CA2522345A1

CA2522345A1 - Glycopegylation methods and proteins/peptides produced by the methods

Info

Publication number: CA2522345A1
Application number: CA002522345A
Authority: CA
Inventors: Shawn De Frees; David Zopf; Robert Bayer; Caryn Bowe; David Hakes; Xi Chen
Original assignee: Neose Technologies, Inc.; Shawn De Frees; David Zopf; Robert Bayer; Caryn Bowe; David Hakes; Xi Chen; Novo Nordisk A/S
Current assignee: Novo Nordisk AS
Priority date: 2003-04-09
Filing date: 2004-04-09
Publication date: 2004-11-18
Also published as: JP2007525941A; CY1114626T1; MXPA05010773A; EP1615945B1; SG155777A1; AU2004236174A1; PL1615945T3; LU92358I2; US20120220517A1; NZ542586A; JP4674702B2; EP2338333B1; US20140294762A1; US20100048456A1; US20070026485A1; IL171109A; EP2338333A3; EP1615945A2; WO2004099231A3; EP1615945A4

Abstract

The invention includes methods and compositions for remodeling a peptide molecule, including the addition or deletion of one or more glycosyl groups to a peptide, and/or the addition of a modifying group to a peptide.

Description

DEMANDE OU BREVET VOLUMINEUX
LA PRESENTE PARTIE DE CETTE DEMANDE OU CE BREVET COMPREND
PLUS D'UN TOME.

NOTE : Pour les tomes additionels, veuillez contacter 1e Bureau canadien des brevets JUMBO APPLICATIONS/PATENTS
THIS SECTION OF THE APPLICATION/PATENT CONTAINS MORE THAN ONE
VOLUME

NOTE: For additional volumes, please contact the Canadian Patent Office NOM DU FICHIER / FILE NAME
NOTE POUR LE TOME / VOLUME NOTE:

TITLE OF THE INVENTION
CiLS~COPE(sYLATION METHODS AND PROTEINS/PEPT1DES PRODUCED BY
THE METHODS
BACKGROUND OF THE INVENTION
Most naturally occurring peptides contain caxbohydrate moieties attached to the peptide via specific linkages to a select number of amino acids along the length of the primary peptide chain. Thus, many naturally occurring peptides are termed "glycopeptides." The variability of the glycosylation pattern on any given peptide has enormous implications for the function of that peptide. For example, the structure of the N-linked glycans on a peptide can impact various characteristics of the peptide, including the protease susceptibility, intracellular trafficking, secretion, tissue targeting, biological half life and antigenicity of the peptide in a cell or organism.
The alteration of one or more of these characteristics greatly affects the efficacy of a peptide in its natural setting, and also affects the efficacy of the peptide as a therapeutic agent in situations where the peptide has been generated for that purpose.
The carbohydrate structure attached to the peptide chain is known as a "glycan"
molecule. The specific glycan structure present on a peptide affects the solubility and aggregation characteristics of the peptide, the folding of the primary peptide chain and therefore its functional or enzymatic activity, the resistance of the peptide to proteolytic attack and the control of proteolysis leading to the conversion of inactive forms of the peptide to active forms. Importantly, terminal sialic acid residues present on the glycan molecule affect the length of the half life of the peptide in the mammalian circulatory system. Peptides whose glycans do not contain terminal sialic acid residues are rapidly removed from the circulation by the liver, an event which negates any potential therapeutic benefit of the peptide.
The glycan structures found in naturally occurring glycopeptides are typically divided into two classes, N-linked and ~-linked glycans.
Peptides expressed in eukaryotic cells are typically N-glycosylated on asparagine residues at sites in the peptide primary structure containing the sequence aspaxagine-X-serine/threonine where X can be any amino acid except proline and aspartic acid. The carbohydrate portion of such peptides is known as an N-linked glycan. The early events of N-glycosylation occur in the endoplasmic reticulum (ER) and are identical in mammals, plants, insects and other higher eukaryotes.
First, an oligosaccharide chain comprising fourteen sugar residues is constructed on a lipid carrier molecule. As the nascent peptide is translated and translocated into the ER, the entire oligosaccharide chain is transferred to the amide group of the asparagine residue in a reaction catalyzed by a membrane bound glycosyltransferase enzyme. The N-linked glycan is further processed both in the ER and in the Golgi apparatus.
The further processing generally entails removal of some of the sugar residues and addition of other sugar residues in reactions catalyzed by glycosidases and glycosyltransferases specific for the sugar residues removed and added.
Typically, the final structures of the N-linked glycans are dependent upon the organism in which the peptide is produced. For example, in general, peptides produced in bacteria are completely unglycosylated. Peptides expressed in insect cells contain high mamiose and paunci-mannose N-linked oligosaccharide chains, among others.
Peptides produced in mammalian cell culture are usually glycosylated differently depending, e.g., upon the species and cell culture conditions. Even in the same species and under the same conditions, a certain amount of heterogeneity in the glycosyl chains is sometimes encountered. Further, peptides produced in plant cells comprise glycan structures that differ significantly from those produced in animal cells. The dilemma in _2_ the art of the production of recombinant peptides, particularly when the peptides are to be used as therapeutic agents, is to be able to generate peptides that are correctly glycosylated, i.e., to be able to generate a peptide having a glycan structure that resembles, or is identical to that present on the naturally occurring form of the peptide.
Most peptides produced by recombinant means comprise glycan structures that are different from the naturally occurring glycans.
A variety of methods have been proposed in the art to customize the glycosylation pattern of a peptide including those described in WO 99/22764, WO
98/58964, WO 99/54342 and U.S. Patent No. 5,047,335, among others.
Essentially, many of the enzymes required for the ivc vitf°o glycosylation of peptides have been cloned and sequenced. In some instances, these enzymes have been used i~ vitro to add specific sugars to an incomplete glycan molecule on a peptide. In other instances, cells have been genetically engineered to express a combination of enzymes and desired peptides such that addition of a desired sugar moiety to an expressed peptide occurs within the cell.
Peptides may also be modified by addition of O-linked glycans, also called mucin-type glycans because of their prevalence on mucinous glycopeptide.
Unlike N-glycans that are linked to asparagine residues and are formed by eh bloc transfer of oligosaccharide from lipid-bound intermediates, O-glycans are linked primarily to serine and threonine residues and are formed by the stepwise addition of sugars from nucleotide sugars (Tanner et al., Biochinz. Biophys. Acta. 906:81-91 (1987);
and Hounsell et al., Glycoco~j. J. 13:19-26 (1996)). Peptide function can be affected by the structure of the O-linked glycans present thereon. For example, the activity of P-selectin ligand is affected by the O-linked glycan structure present thereon.
For a review of O-linked glycan structures, see Schachter and Brockhausen, The Biosynthesis of Branched ~-Linked Glycans, 1989, Society for Experimental Biology, pp. 1-26 (Great Britain). Other glycosylation patterns are formed by linking glycosylphosphatidylinositol to the carboxyl-terminal carboxyl group of the protein (Takeda ct al., ~'re~rds ~i~cher3~. Sci. 20:367-371 (1995); and Udenfriend et al., An~c.
Rev. ~iocher~z. ~~~:593-591 (1995).
Although various techniques currently exist to modify the N-linked glycans of peptides, there exists in the art the need for a generally applicable method of producing peptides having a desired, i.e., a customized glycosylation pattern. There is a particular need ~in the art for the customized iyz vitr~ glycosylation of peptides, where the resulting peptide can be produced at industrial scale. This and other needs are met by the present invention.
The administxation of glycosylated and non-glycosylated peptides for engendering a particular physiological response is well known in the medicinal arts.
Among the best known peptides utilized for this purpose is insulin, which is used to txeat diabetes. Enzymes have also been used for their therapeutic benefits. A
maj ox factor, which has limited the use of therapeutic peptides is the immunogenic nature of 1 S most peptides. In a patient, an immunogenic response to an administered peptide can neutralize the peptide andlor lead to the development of an allergic response in the patient. Other deficiencies of therapeutic peptides include suboptimal potency and rapid clearance rates. The problems inherent in peptide therapeutics are recognized in the art, and various methods of eliminating the problems have been investigated. To provide soluble peptide therapeutics, synthetic polymers have been attached to the peptide backbone.
Polyethylene glycol) ("PEG") is an exemplary polymer that has been conjugated to peptides: The use of PEG to derivatize peptide therapeutics has been demonstrated to reduce the immunogenicity of the peptides and prolong the clearance time from the circulation. For example, U.S. Pat. No. 4,179,337 (Davis et al.) concerns non-immunogenic peptides, such as enzymes and peptide hormones coupled to polyethylene glycol (PEG) or polypropylene glycol. Between 10 and 100 moles of _q_ polymer are used per mole peptide and at least 15% of the physiological activity is maintained.
WO 93/15189 (Veronese et al.) concerns a method to maintain the activity of polyethylene glycol-modified proteolytic enzymes by linking the proteolytic enzyme to a macromolecularized inhibitor. The conjugates are intended for medical applications.
The principal mode of attachment of PEG, and its derivatives, to peptides is a non-specific bonding through a peptide amino acid residue. For example, U.S.
Patent No. 4,088,538 discloses an enzymatically active polymer-enzyme conjugate of an enzyme covalently linked to PEG. Similarly, U.S. Patent No. 4,496,689 discloses a covalently attached complex of oc-1 protease inhibitor with a polymer such as PEG or methoxypoly(ethylene glycol) ("mPEG"). Abuchowski et al. (J. Biol. Chem. 252:

