WO1996040248A1

WO1996040248A1 - Transvacular and intracellular delivery of lipidized proteins

Info

Publication number: WO1996040248A1
Application number: PCT/US1996/009788
Authority: WO
Inventors: Bernard Malfroy-Camine
Original assignee: Eukarion, Inc.
Priority date: 1995-06-07
Filing date: 1996-06-06
Publication date: 1996-12-19
Also published as: AU6266896A

Abstract

The invention provides methods for targeting a protein, such as an antibody, to intracellular compartments in an eukaryotic cell, methods for enhancing organ uptake of proteins, pharmaceutical compositions of modified proteins for use in human therapy, and methods for manufacturing modified proteins. The modified proteins of the invention comprise an attached lipid portion, wherein one or more acyl groups are linked to the protein through a carbohydrate side-chain and various covalent linkage chemistries which are provided. Lipidized antibodies of the invention can be used for diagnostic and therapeutic uses.

Description

TRANSVASCULAR AND INTRACELLULAR DELIVERY OF LIPIDIZED PROTEINS

FIELD OF THE INVENTION The invention provides methods for targeting a protein, such as an antibody, to intracellular compartments in a eukaryotic cell, methods for enhancing organ uptake of proteins, pharmaceutical compositions of modified proteins for use in human therapy, and methods for manufacturing modified proteins. The modified proteins of the invention comprise an attached lipid portion, wherein one or more acyl groups are linked to the protein through a carbohydrate side-chain and various covalent linkage chemistries which are provided.

BACKGROUND OF THE INVENTION Many naturally-occurring or modified proteins have been proposed as diagnostic and/or therapeutic agents for use in humans and domestic animals. However, proteins are generally only poorly transported across vascular endothelial membranes, if at all, and usually cannot traverse cellular membranes to gain access to intracellular compartments. Thus, for example, antibodies can be raised against purified intracellular proteins, such as transcription factors, intracellular enzymes, and cytoarchitectural structural proteins, but such antibodies generally are not able to enter intact cells and bind to the intracellular antigen targets unless the cell membrane is disrupted. The advent of monoclonal antibody technology in the mid 1970's heralded a new age of medicine. For the first time, researchers and clinicians had access to essentially unlimited quantities of uniform antibodies capable of binding to a predetermined antigenic site and having various immunological effector functions. These proteins, known as

"monoclonal antibodies" were thought to hold great promise in, e.g. , the removal of harmful cells, microbial pathogens, and viruses in vivo. Methods allowing the development of specific monoclonal antibodies having binding specificities directed against almost any desired antigenic epitope, including antigens which are located in intracellular compartments in intact cells, promised a cornucopia of medicinal "magic bullets". Unfortunately, the development of appropriate therapeutic products based on monoclonal antibodies, as well as polyclonal antisera, has been severely hampered by a number of drawbacks inherent in the chemical nature of naturally- occurring antibodies. First, antibodies are generally not able to efficiently gain access to intracellular locations, as immunoglobulins are not able to traverse the plasma membrane of cells, and are typically only internalized, if at all, as a consequence of inefficient endocytotic mechanisms. Second, antibodies do not generally cross vascular membranes (e.g., subendothelial basement membrane) , hampering the efficient uptake of antibodies into organs and interstitial spaces. Therefore, therapies for many important diseases could be developed if there were an efficient method to get specific, biologically active immunoglobulin molecules across capillary barriers and into intracellular locations. For example, the life cycle of a retrovirus such as HIV involves intracellular replication wherein several viral-encoded polypeptides essential for production of infectious virions from an infected cell could potentially be inhibited or blocked if specific monoclonal antibodies reactive with the viral-encoded proteins could readily gain access to the intracellular locations where retroviral replication occurs.

Immunoliposomes have been produced as a potential targeted delivery system for delivering various molecules contained in the liposome to a targeted cell. Immunoliposomes employ immunoglobulins as targeting agents, wherein an acylated immunoglobulin is anchored in the lipid bilayer of the liposome to target the liposome to particular cell types that have external antigens that are bound by the acylated immunoglobulin(s) of the immunoliposomes (Connor and Huang (1985) J. Cell Biol. 101: 582; Huang, L. (1985) Biochemistry 24: 29; Babbitt et al. (1984) Biochemistry 23: 3920; Connor et al. (1984) Proc. Natl. Acad. Sci. (U.S.A.) 81: 1715; Huang et al. (1983) J. Biol. Chem. 258: 14034; Shen et al. (1982) Biochim. Biphvs. Acta 689: 31; Huang et al. (1982) Biochim. Biophys. Acta 716: 140; Huang et al. (1981) J. Immunol. Methods 46: 141; and Huang et al. (1980) J. Biol. Chem. 255: 8015) . Immunoliposomes generally contain immunoglobulins which are attached to acyl substituents of a liposome bilayer through a crosslinking agent such as N-hydroxysuccimide and which thus become anchored in the liposome lipid bilayer. Hence, the crosslinked immunoglobulin is linked to the liposome and serves to target the liposomes to specific cell types bearing a predetermined external antigen by binding to the external cellular antigen. While such methods may serve to target liposomes to particular cell types, immunoliposomes suffer from several important drawbacks that have limited their application as drug-delivery vehicles, particularly for delivering proteins to intracellular locations.

Attempts have been made at modifying proteins so as to facilitate their transport across capillary barriers and into cells (EP 0 329 185) , however, no completely satisfactory method has yet been reported in the art. Chemical modification of proteins, such as antibodies, by non-specific "cationization" to enhance transvascular and intracellular delivery of proteins, has been reported (U.S.S.N. 07/693,872) . However, present methods for making cationized immunoglobulins lead to a significant loss of binding affinity (approximately about 90 percent) of a cationized immunoglobulin for binding to its predetermined epitope as compared to the comparable non-cationized immunoglobulin. Generally, cationization involves carbodiimide linkage of a diamine, such as putrescine or hexanediamine, to the carboxylates of aspartate and glutamate residues in the immunoglobulin polypeptide sequence. These chemical modifications of primary amino acids likely disrupt the secondary and tertiary structure of the immunoglobulin sufficiently to account for the loss in binding affinity. Also, present methods produce some degree of cationization in glutamate and aspartate residues located in the variable domain of an immunoglobulin chain, which results in significant loss of binding affinity and/or specificity. Chemical modification of small molecules has also been proposed as a method to augment transport of small bioactive compounds. Feigner (W091/17242) discloses forming lipid complexes consisting of lipid vesicles and bioactive substances contained therein. Feigner et al. (WO91/16024) discloses cationic lipid compounds that are allegedly useful for enhancing transfer of small bioactive molecules in plants and animals. Liposomes and polycationic nucleic acids have been suggested as methods to deliver polynucleotides into cells. Liposomes often show a narrow spectrum of cell specificities, and when DNA is coated externally on to them, the DNA is often sensitive to cellular nucleases. Newer polycationic lipospermines compounds exhibit broad cell ranges (Behr et al., (1989) Proc. Natl. Acad. Sci. USA 86.6982) and DNA is coated by these compounds. In addition, a combination of neutral and cationic lipid has been indicated as a method for transfection of animal cells (Rose et al., (1991) BioTechniαues 10:520) .

Other approaches to enhancing delivery of drugs, particularly across the blood-brain barrier, utilize pharmacologic-based procedures involving drug latentiation or the conversion of hydrophilic drugs into lipid-soluble drugs. The majority of the latentiation approaches involve blocking the hydroxyl, carboxyl and primary amine groups on the drug to make it more lipid-soluble and therefore more easily transported across the blood-brain barrier. Pardridge and Schimmel, U.S. Patent 4,902,505, disclose chimeric peptides for enhancing transport by receptor-mediated transcytosis.

Thuε, there exists a need in the art for methods of facilitating transport of specific proteins, such as antibodies, across capillary barriers and into cells, and for pharmaceutical compositions of such immunoglobulins for treating human and veterinary diseases which are amenable to treatment with intracellular proteins and targeting agents like monoclonal antibodies. SUMMARY OF THE INVENTION The prior art method of increasing antibody transport into cells by attaching a cationic substituent to the primary polypeptide sequence of an immunoglobulin by a relatively nonspecific linkage chemistry have been observed to produce detrimental alterations in the secondary, tertiary, and/or quaternary structure of the protein. These structural alterations apparently cause the loss of binding affinity observed in cationized antibodies. To overcome this, the present invention provides methods wherein lipid substituents are linked to a protein, such as an immunoglobulin, typically by covalent linkage to a carbohydrate side chain of the protein such that the lipid substituent does not substantially destroy the biological activity of the protein (e.g., antigen binding) .

The invention provides methods for producing lipidized proteins, generally by lipidization of a carbohydrate moiety on a glycoprotein or glycopeptide. In general, the methods of the invention are used for attaching a lipid, such as a lipoamine, to a polypeptide, typically by covalent linkage of the lipid to a carbohydrate moiety on a protein, wherein the carbohydrate moiety generally is chemically oxidized and reacted with a lipoamine to form a lipidized protein. The resultant lipidized protein generally has advantageous pharmacokinetic characteristics, such as an increased capacity to cross vascular barriers and access parenchymal cells of various organs and an increased ability to access intracellular compartments. In one aspect of the invention, lipidization of proteins, such as antibodies directed against transcription factors (e.g., Fos, Jun, AP-1, OCT-1, NF-KB) , enhances intranuclear localization of the lipidized protein(s), including for diagnostic detection of antigens in cells, either fixed (e.g., histological sample) or as intact, viable cells, such as cultured cells or cellular explants.

The invention also provides methods for producing lipidized antibodies that are efficiently transported across capillary barriers and internalized into mammalian cells in vivo. The methods of the invention relate to methods for chemically attaching at least one lipid substituent (e.g., lipoamine) to a carbohydrate substituent on an immunoglobulin to produce a carbohydrate-linked lipidized immunoglobulin, wherein the lipidized immunoglobulin is capable of intracellular localization. In alternate embodiments of the invention, at least one lipid substituent (e.g., lipoamine) is covalently attached to a non-carbohydrate moiety on a protein or polypeptide (e.g., by formation of an amide linkage with a Asp or Glu residue side-chain carboxyl substituent or a thioester linkage with a Cys residue) . Also, a fatty acid can be linked to an Arg or Lys residue by the side-chain amine substituents. Similarly, lipid substitutents can be covalently attached to peptidomimetic compounds. Peptide analogs are commonly used in the pharmaceutical industry as non-peptide drugs with properties analogous to those of the template peptide. These types of non-peptide compound are termed "peptide mimetics" or "peptidomimetics" (Fauchere, J. (1986) Adv. Drug Res. 15: 29; Veber and Freidinger (1985) TINS p.392; and Evans et al. (1987) J. Med. Chem 30: 1229, which are incorporated herein by reference) and are usually developed with the aid of computerized molecular modeling. Peptide mimetics that are structurally similar to therapeutically useful peptides may be used to produce an equivalent therapeutic or prophylactic effect. Generally, peptidomimetics are structurally similar to a paradigm polypeptide (i.e., a polypeptide that has a biological or pharmacological activity) , but have one or more peptide linkages optionally replaced by a linkage selected from the group consisting of: -CH₂NH-, -CH₂S-, -CH₂-CH₂-, -CH=CH- (cis and trans), -COCH₂-, -CH(OH)CH₂-, and -CH₂SO-, by methods known in the art and further described in the following references: Spatola, A.F. in "Chemistry and Biochemistry of Amino Acids, Peptides, and Proteins," B. Weinstein, eds., Marcel Dekker, New York, p. 267 (1983); Spatola, A.F., Vega Data (March 1983), Vol. 1, Issue 3, "Peptide Backbone Modifications" (general review); Morley, J.S., Trends Pharm Sci (1980) pp. 463-468 (general review); Hudson, D. et al. , Int J Pept Prot Res (1979) 11:177-185 (-CH₂NH-, CH₂CH₂-) ; Spatola, A.F. et al., Life Sci (1986) 3_8:1243-1249 (-CH₂-S) ; Hann, M.M. , J Chem Soc Perkin Trans I (1982) 307-314 (-CH-CH-, cis and trans) ; Al quist, R.G. et al., J Med Chem (1980) 2_3:1392-1398 (-COCH₂- ); Jennings-White, C. et al., Tetrahedron Lett (1982) 23:2533 (-COCH₂-) ; Szelke, M. et al., European Appln. EP 45665 (1982) CA: 92:39405 (1982) (-CH(OH)CH₂~) ; Holladay, M.W. et al.,

Tetrahedron Lett (1983) 2_4:4401-4404 (-C(OH) CH₂-) ; and Hruby, V.J., Life Sci (1982) 3.1:189-199 (-CH₂-S-) ; each of which is incorporated herein by reference. A particularly preferred non-peptide linkage is -CH₂NH-. Such peptide mimetics may have significant advantages over polypeptide embodiments, including, for example: more economical production, greater chemical stability, enhanced pharmacological properties (half- life, absorption, potency, efficacy, etc.), altered specificity (e.g., a broad-spectrum of biological activities), reduced antigenicity, and others. Lipidization of peptidomimetics usually involves covalent attachment of one or more acyl chains, directly or through a spacer (e.g., an amide group), to non-interfering position(s) on the peptidomimetic that are predicted by quantitative structure-activity data and/or molecular modeling. Such non-interfering positions generally are positions that do not form direct contacts with the macromolecules(s) (e.g., receptors) to which the peptidomimetic binds to produce the therapeutic effect. Lipidization of peptidomimetics should not substantially interfere with the desired biological or pharmacological activity of the peptidomimetic.

