov. Another 5 preferred, non-limiting example of a mathematical algorithm utilized for the comparison of sequences is the algorithm of Myers and Miller, CABIOS (1989). Such an algorithm is incorporated into the ALIGN program (version 2.0) which is part of the GCG sequence alignment software package. When utilizing the ALIGN program for comparing amino acid sequences, a PAM120 weight residue table, a gap length penalty of 12, and a gap penalty of 10 4 can be used. As used herein, the term transgene means a nucleic acid sequence (encoding, e.g., one or more fusion protein polypeptides), which is introduced into the genome of a transgenic organism. A transgene can include one or more transcriptional regulatory sequences and other nucleic acid, such as introns, that may be necessary for optimal 15 expression and secretion of a nucleic acid encoding the fusion protein. A transgene can include an enhancer sequence. A fusion protein sequence can be operatively linked to a tissue specific promoter, e.g., mammary gland specific promoter sequence that results in the secretion of the protein in the milk of a transgenic mammal, a urine specific promoter, or an egg specific promoter. 20 As used herein, the term "transgenic cell" refers to a cell containing a transgene. A transgenic organism, as used herein, refers to a transgenic animal or plant. As used herein, a "transgenic animal" is a non-human animal in which one or more, and preferably essentially all, of the cells of the animal contain a transgene introduced by way of human intervention, such as by transgenic techniques known in the art. The 25 transgene can be introduced into the cell, directly or indirectly by introduction into a precursor of the cell, by way of deliberate genetic manipulation, such as by microinjection or by infection with a recombinant virus. Mammals are defined herein as all animals, excluding humans, that have mammary glands and produce milk.

WO 01/19842 PCT/US00/25558 -21 As used herein, a "dairy animal" refers to a milk producing non-human animal which is larger than a rodent. In preferred embodiments, the dairy animal produce large volumes of milk and have long lactating periods, e.g., cows or goats. As used herein, the language "subject" includes human and non-human animals. 5 The term "non-human animals" of the invention includes vertebrates, e.g., mammals and non-mammals, such as non-human primates, ruminants, birds, amphibians, reptiles and rodents, e.g., mice and rats. The term also includes rabbits. As used herein, a "transgenic plant" is a plant, preferably a multi-celled or higher plant, in which one or more, and preferably essentially all, of the cells of the plant contain a 10 transgene introduced by way of human intervention, such as by transgenic techniques known in the art. As used herein, the term "plant" refers to either a whole plant, a plant part, a plant cell, or a group of plant cells. The class of plants which can be used in methods of the invention is generally as broad as the class of higher plants amenable to transformation 15 techniques, including both monocotyledonous and dicotyledonous plants. It includes plants of a variety of ploidy levels, including polyploid, diploid and haploid. As used herein, the terms "immunoglobulin" and "antibody" refer to a glycoprotein comprising at least two heavy (H) chains and two light (L) chains inter-connected by disulfide bonds. Each heavy chain is comprised of a heavy chain variable region 20 (abbreviated herein as HCVR or VH) and a heavy chain constant region. The heavy chain constant region is comprised of three domains, CHI, CH2 and CH3. Each light chain is comprised of a light chain variable region (abbreviated herein as LCVR or VL) and a light chain constant region. The light chain constant region is comprised of one domain, CL. The VH and VL regions can be further subdivided into regions of hypervariability, termed 25 complementarity determining regions (CDR), interspersed with regions that are more conserved, termed framework regions (FR). Each VH and VL is composed of three CDRs and four FRs, arranged from amino-terminus to carboxy-terminus in the following order: FR1, CDR1, FR2, CDR2, FR3, CDR3, FR4. The variable regions of the heavy and light chains contain a binding domain that interacts with an antigen. The constant regions of the 30 antibodies may mediate the binding of the immunoglobulin to host tissues or factors, WO 01/19842 PCT/US00/25558 - 22 including various cells of the immune system (e.g., effector cells) and the first component (Clq) of the classical complement system. The term "antigen-binding portion" of an antibody (or simply "antibody portion"), as used herein, refers to one or more fragments of an antibody that retain the ability to 5 specifically bind to an antigen (e.g. a target antigen). It has been shown that the antigen binding function of an antibody can be performed by fragments of a full-length antibody. Examples of binding fragments encompassed within the term "antigen-binding portion" of an antibody include (i) a Fab fragment, a monovalent fragment consisting of the VL, VH, CL and CH1 domains; (ii) a F(ab') 2 fragment, a bivalent fragment comprising two Fab 10 fragments linked by a disulfide bridge at the hinge region; (iii) a Fd fragment consisting of the VH and CH 1 domains; (iv) a Fv fragment consisting of the VL and VH domains of a single arm of an antibody, (v) a dAb fragment (Ward et al., (1989) Nature 341:544-546), which consists of a VH domain; and (vi) an isolated complementarity determining region (CDR). Furthermore, although the two domains of the Fv fragment, VL and VH, are coded 15 for by separate genes, they can be joined, using recombinant methods, by a synthetic linker that enables them to be made as a single protein chain in which the VL and VH regions pair to form monovalent molecules (known as single chain Fv (scFv); see e.g., Bird et al. (1988) Science 242:423-426; and Huston et al. (1988) Proc. Natl. Acad Sci. USA 85:5879-5883). Such single chain antibodies are also intended to be encompassed within the term "antigen 20 binding portion" of an antibody. These antibody fragments are obtained using conventional techniques known to those with skill in the art, and the fragments are screened for utility in the same manner as are intact antibodies. The term "monoclonal antibody" as used herein refers to an antibody molecule of single molecular composition. A monoclonal antibody composition displays a single 25 binding specificity and affinity for a particular epitope. Accordingly, the term "human monoclonal antibody" refers to antibodies displaying a single binding specificity which have variable and constant regions derived from human germline immunoglobulin sequences. In one embodiment, the human monoclonal antibodies are produced by a hybridoma which includes a B cell obtained from a transgenic non-human animal, e.g., a WO 01/19842 PCT/US00/25558 - 23 transgenic mouse, having a genome comprising a human heavy chain transgene and a light chain transgene fused to an immortalized cell. The term "recombinant human antibody", as used herein, is intended to include all human antibodies that are prepared, expressed, created or isolated by recombinant means, 5 such as antibodies isolated from an animal (e.g., a mouse) that is transgenic for human immunoglobulin genes; antibodies expressed using a recombinant expression vector transfected into a host cell, antibodies isolated from a recombinant, combinatorial human antibody library, or antibodies prepared, expressed, created or isolated by any other means that involves splicing of human immunoglobulin gene sequences to other DNA sequences. 10 Such recombinant human antibodies have variable and constant regions derived from human germline immunoglobulin sequences. In certain embodiments, however, such recombinant human antibodies are subjected to in vitro mutagenesis (or, when an animal transgenic for human Ig sequences is used, in vivo somatic mutagenesis) and thus the amino acid sequences of the VH and VL regions of the recombinant antibodies are sequences that, 15 while derived from and related to human germline VH and VL sequences, may not naturally exist within the human antibody germline repertoire in vivo. A nucleic acid is "operably linked" when it is placed into a functional relationship with another nucleic acid sequence. For instance, a promoter or enhancer is operably linked to a coding sequence if it affects the transcription of the sequence. With respect to 20 transcription regulatory sequences, operably linked means that the DNA sequences being linked are contiguous and, where necessary to join two protein coding regions, contiguous and in reading frame. The terms "vector" or "construct", as used herein, is intended to refer to a nucleic acid molecule capable of transporting another nucleic acid to which it has been linked. One 25 type of vector is a "plasmid", which refers to a circular double stranded DNA loop into which additional DNA segments may be ligated. Another type of vector is a viral vector, wherein additional DNA segments may be ligated into the viral genome. Certain vectors are capable of autonomous replication in a host cell into which they are introduced (e.g., bacterial vectors having a bacterial origin of replication and episomal mammalian vectors). 30 Other vectors (e.g., non-episomal mammalian vectors) can be integrated into the genome of WO 01/19842 PCT/US00/25558 - 24 a host cell upon introduction into the host cell, and thereby are replicated along with the host genome. Moreover, certain vectors are capable of directing the expression of genes to which they are operatively linked. Such vectors are referred to herein as "recombinant expression vectors" (or simply, "expression vectors"). In general, expression vectors of 5 utility in recombinant DNA techniques are often in the form of plasmids. In the present specification, "plasmid" and "vector" may be used interchangeably as the plasmid is the most commonly used form of vector. However, the invention is intended to include such other forms of expression vectors, such as viral vectors (e.g., replication defective retroviruses, adenoviruses and adeno-associated vectors. 10 The term "recombinant host cell" (or simply "host cell"), as used herein, is intended to refer to a cell into which a recombinant expression vector has been introduced. It should be understood that such terms are intended to refer not only to the particular subject cell but to the progeny of such a cell. Because certain modifications may occur in succeeding generations due to either mutation or environmental influences, such progeny may not, in 15 fact, be identical to the parent cell, but are still included within the scope of the term "host cell" as used herein. Other features and advantages of the invention will be apparent from the following detailed description, and from the claims. 20 Detailed Description The drawings are first described. Figure 1A is a schematic diagram of a construct containing the genomic sequence of the light chain (LC) of humanized anti-carcinoembryonic antigen antibody 431. The location of the signal peptide sequence (s) and the light chain variable (Vk) and the Ck 25 regions are also indicated. The location of the restriction enzyme sites is also indicated. Figure 1B depicts the nucleotide and amino acid sequence for the light chain of humanized anti-carcinoembryonic antigen antibody 431. The location of the restriction enzyme sites is indicated. Figure 2 depicts the nucleotide sequence for a Sal I insert containing the coding 30 sequences for light chain of humanized anti-carcinoembryonic antigen antibody 431.