(1977) discloses the covalent attachment of mPEG to an amine group of bovine serum albumin. U.S. Patent No. 4,414,147 discloses a method of rendering interferon less hydrophobic by conjugating it to an anhydride of a dicarboxylic acid, such as I S polyethylene succinic anhydride). PCT WO 87/00056 discloses conjugation of PEG
and poly(oxyethylated) polyols to such proteins as interferon-(3, interleukin-2 and immunotoxins. EP 154,316 discloses and claims chemically modified lymphokines, such as IL-2 containing PEG bonded directly to at least one primary amino group of the lyrnphokine. U.S. Patent No. 4,055,635 discloses pharmaceutical compositions of a water-soluble complex of a proteolytic enzyme linked covalently to a polymeric substance such as a polysaccharide.
Another mode of attaching PEG to peptides is through the non-specific oxidation of glycosyl residues on a peptide. The oxidized sugar is utilized as a locus for attaching a PEG moiety to the peptide. For example, M'Timkulu (WO
94/05332) discloses the use of a hydrazine- or amino-PEG to add PEG to a glycoprotein.
The glycosyl moieties are randomly oxidized to the corresponding aldehydes, which are subsequently coupled to the amino-PEG. See also, Bona et al. (WO 96/40731), where a PEG is added to an immunoglobulin molecule by enzymatically oxidizing a glycan on the imnunoglobulin and then contacting the glycan with an amino-PEG molecule.
In each of the methods described above, polyethylene glycol) is added in a random, non-specific manner to reactive residues on a peptide backbone. For the production of therapeutic peptides, it is clearly desirable to utilize a derivatization strategy that results in the formation of a specifically labeled, readily characterizable, essentially homogeneous product.
Two principal classes of enzymes are used in the synthesis of carbohydrates, glycosyltransferases (e.g., sialyltransferases, oligosaccharyltransferases, N-acetylglucosaminyltransferases), and glycosidases. The glycosidases are further classified as exoglycosidases (e.g., ~3-mannosidase, (3-glucosidase), and endoglycosidases (e.g., Endo-A, Endo-M). Each of these classes of enzymes has been successfully used synthetically to prepare carbohydrates. For a general review, see, Crout et al., Cu~y~. Opi~c. Cherrc. Biol. 2: 98-111 (1998).
Glycosyltransferases modify the oligosaccharide structures on peptides.
Glycosyltransferases are effective for producing specific products with good stereochemical and regiochemical control. Glycosyltransferases have been used to prepare oligosaccharides and to modify terminal N- and O-linked carbohydrate structures, particularly on peptides produced in mammalian cells. For example, the terminal oligosacchaxides of glycopeptides have been completely sialylated and/or fucosylated to provide more consistent sugar structures, which improves glycopeptide pharmacodynamics and a variety of other biological properties. For example, [3-1,4-galactosyltransferase is used to synthesize lactosamine, an illustration of the utility of glycosyltransferases in the synthesis of carbohydrates (see, e.g., Wong et al., J: O~g:
Chem. 47: 5416-5418 (1982)). Moreover, numerous synthetic procedures have made use of a-sialyltransferases to transfer sialic acid from cytidine-5'-monophospho-N-acetylneuraminic acid to the 3-~I-I or 6-~H of galactose (see, e.g:, Kevin et al., Chenz.
Euf~. .J: 2: 1359-1362 (1996)). Fucosyltransferases axe used in synthetic pathways to transfer a fucose unit from guanosine-5'-diphosphofucose to a specific hydroxyl of a saccharide acceptor. For example, Ichikawa prepared sialyl Lewis-X by a method that involves the fucosylation of sialylated la.ctosamine with a cloned fucosyltransferase (Ichikawa et al., .I. Arn. Chern. Soc. 114: 9283-9298 (1992)). For a discussion of recent advances in glycoconjugate synthesis for therapeutic use see, Koeller et al., Nature Biotechnology 18: 835-841 (2000). See also, U.S. Patent No. 5,876,980;
6,030,815;
5,728,554; 5,922,577; and WO/9831826.
Glycosidases can also be used to prepare saccharides. Glycosidases normally catalyze the hydrolysis of a glycosidic bond. However, under appropriate conditions, they can be used to form this linkage. Most glycosidases used for carbohydrate synthesis are exoglycosidases; the glycosyl transfer occurs at the non-reducing terminus I5 of the substrate. The glycosidase binds a glycosyl donor in a glycosyl-enzyme intermediate that is either intercepted by water to yield the hydrolysis product, or by an acceptor, to generate a new glycoside or oligosaccharide. An exemplary pathway using an exoglycosidase is the synthesis of the core trisaccharide of all N-linked glycopeptides, including the (3-mannoside linkage, which is formed by the action of j3-mannosidase (Singh et al., Chem. Comrnun. 993-994 (1996)).
In another exemplary application of the use of a glycosidase to form a glycosidic linkage, a mutant glycosidase has been prepared in which the normal nucleophilic amino acid within the active site is changed to a non-nucleophilic amino acid. The mutant enzyme does not hydrolyze glycosidic linkages, but can still form them. Such a mutant glycosidase is used to prepare oligosaccharides using an a-glycosyl fluoride donor and a glycoside acceptor molecule (Withers et al., TJ.S. Patent No. 5,716,812).
Although their use is less common them that of the exoglycosidases, endoglycosidases are also utilized to prepare carbohydrates. Methods based on the use of endoglycosidases have the advantage that an oligosaccharide, rather than a monosaccharide, is transferred. Oligosaccharide fragments have been added to substrates using endo-(3-N-acetylglucosamines such as endo-F, e~cdo-M (Wang et al., Tetrahedron Lett. 37: 1975-1978); and Haneda et al., Canbohydr~. Res. 292: 61-(1996)).
In addition to their use in preparing carbohydrates, the enzymes discussed above are applied to the synthesis of glycopeptides as well. The synthesis of a homogenous glycoform of ribonuclease B has been published (Witte I~. et al., J. Am. Chem.
Soc.
119: 2114-2118 (1997)). The high mannose core of ribonuclease B was cleaved by treating the glycopeptide with endoglycosidase H. The cleavage occurred specifically between the two core GIcNAc residues. The tetrasaccharide sialyl Lewis X was then enzymatically rebuilt on the remaining GIcNAc anchor site on the now homogenous protein by the sequential use of [3-1,4-galactosyltransferase, a-2,3-sialyltransferase and a-1,3-fucosyltransferase V. However, while each enzymatically catalyzed step proceeded in excellent yield, such procedures have not been adapted for the generation of glycopeptides on an industrial scale.
Methods combining both chemical and enzymatic synthetic elements are also known in the art. For example, Yamamoto and coworkers (Carbohydr. Res. 305:

422 (1998)) reported the chemoenzymatic synthesis of the glycopeptide, glycosylated Peptide T, using an endoglycosidase. The N-acetylglucosaminyl peptide was synthesized by purely chemical means. The peptide was subsequently enzymatically elaborated with the oligosaccharide of human transferrin peptide. The saccharide _g_ portion was added to the peptide by treating it with an endo-(3-N-acetylglucosaminidase. The resulting glycosylated peptide was highly stable and resistant to proteolysis when compared to the peptide T and N-acetylglucosaminyl peptide T.
The use of glycosyltransferases to modify peptide structure with reporter groups has been explored. For example, Brossmer et al. (U.S. Patent No. 5,405,753) discloses the formation of a fluorescent-labeled cytidine monophosphate ("CMP") derivative of sialic acid and the use of the fluorescent glycoside in an assay for sialyl transferase activity and for the fluorescent-labeling of cell surfaces, glycoproteins and peptides.
Gross et al. (Ahalyt. Biochem. 186: 127 (1990)) describe a similar assay. Bean et al.
(U.S. Patent No. 5,432,059) discloses an assay for glycosylation deficiency disorders utilizing reglycosylation of a deficiently glycosylated protein. The deficient protein is reglycosylated with a fluorescent-labeled CMP glycoside. Each of the fluorescent sialic acid derivatives is substituted with the fluorescent moiety at either the 9-position or at the amine that is normally acetylated in sialic acid. The methods using the fluorescent sialic acid derivatives are assays for the presence of glycosyltransferases or for non-glycosylated or improperly glycosylated glycoproteins. The assays are conducted on small amounts of enzyme or glycoprotein in a sample of biological origin. The enzymatic derivatization of a glycosylated or non-glycosylated peptide on a preparative or industrial scale using a modified sialic acid has not been disclosed or suggested in the prior art.
Considerable effort has also been directed towards the modification of cell surfaces by altering glycosyl residues presented by those surfaces. For example, Fukuda and coworkers have developed a method for attaching glycosides of defined structure onto cell surfaces. The method exploits the relaxed substrate specificity of a fucosyltransferase that can transfer fucose and fucose analogs bearing diverse glycosyl substrates (Tsuboi et al., ..J:: Bi~l. Chena. 271: 27213 (1996)).

Enzymatic methods have also been used to activate glycosyl residues on a glycopeptide towards subsequent chemical elaboration. The glycosyl residues are typically activated using galactose oxidase, which converts a terrninal galactose residue to the corresponding aldehyde. The aldehyde is subsequently coupled to an amine-s containing modifying group. For example, Casares et al. (Natm°e Bi~tech. 19: 142 (2001)) have attached doxorubicin to the oxidized galactose residues of a recombinant MHCII-peptide chimera.
Glycosyl residues have also been modified to contain ketone groups. For example, Mahal and co-workers (Science 276: 1125 (1997)) have prepared N-levulinoyl mannosamine ("ManLev"), which has a ketone functionality at the position normally occupied by the acetyl group in the natural substrate. Cells were treated with the ManLev, thereby incorporating a ketone group onto the cell surface. See, also Saxon et al., Science 287: 2007 (2000); Hang et al., J. Arn. Chem. Soc. 123:

(2001); Yarema et al., J. Biol. Chem. 273: 31168 (1998); and Charter et al., Glycobiology 10: 1049 (2000).
The methods of modifying cell surfaces have not been applied in the absence of a cell to modify a glycosylated or non-glycosylated peptide. Further, the methods of cell surface modification are not utilized for the enzymatic incorporation preformed modified glycosyl donor moiety into a peptide. Moreover, none of the cell surface modification methods are practical for producing glycosyl-modified peptides on an industrial scale.
Despite the efforts directed toward the enzymatic elaboration of saccharide structures, there remains still a need for an industrially practical method for the modification of glycosylated and non-glycosylated peptides with modifying groups such as water-soluble polymers, therapeutic moieties, biomolecules and the like. Of particular interest are methods in which the modified peptide has improved properties, which enhance its use as a therapeutic or diagnostic agent. The present invention fulfills these and other needs.