The invention also relates to therapeutic and diagnostic compositions of lipidized proteins, such as lipidized antibodies, that can cross vascular membranes and enter the intracellular compartment, particularly lipidized antibodies that bind to intracellular immunotherapeutic targets, such as viral-encoded gene products that are essential components of a viral life cycle (e.g., HIV-1 Tat protein) , to intracellular antigens that are biologically active (e.g., an oncogene protein such as c-fos, c-src, c-myc, c-lck (p56) , c-fyn (p59) , and c-abl) , and/or to transmembrane or extracellular antigens (e.g., polypeptide hormone receptors such as an IL-2 receptor, PDGF receptor, EGF receptor, NGF receptor, GH receptor, or TNF receptor) . Other proteins which can be targeted by lipidized antibodies include, but are not limited to, the following: c-ras p21, c-her-2 protein, c-raf, any of the various G proteins and/or G-protein activating proteins (GAPs) , transcription factors such as NF-AT, calcineurin, and cis-trans prolyl isomerases. The lipidized antibodies can be used to localize a diagnostic reagent, such as a radiocontrast agent or magnetic resonance imaging component, to a specific location in the body, such as a specific organ, tissue, body compartment, cell type, neoplasm, or other anatomical structure (e.g., a pathological lesion). The lipidized antibodies can also be used to localize linked therapeutic agents, such as chemotherapy drugs, radiosensitizing agents, radionuclides, antibiotics, and other agents, to specific locations in the body. Alternatively, the lipidized antibodies of the invention can be used therapeutically for neutralizing (i.e, binding to and thereby inactivating) an intracellular target antigen, such as HIV-1 Tat protein, a transmembrane or membrane-associated antigen target (e.g., -glutamyltranspeptidase, c-ras^H p21, rasGAP) or an extracellular antigen target (i.e., β-amyloid protein deposits in the brain of an Alzheimer's disease patient). Lipidized antibodies can traverse the blood-brain barrier and react with extracellular antigen targets that are generally inaccessible to immunoglobulins which circulate in the blood or lymphatic system. Lipidized antibodies can also react with intracellular portions on transmembrane proteins, such as cytoplasmic tails of viral envelope glycoproteins or protein kinase domains of protooncogene proteins (c-src, c-abl) , and thus inhibit production of infectious enveloped virus or kinase activity, respectively. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 shows structural formulae representing various lipoamines that can be used in the invention. The righthand column exemplifies branched-chain lipoamines and the lefthand column exemplifies straight-chain lipoamines.

Figure 2 is a schematic representation of (1) a glycosylated antibody comprising an immunoglobulin tetramer (two light chains associated with two heavy chains) , and (2) a schematic representation of carbohydrate-linked lipidized immunoglobulins of the invention. For example but not limitation, branched-chain lipoamide substituents are shown attached to partially oxidized carbohydrate sidechains of an immunoglobulin tetramer. Such carbohydrate sidechains may be located in the C_H, V_H, C_L, and/or V_L regions. Figure 3 shows the beneficial effect of a lipidized anti-Tat immunoglobulin on the in vitro survival of cells infected with HIV-1 as compared to the lack of effect of the native (i.e., non-lipidized) anti-Tat immunoglobulin.

Fig. 4 shows that the lipidized anti-Tat antibody significantly inhibited CAT activity (by approximately 75%) , whereas native (unlipidized) anti-Tat antibody, lipidized anti-gpl20 antibody, or rsCD4 were far less effective in inhibiting CAT activity in HLCD4-CAT cells.

Fig. 5 shows the effect of lipidized anti-Tat antibody on CAT expression in HIV-1 LTR-transfected HeLa cells.

Fig. 6A and 6B show the effect of lipidized anti-Tat antibody on HIV-1 replication in living cells.

Fig. 7 shows the effect of lipidized anti-Tat antibody on HIV-1 replication in living cells.

DETAILED DESCRIPTION Definitions

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Although any methods and materials similar or equivalent to those described herein can be used in the practice or testing of the present invention, the preferred methods and materials are described. For purposes of the present invention, the following terms are defined below. As used herein, the twenty conventional amino acids and their abbreviations follow conventional usage (Immunology - A Synthesis. 2nd Edition, E.S. Golub and D.R. Gren, Eds., Sinauer Associates, Sunderland, Massachusetts (1991) , which is incorporated herein by reference) . The term "corresponds to" is used herein to mean that a polynucleotide sequence is homologous (i.e., is identical, not strictly evolutionarily related) to all or a portion of a reference polynucleotide sequence, or that a polypeptide sequence is identical to a reference polypeptide sequence. In contradistinction, the term "complementary to" is used herein to mean that the complementary sequence is homologous to all or a portion of a reference polynucleotide sequence. For illustration, the nucleotide sequence "TATAC" corresponds to a reference sequence "TATAC" and is complementary to a reference sequence "GTATA" .

The terms "substantial similarity" or "substantial identity" as used herein denotes a characteristic of a polypeptide sequence or nucleic acid sequence, wherein the polypeptide sequence has at least 50 percent sequence identity compared to a reference sequence, and the nucleic acid sequence has at least 70 percent sequence identity compared to a reference sequence. The percentage of sequence identity is calculated excluding small deletions or additions which total less than 25 percent of the reference sequence. The reference sequence may be a subset of a larger sequence, such as a constant region domain of a constant region immunoglobulin gene; however, the reference sequence is at least 18 nucleotides long in the case of polynucleotides, and at least 6 amino residues long in the case of a polypeptide. The term "naturally-occurring" as used herein as applied to an object refers to the fact that an object can be found in nature. For example, a polypeptide or polynucleotide sequence that is present in an organism (including viruses) that can be isolated from a source in nature and which has not been intentionally modified by man in the laboratory is naturally-occurring. A lipoprotein (e.g., a naturally- occuring isoprenylated or myristylated protein) that can be isolated from an organism that is found in nature and has not been engineered by man is a naturally-occurring lipoprotein. "Glycosylation sites" refer to amino acid residues which are recognized by a eukaryotic cell as locations for the attachment of sugar residues. The amino acids where carbohydrate, such as oligosaccharide, is attached are typically asparagine (N-linkage) , serine (O-linkage) , and threonine (O-linkage) residues. The specific site of attachment is typically signaled by a sequence of amino acids, referred to herein as a "glycosylation site sequence". The glycosylation site sequence for N-linked glycosylation is: -Asn-X-Ser- or -Asn-X-Thr-, where X may be any of the conventional amino acids, other than proline. The predominant glycosylation site sequence for O-linked glycosylation is: -(Thr or Ser) -X-X-Pro-, where X is any conventional amino acid. The recognition sequence for glycosaminoglycans (a specific type of sulfated sugar) is -Ser-Gly-X-Gly-, where X is any conventional amino acid. The terms "N-linked" and "O- linked" refer to the chemical group that serves as the attachment site between the sugar molecule and the amino acid residue. N-linked sugars are attached through an amino group; O-linked sugars are attached through a hydroxyl group. However, not all glycosylation site sequences in a protein are necessarily glycosylated; some proteins are secreted in both glycosylated and nonglycosylated forms, while others are fully glycosylated at one glycosylation site sequence but contain another glycosylation site sequence that is not glycosylated. Therefore, not all glycosylation site sequences that are present in a polypeptide are necessarily glycosylation sites where sugar residues are actually attached. The initial N- glycosylation during biosynthesis inserts the "core carbohydrate" or "core oligosaccharide" (Proteins, Structures and Molecular Principles. (1984) Creighton (ed.) , W.H. Freeman and Company, New York, which is incorporated herein by reference) .

As used herein, "glycosylating cell" is a cell capable of glycosylating proteins, particularly eukaryotic cells capable of adding an N-linked "core oligosaccharide" containing at least one mannose residue and/or capable of adding an O-linked sugar, to at least one glycosylation site sequence in at least one polypeptide expressed in said cell, particularly a secreted protein. Thus, a glycosylating cell contains at least one enzymatic activity that catalyzes the attachment of a sugar residue to a glycosylating site sequence in a protein or polypeptide, and the cell actually glycosylates at least one expressed polypeptide. For example but not for limitation, mammalian cells are typically glycosylating cells. Other eukaryotic cells, such as insect cells and yeast, may be glycosylating cells.

As used herein, the term "antibody" refers to a protein consisting of one or more polypeptides substantially encoded by genes of the immunoglobulin superfamily (e.g., see The Immunoglobulin Gene Superfamily, A.F. Williams and A.N. Barclay, in Immunoglobulin Genes. T. Honjo, F.W. Alt, and T.H. Rabbitts, eds., (1989) Academic Press: San Diego, CA, pp.361- 387, which is incorporated herein by reference) . For example, but not for limitation, an antibody may comprise part or all of a heavy chain and part or all of a light chain, or may comprise only part or all of a heavy chain. The recognized immunoglobulin genes include the kappa, lambda, alpha, gamma (IgG-_L, IgG₂, IgG₃, IgG₄) , delta, epsilon and mu constant region genes, as well as the myriad immunoglobulin variable region genes. Full-length immunoglobulin "light chains" (about 25 Kd or 214 amino acids) are encoded by a variable region gene at the NH2-terminus (about 110 amino acids) and a kappa or lambda constant region gene at the COOH - terminus. Full-length immunoglobulin "heavy chains" (about 50 Kd or 446 amino acids) , are similarly encoded by a variable region gene (about 116 amino acids) and one of the other aforementioned constant region genes, e.g. , gamma (encoding about 330 amino acids) . Antibodies include, but are not limited to, the following: immunoglobulin fragments (e.g., Fab, F(ab)₂), Fv) , single chain immunoglobulins, chimeric immunoglobulins, humanized antibodies, primatized antibodies, and various light chain- heavy chain combinations) . Antibodies can be produced in glycosylating cells (e.g., human lymphocytes, hybridoma cells, yeast, etc.), in non-glycosylating cells (e.g., in E. coli ) , or synthesized by chemical methods or produced by .in vitro translation systems using a polynucleotide template to direct translation.

As used herein, a "lipidized antibody" is an antibody which has been modified by lipid derivatization (e.g., by covalent attachment of a lipoamine, such as glycyldioctadecylamide, dilauroylphosphatidylethanolamine, or dioctadecylamidoglycylspermidine) of one or more carbohydrate moieties attached to an immunoglobulin at a glycosylation site. Generally, the lipid substituent, such as a lipoamine, is covalently attached through a naturally-occurring carbohydrate moiety at a naturally-occurring glycosylation site. However, it is possible to produce immunoglobulins that have altered glycosylation site sequences (typically by site- directed mutagenesis of polynucleotides encoding immunoglobulin chains) and/or altered glycosylation patterns (e.g., by expression of immunoglobulin-encoding polynucleotides in glycosylating cells other than lymphocytes or in lymphocytes of other species) . Lipid substituents can be attached to one or more naturally-occurring or non- naturally-occurring carbohydrate moiety on an immunoglobulin chain. When an antibody is produced by direct polypeptide synthesis or by biosynthesis in a non-glycosylating cell (e.g., a phage display library), it will generally be necessary to attach a carbohydrate substituent by chemical or enzymatic modification for subsequent lipidization (alternatively, the carbohydrate may be lipidized prior to attachment to the immunoglobulin) . The term "lipidated antibody" may be substituted to "lipidized antibody". As used herein, a "lipidized protein" refers to a protein (including multimeric proteins, glycoproteins, and polypeptides of various sizes) that has been modified by attachment of lipid (e.g., lipoamine), generally through a carbohydrate moiety. A lipidized protein is generated by derivatizing a protein such that the resultant lipidized protein is distinct from naturally-occurring lipid-linked proteins and lipoproteins. For proteins that are biologically active (e.g., enzymes, receptors, transcription factors), lipidization should not substantially destroy the biological activity (e.g., at least about 15 percent of a native biological activity should be preserved in the lipidized protein) . Lipidized peptidomimetics should retain at least about 25 to 95 percent of the pharmacologic activity of a corresponding non-lipidized peptidomimetic.

"Alkyl" refers to a fully saturated aliphatic group which may be cyclic, branched or straightchain. Alkyl groups include those exemplified by methyl, ethyl, cyclopropyl, cyclopropylmethyl, sec-butyl, heptyl, and dodecyl. All of the above can either be unsubstituted or substituted with one or more non-interfering substitutents, e.g., halogen; C_J-C_^J alkoxy; C₁-C₄ acyloxy; formyl; alkylenedioxy; benzyloxy; phenyl or benzyl, each optionally substituted with from 1 to 3 substituents selected from halogen, ^-^ alkoxy or C_}-^ acyloxy. The term "non-interfering" characterizes the substituents as not adversely affecting any reactions to be performed in accordance with the process of this invention. If more than one alkyl group is present in a given molecule, each may be independently selected from "alkyl" unless otherwise stated.