WO 01/19842 PCT/US00/25558 -25 Figure 3 is a schematic diagram of a construct (Bc 458) which includes the Sal I insert containing the coding sequences for light chain of humanized anti-carcinoembryonic antigen antibody 431. Also indicated is the location of the silencer, 5' P-casein untranslated region, the light chain coding region, and the 3' P-casein untranslated region. 5 Figure 4A is a schematic diagram of a construct containing the genomic sequence of the heavy chain (HC) of humanized anti-carcinoembryonic antigen antibody 431 linked to the 1-glucuronidase sequence. The location of the signal peptide sequence (s) and the heavy chain variable (Vh) and CH1 are also indicated. The location of the restriction enzyme sites is also indicated. 10 Figure 4B depicts the nucleotide and amino acid sequence for the heavy chain of humanized anti-carcinoembryonic antigen antibody 431. The location of the restriction enzyme sites is indicated. Figure 5 depicts the nucleotide and amino acid sequence for the mutant heavy chain of humanized anti-carcinoembryonic antigen antibody 431. The mutant heavy chain lacks 15 the hinge region. The location of the restriction enzyme sites is indicated. Figure 6 is a schematic diagram of a construct (Bc 454) containing the mutant heavy chain of humanized anti-carcinoembryonic antigen antibody 431 linked to the 3 glucuronidase sequence. The location of the silencer, 5' P-casein untranslated region, the heavy chain mutant/p3-glucuronidase fusion coding region, and the 3' P3-casein untranslated 20 region. The location of the restriction enzyme sites is also indicated. Figure 7 is an overview of the construction of the heavy chain mutants. Figure 8 is an enlarged view of the mutations to 3-glucuronidase The present invention provides, at least in part, transgenically produced fusion 25 proteins wherein one member of the fusion protein assembles into a multimer and the other member is chosen, or modified, to promote assembly into the optimal number of subunits. In one embodiment, the fusion protein includes an immunoglobulin subunit (e.g., an immunoglobulin heavy or light chain) fused to a toxin (e.g., a subunit of an enzyme). The immunogloblulin-enzyme fusion proteins described herein serve to target a cytotoxic agent 30 (e.g. the enzyme) to an undesirable cell, e.g., a tumor cell. For example, the fusion proteins WO 01/19842 PCT/US00/25558 - 26 described in the Examples below, (i.e., an antibody against carcinoembryonic antigen (CEA) fused to an enzyme, e.g., glucuronidase) can be used to target, to a tumor cell. After allowing sufficient time for the immunoglobulin-enzyme fusion to localize at the tumor site, a non-toxic prodrug can be administered. This prodrug is converted to a highly cytotoxic 5 drug by the action of the targeted enzyme localizedeat the tumor site, permitting to achieve therapeutic levels of the drug without unacceptable toxicity for the patients. Production of Immunoglobulins A monoclonal antibody against a target antigen, e.g., a cell surface protein (e.g., 10 receptor) on a cell can be produced by a variety of techniques, including conventional monoclonal antibody methodology e.g., the standard somatic cell hybridization technique of Kohler and Milstein, Nature 256: 495 (1975). Although somatic cell hybridization procedures are preferred, in principle, other techniques for producing monoclonal antibody can be employed e.g., viral or oncogenic transformation of B lymphocytes. 15 The preferred animal system for preparing hybridomas is the murine system. Hybridoma production in the mouse is a very well-established procedure. Immunization protocols and techniques for isolation of immunized splenocytes for fusion are known in the art. Fusion partners (e.g., murine myeloma cells) and fusion procedures are also known. Human monoclonal antibodies (mAbs) directed against human proteins can be 20 generated using transgenic mice carrying the complete human immune system rather than the mouse system. Splenocytes from these transgenic mice immunized with the antigen of interest are used to produce hybridomas that secrete human mAbs with specific affinities for epitopes from a human protein (see, e.g., Wood et al. International Application WO 91/00906, Kucherlapati et al. PCT publication WO 91/10741; Lonberg et al. International 25 Application WO 92/03918; Kay et al. International Application 92/03917; Lonberg, N. et al. 1994 Nature 368:856-859; Green, L.L. et al. 1994 Nature Genet. 7:13-21; Morrison, S.L. et al. 1994 Proc. Natl. Acad Sci. USA 81:6851-6855; Bruggeman et al. 1993 Year Immunol 7:33-40; Tuaillon et al. 1993 PNAS 90:3720-3724; Bruggeman et al. 1991 Eur J Immunol 21:1323-1326).

WO 01/19842 PCT/US00/25558 - 27 Monoclonal antibodies can also be generated by other methods known to those skilled in the art of recombinant DNA technology. An alternative method, referred to as the "combinatorial antibody display" method, has been developed to identify and isolate antibody fragments having a particular antigen specificity, and can be utilized to produce 5 monoclonal antibodies (for descriptions of combinatorial antibody display see e.g., Sastry et al. 1989 PNAS 86:5728; Huse et al. 1989 Science 246:1275; and Orlandi et al. 1989 PNAS 86:3833). After immunizing an animal with an immunogen as described above, the antibody repertoire of the resulting B-cell pool is cloned. Methods are generally known for obtaining the DNA sequence of the variable regions of a diverse population of 10 immunoglobulin molecules by using a mixture of oligomer primers and PCR. For instance, mixed oligonucleotide primers corresponding to the 5' leader (signal peptide) sequences and/or framework 1 (FRI1) sequences, as well as primer to a conserved 3' constant region primer can be used for PCR amplification of the heavy and light chain variable regions from a number of murine antibodies (Larrick et al.,1991, Biotechniques 11:152-156). A similar 15 strategy can also been used to amplify human heavy and light chain variable regions from human antibodies (Larrick et al., 1991, Methods.: Companion to Methods in Enzymology 2:106-110). In an illustrative embodiment, RNA is isolated from B lymphocytes, for example, peripheral blood cells, bone marrow, or spleen preparations, using standard protocols (e.g., 20 U.S. Patent No. 4,683,202; Orlandi, et al. PNAS (1989) 86:3833-3837; Sastry et al., PNAS (1989) 86:5728-5732; and Huse et al. (1989) Science 246:1275-1281.) First-strand cDNA is synthesized using primers specific for the constant region of the heavy chain(s) and each of the K and X light chains, as well as primers for the signal sequence. Using variable region PCR primers, the variable regions of both heavy and light chains are amplified, each 25 alone or in combinantion, and ligated into appropriate vectors for further manipulation in generating the display packages. Oligonucleotide primers useful in amplification protocols may be unique or degenerate or incorporate inosine at degenerate positions. Restriction endonuclease recognition sequences may also be incorporated into the primers to allow for the cloning of the amplified fragment into a vector in a predetermined reading frame for 30 expression.

WO 01/19842 PCT/US00/25558 - 28 The V-gene library cloned from the immunization-derived antibody repertoire can be expressed by a population of display packages, preferably derived from filamentous phage, to form an antibody display library. Ideally, the display package comprises a system that allows the sampling of very large variegated antibody display libraries, rapid sorting 5 after each affinity separation round, and easy isolation of the antibody gene from purified display packages. In addition to commercially available kits for generating phage display libraries (e.g., the Pharmacia Recombinant Phage Antibody System, catalog no. 27-9400-01; and the Stratagene SurfZAPTM phage display kit, catalog no. 240612), examples of methods and reagents particularly amenable for use in generating a variegated antibody 10 display library can be found in, for example, Ladner et al. U.S. Patent No. 5,223,409; Kang et al. International Publication No. WO 92/18619; Dower et al. International Publication No. WO 91/17271; Winter et al. International Publication WO 92/20791; Markland et al. International Publication No. WO 92/15679; Breitling et al. International Publication WO 93/01288; McCafferty et al. International Publication No. WO 92/01047; Garrard et al. 15 International Publication No. WO 92/09690; Ladner et al. International Publication No. WO 90/02809; Fuchs et al. (1991) Bio/Technology 9:1370-1372; Hay et al. (1992) Hum Antibod Hybridomas 3:81-85; Huse et al. (1989) Science 246:1275-1281; Griffths et al. (1993) EMBO J 12:725-734; Hawkins et al. (1992) JMol Biol 226:889-896; Clackson et al. (1991) Nature 352:624-628; Gram et al. (1992) PNAS 89:3576-3580; Garrad et al. (1991) 20 Bio/Technology 9:1373-1377; Hoogenboom et al. (1991) NucAcidRes 19:4133-4137; and Barbas et al. (1991) PNAS 88:7978-7982. In certain embodiments, the V region domains of heavy and light chains can be expressed on the same polypeptide, joined by a flexible linker to form a single-chain Fv fragment, and the scFV gene subsequently cloned into the desired expression vector or 25 phage genome. As generally described in McCafferty et al., Nature (1990) 348:552-554, complete VH and VL domains of an antibody, joined by a flexible (Gly4-Ser)3 linker can be used to produce a single chain antibody which can render the display package separable based on antigen affinity. Isolated scFV antibodies immunoreactive with the antigen can subsequently be formulated into a pharmaceutical preparation for use in the subject method.