SUMMARY OF THE INVENTION
The invention includes a multitude of methods of remodeling a peptide to have a specific glycan structure attached thereto. Although specific glycan structures are described herein, the invention should not be construed to be limited to any one particular structure. In addition, although specific peptides are described herein, the invention should not be limited by the nature of the peptide described, but rather should encompass any and all suitable peptides and variations thereof.
The description which follows discloses the preferred embodiments of the invention and provides a written description of the claims appended hereto. The invention encompasses any and all variations of these embodiments that are or become apparent following a reading of the present specification.
The invention includes a cell-free, in vitro method, of remodeling a peptide comprising polyethylene glycol), the peptide having the formula:
_~-X1-X2 wherein AA is a terminal or internal amino acid residue of the peptide;
Xl-X2 is a saccharide covalently linked to the AA, wherein Xl is a first glycosyl residue; and X2 is a second glycosyl residue covalently linked to Xl, wherein Xl and X2 are selected from monosaccharyl and oligosaccharyl residues;
the method comprising:

(a) removing X2 or a saccharyl subunit thereof from the peptide, thereby forming a truncated glycan.
In one aspect, the invention further comprises formation of a truncated glycan by removing a Sia residue.
In one embodiment of the invention, a peptide has the formula:
(X~~) X
Man-(X3)a (X6)d AA-GIcNAc-GIcNAc-Man-(X4)b Man-(X5)c I
(X7) a wherein X3, X4, Xs, X6, X7, and X17 , are independently selected monosaccharyl or oligosaccharyl residues; and a, b, c, d, e, and x are independently selected from the integers 0, 1 and 2.
In one aspect of the invention, an oligosaccharyl residue is a member selected from GIcNAc-Gal-Sia and GIcNAc-Gal. In another aspect, at least one oligosaccharide member is selected from a, b, c, d, a and x is 1 or 2. In yet another aspect, the removing of step (a) produces a truncated glycan in which at least one of a, b, c, a and x are 0.

The invention includes a method of remodeling a peptide wherein X3, Xs and X7 are members independently selected from (mannose)~ and (mannose)~ (X8) wherein X8 is a glycosyl moiety selected fiom mono- and oligo-saccharides; and z is an integer between 1 and 20, wherein when z is 3 or greater, each (mannose)Z is independently selected from linear and branched structures.
In one aspect, X4 is selected from the group consisting of GIcNAc and xylose. In another aspect, X3, XS and X7 are (mannose)u, wherein a is selected from the integers between 1 and 20, and when a is 3 or greater, each (mannose)u is independently selected from linear and branched structures.
The invention also includes a method of remodeling a peptide, wherein the peptide has the formula:
Man-(GIcNAc)S
--~-GIcNAc-GIcNAc-Man Man-(GIcNAc)t wherein r, s, and t are integers independently selected from 0 and 1.
In an embodiment of the invention, a peptide has the formula:

( ~ 9~m 2 ~AA-GaINAc-(Gal)f-'~
(~~10)n wherein X9 and Xl° are independently selected monosaccharyl or oligosaccharyl residues and m, n and f are integers independently selected from 0 and 1.
In one aspect, a peptide has the formula:
x(16 Gal Fuc-GIcNAc -AA- GaINAc-Gal-Sia wherein Xi6 is a member selected from:
(Fuc)S (Fuc)S ( i uc);
I
-Sia ; S GIcNAc-Gal-Sia ; and 5 GIcNAc-Gal GIcNAc-Gal-Sia wherein s and i are integers independently selected from 0 and 1.
In another aspect, a peptide has the formula:

(FUC)j (~~C~e~C)g-(~13)h AA G i ~~IAC-~G~~)p-~~C94)i (~15)~
wherein Xi3, X14, and Xls are independently selected glycosyl residues; and g, h, i, j, k, and p are independently selected from the integers 0 and 1.
In yet another aspect of the invention, at least one of g, h, i, j, k and p is 1. In another aspect, X14 and Xls are members independently selected from GIcNAc and Sia and i and k are independently selected from the integers 0 and 1. In still another aspect, at least one of i and k is 1, and if k is 1, g, h, and j axe 0.
The invention also includes a method of remodeling a peptide, wherein the method comprises contacting the truncated glycan with at least one glycosyltransferase and at least one glycosyl donor under conditions suitable to transfer the at least one glycosyl donor to the truncated glycan, thereby remodeling the peptide 1 S comprising polyethylene glycol).
In one aspect, a glycosyl donor comprises a modifying group covalently linked thereto.
The invention also includes a method of remodeling a peptide, the method comprising removing Xl, thereby exposing AA. In one aspect, a method includes contacting AA with at least one glycosyltransferase and at least one glycosyl donor under conditions suitable to transfer said at least one glycosyl donor to AA, thereby remodeling said peptide comprising polyethylene glycol).
In one aspect, at least one glycosyl donor comprises a modifying group covalently linked thereto. In another aspect, a modifying group is polyethylene glycol). In one embodiment, a polyethylene glycol) has a molecular weight distribution that is essentially homodisperse.
The invention includes a method of remodeling a peptide, wherein, prior to contacting the truncated glycan with at least one glycosyltransferase and at least one glycosyl donor under conditions suitable to transfer the at least one glycosyl donor to the truncated glycan, thereby remodeling the peptide comprising polyethylene glycol), a group added to the saccharide during post-translational modification is removed.
In one aspect, a removed group is a member selected from phosphate, sulfate, carboxylate and esters thereof.
The invention includes a method of remodeling a peptide wherein a peptide has the formula:
-AA--Z-X~-X2 wherein Z is a member selected from O, S, NH and a cross-linker.
The invention also includes a method of remodeling a peptide, wherein the peptide has the formula:

(X11 ~x (~~12~ r wherein Xll and X12 are independently selected glycosyl moieties; and r and x are integers independently selected from 0 and 1.
In one aspect of the invention, Xl l and X12 are (mannose)q, wherein q is selected from the integers between 1 and 20, and when q is three or greater, (mannose)q is selected from linear and branched structures.
The invention includes a pharmaceutical composition comprising a pharmaceutically acceptable diluent and a remodeled peptide according to a cell-free, in vitro method of remodeling a peptide comprising polyethylene glycol), the peptide having the formula:
AA--X1-X~
wherein AA is a terminal or internal amino acid residue of the peptide;
Xl-X2 is a saccharide covalently linked to the AA, wherein Xl is a first glycosyl residue; and Xa is a second glycosyl residue covalently linked to Xl, wherein Xl and Xa are selected from monosaccharyl and oligosaccharyl residues;
-1 ~-the method comprising:
(a) removing X2 or a saccharyl subunit thereof from the peptide, thereby forming a truncated glycan.
The invention also includes a cell-free, in vitro method of remodeling a peptide comprising polyethylene glycol), the peptide having the formula:
AA X~
a wherein AA is a terminal or internal amino acid residue of the peptide;
Xl is a glycosyl residue covalently linked to the AA, selected from monosaccharyl and oligosaccharyl residues; and a is an integer selected from 0 and 1, the method comprising:
contacting the peptide with at least one glycosyltransferase and at least one glycosyl donor under conditions suitable to transfer at least one glycosyl donor to the truncated glycan, thereby remodeling the peptide.
In one aspect, at least one glycosyl donor comprises a modifying group covalently linked thereto. In another aspect, the modifying group is polyethylene glycol). In yet another aspect, the polyethylene glycol) has a molecular weight distribution that is essentially homodisperse.
The invention also includes a pharmaceutical composition comprising a pharmaceutically acceptable diluent and a remodeled peptide according to a cell-free, in vitro method of remodeling a peptide comprising polyethylene glycol), the peptide having the formula:
~1 a Wherein AA is a terminal or internal amino acid residue of the peptide;
Xl is a glycosyl residue covalently linked to the AA, selected from monosaccharyl and oligosaccharyl residues; and a is an integer selected from 0 and l, the method comprising:
contacting the peptide with at least one glycosyltransferase and at least one glycosyl donor under conditions suitable to transfer at least one glycosyl donor to the truncated glycan, thereby remodeling the peptide.