"Alkylene" refers to a fully saturated divalent radical containing only carbon and hydrogen, and which may be a branched or straight chain radical. This term is further exemplified by radicals such as methylene, ethylene, n-propylene, t-butylene, i-pentylene, n-heptylene, and the like. All of the above can either be unsubstituted or substituted with one or more non-interfering sustituents, e, ., halogen; C_λ~C₄ alkoxy; ^-^ acyloxy; formyl; ε^" ylenedioxy; benzyloxy; phenyl or benzyl, each optionally substituted with from 1 to 3 substituents selected from halogen, C_j-^ alkoxy or C^C acyloxy. The term "non- interfering" characterizes the substituents as not adversely affecting any reactions to be performed in accordance with the process of this invention. If more than one alkylene group is present in a given molecule, each may be independently selected from "alkylene" unless otherwise stated. "Aryl", denoted by Ar, includes monocyclic or condensed carbocyclic aromatic groups having from 6 to 20 carbon atoms. Aryl groups include those exemplified by phenyl and naphthyl. These groups may be substituted with one or more non-interfering substituents, e.g. , those selected from lower alkyl; lower alkenyl; lower alkynyl; lower alkoxy; lower alkylthio; lower alkylsulfinyl; lower alkylsulfonyl, dialkylamine; halogen; hydroxy; phenyl; phenyloxy; benzyl; benzoyl; and nitro. Each substituent may be optionally substituted with additional non-interfering substituents. "Amino" refers to the group -NH₂.

"Alkylcarbonyl" refers to the group -(CHR₁)-C0- wherein R_λ is further designated the α-position. R₂ may be hydrogen, alkyl, or an amino group. Preferably R is an amino group.

Description of the Preferred Embodiments

In accordance with the present invention, novel methods for chemically modifying proteins, such as antibodies, to facilitate passage across capillary barriers and into cells are provided. In general, the methods include the covalent attachment of at least one non-interfering lipid substituent (e.g., glycyldioctadecylamide, glycyldiheptadecylamide, glycyldihexadecylamide, dilauroylphosphatidylethanolamine, and glycyldioctadecadienoylamide) to a reactive site in the protein molecule (e.g., a periodate-oxidized carbohydrate moiety) . Various non-interfering lipid substituents may be attached to proteins to produce lipidized proteins, such as lipidized antibodies of the invention. For example but not for limitation, the following examples of lipids may be conjugated to a protein of interest to yield a lipidized protein: lipoamines, lipopolyamines, and fatty acids (e.g., stearic acid, oleic acid, and others) . Generally, the lipid will be attached by a covalent linkage to a carbohydrate linked to the protein (e.g., a carbohydrate side chain of a glycoprotein) . Naturally-occurring carbohydrate side-chains are preferably used for linkage to a lipoamine, although novel glycosylation sites may be engineered into a polypeptide by genetic manipulation of an encoding polynucleotide, and expression of the encoding polynucleotide in a glycosylating cell to produce a glycosylated polypeptide.

Glycosylated proteins can be lipidized to enhance transvascular transport, organ uptake, and intracellular localization of the lipidized protein, including intranuclear localization. Generally, a glycosylated polypeptide, such as an antibody, is chemically oxidized with an oxidizing agent (e.g., periodate) to yield pendant carboxyl and/or aldehyde groups, and reacted with a lipoamine to form a covalent (amide or imide, respectively) bond linking the lipoamine to the protein. Typically, the oxidation of the carbohydrate side- chain is a partial oxidation producing at least one reactive carboxyl or aldehyde group, although generally chemical oxidation methods will produce some molecules that are partially oxidized and others that are either unoxidized or completely oxidized. However, in order to be lipidized by reaction with a lipoamine, the glycoprotein must be oxidized to produce at least one pendant aldehyde group that can react with a lipoamine, although it may be possible to produce lipidized proteins through linkage to pendant carboxyl groups as well. A pendant carboxyl or aldehyde group of an oxidized glycoprotein is a carboxyl or aldehyde group having a carbonyl carbon derived from an oxidized oligosaccharide and which is covalently attached to the protein, either directly or through a spacer (e.g., an unoxidized portion of a N- or O-linked carbohydrate side-chain) . Preferably, N-linked and O-linked carbohydrate chains are incompletely oxidized to generate a multiplicity of reactive aldehyde and carboxyl groups at each glycosylation position for subsequent reaction with lipoamines. Most usually, glycoproteins having one or more complex N-linked oligosaccharides, such as those having a branched (mannose)₃(/3-N-acetylglucosamino) core, are partially oxidized by limited reaction with a suitable oxidant, generally periodate. Linked oligosaccharides containing N- acetylglucosamine (NAG) , mannose, galactose, fucose (6- deoxygalactose) , N-acetylneuraminic acid (sialic acid) , glucose, N-acetylmuramic acid, N-acetylgalactosamine, xylose, or combinations of these monosaccharide units can be oxidized and reacted with lipoamines to produce lipidized proteins, more specifically carbohydrate-linked lipidized proteins. Glycoproteins containing linked oligosaccharides with monosaccharide units other than those specifically listed above for exemplification, including non-naturally occurring monosaccharides, can also be oxidized and covalently linked to a lipoamine to form a lipidized protein. Lipoamines are molecules having at least one acyl group and at least one free amine (i.e., a primary or second¬ ary amine) . It is believed that the invention can also be practiced with lipoamines that have tertiary amines which comprise at least one substituent that can be displaced by reaction with an oxidized carbohydrate. Examples of lipoamines having a primary amine are shown in Fig. 1. For example, the invention can produce lipidized proteins by reacting a glycoprotein with a straight-chain lipoamine of the formula:

NH₂-R-(CH₂)_n-CH₃

where R is: a disubstituted alkyl (alkylene) , preferably ethylene (-CH₂-) ; a 1,4-disubstituted cyclohexyl; a disubstituted aryl (arylene); preferably a 1,4-disubstituted phenyl (phenylene) ; an a ido group of the formula -(CHR-^-CO- NH- wherein R is hydrogen or an amino group; alkylcarbonyl, preferably α-amino substituted alkylcarbonyl; or a phosphate diester, preferably of the formula -CH₂-0-P0₂-0-. n is an integer which is typically 1 to 50, preferably about 5 to 30, more preferably about 10 to 25, and most usually about 15 to 20. In general, n is selected at the discretion of the practitioner according to the following guideline: when the molecule to be lipidized is large (i.e., a protein of more than about 10 kD) it is preferred than n is at least about 8 to 12 or more to increase the hydrophobicity of the resulting lipidized protein; when the molecule to be lipidized is small (e.g., an oligopeptide) n can typically be in the range 2 to 18, but may be larger if additional hydrophobicity of the lipidized molecule is desired.

The invention can also be practiced with branched- chain lipoamines, which, for example, can include lipoamines of the formula:

NH₂-R'-(CH₂)_n-CH₃

(CH₂)_m

CH,

where R' is: a trisubstituted alkyl, preferably - CH₂-CH< or 1,2,4-trisubstituted cyclohexyl; a trisubstituted aryl, preferably 1,2,4-trisubstituted phenyl; an amido group of the formula -(CHR-L) -CO-N< wherein R_χ is hydrogen or an amino group; an imino group of the formula -CHR₂-NH-CH< wherein R₂ is hydrogen or an amino group or an imino group of the formula - CH₂-N<; or a phosphate diester, preferably of the formula -CH₂- CH₂-0-P0₂-0-CH₂-CH(C0₂-)₂. m and n are selected independently and are integers which are typically 1 to 50, preferably about 5 to 30, more preferably about 10 to 25, and most usually about 15 to 20. In general, n is selected at the discretion of the practitioner according to the following guideline: when the molecule to be lipidized is large (i.e., a protein of more than about 10 kD) it is preferred than m and/or n is at least about 8 to 12 or more to increase the hydrophobicity of the resulting lipidized protein; when the molecule to be lipidized is small (e.g., an oligopeptide) n can typically be in the range 2 to 18, but may be larger if additional hydrophobicity of the lipidized molecule is desired.

Essentially any glycoprotein can be lipidized according to the methods of the invention by reacting a lipoamine with an oxidized carbohydrate side-chain. Fig. 2 shows schematically a glycosylated antibody and a carbohydrate-linked lipidized antibody of the invention, respectively. Non-glycosylated proteins may be conjugated to a lipid by linkage through a suitable crosslinking agent (e.g., by carbodiimide linkage chemistry).

In accordance with the present invention, novel lipidized antibodies capable of specifically binding to predetermined intracellular epitopes with strong affinity are provided. These antibodies readily enter the intracellular compartment and have binding affinities of at least about 1 x 10⁶ M^"1, preferably 1 x 10⁷ M^"1 to 1 x 10⁸ M^"1, more preferably at least about 1 x 10⁹ M^"1 or stronger. The lipidized antibodies typically have a lipid substituent attached to a naturally-occurring carbohydrate side chain on a donor immunoglobulin chain, which composes an antibody specifically reactive with an intracellular, transmembrane, or extracellular epitope. Since carbohydrates are located on the Fc portion of immunoglobulins, chemical modification of the carbohydrate residues by lipidization would be unlikely to produce a substantial loss of affinity of the antibodies for their antigens (Rodwell et al. (1986) Proc. Natl. Acad. Sci. (U.S.A. ) 83: 2632) . The lipidized antibodies generally retain substantial affinity for their antigen, and the avidity can be readily measured by any of several antibody-antigen binding assays known in the art. The antibodies can be produced economically in large quantities and find use, for example, in the treatment of various human disorders by a variety of techniques.

One form of immunoglobulin constitutes the basic structural unit of an antibody. This form is a tetramer and consists of two identical pairs of immunoglobulin chains, each pair having one light and one heavy chain. In each pair, the light and heavy chain variable regions are together responsible for binding to an antigen, and the constant regions are responsible for the antibody effector functions. In addition to antibodies, immunoglobulins may exist in a variety of other forms including, for example, Fv, Fab, and (Fab¹) _{2 ι} as well as bifunctional hybrid antibodies, fusion proteins (e.g., bacteriophage display libraries), and other forms (e.g.. Lanzavecchia et al., Eur. J. Immunol. 17. 105 (1987)) and in single chains (e.g. , Huston et al. , Proc. Natl. Acad. Sci. U.S.A.. 85. 5879-5883 (1988) and Bird et al..

Science. 242. 423-426 (1988)). (See, generally, Hood et al. _f "Immunology", Benjamin, N.Y., 2nd ed. (1984), and Hunkapiller and Hood, Nature. 323. 15-16 (1986)).

Antibodies can be produced in glycosylating cells (e.g., human lymphocytes, hybridoma cells, yeast, etc.), in non-glycosylating cells (e.g., in E. coli ) , or synthesized by chemical methods or produced by in vitro translation systems using a polynucleotide template to direct translation. One source of hybridoma cell lines and immunoglobulin-encoding polynucleotides is American Type Culture Collection, Rockville, MD. Methods for expression of heterologous proteins in recombinant hosts, chemical synthesis of polypeptides, and in vitro translation are well known in the art and are described further in Maniatis et al., Molecular Cloning: A Laboratory Manual (1989), 2nd Ed., Cold Spring

Harbor, N.Y.; Berger and Kimmel, Methods in Enzymology. Volume 152. Guide to Molecular Cloning Techniques (1987) , Academic Press, Inc. , San Diego, CA; Merrifield, J. (1969) J. A . Chem. Soc. 91: 501; Chaiken I.M. (1981) CRC Crit. Rev. Biochem. 11: 255; Kaiser et al. (1989) Science 243: 187; Merrifield, B. (1986) Science 232: 342; Kent, S.B.H. (1988) Ann. Rev. Biochem. 57: 957; and Offord, R.E. (1980) Semisynthetic Proteins. Wiley Publishing, which are incorporated herein by reference) . Antibodies that are produced in non-glycosylating cells can be conjugated to a lipid by use of a bifunctional crosslinking agent or preferably post-translationally glycosylated in a glycosylation system such as purified canine pancreatic microsomes (Mueckler and Lodish (1986) Cell 44 : 629 and Walter, P. (1983) Meth. Enzy ol. 96: 84, which are incorporated herein by reference) . Alternatively, polynucleotides that encode antibodies may be isolated from screened prokaryotic expression libraries, such as combinatorial antibody fragment display libraries, and subsequently expressed in glycosylating cells to produce glycosylated antibodies. According to these methods, glycosylated antibodies may be obtained, having naturally- occurring and/or non-naturally-occurring glycosylation patterns. Such glycosylated antibodies can be lipidized according to the methods of the invention.