WO 01/19842 PCT/US00/25558 - 29 Once displayed on the surface of a display package (e.g., filamentous phage), the antibody library is screened with the target antigen, or peptide fragment thereof, to identify and isolate packages that express an antibody having specificity for the target antigen. Nucleic acid encoding the selected antibody can be recovered from the display package 5 (e.g., from the phage genome) and subcloned into other expression vectors by standard recombinant DNA techniques. Specific antibody molecules with high affinities for a surface protein can be made according to methods known to those in the art, e.g, methods involving screening of libraries (Ladner, R.C., et al., U.S. Patent 5,233,409; Ladner, R.C., et al., U.S. Patent 10 5,403,484). Further, the methods of these libraries can be used in screens to obtain binding determinants that are mimetics of the structural determinants of antibodies. In particular, the Fv binding surface of a particular antibody molecule interacts with its target ligand according to principles of protein-protein interactions, hence sequence data for VH and VL (the latter of which may be of the K or k chain type) is the basis for protein 15 engineering techniques known to those with skill in the art. Details of the protein surface that comprises the binding determinants can be obtained from antibody sequence information, by a modeling procedure using previously determined three-dimensional structures from other antibodies obtained from NMR studies or crytallographic data. See for example Bajorath, J. and S. Sheriff, 1996, Proteins: Struct., Funct., and Genet. 24 (2), 20 152-157; Webster, D.M. and A. R. Rees, 1995, "Molecular modeling of antibody combining sites,"in S. Paul, Ed., Methods in Molecular Biol. 51, Antibody Engineering Protocols, Humana Press, Totowa, NJ, pp 17-49; and Johnson, G., Wu, T.T. and E.A. Kabat, 1995, "Seqhunt: A program to screen aligned nucleotide and amino acid sequences," in Methods in Molecular Biol.51, op. cit., pp 1-15. 25 In one embodiment, a variegated peptide library is expressed by a population of display packages to form a peptide display library. Ideally, the display package comprises a system that allows the sampling of very large variegated peptide display libraries, rapid sorting after each affinity separation round, and easy isolation of the peptide-encoding gene from purified display packages. Peptide display libraries can be in, e.g., prokaryotic 30 organisms and viruses, which can be amplified quickly, are relatively easy to manipulate, WO 01/19842 PCT/US00/25558 -30 and which allows the creation of large number of clones. Preferred display packages include, for example, vegetative bacterial cells, bacterial spores, and most preferably, bacterial viruses (especially DNA viruses). However, the present invention also contemplates the use of eukaryotic cells, including yeast and their spores, as potential 5 display packages. Phage display libraries are described above. Other techniques include affinity chromatography with an appropriate "receptor", e.g., a target antigen, followed by identification of the isolated binding agents or ligands by conventional techniques (e.g., mass spectrometry and NMR). Preferably, the soluble receptor is conjugated to a label (e.g., fluorophores, colorimetric enzymes, radioisotopes, 10 or luminescent compounds) that can be detected to indicate ligand binding. Alternatively, immobilized compounds can be selectively released and allowed to diffuse through a membrane to interact with a receptor. Combinatorial libraries of compounds can also be synthesized with "tags" to encode the identity of each member of the library (see, e.g., W.C. Still et al., International 15 Application WO 94/08051). In general, this method features the use of inert but readily detectable tags, that are attached to the solid support or to the compounds. When an active compound is detected, the identity of the compound is determined by identification of the unique accompanying tag. This tagging method permits the synthesis of large libraries of compounds which can be identified at very low levels among to total set of all compounds 20 in the library. The term modified antibody is also intended to include antibodies, such as monoclonal antibodies, chimeric antibodies, and humanized antibodies which have been modified by, e.g., deleting, adding, or substituting portions of the antibody. For example, an antibody can be modified by deleting the hinge region, thus generating a monovalent 25 antibody. Any modification is within the scope of the invention so long as the antibody has at least one antigen binding region specific. Chimeric mouse-human monoclonal antibodies (i.e., chimeric antibodies) can be produced by recombinant DNA techniques known in the art. For example, a gene encoding the Fc constant region of a murine (or other species) monoclonal antibody molecule is 30 digested with restriction enzymes to remove the region encoding the murine Fc, and the WO 01/19842 PCT/US00/25558 -31 equivalent portion of a gene encoding a human Fc constant region is substituted. (see Robinson et al., International Patent Publication PCT/US86/02269; Akira, et al., European Patent Application 184,187; Taniguchi, M., European Patent Application 171,496; Morrison et al., European Patent Application 173,494; Neuberger et al., International 5 Application WO 86/01533; Cabilly et al. U.S. Patent No. 4,816,567; Cabilly et al., European Patent Application 125,023; Better et al. (1988 Science 240:1041-1043); Liu et al. (1987) PNAS 84:3439-3443; Liu et al., 1987, J. Immunol. 139:3521-3526; Sun et al. (1987) PNAS 84:214-218; Nishimura et al., 1987, Canc. Res. 47:999-1005; Wood et al. (1985) Nature 314:446-449; and Shaw et al., 1988, J. Natl Cancer Inst. 80:1553-1559). 10 The chimeric antibody can be further humanized by replacing sequences of the Fv variable region which are not directly involved in antigen binding with equivalent sequences from human Fv variable regions. General reviews of humanized chimeric antibodies are provided by Morrison, S. L., 1985, Science 229:1202-1207 and by Oi et al., 1986, BioTechniques 4:214. Those methods include isolating, manipulating, and 15 expressing the nucleic acid sequences that encode all or part of immunoglobulin Fv variable regions from at least one of a heavy or light chain. Sources of such nucleic acid are well known to those skilled in the art and, for example, may be obtained from 7E3, an anti GPIIbIIIa antibody producing hybridoma. The recombinant DNA encoding the chimeric antibody, or fragment thereof, can then be cloned into an appropriate expression vector. 20 Suitable humanized antibodies can alternatively be produced by CDR substitution U.S. Patent 5,225,539; Jones et al. 1986 Nature 321:552-525; Verhoeyan et al. 1988 Science 239:1534; and Beidler et al. 1988 J Immunol. 141:4053-4060. All of the CDRs of a particular human antibody may be replaced with at least a portion of a non-human CDR or only some of the CDRs may be replaced with non-human 25 CDRs. It is only necessary to replace the number of CDRs required for binding of the humanized antibody to the Fc receptor. An antibody can be humanized by any method, which is capable of replacing at least a portion of a CDR of a human antibody with a CDR derived from a non-human antibody. Winter describes a method which may be used to prepare the humanized antibodies of the 30 present invention (UK Patent Application GB 2188638A, filed on March 26, 1987), the WO 01/19842 PCT/US00/25558 -32 contents of which is expressly incorporated by reference. The human CDRs may be replaced with non-human CDRs using oligonucleotide site-directed mutagenesis. Also within the scope of the invention are chimeric and humanized antibodies in which specific amino acids have been substituted, deleted or added. In particular, preferred 5 humanized antibodies have amino acid substitutions in the framework region, such as to improve binding to the antigen. For example, in a humanized antibody having mouse CDRs, amino acids located in the human framework region can be replaced with the amino acids located at the corresponding positions in the mouse antibody. Such substitutions are known to improve binding of humanized antibodies to the antigen in some instances. 10 Antibodies in which amino acids have been added, deleted, or subsituted are referred to herein as modified antibodies or altered antibodies. Target Antigens In preferred embodiments, the first member of the fusion proteins of the present 15 invention is a targeting agent, e.g., a polypeptide having a high affinity for a target, e.g., an antibody, a ligand, or an enzyme. Accordingly, the fusion proteins of the invention can be used to selectively direct (e.g., localize) the second member of the fusion protein to the vicinity of an undesirable cell. For example, the first member can be an immunoglobulin that interacts with (e.g., 20 binds to a target antigen). In certain embodiments, the target antigen is present on the surface of a cell, e.g., an aberrant cell such a hyperproliferative cell (e.g., a cancer cell). Exemplary target antigens include carcinoembryonic antigen (CEA), TAG-72, her-2/neu, epidermal growth factor receptor, transferrin receptor, among others. As used herein, "target cell" shall mean any undesirable cell in a subject (e.g., a 25 human or animal) that can be targeted by a fusion protein of the invention. Exemplary target cells include tumor cells, such as carcinoma or adenocarcinoma-derived cells (e.g., colon, breast, prostate, ovarian and endometrial cancer cells) (Thor, A. et al. (1997) Cancer Res 46: 3118; Soisson A. P. et al. (1989) Am. J. Obstet. Gynecol. :1258-63). The term "carcinoma" is art recognized and refers to malignancies of epithelial or endocrine tissues 30 including respiratory system carcinomas, gastrointestinal system carcinomas, genitourinary WO 01/19842 PCT/US00/25558 -33 system carcinomas, testicular carcinomas, breast carcinomas, ovarian carcinomas, prostatic carcinomas, endocrine system carcinomas, and melanomas. Exemplary carcinomas include those forming from tissue of the cervix, lung, prostate, breast, head and neck, colon and ovary. The term also includes carcinosarcomas, e.g., which include malignant tumors 5 composed of carcinomatous and sarcomatous tissues. An "adenocarcinoma" refers to a carcinoma derived from glandular tissue or in which the tumor cells form recognizable glandular structures. The term "sarcoma" is art recognized and refers to malignant tumors of mesenchymal derivation. 10 Production of Fusion Proteins The first and second members of the fusion protein can be linked to each other, preferably via a linker sequence. The linker sequence should separate the first and second members of the fusion protein by a distance sufficient to ensure that each member properly 15 folds into its secondary and tertiary structures. Preferred linker sequences (1) should adopt a flexible extended conformation, (2) should not exhibit a propensity for developing an ordered secondary structure which could interact with the functional first and second members, and (3) should have minimal hydrophobic or charged character, which could promote interaction with the functional protein domains. Typical surface amino acids in 20 flexible protein regions include Gly, Asn and Ser. Permutations of amino acid sequences containing Gly, Asn and Ser would be expected to satisfy the above criteria for a linker sequence. Other near neutral amino acids, such as Thr and Ala, can also be used in the linker sequence. A linker sequence length of 20 amino acids can be used to provide a suitable 25 separation of functional protein domains, although longer or shorter linker sequences may also be used. The length of the linker sequence separating the first and second members can be from 5 to 500 amino acids in length, or more preferably from 5 to 100 amino acids in length. Preferably, the linker sequence is from about 5-30 amino acids in length. In preferred embodiments, the linker sequence is from about 5 to about 20 amino acids, and is 30 advantageously from about 10 to about 20 amino acids. Amino acid sequences useful as WO 01/19842 PCT/US00/25558 -34 linkers of the first and second member include, but are not limited to, (SerGly4)y wherein y is greater than or equal to 8, or Gly4SerGly 5 Ser. A preferred linker sequence has the formula (SerGly 4

)

4 . Another preferred linker has the sequence ((Ser-Ser-Ser-Ser-Gly) 3 Ser-Pro). 5 The first and second members can be directly fused without a linker sequence. Linker sequences are unnecessary where the proteins being fused have non-essential N-or C-terminal amino acid regions which can be used to separate the functional domains and prevent steric interference. In preferred embodiments, the C-terminus of first member can be directly fused to the N-terminus of second, or viceversa. 10 Recombinant Production A fusion protein of the invention can be prepared with standard recombinant DNA techniques using a nucleic acid molecule encoding the fusion protein. A nucleotide sequence encoding a fusion protein can be synthesized by standard DNA synthesis methods. 15 A nucleic acid encoding a fusion protein can be introduced into a host cell, e.g., a cell of a primary or immortalized cell line. The recombinant cells can be used to produce the fusion protein. A nucleic acid encoding a fusion protein can be introduced into a host cell, e.g., by homologous recombination. In most cases, a nucleic acid encoding the fusion protein is incorporated into a recombinant expression vector. 20 The nucleotide sequence encoding a fusion protein can be operatively linked to one or more regulatory sequences, selected on the basis of the host cells to be used for expression. The term "operably linked" means that the sequences encoding the fusion protein compound are linked to the regulatory sequence(s) in a manner that allows for expression of the fusion protein. The term "regulatory sequence" refers to promoters, 25 enhancers and other expression control elements (e.g., polyadenylation signals). Such regulatory sequences are described, for example, in Goeddel; Gene Expression Technology: Methods in Enzymology 185, Academic Press, San Diego, CA (1990), the content of which are incorporated herein by reference. Regulatory sequences include those that direct constitutive expression of a nucleotide sequence in many types of host cells, those that 30 direct expression of the nucleotide sequence only in certain host cells (e.g., tissue-specific regulatory sequences) and those that direct expression in a regulatable manner (e.g., only in WO 01/19842 PCT/US00/25558 -35 the presence of an inducing agent). It will be appreciated by those skilled in the art that the design of the expression vector may depend on such factors as the choice of the host cell to be transformed, the level of expression of fusion protein desired, and the like. The fusion protein expression vectors can be introduced into host cells to thereby produce fusion 5 proteins encoded by nucleic acids. Recombinant expression vectors can be designed for expression of fusion proteins in prokaryotic or eukaryotic cells. For example, fusion proteins can be expressed in bacterial cells such as E. coli, insect cells (e.g., in the baculovirus expression system), yeast cells or mammalian cells. Some suitable host cells are discussed further in Goeddel, Gene 10 Expression Technology: Methods in Enzymology 185, Academic Press, San Diego, CA (1990). Examples of vectors for expression in yeast S. cerevisiae include pYepSecl 1 (Baldari et al., (1987) EMBO J. 6:229-234), pMFa (Kurjan and Herskowitz, (1982) Cell 30:933-943), pJRY88 (Schultz et al., (1987) Gene 54:113-123), and pYES2 (Invitrogen Corporation, San Diego, CA). Baculovirus vectors available for expression of fusion 15 proteins in cultured insect cells (e.g., Sf 9 cells) include the pAc series (Smith et al., (1983) Mol. Cell. Biol. 3:2156-2165) and the pVL series (Lucklow, V.A., and Summers, M.D., (1989) Virology 170:31-39). Examples of mammalian expression vectors include pCDM8 (Seed, B., (1987) Nature 329:840) and pMT2PC (Kaufman et al. (1987), EMBO J. 6:187-195). When used 20 in mammalian cells, the expression vector's control functions are often provided by viral regulatory elements. For example, commonly used promoters are derived from polyoma, Adenovirus 2, cytomegalovirus and Simian Virus 40. In addition to the regulatory control sequences discussed above, the recombinant expression vector can contain additional nucleotide sequences. For example, the 25 recombinant expression vector may encode a selectable marker gene to identify host cells that have incorporated the vector. Moreover, to facilitate secretion of the fusion protein from a host cell, in particular mammalian host cells, the recombinant expression vector can encode a signal sequence operatively linked to sequences encoding the amino terminus of the fusion protein such that upon expression, the fusion protein is 30 synthesized with the signal sequence fused to its amino terminus. This signal sequence WO 01/19842 PCT/US00/25558 -36 directs the fusion protein into the secretory pathway of the cell and is then cleaved, allowing for release of the mature fusion protein (i.e., the fusion protein without the signal sequence) from the host cell. Use of a signal sequence to facilitate secretion of proteins or peptides from mammalian host cells is known in the art. 5 Vector DNA can be introduced into prokaryotic or eukaryotic cells via conventional transformation or transfection techniques. As used herein, the terms "transformation" and "transfection" refer to a variety of art-recognized techniques for introducing foreign nucleic acid (e.g., DNA) into a host cell, including calcium phosphate or calcium chloride co precipitation, DEAE-dextran-mediated transfection, lipofection, electroporation, 10 microinjection and viral-mediated transfection. Suitable methods for transforming or transfecting host cells can be found in Sambrook et al. (Molecular Cloning: A Laboratory Manual, 2nd Edition, Cold Spring Harbor Laboratory press (1989)), and other laboratory manuals. Often only a small fraction of mammalian cells integrate the foreign DNA into their 15 genome. In order to identify and select these integrants, a gene that encodes a selectable marker (e.g., resistance to antibiotics) can be introduced into the host cells along with the gene encoding the fusion protein. Preferred selectable markers include those that confer resistance to drugs, such as G418, hygromycin and methotrexate. Nucleic acid encoding a selectable marker can be introduced into a host cell on the same vector as that encoding the 20 fusion protein or can be introduced on a separate vector. Cells stably transfected with the introduced nucleic acid can be identified by drug selection (e.g., cells that have incorporated the selectable marker gene will survive, while the other cells die). A recombinant expression vector can be transcribed and translated in vitro, for example using T7 promoter regulatory sequences and T7 polymerase. 25 Transgenic Mammals Methods for generating non-human transgenic animals are described herein. DNA constructs can be introduced into the germ line of a mammal to make a transgenic mammal. For example, one or several copies of the construct can be incorporated into the genome of a 30 mammalian embryo by standard transgenic techniques.