BRIEF DESCRIPTI~N ~F THE DRAWINGS
For the purpose of illustrating the invention, there are depicted in the drawings certain embodiments of the invention. However, the invention is not limited to the precise arrangements and instrumentalities of the embodiments depicted in the drawings.
Figure 1 is a scheme depicting a trimannosyl core glycan (left side) and the enzymatic process for the generation of a glycan having a bisecting GIcNAc (right side).
Figure 2 is a scheme depicting an elemental trimannosyl core structure and complex chains in various degrees of completion. The ivy vitf°o enzymatic generation of an elemental trimannosyl core structure from a complex carbohydrate glycan structure which does not contain a bisecting GIcNAc residue is shown, as is the generation of a glycan structure therefrom which contains a bisecting GIcNAc. Symbols: squares: GIcNAc; light circles:
Man; dark circles: Gal; triangles: NeuAc.
Figure 3 is a scheme for the enzymatic generation of a sialylated glycan structure (right side) beginning with a glycan having a trimannosyl core and a bisecting GIcNAc (left side).
Figure 4 is a scheme of a typical high mannose containing glycan structure (left side) and the enzymatic process for reduction of this structure to an elemental trimannosyl core structure. In this scheme, X is mannose as a monosaccharide, an oligosaccharide or a polysaccharide.
Figure 5 is a diagram of a fucose and xylose containing N-linked glycan structure typically produced in plant cells.
Figure 6 is a diagram of a fucose containing N-linked glycan structure typically produced in insect cells. Note that the glycan may have no core fucose, it amy have a single core fucose with either linkage, or it may have a single core fucose having a preponderance of one linkage.
Figure 7 is a scheme depicting a variety of pathways for the tximming of a high mannose structure and the synthesis of complex sugar chains therefrom.
Symbols: squares:
GIcNAc; circles: Man; diamonds: fucose; pentagon: xylose.
Figure 8 is a scheme depicting i~c vity~o strategies for the synthesis of complex structures from an elemental trimannosyl core structure. Symbols: Squares:
GIcNAc; light circles: Man; dark circles: Gal; dark triangles: NeuAc; GnT: N-acetyl glucosazninyltransfexase; GaIT: galactosyltxansferase; ST: sialyltransferase.
Figure 9 is a scheme depicting two in vi~°~ strategies for the synthesis of monoantennary glycans, and the optional glycoPEGylation of the same. Dark squares:
GIcNAc; dark circles: Man; light circles: Gal; dark triangles: sialic acid.
Figure 10 is a scheme depicting two in vitr°~ strategies for the synthesis of monoantennaxy glycans, and the optional glycoPEGylation of the same. Dark squares:
GIcNAc; dark circles: Man; light circles: Gal; dark triangles: sialic acid.
Figure 11 is a scheme depicting various complex structures, which may be synthesized from an elemental trimannosyl core structure. Symbols: Squares:
GIcNAc; light circles: Man; dark circles: Gal; triangles: NeuAc; diamonds: fucose; FT and FucT:
fucosyltransferase; GaIT: galactosyltransfexase; ST: sialyltransferase; Le:
Lewis antigen;
SLe: sialylated Lewis antigen.
Figure 12 is an exemplary scheme for preparing O-linked glycopeptides originating with serine or threonine. Optionally, a water soluble polymer (WSP) such as polyethylene glycol) is added to the final glycan structure.
Figure 13 is a series of diagrams depicting the four types of O-glycan structures, termed cores 1 through 4. The core structure is outlined in dotted lines.
Figure 14, comprising Figure 14A and Figure 14B, is a series of schemes showing an exemplary embodiment of the invention in which carbohydrate residues comprising complex carbohydrate structures and/or high mannose high mannose structures are trimmed back to the first generation biantennary structure. Optionally, fucose is added only after reaction with GnT I. A modified sugar bearing a water-soluble polymer (WSP) is then conjugated to one or more of the sugar residues exposed by the trimming back process.
Figure 15 is a scheme similar to that shown in Figure 4, in which a high mannose or complex structure is "trimmed back" to the mannose beta-linked core and a modified sugar bearing a water soluble polymer is then conjugated to one or more of the sugar residues exposed by the trimming back process. Sugars axe added sequentially using glycosyltransferases.
Figure 16 is a scheme similar to that shown in Figure 4, in which a high mannose or complex structure is trimmed back to the GIcNAc to which the first mannose is attached, and a modified sugar bearing a water soluble polymer is then conjugated to one or more of the sugar residues exposed by the trimming back process. Sugars are added sequentially using glycosyltransferases.
Figure 17 is a scheme similar to that shown in Figure 4, in which a high mannose or cpomplex structure is trimmed back to the first GIcNAc attached to the Asn of the peptide, following which a water soluble polymer is conjugated to one or more sugar residues which have subsequently been added on. Sugars are added sequentially using glycosyltransferases.
Figure 18, comprising Figure 18A and 18B, is a scheme in which an N-linked carbohydrate is optionally trirmned back from a high mannose or cpmplex structure, and subsequently derivatized with a modified sugar moiety (Gal or GIcNAc) bearing a water-soluble polymer.
Figure 19, comprising Figure 19A and 198, is a scheme in which an N-linked carbohydrate is trimmed back from a high mannose or complex structure and subsequently derivatized with a sialic acid moiety bearing a watex-soluble polymer. Sugars are added sequentially using glycosyltransferases.
Figure 20 is a scheme in which an N-linked carbohydrate is optionally trimmed back from a high mannose oor complex structure and subsequently derivatized with one or more sialic acid moieties, and terminated with a sialic acid derivatized with a water-soluble polymer. Sugars are added sequentially using glycosyltransferases.
Figure 21 is a scheme in which an O-linked saccharide is "trimmed back" and subsequently conjugated to a modified sugar bearing a water-soluble polymer.
In the exemplary scheme, the carbohydrate moiety is "trimmed back" to the first generation of the biantennary structure.
Figure 22 is an exemplary scheme for trimming back the carbohydrate moiety of an O-linked glycopeptide to produce a mannose available for conjugation with a modified sugar having a water-soluble polymer attached thereto.
Figure 23, comprising Figure 23A to Figure 23C, is a series of exemplary schemes.
Figure 23A is a scheme that illustrates addition of a PEGylated sugar, followed by the addition of a non-modified sugar. Figure 23B is a scheme that illustrates the addition of more that one kind of modified sugar onto one glycan. Figure 23C is a scheme that illustrates the addition of different modified sugars onto O-linked glycans and N-linked glycans.

Figure 24 is a diagram of various methods of improving the therapeutic function of a peptide by glycan remodeling, including conjugation.
Figure 25 is a set of schemes for glycan remodeling of a therapeutic peptide to treat Gaucher Disease.
Figure 26 is a scheme for glycan remodeling to generate glycans having a terminal mannose-6-phosphate moiety.
Figure 27 is a diagram illustrating the array of glycan structures found on CH~-produced glucocerebrosidase (CerezymeTM) after sialylation.
Figure 28, comprising Figure 28A to Figure 28Z and Figure 28AA to Figure 28CC, is a list of peptides useful in the methods of the invention.
Figure 29, comprising Figures 29A to 29G, provides exemplary schemes for remodeling glycan structures on granulocyte colony stimulating factor (G-CSF).
Figure 29A
is a diagram depicting the G-CSF peptide indicating the amino acid residue to which a glycan is bonded, and an exemplary glycan formula linked thereto. Figure 29B to 29G
are diagrams of contemplated remodeling steps of the glycan of the peptide in Figure 29A
based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 30, comprising Figures 30A to 30EE sets forth exemplary schemes for remodeling glycan structures on interferon-alpha. Figure 30A is a diagram depicting the interferon-alpha isoform 14c peptide indicating the amino acid residue to which a glycan is bonded, and an exemplary glycan formula linked thereto. Figure 30B to 30D are diagrams of contemplated remodeling steps of the glycan of the peptide in Figure 30A based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 30E is a diagram depicting the interferon-alpha isoform I4c peptide indicating the amino acid residue to which a glycan is linked, and an exemplary glycan formula linked thereto. Figure 30F to 30N are diagrams of contemplated remodeling steps of the glycan of the peptide in Figure 30E based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 300 is a diagram depicting the interferon-alpha isoform 2a or 2b peptides indicating the amino acid residue to which a glycan is linked, and an exemplary glycan formula linked thereto. Figure 30P to 30W are diagrams of contemplated remodeling steps of the glycan of the peptide in Figure 30~ based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 30X is a diagram depicting the interferon-alpha-mucin fusion peptides indicating the residues) which is linked to glycans contemplated for remodeling, and exemplary glycan formulas linked thereto. Figure 30Y to 30AA are diagrams of contemplated remodeling steps of the glycan of the peptides in Figure 30~ based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 30BB is a diagram depicting the interferon-alpha-mucin fusion peptides and interferon-alpha peptides indicating the residues) which bind to glycans contemplated for remodeling, and formulas for the glycans. Figure 30CC to 30EE
are diagrams of contemplated remodeling steps of the glycan of the peptides in Figure 308B
based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 31, comprising Figures 31A to 31 S, sets forth exemplary schemes for remodeling glycan structures on interferon-beta. Figure 31A is a diagram depicting the interferon-beta peptide indicating the amino acid residue to which a glycan is linked, and an exemplary glycan formula linked thereto. Figure 31B to 31 O are diagrams of contemplated remodeling steps of the glycan of the peptide in Figure 31A based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 31P
is a diagram depicting the interferon-beta peptide indicating the amino acid residue to which a glycan is linked, and an exemplary glycan formula linked thereto. Figure 31 Q to 31 S
are diagrams of contemplated remodeling steps of the glycan of the peptide in Figure 31 P
based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 32, comprising Figures 32A to 32D, sets forth exemplary schemes for remodeling glycan structures on Factor VII and Factor VIIa. Figure 32A is a diagram depicting the Factor-VII and Factor-VIIa peptides A (solid line) and B (dotted line) indicating the residues which bind to glycans contemplated for remodeling, and the formulas for the glycans. Figure 32B to 32D are diagrams of contemplated remodeling steps of the glycan of the peptide in Figure 32A based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 33, comprising Figures 33A to 33G, sets forth exemplary schemes for remodeling glycan structures on Factor IX. Figure 33A is a diagram depicting the Factor-IX
peptide indicating residues which bind to glycans contemplated for remodeling, and formulas of the glycans. Figure 33B to 33G axe diagrams of contemplated remodeling steps of the glycan of the peptide in Figure 33A based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 34, comprising Figures 34A to 34J, sets forth exemplary schemes for remodeling glycan structures on follicle stimulating hormone (FSH), comprising o, and ~3 subunits. Figure 34A is a diagram depicting the Follicle Stimulating Hormone peptides FSHa, and FSH(3 indicating the residues which bind to glycans contemplated for remodeling, and exemplary glycan formulas linked thereto. Figure 34B to 34J are diagrams of contemplated remodeling steps of the glycan of the peptides in Figure 34A
based on the type of cell the peptides are expressed in and the desired remodeled glycan structures.
Figure 35, comprising Figures 35A to 35AA, sets forth exemplary schemes for remodeling glycan structures on Erythropoietin (EPO). Figure 35A is a diagram depicting the EPO peptide indicating the residues which bind to glycans contemplated for remodeling, and formulas for the glycans. Figure 35B to 35S are diagrams of contemplated remodeling steps of the glycan of the peptide in Figure 35A based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 35T is a diagram depicting the EPO peptide indicating the residues which bind to glycans contemplated for remodeling, and formulas for the glycans. Figure 35U to 35W are diagrams of contemplated remodeling steps of the glycan of the peptide in Figure 35T based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 35X is a diagram depicting the EPO peptide indicating the residues which bind to glycans contemplated for remodeling, and formulas for the glycans. Figure 35Y to 35AA are diagrams of contemplated remodeling steps of the glycan of the peptide in Figure 35X based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 36, comprising Figures 36A to 36I~ sets forth exemplary schemes for remodeling glycan structures on Granulocyte-Macrophage Colony Stimulating Factor (GM-CSF). Figure 36A is a diagram depicting the GM-CSF peptide indicating the residues which bind to glycans contemplated for remodeling, and formulas for the glycans.
Figure 36B to 36G are diagrams of contemplated remodeling steps of the glycan of the peptide in Figure 36A based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 36H is a diagram depicting the GM-CSF peptide indicating the residues which bind to glycans contemplated for remodeling, and formulas for the glycans. Figure 36I