Glycosylation of immunoglobulins has been shown to have significant effects on their effector functions, structural stability, and rate of secretion from antibody- producing cells (Leatherbarrow et al., Mol. Immunol. 22: 407 (1985)). The carbohydrate groups responsible for these properties are generally attached to the constant (C) regions of the antibodies. For example, glycosylation of IgG at asparagine 297 in the C_H2 domain is required for full capacity of IgG to activate the classical pathway of complement- dependent cytolysis (Tao and Morrison, J. Immunol. 143 : 2595 (1989)). Glycosylation of IgM at asparagine 402 in the C_H3 domain is necessary for proper assembly and cytolytic activity of the antibody (Muraoka and Shulman, J. Immunol. 142: 695

(1989)) . Removal of glycosylation sites as positions 162 and 419 in the C_H1 and C_H3 domains of an IgA antibody lead to intracellular degradation and at least 90% inhibition of secretion (Taylor and Wall, Mol. Cell. Biol. 8_.: 4197 (1988)). Glycosylation of immunoglobulins in the variable (V) region has also been observed. Sox and Hood, Proc. Natl. Acad. Sci. USA 66: 975 (1970) , reported that about 20% of human antibodies are glycosylated in the V region. Glycosylation of the V domain is believed to arise from fortuitous occurrences of the N-linked glycosylation signal Asn-Xaa-Ser/Thr in the V region sequence and has not been recognized in the art as playing an important role in immunoglobulin function.

Therefore, it is generally preferred that lipidization is performed on antibodies having naturally- occurring glycosylation patterns. If glycosylation sites are engineered into an antibody, it is preferred that novel glycosylation site be introduced in a constant region or variable region framework region, which are less likely to adversely affect the antigen binding activity of the antibody. It is generally most preferred that novel glycosylation sites which are engineered into an antibody are placed in a constant region.

Alternatively, polypeptide fragments comprising only a portion of a primary antibody structure and having a carbohydrate side chain that may be derivatized with a lipid substituent (e.g., lipoamine) can be produced, which fragments possess one or more immunoglobulin activities (e.g., antigen binding activity) . These polypeptide fragments may be produced by proteolytic cleavage of intact antibodies by methods well known in the art, or by site-directed mutagenesis at the desired locations in expression vectors containing sequences encoding immunoglobulin proteins, such as after Crl to produce Fab fragments or after the hinge region to produce (Fab¹)₂ fragments. Single chain antibodies may be produced by joining V_L and V_H with a DNA linker (see, Huston et al., op. cit., and Bird et al., op. cit.). Also because like many genes, the immunoglobulin-related genes contain separate functional regions, each having one or more distinct biological activities, the genes may be fused to functional regions from other genes having novel properties. Nucleic acid sequences for producing immunoglobulins for the present invention are capable of ultimately expressing the desired antibodies and can be formed from a variety of different polynucleotides (genomic or cDNA, RNA, synthetic oligonucleotides, etc.) and components (e.g., V, J, D, and C regions), as well as by a variety of different techniques. Joining appropriate synthetic and genomic sequences is presently the most common method of production, but cDNA sequences may also be utilized (see. European Patent Publication No. 0239400 and L. Reichmann et al.. Nature, 332. 323-327 (1988)).

Immunoglobulins and/or DNA sequences encoding immunoglobulin chains may be obtained, for example, by hybridoma clones which can be produced according to methods known in the art (Kohler and Milstein (1976) Eur. J. Immunol. _5: 511, incorporated herein by reference) or can be obtained from several sources ("ATCC Catalog of Cell Lines and Hybridomas", American Type Culture Collection, Rockville, MD, which is incorporated herein by reference) . DNA sequences encoding immunoglobulin chains can be obtained by conventional cloning methods known in the art and described in various publications, for example, Maniatis et al., Molecular Cloning: A Laboratory Manual. 2nd Ed., (1989), Cold Spring Harbor, N.Y. and Berger and Kimmel, Methods in Enzymology. Volume 152, Guide to Molecular Cloning Techniques (1987) , Academic Press, Inc., San Diego, CA, which are incorporated herein by reference. As stated previously, the DNA sequences will be expressed in hosts, typically glycosylating cells, after the sequences have been operably linked to (i.e., positioned to ensure the functioning of) an expression control sequence. These expression vectors are typically replicable in the host organisms either as episomes or as an integral part of the host chromosomal DNA. Commonly, expression vectors will contain selection markers, e.g., tetracycline-resistance or G418-resistance, to permit detection of those cells transformed with the desired DNA sequences (see, e.g., U.S. Patent 4,704,362) .

E . coli is one prokaryotic host useful particularly for cloning the DNA sequences of the present invention. Other microbial hosts suitable for use include bacilli, such as Bacillus subtiliε, and other Enterobacteriaceae, such as Salmonella , Serratia , and various Pseudomonas species. In these prokaryotic hosts, one can also make expression vectors, which will typically contain expression control sequences compatible with the host cell (e.g., an origin of replication) . In addition, any number of a variety of well- known promoters will be present, such as the lactose promoter system, a tryptophan (trp) promoter system, a β-galactosidase promoter system, or a promoter system from phage lambda. The promoters will typically control expression, optionally with an operator sequence, and have ribosome binding site sequences and the like, for initiating and completing transcription and translation. Proteins, such as antibodies, that are expressed in non-glycosylating cells can be post-translationally glycosylated in a glycosylation system (Mueckler and Lodish, op.cit.. which is incorporated herein by reference.

Other microbes, such as yeast, may also be used for expression. Saccharomyces is a preferred host glycosylating cell, with suitable vectors having expression control sequences, such as promoters, including 3-phosphoglycerate kinase or other glycolytic enzymes, and an origin of replication, termination sequences and the like as desired.

In addition to microorganisms, mammalian tissue cell culture may also be used to express and produce the polypeptides of the present invention (see. Winnacker, "From Genes to Clones," VCH Publishers, N.Y., N.Y. (1987)). Eukaryotic cells are actually preferred, because a number of suitable host cell lines capable of secreting intact immunoglobulins have been developed in the art, and include the CHO cell lines, various COS cell lines, HeLa cells, preferably myeloma cell lines, etc, and transformed B-cells or hybridomas. Expression vectors for these cells can include expression control sequences, such as an origin of replication, a promoter, an enhancer (Queen et al.. Immunol♦ Rev.. 89, 49-68 (1986)), and necessary processing information sites, such as ribosome binding sites, RNA splice sites, polyadenylation sites, and transcriptional terminator sequences. Preferred expression control sequences are promoters derived from immunoglobulin genes, SV40, Adenovirus, cytomegalovirus, Bovine Papilloma Virus, and the like.

The vectors containing the DNA segments of interest (e.g., the heavy and light chain encoding sequences and expression control sequences) can be transferred into the host cell by well-known methods, which vary depending on the type of cellular host. For example, calcium chloride transfection is commonly utilized for prokaryotic cells, whereas calcium phosphate treatment or electroporation may be used for other cellular hosts. (See, generally, Maniatis et al.. Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Press, (1982) .) Once expressed, the whole antibodies, their dimers, individual light and heavy chains, or other immunoglobulin forms of the present invention, can be purified according to standard procedures of the art, including ammonium sulfate precipitation, affinity columns, column chromatography, gel electrophoresis and the like (see, generally, R. Scopes, "Protein Purification", Springer-Verlag, N.Y. (1982)). Substantially pure immunoglobulins of at least about 90 to 95% homogeneity are preferred, and 98 to 99% or more homogeneity most preferred, for pharmaceutical uses. Once purified, partially or to homogeneity as desired, the polypeptides may then be used therapeutically (including extracorporeally) or in developing and performing assay procedures, immunofluorescent stainings, and the like. (See, generally, Immunological Methods, Vols. I and II, Lefkovits and Perniε, eds., Academic Press, New York, N.Y. (1979 and 1981)). In the methods of the invention, intact immunoglobulins or their binding fragments, such as Fab, are preferably used. Typically, lipidized antibodies will be of the human IgM or IgG isotypes, but other mammalian constant regions may be utilized as desired. Lipidized antibodies of the IgA, IgG, IgM, IgE, IgD classes may be produced. Preferably, the lipidized antibodies of the invention are human, urine, bovine, equine, porcine, or non-human primate antibodies, more preferably human or murine antibodies. The invention can be used to produce lipidized antibodies of various types, including but not limited to: chimeric antibodies, humanized antibodies, pri atized antibodies, F_v fragments, toxin-antibody conjugates, isotope-antibody conjugates, and imaging agent-antibody conjugates. For in vivo imaging, lipidized antibodies are suitably labeled with a diagnostic label, administered to the patient, and their location determined at various times following administration. Various methods of labeling antibodies with diagnostic reporters (e.g., with Tc", other radioligands, radiocontrast agents or radio-opaque dye) are known in the art.

Proteins and oligopeptides (i.e., polypeptides comprising from 2 to about 50 amino acid residues attached in peptidyl linkage) other than immunoglobulins can be lipidized according to the methods the invention. Naturally-occurring glycoproteins (e.g., 7-glutamyltranspeptidase, thrombomodulin, glucose transporter proteins) are preferred substrates for lipidization through carbohydrate linkage, although substantially any polypeptide can be lipidized by covalent attachment through a crosslinking agent (e.g., N- hydroxysuccimide) to a suitable amino acid side chain. In alternate embodiments of the invention, at least one lipid substituent (e.g., lipoamine) iε covalently attached to a non- carbohydrate moiety on a protein or polypeptide (e.g., by formation of an amide linkage with a Asp or Glu residue side- chain carboxyl substituent or a thioester linkage with a Cys residue) . Also, a fatty acid can be linked to an Arg or Lys residue by the side-chain amine substituents. Examples of non-glycosylated proteinε which may be lipidized for enhancing transvascular and intracellular transport include, but are not limited to, the following proteinε: c-fos, c-myc, c-src, NF- AT, and HMG CoA reductase. Naturally-occurring lipoproteinε, such as native proteins which undergo physiological farnesylation, geranylgeranylation, and pal itylation are natural products and are not defined herein as "lipidized proteins".

The lipidized antibodies and pharmaceutical compositions thereof are particularly useful for parenteral administration, i.e., subcutaneously, intramuscularly or in¬ travenously. The compositions for parenteral administration will commonly comprise a solution of the immunoglobulin or a cocktail thereof dissolved in an acceptable carrier, preferably an aqueous carrier. A variety of aqueous carriers can be used, e.g., water, buffered water, 0.4% saline, 0.3% glycine and the like. These solutions are sterile and generally free of particulate matter. These compositionε may be sterilized by conventional, well known sterilization techniques. The compositions may contain pharmaceutically acceptable auxiliary substances as required to approximate physiological conditions such as pH adjuεting and buffering agents, toxicity adjusting agents and the like, for example sodium acetate, sodium chloride, potasεium chloride, calcium chloride, εodium lactate, human albumin, etc. The concentration of antibody in these formulations can vary widely, i.e., from less than about 0.5%, usually at or at least about 1% to as much as 15 or 20% by weight and will be selected primarily based on fluid volumes, viscoεitieε, etc. , in accordance with the particular mode of administration selected. Thus, a typical pharmaceutical composition for injection could be made up to contain 1 ml εterile buffered water, and 1-10 mgε of lipidized immunoglobulin. A typical composition for intravenous infusion could be made up to contain 250 ml of sterile Ringer'ε εolution, and 150 mg of antibody. Actual methods for preparing parenterally administrable compositions will be known or apparent to those skilled in the art and are described in more detail in, for example, Remington's Pharmaceutical Science. 15th ed. , Mack Publishing Company, Easton, Pennsylvania (1980) , which is incorporated herein by reference.

The lipidized proteins and antibodies of this invention can be frozen or lyophilized for storage and reconstituted in a suitable carrier prior to use. This technique has been shown to be effective with conventional immune globulins and art-known lyophilization and reconsti- tution techniques can be employed. It will be appreciated by those skilled in the art that lyophilization and recon- stitution can lead to varying degrees of activity loss (e.g., with conventional immune globulins, IgM antibodies tend to have greater activity loss than IgG antibodies) and that use levels may have to be adjusted to compensate. The compositionε containing the present lipidized proteins (e.g., antibodies) or a cocktail thereof can be administered for prophylactic and/or therapeutic treatments. In therapeutic application, compositions are administered to a patient in an amount sufficient to cure or at leaεt partially arreεt the diεease and its complicationε. An amount adequate to accomplish this is defined as a "therapeutically effective dose." Amounts effective for this use will depend upon the severity of the infection and the general state of the patient'ε own immune system, but generally range from about 1 to about 200 mg of antibody per dose, with dosageε of from 5 to 25 mg being more commonly uεed. It muεt be kept in mind that the materialε of this invention may generally be employed in serious diseaεe states, that is life-threatening or potentially life-threatening situations. In prophylactic applications, compositionε containing the preεent immunoglobulins or a cocktail thereof are administered to a patient not already in a disease state to enhance the patient's resistance. Such an amount is defined to be a "prophylactically effective dose." In this use, the preciεe amountε again depend upon the patient'ε εtate of health and general level of immunity, but generally range from 0.1 to 25 mg per dose.

Single or multiple administrationε of the compoεitions can be carried out with dose levels and pattern being selected by the treating physician. In any event, the pharmaceutical formulations should provide a quantity of the lipidized proteins and/or lipidized antibody(ies) of this in¬ vention sufficient to effectively treat the patient.