WO 01/19842 PCT/US00/25558 -37 It is often desirable to express the transgenic protein in the milk of a transgenic mammal. Mammals that produce large volumes of milk and have long lactating periods are preferred. Preferred mammals are ruminants, e.g., cows, sheep, camels or goats, e.g., goats of Swiss origin, e.g., the Alpine, Saanen and Toggenburg breed goats. Other preferred 5 animals include oxen, rabbits and pigs. In an exemplary embodiment, a transgenic non-human animal is produced by introducing a transgene into the germline of the non-human animal. Transgenes can be introduced into embryonal target cells at various developmental stages. Different methods are used depending on the stage of development of the embryonal target cell. The specific 10 line(s) of any animal used should, if possible, be selected for general good health, good embryo yields, good pronuclear visibility in the embryo, and good reproductive fitness. Introduction of the fusion protein transgene into the embryo can be accomplished by any of a variety of means known in the art such as microinjection, electroporation, or lipofection. For example, a fusion protein transgene can be introduced into a mammal by 15 microinjection of the construct into the pronuclei of the fertilized mammalian egg(s) to cause one or more copies of the construct to be retained in the cells of the developing mammal(s). Following introduction of the transgene construct into the fertilized egg, the egg can be incubated in vitro for varying amounts of time, or reimplanted into the surrogate host, or both. One common method is to incubate the embryos in vitro for about 1-7 days, 20 depending on the species, and then reimplant them into the surrogate host. The progeny of the transgenically manipulated embryos can be tested for the presence of the construct by Southern blot analysis of a segment of tissue. An embryo having one or more copies of the exogenous cloned construct stably integrated into the genome can be used to establish a permanent transgenic mammal line carrying the 25 transgenically added construct. Litters of transgenically altered mammals can be assayed after birth for the incorporation of the construct into the genome of the offspring. This can be done by hybridizing a probe corresponding to the DNA sequence coding for the fusion protein or a segment thereof onto chromosomal material from the progeny. Those mammalian progeny 30 found to contain at least one copy of the construct in their genome are grown to maturity.

WO 01/19842 PCT/US00/25558 -38 The female species of these progeny will produce the desired protein in or along with their milk. The transgenic mammals can be bred to produce other transgenic progeny useful in producing the desired proteins in their milk. Transgenic females may be tested for protein secretion into milk, using an art-known 5 assay technique, e.g., a Western blot or enzymatic assay. Other Transgenic Animals Fusion protein can be expressed from a variety of transgenic animals. A protocol for the production of a transgenic pig can be found in White and Yannoutsos, Current Topics in 10 Complement Research: 64th Forum in Immunology, pp. 88-94; US Patent No. 5,523,226; US Patent No. 5,573,933; PCT Application WO93/25071; and PCT Application WO95/04744. A protocol for the production of a transgenic mouse can be found in US Patent No. 5,530,177. A protocol for the production of a transgenic rat can be found in Bader and Ganten, Clinical and Experimental Pharmacology and Physiology, Supp. 3:S81 15 S87, 1996. A protocol for the production of a transgenic cow can be found in Transgenic Animal Technology, A Handbook, 1994, ed., Carl A. Pinkert, Academic Press, Inc. A protocol for the production of a transgenic sheep can be found in Transgenic Animal Technology, A Handbook, 1994, ed., Carl A. Pinkert, Academic Press, Inc. A protocol for the production of a transgenic rabbit can be found in Hammer et al., Nature 315:680-683, 20 1985 and Taylor and Fan, Frontiers in Bioscience 2:d298-308, 1997. Production of Transgenic Protein in the Milk of a Transgenic Animal Milk Specific Promoters 25 Useful transcriptional promoters are those promoters that are preferentially activated in mammary epithelial cells, including promoters that control the genes encoding milk proteins such as caseins, beta lactoglobulin (Clark et al., (1989) Bio/Technology7: 487 492), whey acid protein (Gorton et al. (1987) Bio/Technology 5: 1183-1187), and lactalbumin (Soulier et al., (1992) FEBS Letts. 297: 13). The alpha, beta, gamma or kappa 30 casein gene promoter of any mammalian species can be used to provide mammary WO 01/19842 PCT/US00/25558 -39 expression; a preferred promoter is the goat beta casein gene promoter (DiTullio, (1992) Bio/Technology 10:74-77). Milk-specific protein promoter or the promoters that are specifically activated in mammary tissue can be isolated from cDNA or genomic sequences. Preferably, they are genomic in origin. 5 DNA sequence information is available for mammary gland specific genes listed above, in at least one, and often in several organisms. See, e.g., Richards et al., J Biol. Chem. 256, 526-532 (1981) (a-lactalbumin rat); Campbell et al., Nucleic Acids Res. 12, 8685-8697 (1984) (rat WAP); Jones et al., J Biol. Chem. 260, 7042-7050 (1985) (rat 13 casein); Yu-Lee & Rosen, J Biol. Chem. 258, 10794-10804 (1983) (rat y-casein); Hall, 10 Biochem. J. 242, 735-742 (1987) (a-lactalbumin human); Stewart, Nucleic Acids Res. 12, 389 (1984) (bovine asl1 and K casein cDNAs); Gorodetsky et al., Gene 66, 87-96 (1988) (bovine 1 casein); Alexander et al., Eur. J. Biochem. 178, 395-401 (1988) (bovine K casein); Brignon et al., FEBSLett. 188, 48-55 (1977) (bovine aS2 casein); Jamieson et al., Gene 61, 85-90 (1987), Ivanov et al., Biol. Chem. Hoppe-Seyler 369, 425-429 (1988), Alexander et 15 al., Nucleic Acids Res. 17, 6739 (1989) (bovine 03 lactoglobulin); Vilotte et al., Biochimie 69, 609-620 (1987) (bovine ax-lactalbumin). The structure and function of the various milk protein genes are reviewed by Mercier & Vilotte, J. Dairy Sci. 76, 3079-3098 (1993) (incorporated by reference in its entirety for all purposes). If additional flanking sequence are useful in optimizing expression, such sequences can be cloned using the existing 20 sequences as probes. Mammary-gland specific regulatory sequences from different organisms can be obtained by screening libraries from such organisms using known cognate nucleotide sequences, or antibodies to cognate proteins as probes. Signal Sequences 25 Useful signal sequences are milk-specific signal sequences or other signal sequences which result in the secretion of eukaryotic or prokaryotic proteins. Preferably, the signal sequence is selected from milk-specific signal sequences, i.e., it is from a gene which encodes a product secreted into milk. Most preferably, the milk-specific signal sequence is related to the milk-specific promoter used in the expression system of this invention. The 30 size of the signal sequence is not critical for this invention. All that is required is that the WO 01/19842 PCT/US00/25558 -40 sequence be of a sufficient size to effect secretion of the desired recombinant protein, e.g., in the mammary tissue. For example, signal sequences from genes coding for caseins, e.g., alpha, beta, gamma or kappa caseins, beta lactoglobulin, whey acid protein, and lactalbumin are useful in the present invention. A preferred signal sequence is the goat P-casein signal 5 sequence. Signal sequences from other secreted proteins, e.g., immunoglobulins, or proteins secreted by liver cells, kidney cell, or pancreatic cells can also be used. Insulator Sequences 10 The DNA constructs of the invention further comprise at least one insulator sequence. The terms "insulator", "insulator sequence" and "insulator element" are used interchangeably herein. An insulator element is a control element which insulates the transcription of genes placed within its range of action but which does not perturb gene expression, either negatively or positively. Preferably, an insulator sequence is inserted on 15 either side of the DNA sequence to be transcribed. For example, the insulator can be positioned about 200 bp to about 1 kb, 5' from the promoter, and at least about 1 kb to 5 kb from the promoter, at the 3' end of the gene of interest. The distance of the insulator sequence from the promoter and the 3' end of the gene of interest can be detennrmined by those skilled in the art, depending on the relative sizes of the gene of interest, the promoter 20 and the enhancer used in the construct. In addition, more than one insulator sequence can be positioned 5' from the promoter or at the 3' end of the transgene. For example, two or more insulator sequences can be positioned 5' from the promoter. The insulator or insulators at the 3' end of the transgene can be positioned at the 3' end of the gene of interest, or at the 3'end of a 3' regulatory sequence, e.g., a 3' untranslated region (UTR) or a 25 3' flanking sequence. A preferred insulator is a DNA segment which encompasses the 5' end of the chicken P3-globin locus and corresponds to the chicken 5' constitutive hypersensitive site as described in PCT Publication 94/23046, the contents of which is incorporated herein by reference. 30 WO 01/19842 PCT/US00/25558 -41 DNA Constructs A fusion protein can be expressed from a construct which includes a promoter specific for mammary epithelial cells, e.g., a casein promoter, e.g., a goat beta casein promoter, a milk-specific signal sequence, e.g., a casein signal sequence, e.g., a P-casein 5 signal sequence, and a DNA encoding a fusion protein. A construct can also include a 3' untranslated region downstream of the DNA sequence coding for the non-secreted protein. Such regions can stabilize the RNA transcript of the expression system and thus increases the yield of desired protein from the expression system. Among the 3' untranslated regions useful in the constructs of this invention are 10 sequences that provide a poly A signal. Such sequences may be derived, e.g., from the SV40 small t antigen, the casein 3' untranslated region or other 3' untranslated sequences well known in the art. Preferably, the 3' untranslated region is derived from a milk specific protein. The length of the 3' untranslated region is not critical but the stabilizing effect of its poly A transcript appears important in stabilizing the RNA of the expression sequence. 15 A construct can include a 5' untranslated region between the promoter and the DNA sequence encoding the signal sequence. Such untranslated regions can be from the same control region from which promoter is taken or can be from a different gene, e.g., they may be derived from other synthetic, semi-synthetic or natural sources. Again their specific length is not critical, however, they appear to be useful in improving the level of expression. 20 A construct can also include about 10%, 20%, 30%, or more of the N-terminal coding region of a gene preferentially expressed in mammary epithelial cells. For example, the N-terminal coding region can correspond to the promoter used, e.g., a goat P-casein N terminal coding region. Prior art methods can include making a construct and testing it for the ability to 25 produce a product in cultured cells prior to placing the construct in a transgenic animal. Surprisingly, the inventors have found that such a protocol may not be of predictive value in determining if a normally non-secreted protein can be secreted, e.g., in the milk of a transgenic animal. Therefore, it may be desirable to test constructs directly in transgenic animals, e.g., transgenic mice, as some constructs which fail to be secreted in CHO cells are 30 secreted into the milk of transgenic animals.