to 36K are diagrams of contemplated remodeling steps of the glycan of the peptide in Figure 36H based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 37, comprising Figures 37A to 37N, sets forth exemplary schemes for remodeling glycan structures on interferon-gamma. Figure 37A is a diagram depicting an interferon-gamma peptide indicating the residues which bind to glycans contemplated for remodeling, and exemplary glycan formulas linked thereto. Figure 37B to 37G
are diagrams of contemplated remodeling steps of the peptide in Figure 37A based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 37H
is a diagram depicting an interferon-gamma peptide indicating the residues which bind to glycaus contemplated for remodeling, and exemplary glycan formulas linked thereto.
Figure 37I to 37N are diagrams of contemplated remodeling steps of the peptide in Figure 37H
based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 38, comprising Figures 38A to 38N, sets forth exemplary schemes for remodeling glycan structures on al-antitrypsin (ATT, or a-1 protease inhibitor). Figure 38A
is a diagram depicting an AAT peptide indicating the residues which bind to glycans contemplated for remodeling, and exemplary glycan formulas linked thereto.
Figure 38B to 38F are diagrams of contemplated remodeling steps of the glycan of the peptide in Figure 38A based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 38G is a diagram depicting an AAT peptide indicating the residues which bind to glycans contemplated for remodeling, and exemplary glycan formulas linked thereto.
Figure 38H to 38J are diagrams of contemplated remodeling steps of the peptide in Figure 38G based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 38K is a diagram depicting an AAT peptide indicating the residues which bind to glycans contemplated for remodeling, and exemplary glycan formulas linked thereto.
Figure 38L to 38N are diagrams of contemplated remodeling steps of the peptide in Figure 38K based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 39, comprising Figures 39A to 39J sets forth exemplary schemes for remodeling glycan structures on glucocerebrosidase. Figure 39A is a diagram depicting the glucocerebrosidase peptide indicating the residues which bind to glycans contemplated for _27_ remodeling, and exemplary glycan formulas linked thereto. Figure 39B to 39F
are diagrams of contemplated remodeling steps of the glycan of the peptide in Figure 39A
based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 39G is a diagram depicting the glucocerebrosidase peptide indicating the residues which bind to glycans contemplated for remodeling, and exemplary glycan formulas linked thereto.
Figure 39H to 39K are diagrams of contemplated remodeling steps of the glycan of the peptide in Figure 39G based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 40, comprising Figures 40A to 40W, sets forth exemplary schemes for remodeling glycan structures on Tissue-Type Plasminogen Activator (TPA).
Figure 40A is a diagram depicting the TPA peptide indicating the residues which bind to glycans contemplated fox remodeling, and formulas for the glycans. Figure 40B to 40G
are diagrams of contemplated remodeling steps of the peptide in Figure 40A based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 40H
is a diagram depicting the TPA peptide indicating the residues which bind to glycans contemplated for remodeling, and formulas for the glycans. Figure 40I to 40K are diagrams of contemplated remodeling steps of the peptide in Figure 40H based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 40L is a diagram depicting a mutant TPA peptide indicating the residues which bind to glycans contemplated for remodeling, and the formula for the glycans. Figure 40M to 40~ are diagrams of contemplated remodeling steps of the peptide in Figure 40L based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 40P
is a diagram depicting a mutant TPA peptide indicating the residues which bind to glycans contemplated for remodeling, and formulas for the glycans. Figure 40Q to 40S are diagrams of contemplated remodeling steps of the peptide in Figure 40P based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 40T
is a diagram depicting a mutant TPA peptide indicating the residues which links to glycans contemplated for remodeling, and formulas for the glycans. Figure 40U to 40W are diagrams of contemplated remodeling steps of the peptide in Figure 40T based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
_28_ Figure 41, comprising Figures 41A to 41G, sets forth exemplary schemes for remodeling glycan structures on Interleukin-2 (IL-2). Figure 41A is a diagram depicting the Interleukin-2 peptide indicating the amino acid residue to which a glycan is linked, and an exemplary glycan formula linked thereto. Figure 4~1B to 4~1G are diagrams of contemplated remodeling steps of the glycan of the peptide in Figure 41A based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 42, comprising Figures 42A to 42M, sets forth exemplary schemes for remodeling glycan structures on Factor VIII. Figure 42A are the formulas fox the glycans .
that bind to the N-linked glycosylation sites (A and A') and to the O-linked sites (B) of the Factor VIII peptides. Figure 42B to 42F are diagrams of contemplated remodeling steps of the peptides in Figure 42A based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 42G are the formulas for the glycans that bind to the N-linked glycosylation sites (A and A') and to the O-linked sites (B) of the Factor VIII
peptides. Figure 42H to 42M are diagrams of contemplated remodeling steps of the peptides in Figure 42G based on the type of cell the peptide is expressed in and the desired remodeled glycan structures.
Figure 43, comprising Figures 43A to 43M, sets forth exemplary schemes for remodeling glycan structures on urokinase. Figure 43A is a diagram depicting the urokinase peptide indicating a residue which is Linked to a glycan contemplated for remodeling, and an exemplary glycan formula linked thereto. Figure 43B to 43F are diagrams of contemplated remodeling steps of the peptide in Figure 43A based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 43G is a diagram depicting the urokinase peptide indicating a residue which is linked to a glycan contemplated for remodeling, and an exemplary glycan formula linked thereto. Figure 43H to 43L
are diagrams of contemplated remodeling steps of the peptide in Figure 43G based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 44, comprising Figures 44A to 44J, sets forth exemplary schemes for remodeling glycan structures on human DNase (hDNase). Figure 44A is a diagram depicting the human I)Nase peptide indicating the residues which bind to glycans contemplated for remodeling, and exemplary glycan formulas linked thereto. Figure 44B to 44F
are diagrams of contemplated remodeling steps of the peptide in Figure 44A based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 44G
is a diagram depicting the human L~Nase peptide indicating residues which bind to glycans contemplated for remodeling, and exemplary glycan formulas linked thereto. Figure 44H to 44J are diagrams of contemplated remodeling steps of the peptide in Figure 44~F based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 45, comprising Figures 45A to 45L, sets forth exemplary schemes for remodeling glycan structures on insulin. Figure 45A is a diagram depicting the insulin peptide mutated to contain an N glycosylation site and an exemplary glycan formula linked thereto. Figure 45B to 45D are diagrams of contemplated remodeling steps of the peptide in Figure 45A based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 45E is a diagram depicting insulin-mucin fusion peptides indicating a residues) which is linked to a glycan contemplated for remodeling, and an exemplary glycan formula linked thereto. Figure 45F to 45H are diagrams of contemplated remodeling steps of the peptide in Figure 45E based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 45I is a diagram depicting the insulin-mucin fusion peptides and insulin peptides indicating a residues) which is linked to a glycan contemplated for remodeling, and formulas for the glycan. Figure 45J to 45L
are diagrams of contemplated remodeling steps of the peptide in Figure 45I based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 46, comprising Figures 46A to 46K, sets forth exemplary schemes for remodeling glycan structures on the M-antigen (preS and S) of the Hepatitis B
surface protein (HbsAg). Figure 46A is a diagram depicting the M-antigen peptide indicating the residues which bind to glycans contemplated for remodeling, and formulas for the glycans. Figure 46B to 46G are diagrams of contemplated remodeling steps of the peptide in Figure 46A
based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 46H is a diagram depicting the M-antigen peptide indicating the residues which bind to glycans contemplated for remodeling, and formulas for the glycans. Figure 46I
to 46K are diagrams of contemplated remodeling steps of the peptide in Figure 4~6H based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 47, comprising Figures 47A to 47K, sets forth exemplary schemes for remodeling glycan structures on human growth hormone, including N, V and variants thereof. Figure 47A is a diagram depicting the human growth hormone peptide indicating a residue which is linked to a glycan contemplated for remodeling, and an exemplary glycan formula linked thereto. Figure 47B to 47D are diagrams of contemplated remodeling steps of the glycan of the peptide in Figure 47A based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 47E is a diagram depicting the three fusion peptides comprising the human growth hormone peptide and part or all of a mucin peptide, and indicating a residues) which is linked to a glycan contemplated for remodeling, and exemplary glycan formulas) linked thereto. Figure 47F to 47K are diagrams of contemplated remodeling steps of the glycan of the peptides in Figure 47E
based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 48, comprising Figures 48A to 48G, sets forth exemplary schemes for remodeling glycan structures on a TNF Receptor-IgG Fc region fusion protein (EnbrelTM).
Figure 48A is a diagram depicting a TNF Receptor--IgG Fc region fusion peptide which may be mutated to contain additional N-glycosylation sites indicating the residues which bind to glycans contemplated for remodeling, and formulas fox the glycans. The TNF
receptor peptide is depicted in bold line, and the IgG Fc regions is depicted in regular line. Figure 48B to 48G are diagrams of contemplated remodeling steps of the peptide in Figure 48A
based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 49, comprising Figures 49A to 49D, sets forth exemplary schemes for remodeling glycan structures on an anti-HER2 monoclonal antibody (HerceptinTM). Figure 49A is a diagram depicting an anti-HER2 monoclonal antibody which has been mutated to contain an N-glycosylation sites) indicating a residues) on the antibody heavy chain which is linked to a glycan contemplated for remodeling, and an exemplary glycan formula linked thereto. Figure 49B to 49D are diagrams of contemplated remodeling steps of the glyca.n of the peptides in Figure 49A based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 50, comprising Figures SOA to SOD, sets forth exemplary schemes for remodeling glycan structures on a monoclonal antibody to Protein F of Respiratory Syncytial Virus (SynagisTM). Figure SOA is a diagram depicting a monoclonal antibody to Protein F
peptide which is mutated to contain an N-glycosylation sites) indicating a residues) which is linked to a glycan contemplated for remodeling, and an exemplary glycan formula linked thereto. Figure SOB to SOD are diagrams of contemplated remodeling steps of the peptide in Figure SOA based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 51, comprising Figures S1A to S1D, sets forth exemplary schemes for remodeling glycan structures on a monoclonal antibody to TNF-a. (RemicadeTM).
Figure S 1A
is a diagram depicting a monoclonal antibody to TNF-oe. which has an N-glycosylation sites) indicating a residue which is linked to a glycan contemplated for remodeling, and an exemplary glycan formula linked thereto. Figure 51 B to 51 D are diagrams of contemplated remodeling steps of the peptide in Figure S1A based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 52, comprising Figures 52A to 52L, sets forth exemplary schemes for remodeling glycan structures on a monoclonal antibody to glycoprotein IIb/IIIa (ReoproTM).
Figure 52A is a diagram depicting a mutant monoclonal antibody to glycoprotein ITb/TIIa peptides which have been mutated to contain an N-glycosylation sites) indicating the residues) which bind to glycans contemplated for remodeling, and exemplary glycan formulas linked thereto. Figure 52B to 52D are diagrams of contemplated remodeling steps based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 52E is a diagram depicting monoclonal antibody to glycoprotein TIb/IIIa-mucin fusion peptides indicating the residues which bind to glycans contemplated for remodeling, and exemplary glycan formulas linked thereto. Figure 52F to 52H
are diagrams of contemplated remodeling steps based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 52I is a diagram depicting monoclonal antibody to glycoprotein IIb/IIIa- mucin fusion peptides and monoclonal antibody to glycoprotein ITb/TIIa peptides indicating the residues which bind to glycans contemplated for remodeling, and exemplary glycan formulas linked thereto. Figure 52J to 52L are diagrams of contemplated remodeling steps based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 53, comprising Figures 53A to 53G, sets forth exemplary schemes for remodeling glycan structures on a monoclonal antibody to CD20 (RituxanTM).
Figure 53A is a diagram depicting monoclonal antibody to CD20 which have been mutated to contain an N--32_ glycosylation sites) indicating the residue which is linked to glycans contemplated for remodeling, and exemplary glycan formulas linked thereto. Figure 53B to 53D
are diagrams of contemplated remodeling steps of the glycan of the peptides in Figure 53A
based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 53E is a diagram depicting monoclonal antibody to CD20 which has been mutated to contain an N-glycosylation sites) indicating the residues) which is linked to glycans contemplated for remodeling, and exemplary glycan formulas linked thereto. Figure 53F to 53G are diagrams of contemplated remodeling steps of the glycan of the peptides in Figure 53E based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 54, comprising Figures 54A to 540, sets forth exemplary schemes for remodeling glycan structures on anti-thrombin III (AT III). Figure 54A is a diagram depicting the anti-thrombin III peptide indicating the amino acid residues to which an N-linked glycan is linked, and an exemplary glycan formula linked thereto.
Figure 54B to 54G
are diagrams of contemplated remodeling steps of the glycan of the peptide in Figure 54A
based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 54H is a diagram depicting the anti-thrombin III peptide indicating the amino acid residues to which an N-linked glycan is linked, and an exemplary glycan formula linked thereto. Figure 54I to 54K are diagrams of contemplated remodeling steps of the glycan of the peptide in Figure 54H based on the type of cell the peptide is expressed in and the desired remodeled glycan structure. Figure 54L is a diagram depicting the anti-thrombin III peptide indicating the amino acid residues to which an N-linked glycan is linked, and an exemplary glycan formula linked thereto. Figure 54M to 540 are diagrams of contemplated remodeling steps of the glycan of the peptide in Figure 54L based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 55, comprising Figures SSA to SSJ, sets forth exemplary schemes for remodeling glycan structures on subunits cc and (3 of human Chorionic Gonadotropin (hCG).
Figure SSA is a diagram depicting the hCGa and hCG(3 peptides indicating the residues which bind to N-linked glycans (A) and O-linked glycans (B) contemplated for remodeling, and formulas for the glycans. Figure SSB to SSJ are diagrams of contemplated remodeling steps based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.