For diagnostic purposeε, the lipidized antibodies may either be labeled or unlabeled. Unlabeled antibodies can be used in combination with other labeled antibodies (εecond antibodieε) that are reactive with the lipidized antibody, εuch as antibodies εpecific for human immunoglobulin constant regions. Alternatively, the lipidized antibodies can be directly labeled. A wide variety of labels may be employed, such as radionuclides, enzymes, enzyme subεtrates, enzyme cofactors, enzyme inhibitors, ligands (particularly haptens) , radiocontrast agentε, metal chelateε, etc. Numerous types of diagnostic imaging applications are available and are well known to those skilled in the art. For example, but not for limitation, an antibody that binds to a tumor antigen (e.g., anti-CEA antibodies) may be lipidized and conjugated to a radiocontrast agent or magnetic imaging material, injected into a human patient, and detected so as to localize the position of a tumor or metastatic lesion.

The lipidized immunoglobulins of the present invention can be used for diagnoεiε and therapy. By way of illuεtration and not limitation, they can be uεed to treat cancer, autoimmune diseaεeε, or viral infectionε. For treatment of cancer, the antibodieε will typically bind to an antigen expreεsed preferentially in certain cancer cells, εuch aε c-myc gene product and otherε well known to thoεe skilled in the art. Preferably, the lipidized antibody will bind to a mutant protein, such aε a c-raε oncogene product having a pathogenic (e.g., neoplaεtic) εequence, εuch aε a εubstitution at position 12, 13, 59, or 61 of the protein (e.g., a Ser at position 12 of p21^ras) . For treatment of autoimmune disease, the antibodies will typically bind to an critical regulatory protein expresεed primarily in activated T-cells, such as NF- AT, and many other intracellular proteins well known to those skilled in the art (e.g., see Fundamental Immunology. 2nd ed. , W.E. Paul, ed. , Raven Press: New York, NY, which is incorporated herein by reference) . For treatment of viral infectionε, the antibodieε will typically bind to a protein expressed in cells infected by a particular virus such as the various viral encoded polymerases and HIV-1 Tat, and many other viral proteins well known to those skilled in the art (e.g., see Virology. 2nd ed., B.N. Fields et al., eds., (1990), Raven Presε: New York, NY, which iε incorporated herein by reference) .

Diagnostic uses of the lipidized antibodies include intracellular staining using labeled, lipidized antibodieε which bind to intracellular antigenε (e.g., ras) either in fixed histological specimens or in intact, viable cells (e.g., cell cultureε or cellular explantε) .

The invention alεo provideε kits comprising a vial containing a labeled lipidized antibody; such kits can be εold for practicing diagnoεtic aεεays on intracellular antigens (as well as being capable of uses of unlipidized antibodies; staining extracellular antigenε) according to the methodε deεcribed. Vialε of purified lipidized antibodieε can be εold to the clinical lab or scientist as commercial reagents, juεt like conventional diagnoεtic productε or laboratory biologicalε (e.g. , reεtriction enzymeε, Taq polymeraεe, monoclonal antibodieε) , categorieε which have sufficient utility to have merited the granting of numerous U.S. patents. Such reagents which can be uεed in reεearch and diagnoεtics have been patented; for example, the restriction endonucleaseε Aεcl, Fεel, Pmel, Xcyl, Spll, Srfl, and Apol, among otherε, have been patented in the U.S. (see, U.S. Patent Nos. 5,061,628, 5,196,330, 4,588,689, 4,886,756, 5,300,432, 5,200,336, and 5,200,337), among others. Kitε can alεo be εupplied for use with the subject lipidized antibodieε in the protection againεt or detection of a cellular activity or for the presence of a selected cell intracellular protein or the diagnoεiε of diεeaεe. Thuε, the εubject compoεition of the preεent invention may be provided, uεually in a lyophilized form in a container, either alone or in conjunction with additional antibodieε specific for the desired cell type. The lipidized antibodies, which may be conjugated to a label or toxin, or unconjugated, are included in the kits with bufferε, εuch aε Tris, phosphate, carbonate, etc., stabilizerε, biocideε, inert proteins, e.g., serum albumin, or the like, and a set of instructionε for use. Generally, these materials will be present in lesε than about 5% wt. based on the amount of active antibody, and usually present in total amount of at least about 0.001% wt. based again on the antibody concentration. Frequently, it will be desirable to include an inert extender or excipient to dilute the active ingredients, where the excipient may be present in from about 1 to 99% wt. of the total composition. Where a second antibody capable of binding to the lipidized antibody is employed in an assay, this will usually be present in a separate vial. The second antibody is typically conjugated to a label and formulated in an analogous manner with the antibody formulations described above, as well as typically also being lipidized itself.

The lipidized antibodies of the present invention are also suited for use in improved diagnostic methods and protein purification methods. For example, many intracellular proteinε are unstable (e.g. , short half-life, susceptible to proteolysiε) or prone to aggregation (e.g., j3-amyloid protein) making purification and/or diagnostic detection difficult. Lipidized antibodies are able to penetrate living cells and bind to specific intracellular target antigens; such antibody- antigen binding may stabilize the target antigen and block enzymes involved in degradation of the target antigen (e.g., proteaεeε, ubiquitin-conjugating enzymeε, glycoεidases) facilitating detection and/or purification of the target antigen.

In one variation of the invention, a lipidized antibody which specifically binds to an intracellular target antigen is contacted with live cellε compriεing the intracellular target antigen under physiological conditions (e.g., cell culture conditions, εomatic conditions) and incubated for a suitable binding period (e.g., from about 10 minutes to several hours) . The lipidized antibody specifically binds to the target antigen forming an antigen- antibody complex which is leεε εuεceptible to degradation and/or aggregation that is the target antigen itself.

Typically, the cells are then fixed and permeabilized and the antigen-antibody complex, comprising the target antigen bound to the lipidized antibody, iε detected, usually with a labeled secondary antibody that specifically binds the the lipidized antibody. Examples of preferred labels attached to the secondary antibody are: FITC, rhodamine, horseradish peroxidase conjugates, alkaline phosphataεe conjugateε, β- galctoεidaεe conjugateε, biotinyl moieties, radioisotopeε, and the like. In some embodiments, the εecondary antibody may be lipidized and the fixation and/or permeabilization stepε may be omitted and replaced with εubεtantial waεhing of the cell εample to remove non-εpecific εtaining. It may alεo be poεεible to uεe a lipidized, labeled primary antibody directly and omit the εecond antibody. Labelled protein A may alεo be substituted for a secondary antibody for the detection of the primary (lipidized) antibody. Lipidized antibodieε may alεo be uεed for intracellular therapy, εuch as for binding to a predetermined intracellular target antigen and modifying a biochemical property of the target antigen. For example, multi-subunit proteins, such as heteromultimeric proteins (e.g., transcription factors, G-proteins) or homodimeric proteins

(e.g., polymerized tubulin) may poεεess a biochemical activity (e.g., GTPase activity) or other activity that requires intermolecular interaction(s) that may be blocked by a lipidized antibody that specifically binds to one or more subunitε and prevents functional interaction of the subunits. For example, a lipidized anti-Fos antibody which binds to a portion of Fos (e.g., leucine zipper) required for binding to Jun to form a transcriptionally active AP-1 transcription factor (Fos/Jun heterodimer) may block formation of functional AP-1 and inhibit AP-1-mediated gene transcription. Also for example, a lipidized anti-ras antibody may bind to an epitope of ras which is required for its proper signal transduction function (e.g., a GTP/GDP-binding site, a portion of ras that binds an acceεsory protein such aε GAP, or the like) , therebymodifying the activity of intracellular ras in living cells.

The following exampleε are offered by way of illustration, not by way of limitation.

EXPERIMENTAL EXAMPLES

EXAMPLE 1 Preparation of a Lipidized Bovine IgG

Glycyldioctadecylamide was obtained by linking a glycine residue to dioctadecylamine according to the method described by Behr et al. (1989) Proc. Natl. Acad. Sci. (U.S.A. ) 86: 6982, which is incorporated herein by reference. Benzyloxycarbonyl-glycyl-p-nitrophenol at 1 equivalent and triethylamine at 1.1 equivalents in CH₂C1₂ are reacted for 5 hours, followed by addition of H₂, 10% Pd/C in CH₂Cl₂/Et0H and reaction for 1 hour.

Two mg of bovine IgG (Sigma) were disεolved in 400 μl of 300 mM NaHC0₃ in a 1.5 ml Eppendorf vial. Fifty μl of a freshly prepared NaI0₄ solution (42 mg/ml in H₂0) were added and the vial was wrapped in aluminum foil and gently shaken for 90 min. at room temperature. The reaction medium was then loaded on a PD-10 column (Pharmacia) previously equilibrated with 10 mM Na₂C0₃ (fraction 1) , and the column was eluted with 500 μl fractions. Fraction number 7 (between 3 ml and 3.5 ml) contained approximately 1.6 mg of bovine IgG as meaεured using the Bradford protein assay.

A solution of glycyldioctadecylamide in DMSO was prepared (5 mg of the lipid into 1 ml of DMSO, vigorously vortexed for εeveral minuteε) . Under thoεe conditionε the lipid waε not fully diεεolved. Fifty μl of this solution were taken carefully (and did not contain any undisεolved lipid) and were added to 350 μl of fraction 7 obtained aε described above, in an Eppendorf vial. The vial was wrapped in aluminum foil, and the mixture was gently εhaken for 20 h at room temperature.

One hundred μl of a εolution of NaBH₄ (10 mg/ml in H₂0) were then added. After one hour, 40 μl of a solution of ethanolamine (15 μl in 1 ml H20) were added. After an additional 1 h, the reaction mixture was loaded on a PD-10 column previously equilibrated in 100 mM HEPES buffer, pH 8.5. The fraction containing the lipidized IgG (between 3 and 3.5 ml) was collected and stored on ice.

Labelling with ¹ C-acetic anhydride ¹ C-acetic anhydride (500 μCi, Amersham) in benzene

(10x10° cpm/μl) waε used. Two 5 μl aliquots were added to the 500 μl fraction containing the lipidized IgG at 10 min interval in an Eppendorf vial. The reaction was left on ice. A 500 μl solution of native IgG (800 μg in 100 mM HEPES, pH 8.5) was treated the same way.

After 30 min. the vials were warmed to 20 to 25°C, the ¹⁴C-labelled IgGε were εeparated from free ¹⁴C-acetate on a PD-10 column equilibrated with PBS. Radioactivity incorporated waε of approximately 10x10° cpm for 500 μg.

Organ uptake studies Male swisε albinoε mice (20g) were uεed. One hundred μl of ¹⁴C-labeled lipidized IgG or ¹⁴C-labeled control IgG in PBS (approximately 400,000 dpm each) were ad iniεtered intravenously by tail vein injection. Mice were killed after 30 min or 3 h, their blood collected in EDTA-containing tubes, and their brain (minus cerebellum and brainstem) , spleen, one kidney and one liver lobe were dissected. Organs were homogenized in 1 ml 10 mM Tris buffer, pH 7.4, and 500 μl aliquots were counted in a Beckman scintillation counter.

Protein concentration in theεe homogenateε was determined by the Bradford assay (Coomassie blue) . The blood was centrifuged and 20 μl fractions of the plasma were counted. Table I shows the uptake of ¹⁴C in the brain, liver, spleen and kidney, expreεεed as the ration of radioactivity in 1 μg protein of the organs divided by the radioactivity in 1 μl plasma (data expressed as μl/μg protein) .

TABLE I Organ 30 minutes 3 hourε

Control Lipidized Control Lipidized Brain .20 + .03 .32 + .08 .94 + .05 1.75 ± .09

Kidney 1.46 + .31 9.50 + 2.19 1.17 + .04 3.54 ± .22

Liver 1.26 + .26 4.95 + .82 .66 + .04 7.59 ± .27

Spleen 1.08 + .27 4.54 ± .76 1.13 + .03 12.0 ± 2.15

Control groups of 4 and 6 mice at 30 min and 3 hr, respectively. Groupε receiving the lipidized IgG were of 6 mice at both time pointε. Data are meanε ± ε.e.m.

EXAMPLE 2

Inhibition of HIV-1 Cytotoxicity with Lipidized Anti-Tat Antibody

A monoclonal antibody which εpecifically bindε the Tat protein of HIV-1 waε lipidized according to the method described in Example 1, supra. involving periodate oxidation of carbohydrate on the antibody, followed by covalent attachment of glycyldioctadecylamide to yield a carbohydrate- lipidized antibody, which was eluted from the final PD-10 column with PBS. Sup Tl cells were maintained in 24-well plates

(100,000 cells per ml, in 2 ml of modified RPMI 1640 culture medium) . Cells were kept in culture with either: (1) no additional treatment (two controls) , (2) in the presence of added native anti-Tat antibody (15 μg/ml) , or (3) in the presence of the lipidized anti-Tat antibody (11.7 μg/ml) during the first five days of the experiment. At the end of the first day, HIV-1 IIIB was added to one well of control cells and to the cultures containing native anti-Tat antibody- treated cellε or lipidized anti-Tat antibody-treated cells. Viable cells were counted daily. The untreated, HIV-infected cellε grew up to a density of approximately 500,000 cells per ml, and their number began to decrease after approximately eight days due to the cytotoxic effect of the virus. Uninfected cells grew up to a density of approximately 1,000,000 cells per ml. Treatment of infected cells with the native anti-Tat antibody did not protect the cells from the cytotoxic effect of the virus. In contrast the lipidized anti-Tat antibody led to an almost complete protection of the cells from the cytopathic effects of the HIV-1 viruε. Thiε protection continued for at leaεt about 5 days after the treatment with the lipidized antibody was interrupted. The results are presented in Fig. 3.