WO 01/19842 PCT/US00/25558 - 42 Purification from milk The transgenic fusion protein can be produced in milk at relatively high concentrations and in large volumes, providing continuous high level output of normally 5 processed peptide that is easily harvested from a renewable resource. There are several different methods known in the art for isolation of proteins from milk. Milk proteins usually are isolated by a combination of processes. Raw milk first is fractionated to remove fats, for example, by skimming, centrifugation, sedimentation (H.E. Swaisgood, Developments in Dairy Chemistry, I: Chemistry of Milk Protein, Applied 10 Science Publishers, NY, 1982), acid precipitation (U.S. Patent No. 4,644,056) or enzymatic coagulation with rennin or chymotrypsin (Swaisgood, ibid.). Next, the major milk proteins may be fractionated into either a clear solution or a bulk precipitate from which the specific protein of interest may be readily purified. USSN 08/648,235 discloses a method for isolating a soluble milk component, such 15 as a peptide, in its biologically active form from whole milk or a milk fraction by tangential flow filtration. Unlike previous isolation methods, this eliminates the need for a first fractionation of whole milk to remove fat and casein micelles, thereby simplifying the process and avoiding losses of recovery and bioactivity. This method may be used in combination with additional purification steps to further remove contaminants and purify 20 the component of interest. Production of Transgenic Protein in the Eggs of a Transgenic Animal A fusion protein can be produced in tissues, secretions, or other products, e.g., an egg, of a transgenic animal. For example, fusion proteins can be produced in the eggs of a 25 transgenic animal, preferably a transgenic turkey, duck, goose, ostrich, guinea fowl, peacock, partridge, pheasant, pigeon, and more preferably a transgenic chicken, using methods known in the art (Sang et al., Trends Biotechnology, 12:415-20, 1994). Genes encoding proteins specifically expressed in the egg, such as yolk-protein genes and albumin-protein genes, can be modified to direct expression of fusion protein. 30 WO 01/19842 PCT/US00/25558 - 43 Egg Specific Promoters Useful transcriptional promoters are those promoters that are preferentially activated in the egg, including promoters that control the genes encoding egg proteins, e.g., ovalbumin, lysozyme and avidin. Promoters from the chicken ovalbumin, lysozyme or 5 avidin genes are preferred. Egg-specific protein promoters or the promoters that are specifically activated in egg tissue can be from cDNA or genomic sequences. Preferably, the egg-specific promoters are genomic in origin. DNA sequences of egg specific genes are known in the art (see, e.g., Burley et al., "The Avian Egg", John Wiley and Sons, p. 472, 1989, the contents of which are 10 incorporated herein by reference). If additional flanking sequence are useful in optimizing expression, such sequences can be cloned using the existing sequences as probes. Egg specific regulatory sequences from different organisms can be obtained by screening libraries from such organisms using known cognate nucleotide sequences, or antibodies to cognate proteins as probes. 15 Transgenic Plants A fusion protein can be expressed in a transgenic organism, e.g., a transgenic plant, e.g., a transgenic plant in which the DNA transgene is inserted into the nuclear or plastidic genome. Plant transformation is known as the art. See, in general, Methods in Enzymology 20 Vol. 153 ("Recombinant DNA Part D") 1987, Wu and Grossman Eds., Academic Press and European Patent Application EP 693554. Foreign nucleic acid can be introduced into plant cells or protoplasts by several methods. For example, nucleic acid can be mechanically transferred by microinjection directly into plant cells by use of micropipettes. Foreign nucleic acid can also be transferred 25 into a plant cell by using polyethylene glycol which forms a precipitation complex with the genetic material that is taken up by the cell (Paszkowski et al. (1984) EMBO J. 3:2712-22). Foreign nucleic acid can be introduced into a plant cell by electroporation (Frommn et al. (1985) Proc. Natl. Acad Sci. USA 82:5824). In this technique, plant protoplasts are electroporated in the presence of plasmids or nucleic acids containing the relevant genetic 30 construct. Electrical impulses of high field strength reversibly permeabilize biomembranes WO 01/19842 PCT/US00/25558 -44 allowing the introduction of the plasmids. Electroporated plant protoplasts reform the cell wall, divide, and form a plant callus. Selection of the transformed plant cells with the transformed gene can be accomplished using phenotypic markers. Cauliflower mosaic virus (CaMV) can be used as a vector for introducing foreign 5 nucleic acid into plant cells (Hohn et al. (1982) "Molecular Biology of Plant Tumors," Academic Press, New York, pp. 549-560; Howell, U.S. Pat. No. 4,407,956). CaMV viral DNA genome is inserted into a parent bacterial plasmid creating a recombinant DNA molecule which can be propagated in bacteria. The recombinant plasmid can be further modified by introduction of the desired DNA sequence. The modified viral portion of the 10 recombinant plasmid is then excised from the parent bacterial plasmid, and used to inoculate the plant cells or plants. High velocity ballistic penetration by small particles can be used to introduce foreign nucleic acid into plant cells. Nucleic acid is disposed within the matrix of small beads or particles, or on the surface (Klein et al. (1987) Nature 327:70-73). Although 15 typically only a single introduction of a new nucleic acid segment is required, this method also provides for multiple introductions. A nucleic acid can be introduced into a plant cell by infection of a plant cell, an explant, a meristem or a seed with Agrobacterium tumefaciens transformed with the nucleic acid. Under appropriate conditions, the transformed plant cells are grown to form shoots, 20 roots, and develop further into plants. The nucleic acids can be introduced into plant cells, for example, by means of the Ti plasmid ofAgrobacterium tumefaciens. The Ti plasmid is transmitted to plant cells upon infection by Agrobacterium tumefaciens, and is stably integrated into the plant genome (Horsch et al. (1984) "Inheritance of Functional Foreign Genes in Plants," Science 233:496-498; Fraley et al. (1983) Proc. Natl. Acad Sci. USA 25 80:4803). Plants from which protoplasts can be isolated and cultured to give whole regenerated plants can be transformed so that whole plants are recovered which contain the transferred foreign gene. Some suitable plants include, for example, species from the genera Fragaria, Lotus, Medicago, Onobrychis, Trifolium, Trigonella, Vigna, Citrus, Linum, Geranium, 30 Manihot, Daucus, Arabidopsis, Brassica, Raphanus, Sinapis, Atropa, Capsicum, WO 01/19842 PCT/US00/25558 - 45 Hyoscyamus, Lycopersicon, Nicotiana, Solanum, Petunia, Digitalis, Majorana, Ciohorium, Helianthus, Lactuca, Bromus, Asparagus, Antirrhinum, Hererocallis, Nemesia, Pelargonium, Panicum, Pennisetum, Ranunculus, Senecio, Salpiglossis, Cucumis, Browaalia, Glycine, Lolium, Zea, Triticum, Sorghum, and Datura. 5 Plant regeneration from cultured protoplasts is described in Evans et al., "Protoplasts Isolation and Culture," Handbook ofPlant Cell Cultures 1:124-176 (MacMillan Publishing Co. New York 1983); M.R. Davey, "Recent Developments in the Culture and Regeneration of Plant Protoplasts," Protoplasts (1983)-Lecture Proceedings, pp. 12-29, (Birkhauser, Basal 1983); P.J. Dale, "Protoplast Culture and Plant Regeneration of Cereals and Other 10 Recalcitrant Crops," Protoplasts (1983)-Lecture Proceedings, pp. 31-41, (Birkhauser, Basel 1983); and H. Binding, "Regeneration of Plants," Plant Protoplasts, pp. 21-73, (CRC Press, Boca Raton 1985). Regeneration from protoplasts varies from species to species of plants, but generally a suspension of transformed protoplasts containing copies of the exogenous sequence is first 15 generated. In certain species, embryo formation can then be induced from the protoplast suspension, to the stage of ripening and germination as natural embryos. The culture media can contain various amino acids and hormones, such as auxin and cytokinins. It can also be advantageous to add glutamic acid and proline to the medium, especially for such species as corn and alfalfa. Shoots and roots normally develop simultaneously. Efficient regeneration 20 will depend on the medium, on the genotype, and on the history of the culture. If these three variables are controlled, then regeneration is fully reproducible and repeatable. In vegetatively propagated crops, the mature transgenic plants can be propagated by the taking of cuttings or by tissue culture techniques to produce multiple identical plants for trialling, such as testing for production characteristics. Selection of a desirable transgenic 25 plant is made and new varieties are obtained thereby, and propagated vegetatively for commercial sale. In seed propagated crops, the mature transgenic plants can be self crossed to produce a homozygous inbred plant. The inbred plant produces seed containing the gene for the newly introduced foreign gene activity level. These seeds can be grown to produce plants that have the selected phenotype. The inbreds according to this invention can be used WO 01/19842 PCT/US00/25558 -46 to develop new hybrids. In this method a selected inbred line is crossed with another inbred line to produce the hybrid. Parts obtained from a transgenic plant, such as flowers, seeds, leaves, branches, fruit, and the like are covered by the invention, provided that these parts include cells which 5 have been so transformed. Progeny and variants, and mutants of the regenerated plants are also included within the scope of this invention, provided that these parts comprise the introduced DNA sequences. Progeny and variants, and mutants of the regenerated plants are also included within the scope of this invention. Selection of transgenic plants or plant cells can be based upon a visual assay, such as 10 observing color changes (e.g., a white flower, variable pigment production, and uniform color pattern on flowers or irregular patterns), but can also involve biochemical assays of either enzyme activity or product quantitation. Transgenic plants or plant cells are grown into plants bearing the plant part of interest and the gene activities are monitored, such as by visual appearance (for flavonoid genes) or biochemical assays (Northern blots); Western 15 blots; enzyme assays and flavonoid compound assays, including spectroscopy, see, Harborne et al. (Eds.), (1975) The Flavonoids, Vols. 1 and 2, [Acad. Press]). Appropriate plants are selected and further evaluated. Methods for generation of genetically engineered plants are further described in US Patent No. 5,283,184, US Patent No. 5,482,852, and European Patent Application EP 693 554, all of which are hereby incorporated by reference. 20 Embodiments of the invention are further illustrated by the following examples which should not be construed as being limiting. The contents of all cited references (including literature references, issued patents, published patent applications, and co pending patent applications) cited throughout this application are hereby expressly 25 incorporated by reference. Examples 1 and 2 below describe the generation of two constructs: a light chain construct and a heavy chain/p3-glucuronidase fusion constructs. Two plasmids, one 30 containing a clone of an antibody heavy chain/ human (P-glucuronidase fusion protein and WO 01/19842 PCT/US00/25558 - 47 the other containing kappa light chain sequence were received obtained from Behringwerke AG. EXAMPLE 1: Construction of Light Chain (LC) Construct 5 The Example describes the generation of a light chain nucleic acid construct using the light chain nucleotide sequence from a humanized monoclonal antibody against carcinoembryonic antigen (431) subcloned into a mammary specific expression vector (Bc 163) and a commercial mammalian expression vector (pcDNA3). 10 Briefly, a Hind III -Eco RI fragment containing the light chain sequence was subcloned into pGEM3z to facilitate further manipulation. Two mutations were made: a) To create a Sal I, Xho I, and Kozak consensus sequence at the beginning of the coding region; and b) To creation a Sal I site immediately after the termination codon. 15 The original construct contained approximately 1300 bases of unknown sequence. To remove the unknown sequences, the Gapped Heteroduplex method was used to create a Sal I site just after the termination codon. Sac I sites just before the termination codon and near the Eco RI site were used to make the gap, which was filled using Klenow fragment, 20 deoxynucleotides, T4 DNA ligase, and the following oligonucleotide: TGT TAG AGG TCG ACG CCC CAC (SEQ ID NO:21) term Sal I 25 The gapped region (through the termination codon and new Sal I site) was then sequenced to confirm that no changes were made in sequence. A second Nco I site was found in the unknown sequence that was removed for a subsequent step described below. To remove this site, the construct containing the new Sal I site was digested with Eco RI, ends filled with Klenow fragment and deoxynucleotides, 30 and ligated to a Sal I linker, purchased from New England Biolabs following routine experimental procedures. This construct containing two Sal I sites was then digested with Sal I-and religated, removing the unknown sequence containing the second Nco I site.