Figure 56, comprising Figures 56A to 56J, sets forth exemplary schemes for remodeling glycan structures on alpha-galactosidase (Fabra~ymeTM). Figure 56A
is a diagram depicting the alpha-galactosidase A peptide indicating the amino acid residues v~hich bind to N-linked glycans (A) contemplated for remodeling, and formulas for the glycans.
Figure 56B to 56J are diagrams of contemplated remodeling steps based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 57, comprising Figures 57A to 57J, sets fouh exemplary schemes for remodeling glycan structures on alpha-iduronidase (AldurazymeTM). Figure 57A
is a diagram depicting the alpha-iduronidase peptide indicating the amino acid residues which bind to N-linked glycans (A) contemplated for remodeling, and formulas for the glycans.
Figure 57B to 57J are diagrams of contemplated remodeling steps based on the type of cell the peptide is expressed in and the desired remodeled glycan structure.
Figure 58, comprising Figures 58A and 58B, is an exemplary nucleotide and corresponding amino acid sequence of granulocyte colony stimulating factor (G-CSF) (SEQ
ID NOS: 1 and 2, respectively).
Figure 59, comprising Figures 59A and 59B, is an exemplary nucleotide and corresponding amino acid sequence of interferon alpha (IFN-alpha) (SEQ ID NOS:

3 and 4, respectively).
Figure 60, comprising Figures 60A and 60B, is an exemplary nucleotide and corresponding amino acid sequence of interferon beta (IFN-beta) (SEQ ID NOS: 5 and 6, respectively).
Figure 61, comprising Figures 61A and 61B, is an exemplary nucleotide and corresponding amino acid sequence of Factor VIIa (SEQ ID NOS: 7 and 8, respectively).
Figure 62, comprising Figures 62A and 62B, is an exemplary nucleotide and corresponding amino acid sequence of Factor IX (SEQ ID NOS: 9 and 10, respectively).
Figure 63, comprising Figures 63A through 63D, is an exemplary nucleotide and corresponding amino acid sequence of the alpha and beta chains of follicle stimulating hormone (FSH), respectively (SEQ ID NOS: 11 through 14, respectively).
Figure 64, comprising Figures 64A and 64B, is an exemplary nucleotide and corresponding amino acid sequence of erythropoietin (EPO) (SEQ ID NOS: 15 and 16, respectively).

Figure 65 is an amino acid sequence of mature EPO, i.e. 165 amino acids (SEQ
ID
N0:73).
Figure 66, comprising Figures 66A and 66B, is an exemplary nucleotide and corresponding amino acid sequence of granulocyte-macrophage colony stimulating factor (GM-CSF) (SEQ ID NOS: 17 and 18, respectively).
Figure 67, comprising Figures 67A and 67B, is an exemplary nucleotide and corresponding amino acid sequence of interferon gamma (IFN-gamma) (SEQ ID NOS:

and 20, respectively).
Figure 68, comprising Figures 68A and 68B, is an exemplary nucleotide and corresponding amino acid sequence of a-1-protease inhibitor (A-1-PT, or a-antitrypsin) (SEQ
ID NOS: 21 and 22, respectively).
Figure 69, comprising Figures 69A-1 to 69A-2, and 69B, is an exemplary nucleotide and corresponding amino acid sequence of glucocerebrosidase (SEQ ID NOS: 23 and 24, respectively).
Figure 70, comprising Figures 70A and 70B, is an exemplary nucleotide and corresponding amino acid sequence of tissue-type plasminogen activator (TPA) (SEQ ID
NOS: 25 and 26, respectively).
Figure 71, comprising Figures 71A and 71B, is an exemplary nucleotide and corresponding amino acid sequence of Interleukin-2 (IL-2) (SEQ ID NOS: 27 and 28, respectively).
Figure 72, comprising Figures 72A-1 through 72A-4 and Figure 72B-1 through 72B-4, is an exemplary nucleotide and corresponding amino acid sequence of Factor VIII (SEQ
ID NOS: 29 and 30, respectively).
Figure 73, comprising Figures 73A and 73B, is an exemplary nucleotide and corresponding amino acid sequence of urokinase (SEQ ID NOS: 33 and 34, respectively).
Figure 74, comprising Figures 74A and 74B, is an exemplary nucleotide and corresponding amino acid sequence of human recombinant DNase (hrDNase) (SEQ ID
NOS:
39 and 40, respectively).
Figure 75, comprising Figures 75A and 75B, is an exemplary nucleotide and corresponding amino acid sequence of an insulin molecule (SEQ ID NOS: 43 and 44, respectively).

Figure 76, comprising Figures 76A and 768, is an exemplary nucleotide and corresponding amino acid sequence of S-protein from a Hepatitis B virus (HbsAg) (SEQ ID
NOS: 4~5 and 46, respectively).
Figure 77, comprising Figures 77A and 778, is an exemplary nucleotide and corresponding amino acid sequence of human growth hormone (hGH) (SEQ ID NOS:
47 and 48, respectively).
Figure 78, comprising Figuxes 78A and 78D, are exemplary nucleotide and corresponding amino acid sequences of anti-thrombin III. Figures 78A and 78B, are an exemplary nucleotide and corresponding amino acid sequences of "WT" anti-thrombin III
(SEQ ID NOS: 63 and 64, respectively).
Figure 79, comprising Figures 79A to 79D, are exemplary nucleotide and corresponding amino acid sequences of human chorionic gonadotropin (hCG) a and (3 subunits. Figures 79A and 79B are an exemplary nucleotide and corresponding amino acid sequence of the a-subunit of human chorionic gonadotropin (SEQ ID NOS: 69 and 70, respectively). Figures 79C and 79D are an exemplary nucleotide and corresponding amino acid sequence of the beta subunit of human chorionic gonadotrophin (SEQ ID
NOS: 71 and 72, respectively).
Figure 80, comprising Figures 80A and 80B, is an exemplary nucleotide and corresponding amino acid sequence of a-iduronidase (SEQ ID NOS: 65 and 66, respectively).
Figure 81, comprising Figures 81A and 81B, is an exemplary nucleotide and corresponding amino acid sequence of a-galactosidase A (SEQ ID NOS: 67 and 68, respectively).
Figure 82, comprising Figures 82A and 82B, is an exemplary nucleotide and corresponding amino acid sequence of the 75 kDa tumor necrosis factor receptor (TNF-R), which comprises a portion of EnbrelTM (tumor necrosis factor receptor (TNF-R)/IgG fusion) (SEQ ID NOS: 31 and 32, respectively).
Figure 83, comprising Figures 83A and 838, is an exemplary amino acid sequence of the light and heavy chains, respectively, of HerceptinTM (monoclonal antibody (MAb) to Her-2, human epidermal growth factor receptor) (SEQ ID NOS: 35 and 36, respectively).

Figure 84, comprising Figures 84A and 84B, is an exemplary amino acid sequence the heavy and light chains, respectively, of SynagisTM (MAb to F peptide of Respiratory Syncytial Virus) (SEQ ID NOS: 37 and 38, respectively).
Figure 8S, comprising Figures 8SA and BSB, is an exemplary nucleotide and S corresponding amino acid sequence of the non-human vaxiable regions of RemicadeTM (MAb to TNFoc) (SEQ ID NOS: 41 and 42, respectively).
Figure 86, comprising Figures 86A and 86B, is an exemplary nucleotide and corresponding amino acid sequence of the Fc portion of human IgG (SEQ ID NOS:
49 and S0, respectively).
Figure 87 is an exemplary amino acid sequence of the mature variable region light chain of an anti-glycoprotein IIb/IIIa marine antibody (SEQ ID NO: S2).
Figure 88 is an exemplary amino acid sequence of the mature variable region heavy chain of an anti-glycoprotein IIb/IIIa marine antibody (SEQ ID NO: S4).
Figure 89 is an exemplary amino acid sequence of variable region light chain of a 1 S human IgG (SEQ ID NO: S 1 ).
Figure 90 is an exemplary amino acid sequence of variable region heavy chain of a human IgG (SEQ ID NO:S3).
Figure 91 is an exemplary amino acid sequence of a light chain of a human IgG
(SEQ
ID NO:SS).
Figure 92 is an exemplary amino acid sequence of a heavy chain of a human IgG
(SEQ ID NO:S6).
Figure 93, comprising Figures 93A and 93B, is an exemplary nucleotide and corresponding amino acid sequence of the mature variable region of the light chain of an anti-CD20 marine antibody (SEQ ID NOS: S9 and 60, respectively).
2S Figure 94, comprising Figures 94A and 94B, is an exemplary nucleotide a.nd corresponding amino acid sequence of the mature variable region of the heavy chain of an anti-CD20 marine antibody (SEQ ID NOS: 61 and 62, respectively).
Figure 9S, comprising Figures 9SA through 9SE, is the nucleotide sequence of the tandem chimeric antibody expression vector TCAE 8 (SEQ ID NO:S7).