In another experiment, Sup Tl cells were maintained in culture as described in the previous example and kept in culture without any treatment and without any infection, infected with HIV-1 IIIB with no treatment, treated with the native anti-Tat antibody (1 μg/ml) and infected, or treated with the lipidized anti-Tat antibody (1 μg/ml) and infected. In the last three conditions, the virus was added at the end of the first day in culture. In the last two conditionε, the native or lipidized antibody was present from day 1 until day 7.

The data preεented in Table II εhow that, while the native anti-Tat antibody has little if any effect on viable cell number and reverse transcriptase activity, the lipidized antibody induced a significant protection of the cells in culture from the cytopathic effect of the virus and a εignificant decrease in reverse transcriptase activity. The latter suggeεtε that the lipidized antibody could inhibit intracellular HIV-1 replication.

TABLE II

Effect of a Lipidized Anti-Tat Antibody on Viable

Cell Number and Reverse Transcriptase Activity in Sup Tl Cells Infected with HIV-1

Viable Cells (xlO⁶) Days in Culture

Conditions

Untreated, Uninfected 1.15 1.28 1.52 1.63 1.69 1.76 1.72

Untreated, Infected 1.12 1.2 1.18 1.12 0.92 0.64 0.51

Treated with native 1.12 1.2 1.21 1.15 0.96 0.67 0.51 anti-Tat Ab Infected

Treated with lipidized 1.15 1.26 1.3 1.36 1.17 0.99 0.75 anti-Tat Ab Infected

Reverse Transcriptase Activity (cpm/10⁹ cells)

Days in Culture

1 2 3 5 6 7 8 Conditions

Untreated, Uninfected 2 2 1 2 2 1 1 Untreated, Infected 3 2 4 175 264 337 367

Treated with native anti-Tat Ab Infected 2 2 3 151 237 259 331

Treated with lipidized 3 2 2 57 135 179 184 anti-Tab Ab Infected

HIV-1-infected SupTl cells were treated daily with anti-Tat antibody in native or lipidized form or with rsCD4 (all proteins used at 1 μg/ml) starting from Day 1 before addition of HIV-1 virus containing supernatantε until 10 days post infection. Cell numbers and reverse transcriptaεe activity (RT) in the culture medium were determined every day εtarting from Day 2 poεt-infection. By Day 10, the native anti-Tat still had no significant effect on either cell counts or RT activity, whereas the lipidized anti-Tat antibody increased cell viability as compared to untreated, infected cells by approximately 70% and decreased RT activity by approximately the same extent. Cultures were continued for 3 days without further addition of antibodies. Effects of the lipidized anti-Tat persisted for the 3 days, indicating that the lipidized anti-Tat antibodies had accumulated in the cells in amounts high enough to provide sustained protection against viral infection/replication. The magnitude of the effects of the lipidized anti-Tat antibody on cell viability and RT activity were very similar to those observed with rsCD4 at the same dose. Increasing the concentration of the lipidized anti- Tat antibodies to 10 μg/ml did not induce further decrease in RT activity.

Example 3

Ability of Lipidized Anti-Tat to Inhibit the Transcriptional Activity of Tat on the HIV-1 LTR

A HeLa cell line stably transfected with a polynucleotide expressing CD4, the membrane receptor mediating HIV-1 infection, and alεo containing a reporter construct comprising an HIV-1 long terminal repeat (LTR) in operable linkage to and driving transcription of a linked reporter gene (chloramphenicol acetyltransferase; CAT) . These cells (HLCD4- CAT) are susceptible to HIV-1 infection which produces functional Tat protein; the binding of newly synthesized Tat to the HIV-1 LTR leads to transcription of the linked CAT gene. Thus, the magnitude of CAT expression is approximately proportional to the extent of HIV-1 infection and the activity of Tat protein in the cells. Cultured HeLa cells (3 x 10⁵ cells/ml in DMEM with

10% fetal bovine serum) were exposed to the same concentrations (1 or 10 μg/ml) of various antibodies (in native or lipidized form) or recombinant soluble CD4 (rsCD4) for 1 hour and were extensively washed prior to addition of HIV-containing cell culture supematants (100 μl) . Twenty-four hours later the cells were harvested and CAT expression was measured by the method of Ho et al. (1984) . Each experiment was run in quadruplicate and conducted four different time. Fig. 4 shows that the lipidized anti-Tat antibody significantly inhibited CAT activity (by approximately 75%) , whereas native (unlipidized) anti-Tat antibody, lipidized anti-gpl20 antibody, or rsCD4 were far less effective in inhibiting CAT activity. These data indicate that lipidized anti-Tat was able to specifically bind its intracellular target, Tat, and inhibit the target's activity as a transcriptional activator of the LTR/reporter gene construct.

Moreover, the data showing pasεage of the lipidized anti-Tat antibody into HeLa cells indicates that the transport mechanism does not likely require endosome formation, εince HeLa cellε are reported to undergo little if any phagocytoεiε.

Example 4

Preparation of Lipidized Immunoglobulins Reactive with an

Intracellular Protein Glycyldioctadecylamide is obtained by linking a glycine reεidue to dioctadecylamine according to the method described by Behr et al. (1989) Proc. Natl. Acad. Sci. (U.S.A.)

86: 6982, which is incorporated herein by reference.

Benzyloxycarbonyl-glycyl-p-nitrophenol at 1 equivalent and triethylamine at 1.1 equivalents in CH₂C1₂ are reacted for 5 hours, followed by addition of H₂, 10% Pd/C in CH₂Cl₂/EtOH and reaction for 1 hour.

Anti-Human c-Myc Ig Glycosylated murine immunoglobulins that bind specifically to human c-myc protein are prepared by separately culturing the hybridoma cell lines MYC CT9-B7.3 (ATCC CRL 1725), MYC CT 14-G4.3 (ATCC CRL 1727), and MYC 1-9E10.2 (ATCC CRL 1729) in RPMI 1640 with 10 percent fetal bovine serum under specified conditions (Evan et al. (1985) Mol. Cell. Biol. 5_.: 3610, incorporated herein by reference) and the monoclonal antibodieε εecreted are collected and purified by conventional methods known in the art.

About 2 mg of each purified monoclonal antibody are disεolved in 400 μl of 300 mM NaHC0₃ in a 1.5 ml Eppendorf vial. Fifty μl of a freεhly prepared NaI0₄ εolution (42 mg/ml in H20) is added and the vial is wrapped in aluminum foil and gently shaken for 90 min. at room temperature. The reaction medium iε then loaded on a PD-10 column (Pharmacia) previously equilibrated with 10 mM Na₂C0₃ (fraction 1) , and the column is eluted with 500 μl fractions. The fraction(ε) containing at leaεt approximately 500 μg of IgG aε measured using the Bradford protein asεay are collected.

A solution of glycyldioctadecylamide in DMSO is prepared (5 mg of the lipid into 1 ml of DMSO, vigorously vortexed for several minutes) . Under those conditions the lipid is not fully disεolved. Fifty μl of this εolution iε taken carefully and added to 350 μl of the purified IgG fractionε obtained aε deεcribed above, in an Eppendorf vial. The vial iε wrapped in aluminum foil, and the mixture iε gently shaken for 20 h at room temperature.

One hundred μl of a solution of NaBH₄ (10 mg/ml in H₂0) is then added. After one hour, 40 μl of a solution of ethanolamine (15 μl in 1 ml H20) is added. After an additional 1 h, the reaction mixture is loaded on a PD-10 column previously equilibrated in PBS. The fraction containing the lipidized murine anti-human-myc IgG (between 3 and 3.5 ml) is collected and stored on ice.

Anti-HMG CoA Reductase Ig Glycoεylated murine immunoglobulinε that bind specifically to the intracellular enzyme HMG CoA reductase are prepared by separately culturing the hybridoma cell line A9 (ATCC CRL 1811) in DMEM with 4.5 g/1 glucose, 5% horse serum and 2.5% fetal bovine serum as described (Goldstein et al. (1983) J. Biol. Chem. 258: 8450, incorporated herein by reference) and the monoclonal antibodies secreted are collected and purified by conventional methods known in the art.

About 2 mg of each purified monoclonal antibody are disεolved in 400 μl of 300 mM NaHC0₃ in a 1.5 ml Eppendorf vial. Fifty μl of a freεhly prepared NaI0₄ εolution (42 mg/ml in H20) iε added and the vial is wrapped in aluminum foil and gently shaken for 90 min. at room temperature. The reaction medium is then loaded on a PD-10 column (Pharmacia) previously equilibrated with 10 mM Na₂C0₃ (fraction 1) , and the column is eluted with 500 μl fractions. The fraction(ε) containing at 41 least approximately 500 μg of IgG as measured using the Bradford protein asεay are collected.

A εolution of glycyldioctadecylamide in DMSO iε prepared (5 mg of the lipid into 1 ml of DMSO, vigorouεly vortexed for several minuteε) . Under thoεe conditionε the lipid is not fully dissolved. Fifty μl of this εolution iε taken carefully and added to 350 μl of the purified IgG fractionε obtained aε deεcribed above, in an Eppendorf vial. The vial iε wrapped in aluminum foil, and the mixture iε gently shaken for 20 h at room temperature.

One hundred μl of a solution of NaBH₄ (10 mg/ml in H₂0) is then added. After one hour, 40 μl of a solution of ethanolamine (15 μl in 1 ml H20) is added. After an additional 1 h, the reaction mixture is loaded on a PD-10 column previously equilibrated in PBS. The fraction containing the lipidized anti-HMG CoA reductase IgG (between 3 and 3.5 ml) is collected and stored on ice.

Example 5 Preparation of Lipidized Immunoglobulins Reactive with a Transmembrane Protein

Glycyldioctadecylamide is obtained by linking a glycine residue to dioctadecylamine according to the method described by Behr et al. (1989) Proc. Natl. Acad. Sci. (U.S.A.) J6: 6982, which is incorporated herein by reference.

Benzyloxycarbonyl-glycyl-p-nitrophenol at 1 equivalent and triethylamine at l.l equivalents in CH₂C1₂ are reacted for 5 hours, followed by addition of H₂, 10% Pd/C in CH₂Cl₂/EtOH and reaction for 1 hour.

Anti-Ras Ig Glycosylated murine immunoglobulins that bind specifically to ras oncogene protein are prepared by separately culturing the hybridoma cell line 142-24E5 (ATCC HB 8679; U.S. Pats. 5,015,571 and 5,030,565, incorporated herein by reference) in DMEM with 4.5 g/1 glucose, 2mM L-glutamine, ImM sodium pyruvate, non-esεential amino acids, IxBME vitamins, 0.1 mM hypoxanthine, 0.032mM thymidine, 0.05 mg/ml gentamicin, and 10% fetal bovine serum, and the hybridoma cell MX (ATCC HB 9158) in Iscove's DMEM with 1% L-glutamine and HT and 10 percent fetal bovine serum under specified conditions (U.S. Patent 4,820, 631, incorporated herein by reference) and the monoclonal antibodies secreted from the hybridoma lines are collected and purified by conventional methods known in the art.

About 2 mg of each purified monoclonal antibody are dissolved in 400 μl of 300 mM NaHC0₃ in a 1.5 ml Eppendorf vial. Fifty μl of a freshly prepared NaI0₄ solution (42 mg/ml in H20) is added and the vial is wrapped in aluminum foil and gently shaken for 90 min. at room temperature. The reaction medium is then loaded on a PD-10 column (Pharmacia) previously equilibrated with 10 mM Na₂C0₃ (fraction 1) , and the column is eluted with 500 μl fractionε. The fraction(ε) containing at least approximately 500 μg of IgG aε meaεured uεing the Bradford protein aεεay are collected.

A solution of glycyldioctadecylamide in DMSO is prepared (5 mg of the lipid into 1 ml of DMSO, vigorously vortexed for several minutes) . Under those conditions the lipid is not fully disεolved. Fifty μl of this solution is taken carefully and added to 350 μl of the purified IgG fractions obtained as described above, in an Eppendorf vial. The vial is wrapped in aluminum foil, and the mixture is gently shaken for 20 h at room temperature.

One hundred μl of a solution of NaBH₄ (10 mg/ml in H₂0) is then added. After one hour, 40 μl of a solution of ethanolamine (15 μl in 1 ml H20) is added. After an additional 1 h, the reaction mixture iε loaded on a PD-10 column previouεly equilibrated in PBS. The fraction containing the lipidized murine anti-raε IgG (between 3 and 3.5 ml) is collected and stored on ice.