WO 01/19842 PCT/US00/25558 -48 A Sal I site and Kozak consensus sequence were then inserted immediately before the initial methionine codon (instead of simply changing the Hind III site) because there were several ATG sequences prior to the correct starting codon that could possibly have been used as alternative start sites. While these ATG sequences do not seem to be a 5 problem in tissue culture, the safest route was to remove this region. These ATG sequences were removed by excising the Hind III Nco I site and replacing it with a Hind III --Nco I adapter containing Sal I and Xho I and a Kozak consensus sequence. The replaced region was also confirmed by sequencing. The sequence changes were as follows: 10 The original 5' prime region had the nucleotide sequence (ATG sequences are capitalized; ATG corresponding to initial methionine is indicated in bold): aagctt ATG aat ATG caaatcctgctc ATG aat ATG caaatcctctga 15 atctac ATG gtaaatataggtttgtctataccacaaacagaaaaac ATG agat cacagttctctctacagttactgagcacacaggacctcacc ATG (SEQ ID NO:22) The original sequence was replace with the following replacement sequence: 20 Hind III Sal I XhoI AAGCTT GTCGAC CTCGAG CCACCATG Kozak (consensus sequence) (SEQ ID NO:23) 25 The Sal I fragment containing the entire coding region of the light chain was then subcloned into the Xho I site ofBcl63, a mammary specific expression vector and pcDNA3, a commercial mammalian expression vector. Orientation was determined by restriction enzyme analysis and/or sequencing. Figure lA is a schematic diagram of the light chain construct (431A). The nucleotide and amino acid sequences are shown in Figure lB. 30 Figure 2 depicts the nucleotide sequence for a Sal I insert containing the coding sequences WO 01/19842 PCT/US00/25558 -49 for light chain of humanized anti-carcinoembryonic antigen antibody 431. Shown as Figure 3 is a schematic diagram of a construct (Bc 458) which includes the Sal I insert containing the coding sequences for light chain of humanized anti-carcinoembryonic antigen antibody 431. Also indicated is the location of the silencer, 5' P3-casein untranslated region, the light 5 chain coding region, and the 3' 1-casein untranslated region. EXAMPLE 2: Construction of Heavy Chain/P3-Glucuronidase Fusion Construct The Example describes the generation of a heavy chain/P3-glucuronidase fusion 10 construct using the heavy chain nucleotide sequence from a humanized monoclonal antibody against carcinoembryonic antigen (431) subcloned into a mammary specific expression vector (Bcl163) and a commercial mammalian expression vector (pcDNA3). The Hind III -Xba I fragment containing the heavy chain/p3-glucuronidase fusion sequence was subcloned into pGEM3z to facilitate further manipulation. Three mutations 15 were made to the coding region of the heavy chain/P-glucuronidase fusion construct: a) To create a Sal I, Xho I, and Kozak consensus sequence at the beginning of the coding region; b) to change the sequence at the internal Sal I site while retaining thecorrect amino acid sequence; and 20 c) to create a Sal I site immediately after the termination codon. The signal sequence that was used for the light chain was also used for the heavy chain. Again, the region between the Hind III and Nco I sites was removed and replaced with the same set of oligonucleotides used in the light chain to create a Sal I site and Kozak consensus sequence immediately before the initial methionine 25 codon. (see above). The internal Sal I site had to be changed for the purpose of subcloning the fragment into a beta casein expression vector. Asn Gly Val Asp Thr Leu (SEQ ID NO:24) 30 original sequence AAT GGG GTC GAC ACG CTA (SEQ ID NO:25) new sequence GTG GAT (SEQ ID NO:26) WO 01/19842 PCT/US00/25558 - 50 Val Asp The 3-prime flanking sequence contained two polyadenylation signal sites and a string of 16 adenine residues between the translational stop codon and the Xba I site. To 5 remove these sequences, a Sal I site was inserted just after the stop codon. Phe Thr * * * original sequence TTT ACT TGA GCA AGA CTG (SEQ ID NO:27) new sequence TTT ACT TGA GGT CGA CTG (SEQ ID NO:28) 10 Sal I The Gapped Heteroduplex method was used to make the changes above. The original plan was to gap the DNA between the Not I and Xba I sites and change the internal Sal I site and add the 3-prime Sal I site at the same time. This proved difficult to 15 accomplish so the 3-prime Sal I site was added first and a new gap was made between the two Bgl II sites to change the internal Sal I site. The gapped regions were then sequenced in entirety to confirm that no changes were made to the sequence. The only difference found was in the fourth intron, 1673 bases from the initial ATG. A cytosine was found in both the mutated and the original plasmid instead of adenine, as shown in the printed sequence 20 above. The Sal I fragment containing the entire coding region of the heavy chain glucuronidase fusion protein was then subcloned into the Xho I site of Bc 163, a mammary specific expression vector and pcDNA3, a commercial mammalian expression vector. Orientation was determined by restriction enzyme analysis and/or sequencing. Figure 4A is a schematic diagram of the light chain construct (431A). The nucleotide and 25 amino acid sequences are shown in Figure 4B. EXAMPLE 3: Generation of Linked Construct This Example described the generation of a construct which includes the light chain 30 and the heavy chain/13-glucuronidase fusion, along with their corresponding upstream and downstream beta casein sequences ligated together into a single cosmid.