Figure 96, comprising Figures 96A through 96E, is the nucleotide sequence of the tanderu chimeric antibody expression vector TCAE 8 containing the light and heavy variable domains of the anti-CD20 marine antibody (SEQ ID N~:58).
Figure 97, comprising Figures 97A to 97C, are graphs depicting 2-AA I~PLC
analysis of glycans released by PNGaseF from myeloma-expressed Cri-IgG 1 antibody. The structure of the glycans is determined by retention time: the GO glycoform elutes at 30 min., the G1 glycoform elutes at ~ 33 min., the G2 glycoform elutes at about approximately 37 min. and the S 1-G1 glycoform elutes at ~ 70 min. Figuxe 97A depicts the analysis of the DEAF
antibody sample. Figure 97B depicts the analysis of the SPA antibody sample.
Figure 97C
depicts the analysis of the Fc antibody sample. The percent area under the peaks for these graphs is surmnarized in Table 14.
Figure 98, comprising Figures 98A to 98C, are graphs depicting the. MALDI
analysis of glycans released by PNGaseF from myeloma-expressed Cri-IgGl antibody. The glycans were derivatized with 2-AA and then analyzed by MALDI. Figure 98A depicts the analysis of the DEAF antibody sample. Figure 98B depicts the analysis of the SPA
antibody sample.
Figure 98C depicts the analysis of the Fc antibody sample.
Figure 99, comprising Figures 99A to 99D, are graphs depicting the capillary electrophoresis analysis of glycans released from Cri-IgGl antibodies that have been glycoremodeled to contain M3N2 glycoforms. A graph depicting the capillary electrophoresis analysis of glycan standards derivatized with APTS is shown in Figure 99A.
Figure 99B depicts the analysis of the DEAE antibody sample. Figure 99C
depicts the analysis of the SPA antibody sample. Figure 99D depicts the analysis of the Fc antibody sample. The percent area under the peaks for these graphs is summarized in Table 15.
Figure 100, comprising Figuxes 100A to 100D, are graphs depicting the capillary electrophoresis analysis of glycans released from Cri-IgGl antibodies that have been glycoremodeled to contain GO glycoforms. A graph depicting the capillary electrophoresis analysis of glycan standards derivatized with APTS is shown in Figure 100A.
Figure 100B
depicts the analysis of the DEAF antibody sample. Figure 100C depicts the analysis of the SPA antibody sample. Figure 100D depicts the analysis of the Fc antibody sample. The percent area under the peaks for these graphs is summarized in Table 16.

Figure 101, comprising Figures lOlA to lOlC, are graphs depicting 2-AA HPLC
analysis of glycans released from Cri-IgGl antibodies that have been glycoremodeled to contain GO glycoforms. The released glycans were labeled with 2AA and separated by HPLC on a NH2P-50 4~D amino column. Figure 1 OlA depicts the analysis of the DEAF
antibody sample. Figure lOlB depicts the analysis of the SPA antibody sample.
Figure 101 C depicts the analysis of the Fc antibody sample. The percent area under the peaks for these graphs is summarized in Table 16 Figure 102, comprising Figures 102A to 102C, are graphs depicting the MALDI
analysis of glycans released from Cri-IgGl antibodies that have been glycoremodeled to contain GO glycoforms. The released glycans were derivatized with 2-AA and then analyzed by MALDI. Figure 102A depicts the analysis of the DEAE antibody sample. Figure depicts the analysis of the SPA antibody sample. Figure 102C depicts the analysis of the Fc antibody sample.
Figure 103, comprising Figures 103A to 103D, are graphs depicting the capillary electrophoresis analysis of glycans released from Cri-IgGl antibodies that have been glycoremodeled to contain G2 glycoforms. A graph depicting the capillary electrophoresis analysis of glycan standards derivatized with APTS is shown in Figure 103A.
Figure 103B
depicts the analysis of the DEAE antibody sample. Figure 103C depicts the analysis of the SPA antibody sample. Figure 103D depicts the analysis of the Fc antibody sample. The percent area under the peaks for these graphs is summarized in Table 17.
Figure 104, comprising Figures 104A to I04C, are graphs depicting the 2-AA
HPLC
analysis of glycans released from remodeled Cri-IgGl antibodies that have been glycoremodeled to contain G2 glycoforms. The released glycans were labeled with 2AA and then separated by HPLC on a NH2P-50 4D amino column. Figure 104A depicts the analysis of the DEAF antibody sample. Figure 104B depicts the analysis of the SPA
antibody sample.
Figure 104C depicts the analysis of the Fc antibody sample. The percent area under the peaks for these graphs is sunnnarized in Table 17.
Figure 105, comprising Figures lOSA to lOSC, are graphs depicting MALDI
analysis of glycans released frown Cri-IgGl antibodies that have been glycoremodeled to contain G2 glycoforms. The released glycans were derivatized with 2-AA and then analyzed by IVIALDI. Figure l OSA depicts the analysis of the DEAF antibody sample. Figure l OSB