Hybridoma cell lines referred to in the above exampleε may be obtained from American Type Culture Collection, Rockville, MD (ATCC Cell Lineε and Hybridomaε (1992) 7th Ed, which iε incorporated herein by reference) .

Example 6 Lipidization of a Transmembrane Enzyme

The enzyme gamma-glutamyltranspeptidase (GGT: EC 2.3.2.2) is a widely distributed enzyme that catalyzes the degradation of glutathione and other γ-glutamyl compounds by hydrolysiε of the 7-glutamyl moiety or by itε transfer to a suitable acceptor. GGT is a heterodimeric glycoprotein, which is syntheεized as a precursor protein that is glycoεylated and cleaved into the two εubunitε of the mature enzyme. GGT iε anchored to the cell membrane through the N-terminal portion of itε heavy εubunit. The active εite of the enzyme lieε on the extracellular portion of the molecule, which iε heavily glycoεylated.

GGT iε εeparately purified from rat kidney and a cultured human hepatoma cell line according to procedures described previously in the art (Barouki et al. (1984) J. Biol. Chem. 259: 7970; Curthoys and Hughey (1979) Enzyme 24 : 383; Matsuda et al. (1983) J. Biochem. 93: 1427; Taniguchi et al. (1985) J. Natl. Cancer Inst. 75: 841; Tate and Meister (1985) Methods Enzvmol. 113 : 400; and Toya et al. (1983) Ann. N.Y. Acad. Sci. 417: 86, which are incorporated herein by reference) .

About 1 mg of each of the purified rat and human GGT preparations are disεolved in 400 μl of 300 mM NaHC0₃ in a 1.5 ml Eppendorf vial. Fifty μl of a freεhly prepared NaI0₄ εolution (42 mg/ml in H20) is added and the vial is wrapped in aluminum foil and gently shaken for 60 min. at room temperature. The reaction medium iε then loaded on a PD-10 column (Pharmacia) previouεly equilibrated with 10 mM Na₂C0₃ (fraction 1) , and the column iε eluted with 500 μl fractionε. The fraction(ε) containing at least approximately 100 μg of GGT as meaεured uεing the Bradford protein assay are collected.

A solution of glycyldioctadecylamide in DMSO iε prepared (5 mg of the lipid into 1 ml of DMSO, vigorouεly vortexed for εeveral minuteε) . Under thoεe conditionε the lipid iε not fully diεsolved. Fifty μl of this solution is taken carefully and added to 350 μl of the purified GGT fractions obtained as described above, in an Eppendorf vial. The vial is wrapped in aluminum foil, and the mixture is gently shaken for 20 h at room temperature.

One hundred μl of a solution of NaBH₄ (10 mg/ml in H₂0) is then added. After one hour, 40 μl of a solution of ethanolamine (15 μl in 1 ml H20) is added. After an additional 1 h, the reaction mixture is loaded on a PD-10 column previously equilibrated in PBS. The fraction containing the lipidized human and rat GGT (between 3 and 3.5 ml) is collected and stored on ice.

The lipidized human and rat GGT fractions are aεεayed for γ-glutamyltranεpeptidase activity by conventional assay procedures (Tate and Meister (1983) op.cit.. incorporated herein by reference) and the εpecific activity of the lipidized human GGT and lipidized rat GGT iε determined.

The lipidized human and rat GGT iε radiolabeled by iodination with ¹²⁵ι uεing chloramine T and approximately 50 μg of the radioiodinated lipidized GGT iε adminiεtered to ratε by intraperitoneal injection. After 24 hourε, the ratε are sacrificed and tisεue εampleε removed for autoradiography to determine the pattern of localization of the lipidized GGT in the variouε organε.

Example 7 Lipidization of an Anti-Actin Antibody and Intracellular Immunostaining In order to demonstrate that lipidized antibodies can localize intracellularly in living cells and bind intracellular targets, an anti-actin antibody was lipidized and evaluated for its ability to penetrate cultured Swiεs 3T3 fibroblasts and bind to the cytoskeletal protein actin. Native anti-actin antibody (unlipidized) was used as a control.

Protein A-purified rabbit anti-actin polyclonal antibodies were lipidized according to the following procedure. A lipoamine, glycyldioctadecylamide, was covalently linked to the carbohydrate moieties of the anti-actin antibodies by periodate oxidation-sodium borohydride reduction. Antibodies were dissolved in 0.8 ml of 300 mM NaHC0₃ at a concentration of approximately 0.2 to 1.0 mg/ml. Fifty μl of a freshly prepared aqueous solution of NaI0₄ (42 mg/ml) were added and the incubation vials were wrapped in aluminum foil and gently εhaken for 90 minuteε at room temperature. The reaction mixture was then purified on a PD-10 column (Pharmacia, Piscataway, NJ) equilibrated in and eluted with 10 mM Na₂C0₃. Fifty μl of a 10 mg/ml εolution of glycyldioctadecylamide in benzene are added to the fraction containing the antibodieε (e.g., aε determined by A₂₈₀ monitoring, Bradford aεεay) and the resulting reaction was incubated for 20 hours at room temperature with gentle shaking. One hundred μl of a freshly prepared aqueous εolution of NaBH (10 mg/ml) was then added and inubated at room temperature for one hour, followed by addition of 50 μl of ethanolamine solution (15 μl ethanolamine dissolved in 1 ml of H₂0) . After an additional hour at room temperature, the resultant lipidized antibodies were purified by chromatography on a PD-10 column equilibrated with phosphate-buffered saline. ELISA Aεεayε

Lipidized anti-actin and lipidized anti-Tat (εupra) were evaluated for their binding affinity for εpecific antigen relative to native (unlipidized) anti-actin or anti-Tat antibody by ELISA aεεay. Lipidization of either the anti-actin antibody or the anti-Tat antibody did not produce a meaεurable loεε of affinity of the antibodieε for their reεpective antigenε aε compared to their native (unlipidized) antibody. Intracellular Immunostaininq

To demonstrate that lipidized anti-actin antibodies are able to bind intracellular actin in live cells, lipidized anti-actin antibody or native anti-actin antibody were contacted with cultured Swiss 3T3 cells for 1 hour, followed by extensive washing to remove residual anti-actin antibodies. The cells were subεequently fixed and permeabilized and the anti-actin antibodieε were detected with a fluoreεcent-labeled εecondary antibody. While no εpecific εtaining could be detected in cellε preincubated with the native (unlipidized) anti-actin antibody, εpecific actin εtaining (e.g., εtained actin cableε) waε clearly evident in cells preincubated with the lipidized anti-actin antibodies. The staining pattern observed with the lipidized anti-actin antibody applied prior to fixation was similar to that seen using native (unlipidized) anti-actin antibody incubated with the cells following fixation and permeabilization. Theεe data demonstrate that the lipidized anti-actin antibodies were able to reach and bind intracellular actin, and that they could εtill be recognized and bound by the labeled secondary antibodies, indicating that they were functional and subεtantially intact.

Example 8 Lipidized anti-Tat antibody enterε living cellε and blockε HIV-1 viral replication

We have developed a chemical modification of antibodieε, lipidation, which enableε their intracellular delivery into living cellε. Intracellular localization of lipidated antibodies was demonstrated by confocal microscopy and by measuring cellular uptake of 125-I-labeled lipidated antibodies. Functionally, a lipidated monoclonal antibody directed against the Tat protein from HIV-1 inhibited viral replication as shown be increaεed viability of infected cellε and, concomitantly, decreased reverse transcriptase activity. The antibody in its native form had no εuch effect. Theεe data εhow that lipidated antibodieε can reach, εelectively bind and functionally inhibit intracellular targets. Lipidation may help to facilitate the development of intracellular immunotherapy for AIDS.

The HIV-1 regulatory protein Tat is esεential for viral replication functioning aε a tranεactivator of the viral promoter when bound to the TAR εequence of nascent transcripts. Thus, antibodies against Tat could conceivable inhibit viral replication provided they could be efficiently delivered inεide infected cells. However, the use of antibodies to target intracellular proteins has been restricted by the need for prior membrane permeabilization, direct micro-injection, or by transduction of single chain antibody expresεion vectorε. Aε a less invasive approach to antibody delivery into living cells, we have linked a lipoamine, glycyldioctadecylamide to the carbohydrate moieties of antibodies by periodate chemistry. In this study, we have applied lipidation technology to monoclonal antibodies againεt the Tat protein. The intracellular delivery of lipidated anti-Tat antibodieε (L-anti-Tat) waε firεt established by using confocal microscopy to visualize the presence of these antibodies inεide living cellε. SupTl cells, a CD4+lymphocytic cell line, were infected with HIV-1_IIIB, then exposed to fluorescently labeled L-anti-Tat (lμg/ l) . After a 1 hr incubation period excess antibody was washed out, and cells were examined by confocal fluorescence microscopy. Successive images were taken at various depths through cells, from 1 μm to 6 μm.

Fluoreεcence waε obεerved intracellularly, apparently diεtributed throughout the cell with no obviouε localization in any organelle or at the plaεma membrane. Thiε waε moεt clearly illustrated at the medial cross section of the cell.

The staining pattern with L-anti-Tat contrast sharply with that observed in HIV infected cells incubated with fluorescently labeled native anti-Tat antibody which showed a ring-shaped staining pattern suggestive of accumulation at the cell surface with no intracellular delivery.

The L-anti-Tat antibody was clearly present intracellularly but with substantially lower fluorescence intenεity than in infected cellε. A εi ilarly low level of fluoreεcence intenεity waε observed when HIV-infected cells were stained with fluorescently labeled lipidated mouεe IgG₁.

Thuε, the difference in fluoreεcence intenεity between infected and uninfected SupTl cellε with L-anti-Tat was not due to nonspecific membrane damage induced by the virus, because it was not observed with L-IgG. Instead, the enhanced intensity in infected cellε εuggeεts intracellular trapping of antigen- bound L-anti-Tat.

Intracellular delivery of lipidated antibodies was further eεtabliεhed by quantitating the uptake of ¹²⁵I-L-anti- Tat in infected and uninfected cellε (Table III) . While uptake of ¹²⁵I-n-anti-Tat waε very low (<0.5% of added antibody) and did not depend on whether cells were infected, uptake of L- anti-Tat was greatly increased, to 4.8% and 32.8% in uninfected and infected cells, respectively. Uptake of a lipidated control antibody (L-lgG) , while markedly higher than that of n- anti-Tat, was the same (approximately 4% of added antibody) independent of HIV1 infection. These experiments demonstrate efficient uptake of L-antibodies and consistent with the confocal microscopy data, suggest intracellular trapping of antigen-bound L-anti-Tat.

Table III Update of ¹²⁵I-antibodies in uninfected and infected SupTl cellε.

Cells L-anti-Tat L-anti-IgG N-anti-Tat

Infected 45, 70411749 4035±521 345±85

(32.8%) (3.7%) (0.47%)

Uninfected 3124 562 36421385 391147 (4.8%) (4.1%) (0.38%) SupTl cells (30xl0⁶ cells in 0.5 ml) were incubated with ¹²⁵I-L- anti-Tat antibody for 2 hrs at 37°C. Antibodieε were radiolabeled uεing Iodobeadε (Pierce) . One hundred fifty thouεand cpmε (representing 3.3ug, 4.lug and 2.9ug of L-anti- Tat, L-anti-IgG and n-anti-Tat respectively) were added to the cells. After incubation the cells were treated with lOmM trypsin for 20 min at 37°C to remove extracellular antibody. (Complete removal of membrane aεsociated protein by this procedure was confirmed previously by FACS analysis of cells labeled with FITC-anti-CD4 and FITC-anti-CD8 antibodies (data not shown) ) . The Table indicates the total amount of cell- aεεociated, trypsin-inεensitive radioactivity obtained under each condition. The values .presented are the means 1 sd from three experiments.

We next tested the hypothesiε that L-anti-Tat could interfere with intracellular Tat function. This was first tested using a HeLa cell line, HLCD4-CAT, which had been transfected to stably expresε CD4 antigen, the membrane receptor mediating HIV-1 infection, and the reporter gene chloramphenicol acyltranεferase (CAT) under the control of the HIV-1 promoter. Theεe cells are susceptible to HIV-1 infection which results in intracellular synthesis of Tat protein, inducing CAT expresεion. The magnitude of CAT expression is proportional to the extent of HIV-1 infection.