WO 01/19842 PCT/US00/25558 -51 In order to eliminate the possibility of integrating only one chain of a two chain protein, such as an antibody, that has been co-injected into mice or other species, both chains along with their own corresponding upstream and downstream beta casein sequences were ligated together into a single cosmid. 5 To achieve this do this supercosl (Stratagene) was modified by inserting the following oligonucleotides into the Barn HI site: Pvu I Pvu I T3... GAT CAC CGA TCG TCG ACC CCC TCG AGC GAT CGA ... T7 (SEQ ID NO:29) 10 TG GCT AGC AGC TGG GCG AGC TCG CTA GCT ACT AG (SEQ ID NO:30) Sal I XhoI These modifications create a new supercos plamid, designated supercos 334, with unique Sal I and Xho I sites. Pvu I, Not I, and Eco RI sites flank these sites and the Barn HI 15 site is destroyed. The Sal I fragments from Bc174 or Bcl75, containing the modified light chain and heavy chain/p3-glucuronidase coding regions within the beta casein 5-prime and 3-prime flanking regions respectively, were inserted into the Xho I site of supercos 334. Three clones were isolated and prepared. The orientation was determined by restriction enzyme 20 analysis. clone # name insert orientation 1 LC14 LC reverse 2 LC13 LC sense 25 11 HC9 HC reverse The complementary Sal I fragments from Bc 174 and Bc 175 (used above) were then ligated into the Sal I site of the above constructions. (Heavy chain fragment into LC13 and LC 14, light chain fragment into HC9). The resulting ligations were then large enough to 30 package in-vitro into lambda phage particles (Amersham kit N. 334) and were used to infect E. coli XL1 Blue. Three versions were generated and one of each of these clones was isolated and prepared: WO 01/19842 PCT/US00/25558 - 52 clone # name insert orientation 1 Bel 80 HC/LC reverse/reverse 9 Bcl81 HC/LC sense/sense 20 Bc 182 LC/HC reverse/reverse 5 Although made through two different pathways, Bcl81 and Bc182 are essentially the same insert when cut away from the vector. When viewed in the sense direction, they both contain the heavy chain/p-glucuronidase Sal I cassette followed by and linked to the 10 light chain Sal I cassette. Each Sal I cassette contains the 5-prime beta casein promoter region, the antibody coding region, and the 3-prime beta casein flanking sequence. In essence, two species were made: the light chain cassette followed by the heavy chain cassette, or the heavy chain cassette followed by the light chain cassette. 15 EXAMPLE 4: Characterization of the Light Chain and Heavy Chain/b-Glucuronidase Constructs The manipulated DNA fragments were tested in tissue culture using the pcDNA3 constructs described above transfected into cos 7 cells using the standard protocol for 20 Lipofectamine using Opti-MEM (Gibco-BRL). Conditioned media (DMEM +10%FBS) was removed after 48 hours and run on a 10 -20% SDS-PAGE gels for Western blotting. Western Blots were conducted following standard procedures, Briefly, for the heavy chain/beta-glucuronidase, samples were run in triplicate under reducing conditions and electroblotted onto nitrocellulose. The nitrocellulose was then cut into three sections and 25 incubated overnight with each of three monoclonal antibodies: Mab 2149/80, Mab 2156/94, and Mab 2156/215. The secondary antibody used for detection was from Cappel (cat. no. 55570), affinity purified horse radish peroxidase conjugated goat anti-mouse IgG. Detection was with the ECL kit from Amersham. Mab 2149/80 was the only antibody that showed a signal on the western blot. 30 For the light chain, samples were again run under reducing conditions and electroblotted onto nitrocellulose. The nitrocellulose was then incubated overnight with WO 01/19842 PCT/US00/25558 - 53 horse radish peroxidase conjugated goat anti-human Kappa chain antibody (Cappel no. 55233). Detection was with the ECL kit from Amersham. EXAMPLE 5: Production of Transgenic Animals 5 Microinjection fragments were prepared by cutting the beta casein constructs Bc 174 (light chain) and Bc 175 (heavy chain) with Sal I to release the bacterial sequences. Fragments were gel purified then buffer exchanged and concentrated using the Wizard system by Promega. 10 Microinjections of the original nucleotide sequences were tested in the mouse model system using an expression vector containing the goat beta casein upstream and coding sequences. Two separate constructions were made and co-injected into mouse embryos, from which founder lines were identified and tested further. The original DNA sequences were also co-injected with an "insulator" sequence which allows us to produce a higher 15 percentage of high producing animal lines. For example, without the insulator generally one in three lines would be a relatively high producer. With the insulator, in many cases, almost all of the lines produced are high expressing lines. Two sets of injections were carried out as follows: For the first set of injections, 1249 embryos were injected of which 838 survived, 20 and 737 were transferred to pseudopregnant females. From these females 80 live pups were born, of which 8 were transgenic, 7 of which carried both chains. For the second set of injections, 508 embryos were injected of which 435 survived, and 426 were transferred to pseudopregnant females. From these females 44 live pups were born of which 2 were transgenic, both of which carried both chains. 25 Bcl81 was injected over three days. In this set, 840 embryos were injected of which 641 survived, and 618 were transferred to pseudopregnant females. From these females 39 live pups were born, of which 5 were transgenic, 3 of which carried both chains. Due to the repetition of the flanking beta casein sequences, it appears that in some cases recombination occurs deleting one chain or the other. 30 Bcl81 was co injected with the silencer fragment over four days. In this case, 1495 goat embryos were injected of which 1183 survived, and 1073 were transferred to WO 01/19842 PCT/US00/25558 - 54 pseudopregnant goat females. From these, 111 live pups were born and 10 of these were transgenic, six carrying both chains. Two of the pups carry both the silencer fragment and both antibody chains. 5 EXAMPLE 5: Generation of Mutants of the Heavy Chain/P-Glucuronidase fusion Protein In an attempt to increase expression of active molecules, two mutations to the heavy chain fusion protein were carried out. The first mutation was to remove the hinge region of the construct. The second mutation removes the hinge and linker sequence (ala-ala-ala-ala 10 val) (SEQ ID NO:31) at the beginning of the 3-glucuronidase coding sequence, fusing the CH2 portion to 03 glucuronidase. To achieve this, gapped heteroduplex mutagenesis was again used. The construct Behring HC5 (which contains the fusion protein in pGem3Z with both ends modified and an internal Sal I site removed) was linearized with (Xba I). A second aliquot was cut with 15 BstE2 plus Not I. When boiled together and cooled some of each strand anneal forming the heteroduplex containing a single stranded gap, in this case between the BstE2 and Not I sites. Two new constructs were then made, sequencing over the gapped portion to make sure no other mutations were made inadvertently. 20 GTC #403: using the oligonucleotide "Behr hinge-alternate" (in bold below) removes the hinge region and part of the introns immediately preceding and after it. ccaaactctctactcACTCAGCTCA CGCATCCACCtccatcccagatccccgt (SEQ ID NO:32) intron intron 25 GTC #406: using the oligonucleotide "Behr hinge/linker" (in bold below) removes the hinge region and ala-ala-ala-ala-val linker, fusing the CH2 and P-glucuronidase coding regions. 30 agcaacaccaaggtgGACAAGAGAGTT CAGGGCGGGATGctgtacccccaggag CH2 coding sequence P-glucuronidase (SEQ ID NO:33) WO 01/19842 PCT/US00/25558 -55 The mutated fusion protein coding sequence can then be excised using Sal I and subcloned into an appropriate expression vector. High levels of expression of the encoded proteins was obtained with a vector consisting of a silencer (or insulator) fragment followed by the goat beta casein promoter, 5 insert DNA, and goat beta casein 3 prime untranslated regions. Both mutant heavy chains and the light chain have been subcloned into such a vector, Bc450, which is flanked by Sal I sites which release the entire injection fragment. Bc454: Bc450 with heavy chain mutant 403 (minus hinge) 10 Bc456: Bc450 with heavy chain mutant 406 (minus hinge/linker) Bc458: Bc450 with the light chain Figure 5 depicts the nucleotide and amino acid sequence for the mutant heavy chain of humanized anti-carcinoembryonic antigen antibody 431 lacking the hinge region. 15 Figure 6 is a schematic diagram of a construct (Bc 454) containing the mutant heavy chain of humanized anti-carcinoembryonic antigen antibody 431 linked to the [ glucuronidase sequence. The location of the silencer, 5' P-casein untranslated region, the heavy chain mutant/ P3 -glucuronidase fusion coding region, and the 3' P3-casein untranslated region. 20 EXAMPLE 6: Characterization of Transgenic Animals The previous examples describe the testing of the original fusion protein and two heavy chain mutants in the milk expression system. The original fusion proteins were tested 25 both without the insulator and also co injected with a separate insulator fragments. The heavy chain mutants, on the other hand, were tested with the insulator integrated into the construct. Initially, the concentration of the fusion protein produced in milk was estimated by comparing the signal of a sample to that of a standard on a Western blot. Later, experiments 30 measured activity rather than concentration based on Western blots. The activity measurements were more accurate.