HLCD4-CAT cells were exposed for lhr to either L- anti-Tat, n-anti-Tat, L-anti-gpl20, or recombinant soluble CD4 (rsCD4) . The cells were washed extensively prior to addition of HIV-1 virus-containing supernatant. Twenty-four hourε later the cells were harvested and CAT expresεion measured. Neither n-anti-Tat, L-anti-gpl20, nor rsCD4 had any significant effect. In contrast, L-anti-Tat induced a reduction in CAT expression of approximately 75% (Fig. 5) . Lack of an effect by the L- anti-gpl20 suggestε that L-anti-Tat-induced inhibition waε due to a εpecific interaction of the antibody with Tat. Theεe reεultε indicate that L-anti-Tat waε able to enter into the HeLa cellε, bind Tat protein, and interfere with itε functionε. We finally inveεtigated whether L-anti-Tat could block active HIV-1 replication in living cellε. One set of experiments utilized SupTl cells infected with HIV-1_IIIB (Fig. 6A and 6B) . Figure 6A showε that treatment with L-anti-Tat (1 μg/ml) preεerved cell viability to approximately 90% that of infected cells.Concomitantly, RT activity in the L-anti-Tat treated cultures was decreased by 75% (Fig. 6B) . A higher concentration of L-anti-Tat (20μg/ml) did not induce any further decrease in RT activity (data not shown) . The magnitude o the effectε of the L-anti-Tat on cell viability and RT activity were εimilar to thoεe obεerved with rεCD4 in this syεtem. The n-anti-Tat had no εignificant effect on cell viability or RT activity. In other experiments, p24 protein, quantitated by ELISA (detection limit 50 pg/ml) was employed as an indicator of HIV-1 replication. L-Anti-Tat decreased p24 levels in the SupTl cultures, by day 9, to 29 ng/ml. compared to 163 and 147 ng/ml for untreated and n-anti-Tat treated cells, respectively. Several other anti-HIV antibodies were also lipidated and their ability to block HIV-1 replication was studied (data not shown) . These antibodies, anti-integrase, anti-Rev, an anti-pl7 (1 μg/ml) had no inhibitory effect. The lack of effect of theεe antibodieε εuggests that protection by the L-anti-Tat was due to a specific interaction with the Tat protein.

To assess the effect of L-anti-Tat on the reactivation of latent HIV-1 infection, two cell lines latently infected with HIV-1, Jl.l a lymphocytic cell line and OM-l-.l a monocytic cell line were used. Viral replication was induced with TNF-alpha and antibodies were administered simultaneously with the cytokine. After 72 hr. HIV-1 infection was quantitated by ELISA asεay for p24. Aε shown in Fig. 7, cells treated with L-anti-Tat (1 μM) had p24 levels in the 10 ng/ml range. By comparison, those treated with n-anti-Tat, or untreated controls, exhibited p24 levels in the 100 ng/ml range.

Previous εtudies in which cells genetically engineered to expreεs single-chain antibodies against gpl20 or Rev exhibited lower HIV-1 infectivity, elegantly illustrated the concept that intracellular antibodies could neutralize viral proteins. Our data show that antibody lipidation, a simpler and more widely applicable technology, enables their entry into living cells in quantities εufficient to interfere with specific intracellular protein function.

Materials and Methods

SupTl cells were cultured either under control conditions or with cell εupernatant from H9 cells infected with HIV-1_IIIB (10 TCIDS₅₀) . (HIV-1_IIIB (HTLV-IIIB/H9) isolate waε obtained from the NIH AIDS Reεearch and Reference Reagent Program. The viruε was expanded in HUT78 cells and used in culture at a 10 TCIDS₅₀ concentration.) One ug/ l of L-anti- Tat or n-anti-Tat was added to the cultures for lhr at which time the cells were washed and applied to glass slideε.

Recombinant Tat, generated in a baculoviruε expression syεtem, waε used to prepare a mouse anti-Tat monoclonal antibody, by AGMED, Inc. , Bedford, MA. This antibody (subclass IgGl) was purified by protein A Sepharose chromatography. Control antibody (mouεse monoclonal IgGl was from Sigma Immunochemicals (St. Louis, MO) . Other anti-HIV-1 antibodies used in this study were polyclonal antibodies generated in either rabbit or sheep and were obtained from the NIH AIDS Research and Reference Reagent Program.

The lipoamine, glycyldioctadecylamide, waε covalently linked to the carbohydrate moieties of antibodies by periodate oxidation-sodium borohydride reduction. Glycyldioctadecylamide was prepared from glycine and dioctadecylamine esεentially aε described. Antibodies were disεolved in 300 mM NaHC0₃ at a typical concentration of 1 to 5 mg/ml. NaI0₄ as a freshly prepared aqueous stock solution (80 mg/ml) , was added to a final concentration of 8 mg/ml and the incubation vials were wrapped in aluminum foil and mixed on a rotator for 90 in at room temperature. The oxidized antibodies were purified on Econopac 10DG desalting columns (BioRad) equilibrated in, and eluted with 10 mM Na₂C0₃. A 10 mg/ml solution of glycyldioctadecylamide in DMSO was added to yield a final lipoamine concentration of 1.3 mg/ml and the mixture waε incubated for 20 hr at room temperature on the rotator. NaBH₄, aε a freshly prepared aqueous stock εolution (10 mg/ml) waε then added to final concentration of 1.8 mg/ml, followed after 1 hr by 50 μl of ethanolamine (15 μl in 1 ml H₂0) . After an additional 1 hr.the lipidated antibodieε were purified by chromatography on Econopac 10DG columnε equilibrated with 50 mM εodium phosphate, 0.9% NaCl. pH 7.4. Lipidated antibody solutions were than filtered through a 0.2 μ syringe filter (Millex GV, Millipore Corp. )

To eεtimate the number of lipid moieties linked to each antibody molecule, the anti-Tat antibody was chemically modified as described above, with a starting protein concentration of 5 mg/ml, except that ¹⁴C- glycyldioctadecyla ide (obtained from Dalton Chemical Laboratories, Inc., Toronto, Canada) was added to the unlabeled lipamine as a tracer. The number o f lipamine residueε linked to the antibody waε determined by meaεuring the radioactivity incorporated in the antibody by liquid εcintillation counting, and the amount of antibody, determined by a Lowry aεεay. By thiε procedure, the average number of lipoamine moietieε per antibody molecule waε estimated to be about 8. By attaching the lipid molecules to the carbohydrate residueε on the antibody, distant from the hypervariable binding domain, epitope specificity and binding efficiency should not be affected. This was confirmed by ELISA, demonεtrating that lipidation did not affect the ability of the anti-Tat antibody to bind recombinant Tat protein (not εhown) .

HLCD4-CAT cellε were obtained from the NIH AIDS Research and Reference Reagent Program and were contributed by Drε. Felber and Pavlakiε. The cellε were plated at a density of 3xl0⁵ cells/ml in 24 well culture plateε containing DMEM with 10% fetal bovine serum. Cells were treated with either 10 or 20 μg/ml of L-anti-Tat, or 10 μg/ml L-anti-HIV-1 gpl20 antibody, 10 μg/ml n-anti-Tat antibody, or 10 μg/ml rsCD4 (AGMED) . After 1 hr. the cellε were extensively washed, and

100 μl of HIV-l_IIIB-containing (HIV-1_IIIB, 10 TCID₅₀) supernatant were added. After 24 hr the cultures were harvested and CAT activity was measured using the method described by Seed et al. CAT activity was measured by cutting appropriate spotε as identified from an autoradiogram and scintillation counting of incorporated ¹ C. Incorporated radioactivity from uninfected cells, and HIV-1-infected cells either untreated or treated with L-anti-Tat antibody at 10 or 20 μg/ml or L-anti-gpl20 antibody at 10 μg/ml was determined.

Figure 6A & 6B Effect of L-anti-Tat antibody on HIV-1 replication in living cells.

Fig. 6A & 6B. HIV-l_IIIB-infected SupTl cells were cultured in the presence of 1 μg/ml of either native or L-anti- Tat antibody, L-anti-gpl20 antibody, or rεCD4, in 2 ml of RPMI 1640 medium supplemented with 10% fetal bovine serum, per well. The proteins were added 24 h prior to the addition of HIV-1 virus-containing supernatantε (10 TCID₅₀) . Three hundred μl aliquots were taken daily from day 2 through day 13 for measurements of cell number (100 μl) (6A) and RT activity (200 μl) (6B) , and 300 μl of fresh medium containing the appropriate protein was added to each well until day 10 (arrow) . On subεequent dayε, the freεh medium added did not contain any antibody or rεCD4. RT activity waε meaεured by the method of Ho et al. and iε expreεεed aε cpm/10⁵ cellε. The data are means 1 S.D. from four separate experiments.

Fig. 7. Latently infected cell lines Jl.l and OM- 10.1 were obtained from the NIH AIDS Research and Reference

Reagent Program and were contributed by Drε. Folkε and Butera. These cells were cultured at 1 x 10⁶ cells/ml in RPMI 1640 εupple ented with 10% FBS. They were εtimulated with TNFα (10U and 100U for Jl.l and OM-10.1 cells, respectively) for 72 hrs in the presence of L-anti-Tat or n-anti-Tat antibody (1 μg/ml for each) . On hundred μl aliquots from the supematants were harvested and asεayed for p24 by ELISA (detection limit 50 pg/ml) .

The foregoing description of the preferred embodiments of the present invention has been presented for purpoεeε of illuεtration and deεcription. They are not intended to be exhaustive or to limit the invention to the precise form disclosed, and many modifications and variations are possible in light of the above teaching.

Such modifications and variations which may be apparent to a person εkilled in the art are intended to be within the scope of this invention.

Claims

CLAIMS 1. A method for modifying the characteristicε of a protein, comprising the steps of: attaching a lipid substituent to the protein by a covalent linkage of at least one lipoamine residue to a carbohydrate side chain to produce a lipidized protein; and recovering the lipidized protein.

2. A method according to Claim 1, wherein the lipid substituent is a lipoamine.

3. A method according to Claim 2, wherein the step of attaching further comprises the stepε of oxidizing a carbohydrate on a glycoεylated polypeptide to produce an oxidized glycoprotein; and reacting the oxidized glycoprotein with a lipoamine under suitable reaction conditions to form a lipidized protein.

4. A method according to Claim 2, wherein the lipoamine is a straight-chain lipoamine according to the formula:

NH₂-R-(CH₂)_n-CH₃

where R is selected from the group consisting of: disubstituted alkyl (alkylene); 1,4-disubstituted cyclohexyl; disubεtituted aryl (arylene) ; amido group of the formula - (CHR-L) -CO-NH- wherein R_λ iε hydrogen or an amino group; alkylcarbonyl; and phoεphate dieεter; n iε 1-50.

5. A method according to Claim 2, wherein the lipoamine iε a branched-chain lipoamine according to the formula:

NH₂-R'-(CH₂)_n-CH₃ I (CH₂)_m I CH₃

where R' iε: a triεubεtituted alkyl; a trisubstituted aryl; an amido group of the formula -(CHR-_j -CO-N< wherein R_x is hydrogen or an amino group; an imino group of the formula - CHR₂-NH-CH< wherein R₂ is hydrogen or an amino group or an imino group of the formula -CH₂-N<; or a phosphate diester; m is 1-50; n is 1-50; and m and n are selected independently.

6. A method according to claim 5, wherein the branched-chain lipoamine is glycyldioctadecylamide.

7. A method according to Claim 1, wherein the protein is a naturally-occurring glycoprotein.

8. A method according to Claim 1, wherein the protein is encoded by an immunoglobulin superfamily gene.

9. A method for targeting an intracellular protein for binding with an antibody in a cell, comprising contacting the cell with a lipidized antibody which binds εpecifically with the intracellular protein.

10. A method according to Claim 9, wherein the lipidized antibody compriεes at least one lipoamine residue linked to a carbohydrate side chain of an immunoglobulin.

11. A method according to Claim 14, wherein the lipoamine is glycyldioctadecylamide.

12. A method according to Claim 13, wherein the lipidized antibody is administered to a living mammalian cells in vivo.

13. A method according to Claim 12, wherein the lipidized antibody is taken up into the living cells to a greater extent than is a comparable antibody having the same amino acid sequence(s) and the same glycoεylation pattern and lacking lipidization.

14. A compoεition for therapy or prophylaxiε of a diεeaεe, comprising a therapeutically effective dosage of a lipidized protein.

15. A composition according to Claim 14, wherein the lipidized protein is an antibody.

16. A composition according to Claim 15, wherein the antibody binds to an intracellular protein.

17. A composition according to Claim 16, wherein the intracellular protein is a viral-encoded protein.

18. A composition according to Claim 17, wherein the viral- encoded protein is a Tat protein encoded by HIV-1.

19. A composition for prophylaxiε, compriεing a prophylactically effective doεage of a lipidized antibody wherein a lipid εubεtituent is covalently linked to the antibody by a covalent linkage of at least one lipoamine residue to a carbohydrate side chain to produce a lipidized antibody.

20. A compoεition compriεing a lipidized antibody linked to a diagnoεtic reporter, wherein a lipid εubεtituent iε covalently linked to the antibody by a covalent linkage of at leaεt one lipoamine reεidue to a carbohydrate εide chain to produce a lipidized antibody.

21. A compoεition of claim 20, wherein the antibody bindε to an antigen intracellularly in living cells.

22. A composition of claim 22, wherein the antibody binds to the HIV-1 Tat protein intracellularly in HIV-infected human cells.

23. A method for diagnosing a pathological condition, comprising the steps of: administering a lipidized antibody comprising a diagnoεtic reporter to a cell εample; and detecting the presence of cells in which the diagnostic reporter is preferentially localized.