WO 01/19842 PCT/US00/25558 - 56 Except for the first set of constructs, Bcl74 + Bcl 175, estimates of protein concentration by Western blot are rough estimates. Generally, lines that express well appear to be in the 1-2 mg/ml range. Expression data is summarized below, with more detained data sets for each 5 construct attached. Constructs DNA Insulator Western Maximum (HC) ug/ml activity estimated ug/ml Bcl74/Bc175 Original no - 800 20 Bcl81 original, linked No 1000-2000 Na Bcl81 + insulator original, linked Co injected - 1000 100 fragment Bc456/Bc458 Minus hinge/linker yes 1000-2000 8 Bc454/Bc458 Minus hinge yes 1000-2000 800 Essentially the results shown herein indicate that while high levels of protein can be made in milk, most of this protein is not active. Such inactivity may be due to a folding 10 problem or a problem in the assembly of the tetramer. Removal of the hinge and linker also produced a protein with low activity. In contast, substantial amount of enzymatic activity was achieved upon the removal of the hinge alone. Approximately, 8 mg of this protein have been produced in mouse milk. The isolated protein is currently being tested in in vivo studies ("human CEA positive colon 15 cancer metastasis model"). A summary of the data regarding the mice produced and analysis done follows, in table form. A. Bcl 174/175 founders 20 Original DNA without the insulator WO 01/19842 PCT/US00/25558 -57 Line 2 n d Sex PCR Copy # Western Activity Gen. ug/ml ug/ml* LC HC LC HC HC 2 F + + 10 10 --- 0.14 4 F + + 1 1 --- <0.1 10 F + + 10 10 -800 18 142 F + + -800 22 F + + 10 10 --- <0.1 154 F + + -800 23 F + + 50 50 -400 4 200 F + + 0.0 40 M + + 100 100 --- <0.1 62 F + + + + --- <0.1 81 M + --- --- --- n.a. 85 F + + 25 25 --- .3 116 M + + 5 5 216 F + + -800 221 F + + -800 n.a. = not analyzed (line carries only one chain) 5 10 WO 01/19842 PCT/US00/25558 -58 B. Bcl81 founders Original DNA without the insulator; a fusion of the Bc 174 and Bc 175 injection fragments PCR Copy # Western Activity ug/ml ug/ml* (approx.) Line F1 Sex LC HC LC HC HC 6 M + + 12 12 49 F 0.0 50 F 0.0 52 F >1000 39 60 F >1000 41 25 F + + 15 15 0.0 29 F --- + n.a. n.a. 0.0 33 M --- + n.a. n.a. n.a. 36 F + + 100 100 0.0 n.a. = not analyzed 5 WO 01/19842 PCT/US00/25558 - 59 C. Bcl 81 + insulator founders Original DNA (fusion of the Bc 174 and Bc 175 injection fragments) co-injected with the insulator PCR Copy # Western Activity ug/ml ug/ml* (approx.) Line 2 n d Sex LC HC Sil LC HC sil HC gen. 9 M --- + n.a. 0 1 --- n.a. 13 M + --- n.a. 2 0 --- n.a. 15 F + --- n.a. 3 0 --- n.a. 33 F + + n.a. 3 3 --- -800 40 M --- + n.a. 0 2 --- n.a. 58 M + + n.a. 2 10 -- 2-139 F + + --- -800 43 2-140 F + + --- -800 31 66 F + + n.a. 1 1 + 78 F + + n.a. 1 1 --- 0.0 81 M + + n.a. 1 1 --- Not pass 90 M + + n.a. 20 20 + 2-123 F + + + -1000 -100 2-124 F + + + -1000 60 2-126 F + + --- Low 15 5 n.a. = not analyzed WO 01/19842 PCT/US00/25558 -60 D. Bc456 + Bc458 founders Mutation removing the hinge and linker: 1"S Generation 2" Generation Sex Western HC Activity ug/ml. 6 F none 0 8 F 0 13 F none 0 18 F 0 24 F none 0 57 F good 2 65 F good 8 66 F low 0 138 M 175 F 152 F none 0 5 WO 01/19842 PCT/US00/25558 -61 E. Bc454 + Bc458 Mutation removing the hinge only 1 st Generation 2" a Generation 3 ra Sex Western HC Activity Generation ug/ml. 162 died c-section F -- 4 F High 66 180 F Good 177 11 F 133 12 F 185 182 M -- 15 F Good 83 187 M Not passing gene 193 M -- 27 F High 838 57 F 829 58 F 742 59 F 944 60 F 574 61 F 752 62 F 534 201 F Good 416 215 F (died) 219 M -- F 220 M -- F 5 WO 01/19842 PCT/US00/25558 - 62 Example 7: Generation and Characterization of Transgenic Goats The sections outlined below briefly describe the major steps in the production of transgenic goats. 5 Goat Species and breeds: Swiss-origin goats, e.g., the Alpine, Saanen, and Toggenburg breeds, are preferred in the production of transgenic goats. 10 Goat superovulation: The timing of estrus in the donors is synchronized on Day 0 by 6 mg subcutaneous norgestomet ear implants (Syncromate-B, CEVA Laboratories, Inc., Overland Park, KS). Prostaglandin is administered after the first seven to nine days to shut down the endogenous synthesis of progesterone. Starting on Day 13 after insertion of the implant, a total of 18 mg 15 of follicle-stimulating hormone (FSH - Schering Corp., Kenilworth, NJ) is given intramuscularly over three days in twice-daily injections. The implant is removed on Day 14. Twenty-four hours following implant removal the donor animals are mated several times to fertile males over a two-day period (Selgrath, et al., Theriogenology, 1990. pp. 1195-1205). 20 Embryo collection: Surgery for embryo collection occurs on the second day following breeding (or 72 hours following implant removal). Superovulated does are removed from food and water 36 hours prior to surgery. Does are administered 0.8 mg/kg Diazepam (Valium®) IV, 25 followed immediately by 5.0 mg/kg Ketamine (Keteset), IV. Halothane (2.5%) is administered during surgery in 2 L/min oxygen via an endotracheal tube. The reproductive tract is exteriorized through a midline laparotomy incision. Corpora lutea, unruptured follicles greater than 6 mm in diameter, and ovarian cysts are counted to evaluate superovulation results and to predict the number of embryos that should be collected by 30 oviductal flushing. A cannula is placed in the ostium of the oviduct and held in place with a WO 01/19842 PCT/US00/25558 - 63 single temporary ligature of 3.0 Prolene. A 20 gauge needle is placed in the uterus approximately 0.5 cm from the uterotubal junction. Ten to twenty ml of sterile phosphate buffered saline (PBS) is flushed through the cannulated oviduct and collected in a Petri dish. This procedure is repeated on the opposite side and then the reproductive tract is replaced in 5 the abdomen. Before closure, 10-20 ml of a sterile saline glycerol solution is poured into the abdominal cavity to prevent adhesions. The linea alba is closed with simple interrupted sutures of 2.0 Polydioxanone or Supramid and the skin closed with sterile wound clips. Fertilized goat eggs are collected from the PBS oviductal flushings on a stereomicroscope, and are then washed in Ham's F12 medium (Sigma, St. Louis, MO) 10 containing 10% fetal bovine serum (FBS) purchased from Sigma. In cases where the pronuclei are visible, the embryos is immediately microinjected. If pronuclei are not visible, the embryos can be placed in Ham's F12 containing 10% FBS for short term culture at 37 0 C in a humidified gas chamber containing 5% CO2 in air until the pronuclei become visible (Selgrath, et al., Theriogenology, 1990. pp. 1195-1205). 15 Microinjection procedure: One-cell goat embryos are placed in a microdrop of medium under oil on a glass depression slide. Fertilized eggs having two visible pronuclei are immobilized on a flame polished holding micropipet on a Zeiss upright microscope with a fixed stage using 20 Normarski optics. A pronucleus is microinjected with the DNA construct of interest, e.g., a BC355 vector containing the fusion protein gene operably linked to the regulatory elements of the goat beta-casein gene, in injection buffer (Tris-EDTA) using a fine glass microneedle (Selgrath, et al., Theriogenology, 1990. pp. 1195-1205). 25 Embryo development: After microinjection, the surviving embryos are placed in a culture of Ham's F12 containing 10% FBS and then incubated in a humidified gas chamber containing 5% CO2 in air at 37 0 C until the recipient animals are prepared for embryo transfer (Selgrath, et al., Theriogenology, 1990. p. 1195-1205). 30 WO 01/19842 PCT/US00/25558 - 64 Preparation of recipients: Estrus synchronization in recipient animals is induced by 6 mg norgestomet ear implants (Syncromate-B). On Day 13 after insertion of the implant, the animals are given a single non-superovulatory injection (400 I.U.) of pregnant mares serum gonadotropin 5 (PMSG) obtained from Sigma. Recipient females are mated to vasectomized males to ensure estrus synchrony (Selgrath, et al., Theriogenology, 1990. pp. 1195-1205). Embryo Transfer: All embryos from one donor female are kept together and transferred to a single 10 recipient when possible. The surgical procedure is identical to that outlined for embryo collection outlined above, except that the oviduct is not cannulated, and the embryos are transferred in a minimal volume of Ham's F 12 containing 10% FBS into the oviductal lumen via the fimbria using a glass micropipet. Animals having more than six to eight ovulation points on the ovary are deemed unsuitable as recipients. Incision closure and 15 post-operative care are the same as for donor animals (see, e.g., Selgrath, et al., Theriogenology, 1990. pp. 1195-1205). Monitoring of pregnancy and parturition: Pregnancy is determined by ultrasonography 45 days after the first day of standing 20 estrus. At Day 110 a second ultrasound exam is conducted to confirm pregnancy and assess fetal stress. At Day 130 the pregnant recipient doe is vaccinated with tetanus toxoid and Clostridium C&D. Selenium and vitamin E (Bo-Se) are given IM and Ivermectin was given SC. The does are moved to a clean stall on Day 145 and allowed to acclimatize to this environment prior to inducing labor on about Day 147. Parturition is induced at Day 147 25 with 40 mg of PGF2a (Lutalyse®, Upjohn Company, Kalamazoo Michigan). This injection is given IM in two doses, one 20 mg dose followed by a 20 mg dose four hours later. The doe is under periodic observation during the day and evening following the first injection of Lutalyse® on Day 147. Observations are increased to every 30 minutes beginning on the morning of the second day. Parturition occurred between 30 and 40 hours after the first WO 01/19842 PCT/US00/25558 - 65 injection. Following delivery the doe is milked to collect the colostrum and passage of the placenta is confirmed. Verification of the transgenic nature of F 0 animals: 5 To screen for transgenic F 0 animals, genomic DNA is isolated from two different cell lines to avoid missing any mosaic transgenics. A mosaic animal is defined as any goat that does not have at least one copy of the transgene in every cell. Therefore, an ear tissue sample (mesoderm) and blood sample are taken from a two day old F 0 animal for the isolation of genomic DNA (Lacy, et al., A Laboratory Manual, 1986, Cold Springs Harbor, 10 NY; and Herrmann and Frischauf, Methods Enzymology, 1987. 152: pp. 180-183). The DNA samples are analyzed by the polymerase chain reaction (Gould, et al., Proc. Natl. Acad. Sci, 1989. 86:pp. 1934-1938) using primers specific for the fusion protein gene and by Southern blot analysis (Thomas, Proc Natl. Acad. Sci., 1980. 77:5201-5205) using a random primed first member or second member cDNA probe (Feinberg and Vogelstein, 15 Anal. Bioc., 1983. 132: pp. 6-13). Assay sensitivity is estimated to be the detection of one copy of the transgene in 10% of the somatic cells. Generation and Selection of production herd The procedures described above can be used for production of transgenic founder 20 (F 0 ) goats, as well as other transgenic goats. The transgenic F 0 founder goats, for example, are bred to produce milk, if female, or to produce a transgenic female offspring if it is a male founder. This transgenic founder male, can be bred to non-transgenic females, to produce transgenic female offspring. 25 Transmission of transgene and pertinent characteristics Transmission of the transgene of interest, in the goat line is analyzed in ear tissue and blood by PCR and Southern blot analysis. For example, Southern blot analysis of the founder male and the three transgenic offspring shows no rearrangement or change in the copy number between generations. The Southern blots are probed with immunoglobulin 30 enzyme fusion protein cDNA probe. The blots are analyzed on a Betascope 603 and copy WO 01/19842 PCT/US00/25558 - 66 number determined by comparison of the transgene to the goat beta casein endogenous gene. Evaluation of expression levels 5 The expression level of the transgenic protein, in the milk of transgenic animals, is determined using enzymatic assays or Western blots. Other embodiments are within the following claims.

Claims

1. A method of making a fusion protein having: a first member, fused 5 to a second member wherein the first and second members are chosen such that the fusion protein assembles into a complex having a number of subunits which optimizes activity of the multimeric form of the second member.

2. The method of claim 1, wherein the first member, or the fusion protein, 10 assembles into a form having the same number of subunits as are present in an active form of the second member.

3. The method of claim 1, wherein the first member includes an Ig subunit. 15

4. The method of claim 1, wherein the second member, is other than an Ig subunit.

5. The method of claim 1, wherein the first member is has been modified at a site which modulates formation or maintenance of a multimer of subunits. 20

6. The method of claim 1, wherein the first member forms a dimmer.

7. The method of claim 1, wherein the first member includes an Ig subunit, 25 which has been modified to inhibit formation of a multimeric form.

8. The method of claim 7, wherein the modification is a change, insertion, or deletion of one or more amino acid residues, and results in a subunit which does not form a multimer or which forms a lower order multimer that it normally would form. 30 WO 01/19842 PCT/US00/25558 - 68

9. The method of claim 7, wherein hinge region of the immunoglobulin is modified.

10. The method of claim 7, wherein the modification results in a dimeric Ig 5 structure.

11. The method of claim 10, wherein the dimer includes a heavy chain fusion and a light chain fusion. 10

12. The method of claim 1, wherein the second member includes beta glucuronidase.

13. The method of claim 1, wherein the first member is an immunoglobulin (Ig) heavy of light chain, and the second member is human beta-glucuronidase fusion protein. 15

14. The method of claim 1, wherein the fusion protein is produced in a transgenic animal.

15. A method for providing a transgenically produced fusion protein of claim 1, comprising obtaining milk from a transgenic mammal, which includes a fusion protein 20 encoding transgene that result in the expression of the protein-coding sequence of fusion protein in mammary gland epithelial cells, thereby secreting the fusion protein in the milk of the mammal.

16. A nucleic acid construct, which includes: 25 (a) optionally, an insulator sequence; (b) a promoter, e.g., a mammary epithelial specific promoter, e.g., a milk protein promoter; (c) a nucleotide sequence which encodes a signal sequence which can direct the secretion of the fusion protein, e.g. a signal sequence from a milk specific protein, or an 30 immunoglobulin; WO 01/19842 PCT/US00/25558 - 69 (d) optionally, a nucleotide sequence which encodes a sufficient portion of the amino terminal coding region of a secreted protein, e.g. a protein secreted into milk, or an immunoglobulin, to allow secretion, e.g., in the milk of a transgenic mammal, of the fusion protein protein; 5 (e) one or more nucleotide sequences which encode a fusion protein, e.g., a fusion protein as described herein; and (f) optionally, a 3' untranslated region from a mammary epithelial specific gene, e.g., a milk protein gene. 10

17. A nucleic acid construct, which includes a nucleic acid molecule encoding a fusion protein of claim 1.

18. A fusion protein described in claim 1. 15

19. A transgenic animal which includes a transgene that encodes a fusion protein of claim 1.