depicts the analysis of the SPA antibody sample. Figure lOSC depicts the analysis of the Fc antibody sample.
Figure 106, comprising Figures 106A to 106D, are graphs depicting capillary electrophoresis analysis of glycans released from Cri-IgGl antibodies that have been glycoremodeled by GnT-I treatment of M3N2 glycoforms. A graph depicting the capillary electrophoresis analysis of glycan standards derivatized with APTS is shown in Figure 106A.
Figure 106B depicts the analysis of the DEAF antibody sample. Figure 1060 depicts the analysis of the SPA antibody sample. Figure 106D depicts the analysis of the Fc antibody sample.
Figure 107, comprising Figures 107A to 1070, are graphs depicting 2-AA HPLC
analysis of glycans released from Cri-IgGl antibodies that have been remodeled by GnT-I
treatment of M3N2 glycoforms. The released glycans were labeled with 2-AA and separated by HPLC on a NH2P-50 4D amino column. Figure 107A depicts the analysis of the DEAE
antibody sample. Figure 107B depicts the analysis of the SPA antibody sample.
Figure 107C depicts the analysis of the Fc antibody sample.
Figure 108, comprising Figures 108A to 108C, are graphs depicting MALDI
analysis of glycans released from Cri-IgGl antibodies that have been glycoremodeled by GnT-I
treatment of M3N2 glycoforms. The released glycans were derivatized with 2-AA
and then analyzed by MALDI. Figure 108A depicts the analysis of the DEAE antibody sample.
Figure 108B depicts the analysis of the SPA antibody sample. Figure 1080 depicts the analysis of the Fc antibody sample.
Figure 109, comprising Figures 109A to 109D, are graphs depicting capillary electrophoresis of glycans released from Cri-IgGl antibodies that have been glycoremodeled by GnT-I, II and III treatment of M3N2 glycoforms. A graph depicting the capillary electrophoresis analysis of glycan standards derivatized with APTS is shown in Figure 109A.
Figure 109B depicts the analysis of the DEAE antibody sample. Figure 1090 depicts the analysis of the SPA antibody sample. Figure 109D depicts the analysis of the Fc antibody sample. The percent area under the peaks for these graphs is summarized in Table 18.
Figure 110, comprising Figures 110A to 1 l OC, are graphs depicting 2-AA HPLC
analysis of glycans released from Cri-IgGl antibodies that have been glycoremodeled by GnT-I, II and III treatment of M3N2 glycoforms. The released glycans were labeled with 2AA and then separated by HPLC on a NH2P-50 4D amino column. Figure 11 OA
depicts the analysis of the DEAF antibody sample. Figure 1 l OB depicts the analysis of the SPA
antibody sample. Figure 1100 depicts the analysis of the Fc antibody sample.
The percent area under the peaks for these graphs is summarized in Table 1 ~.
Figure 111, comprising Figures 111 A to 111 C, are graphs depicting MALDI
analysis of glycans released from Cri-IgGl antibodies that have been glycoremodeled by galactosyltransferase treatment of NGA2F glycoforms. The released glycans were derivatized with 2-AA and then analyzed by MALDI. Figure 111A depicts the analysis of the DEAE antibody sample. Figure 111 B depicts the analysis of the SPA
antibody sample.
Figure 111 C depicts the analysis of the Fc antibody sample.
Figure 112, comprising 112A to 112D, are graphs depicting 2-AA HPLC analysis of glycans released from Cri-IgGl antibodies containing NGA2F isoforms before GalTl treatment (Figures 112A and 112C) and after GalT1 treatment (Figures 112B and 112D).
Figures 112A and 112B depict the analysis of the DEAF sample of antibodies.
Figures 1120 and 112D depict the analysis of the Fc sample of antibodies. The released glycans were labeled with 2AA and separated by HPLC on a NH2P-50 4D amino column.
Figure 113, comprising 113A to 113C, are graphs depicting 2-AA HPLC analysis of glycans released from Cri-IgGl antibodies that have been glycoremodeled by ST3Ga13 treatment of G2 glycoforms. The released glycans are labeled with 2-AA and then separated by HPLC on a NH2P-50 4D amino column. Figure 113A depicts the analysis of the DEAF
antibody sample. Figure 113B depicts the analysis of the SPA antibody sample.
Figure 1130 depicts the analysis of the Fc antibody sample. The percent area under the peaks for these graphs is summarized in Table 19.
Figure 114, comprising Figures 114A to 1140, are graphs depicting MALDI
analysis of glycans released from Cri-IgGl antibodies that had been glycoremodeled by ST3Gal3 treatment of G2 glycoforms. The released glycans were derivatized with 2-AA
and then analyzed by MALDI. Figure 114A depicts the analysis of the DEAE antibody sample.
Figure 114B depicts the analysis of the SPA antibody sample. Figure 1140 depicts the analysis of the Fc antibody sample.
Figure 115, comprising Figures 115A to 115D, are graphs depicting capillary electrophoresis analysis of glycans released from Cri-IgGl antibodies that had been glycoremodeled by ST6Ga11 treatment of G2 glycoforms. A graph depicting the capillary elecfirophoresis analysis of glycan standards derivatized with APTS is shown in Figure 115A.
Figure 115B depicts the analysis of the DEAE antibody sample. Figure 1150 depicts the analysis of the SPA antibody sample. Figure 11 SD depicts the analysis of the Fc antibody sample.
Figure 116, comprising Figures 116A to 1160, are graphs depicting 2-AA HPLC
analysis of glycans released from Cri-IgGl antibodies that had been glycoremodeled by ST6Gall treatment of G2 glycoforms. The released glycans were labeled with 2-AA and separated by HPLC on a NH2P-50 4D amino column. Figure 116A depicts the analysis of the DEAE antibody sample. Figure 116B depicts the analysis of the SPA antibody sample.
Figure 116C depicts the analysis of the Fc antibody sample.
Figure 117, comprising Figures 117A to 117C, are graphs depicting MALDI
analysis of glycans released from Cri-IgGl antibodies that had been glycoremodeled by ST6Gal1 treatment of G2 glycoforms. The released glycans were derivatized with 2-AA
and then analyzed by MALDI. Figure 117A depicts the analysis of the DEAE antibody sample.
Figure 117B depicts the analysis of the SPA antibody sample. Figure 117C
depicts the analysis of the Fc antibody sample.
Figure 118, comprising Figures 118A to 118E, depicts images of SDS-PAGE
analysis of the glycoremodeled of Cri-IgGl antibodies with different glycoforms under non-reducing conditions. Bovine serum albumin (BSA) was run under reducing conditions as a quantitative standard. Protein molecular weight standards are displayed and their size is indicated in kDa. Figure 118A depicts SDS-PAGE analysis of the DEAE, SPA and Fc Cri-IgGl antibodies glycoremodeled to contain GO and G2 glycoforms. Figure 118B
depicts SDS-PAGE analysis of the DEAE, SPA and Fc Cri-IgGl antibodies glycoremodeled to contain NGA2F (bisecting) and GnT-I-M3N2 (GnTl) glycoforms. Figure 1180 depicts SDS-PAGE analysis of the DEAE, SPA and Fc Cri-IgGl antibodies glycoremodeled to contain S2G2 (ST6Ga11) glycoforms. Figure 118D depicts SDS-PAGE analysis of the DEAE, SPA and Fc Cri-IgGl antibodies glycoremodeled to contain M3N2 glycoforms, and BSA. Figure 118E depicts SDS-PAGE analysis of the DEAE, SPA and Fc Cri-IgGl antibodies glycoremodeled to contain Gal-NGA2F (Gal-bisecting) glycoforms, and BSA.
Figure 119 is an image of an acrylamide gel depicting the results of FACE
analysis of the pre- and post-sialylation of TP10. The BiNAo species has no sialic acid residues. The BiNAI species has one sialic acid residue. The BiNA2 species has two sialic acid residues. Bi = biantennary; NA = neuraminic acid.
Figure 120 is a graph depicting the plasma concentration in ~,g/ml over time of pre-and post-sialylation TP10 injected into rats.
Figure 121 is a graph depicting the area under the plasma concentration-time curve (AUC) in ~.g/hr/ml for pre- and post sialylated TP10.
Figure 122 is an image of an acrylamide gel depicting the results of FACE
glycan analysis of the pre- and post-fucosylation of TP10 and FACE glycan analysis of CHO cell produced TP-20. The BiNA2F2 species has two neuraminic acid (NA) residues and two fucose residues (F).
Figure 123 is a graph depicting the in vitro binding of TP20 (sCRIsLeX) glycosylated in vitro (diamonds) and in vivo in Lecl l CHO cells (squares).
Figure 124 is a graph depicting the analysis by 2-AA HPLC of glycoforms from the GIcNAc-ylation of EPO.
Figure 125, comprising Figures 125A and 125B, are graphs depicting the 2-AA
HPLC
analysis of two lots of EPO to which N-acetylglucosamine was been added.
Figure 125A
depicts the analysis of lot A, and Figure 125B depicts the analysis of lot B.
Figure 126 is a graph depicting the 2-AA HPLC analysis of the products the reaction introducing a third glycan branch to EPO with GnT-V.
Figure 127 is a graph depicting a MALDI-TOF spectrum of the glycans of the EPO
preparation after treatment with GnT-I, GnT-II, GnT-III, GnT-V and GalTl, with appropriate donor groups.
Figure 128 is a graph depicting a MALDI spectrum the glycans of native EPO.
Figure 129 is an image of an SDS-PAGE gel of the products of the PEGylation reactions using CMP-SA-PEG (1 kDa), and CMP-SA-PEG (10 kDa).
Figure 130 is a graph depicting the results of the in vitro bioassay of PEGylated EPO.
Diamonds represent the data from sialylated EPO having no PEG molecules.
Squares represent the data obtained using EPO with PEG (1 kDa). Triangles represent the data obtained using EPO with PEG (10 kDa).

Figure 131 is a diagram of CHO-expressed EPO. The EPO polypeptide is 165 amino acids in length, with a molecular weiglat of 18 kDa without glycosylation. The glycosylated forms of EPO produced in CHO cells have a molecular weight of about 33 kDa to 39 kDa.
The shapes which represent the sugars in the glycan chains are identified in the box at the lower edge of the drawing.
Figure 132 is a diagram of insect cell expressed EPO. The shapes that represent the sugars in the glycan chains are identified in the box at the lower edge of FIG. 131.
Figure 133 is a bar graph depicting the molecular weights of the EPO peptides expressed in insect cells which were remodeled to form complete mono-, bi- and tri-antermary glycans, with optional glycoPEGylation with 1 kDa, 10 kDa or 20 kDa PEG.
EpoetinTM is EPO expressed in mammalian cells without further glycan modification or PEGylation. NESP (AranespTM, Amgen, Thousand Oaks, CA) is a form of EPO having linked glycan sites that is also expressed in mammalian cells without further glycan modification or PEGylation.
Figure 134, comprising Figures 134A and 134B, depicts one scheme for the remodeling and glycoPEGylation of insect cell expressed EPO. Figure 134A
depicts the remodeling and glycoPEGylation steps that remodel the insect expressed glycan to a mono-antermary glycoPEGylated glycan. Figure 134B depicts the remodeled EPO
polypeptide having a completed glycoPEGylated mono-antennary glycan at each N-linked glycan site of the polypeptide. The shapes that represent the sugars in the glycan chains are identified in the box at the lower edge of FIG. 131, except that the triangle represents sialic acid.
Figure 135 is a graph depicting the i~ vitro bioactivities of EPO-SA and EPO-SA-PEG constructs. The in vitro assay measured the proliferation of TF-1 erythroleukemia cells which were maintained for 48 hr in RBMI + FBS 10% + GM-CSF (12 ng/ml) after the EPO
construct was added at 10.0, 5.0, 2.0, 1.0, 0.5, and 0 ~.g/ml. Tri-SA refers to EPO constructs where the glycans are tri-antennary and have SA. Tri-SA 1K PEG refers to EPO
constructs where the glycans are tri-antennary and have Gal and are then glycoPEGylated with SA-PEG
1 kDa. Di-SA l OK PEG refers to EPO constructs where the glycans are bi-antennary and have Gal and are then glycoPEGylated with SA-PEG 10 kDa. Di-SA 1K PEG refers to EPO
constructs where the glycans are bi-antennary and have Gal and are then glycoPEGylated with SA-PEG 1 kDa. Di-SA refers to EPO constructs where the glycans are bi-antermary and are built out to SA. EpogenTM is EPO expressed in CHO cells with no further glycan modification.
Figure 136 is a graph depicting the pharmacokinetics of the EPO constructs in rat.
Rats were bolus injected with (has]-labeled glycoPEGylated and non-glycoPEGylated EPO.
The graph shows the concentration of the radio-labeled EPO in the bloodstream of the rat at 0 to about 72 minutes after injection. "Biant-lOK" refers to EPO with biantennary glycan structures with terminal 10 kDa PEG moieties. "Mono-20K" refers to EPO with monoantennary glycan structures with terminal 20 kDa PEG moieties. NESP refers to the commercially available Aranesp. "Biant-1K" refers to EPO with biantennary glycan 10. structures with terminal 1 kDa PEG moieties. "Biant-SA" refers to EPO with biantennary glycan structures with terminal 1 kDa moieties. The concentration of the EPO
constructs in the bloodstream at 72 hr. is as follows: Biant-lOK, 5.1 cpm/ml; Mono-20K, 3.2 cpm/ml;
NESP, 1 cpm/ml; and Biant-1K, 0.2 cpm/ml; Biant-SA, 0.1 cpm/ml. The relative area under the curve of the EPO constructs is as follows: Biant-lOK, 2.9; Mono-20K, 2.1;
NESP, 1;
Biant-1K, 0.5; and Biant-SA, 0.2.
Figure 137 is a bar graph depicting the ability of the EPO constructs to stimulate reticulocytosis ivy vivo. Each treatment group is composed of eight mice. Mice were given a single subcutaneous injection of 10 ~g protein / kg body weight. The percent reticulocytosis was measured at 96 hr. Tri-antennary-SA2,3(6) construct has the SA molecule bonded in a 2,3 or 2,6 linkage (see, Example 18 herein for preparation) wherein the glycan on EPO is tri-antennary N-glycans with SA-PEG 10 K is attached thereon. Similarly, bi-antennary-l OK
PEG is EPO having a bi-antennary N-glycan with SA-PEG at 10 K PEG attached thereon.
Figure 138 is a bar graph depicting the ability of EPO constructs to increase the hematocrit of the blood of mice in vivo. CD-1 female mice were injected i.p.
with 2.5 ~.g protein/kg body weight. The hematocrit of the mice was measured on day 15 after the EPO
injection. Bi-lk refers to EPO constructs where the glycans are bi-antennary and are built out to the Gal and then glycoPEGylated with SA-PEG 1 kDa. Mono-20k refers to EPO
constructs where the glycans are mono-antennary and are built out to the Gal and then glycoPEGylated with SA-PEG 20 kDa.
Figure 139, comprising Figures 139A and 139B, depicts the analysis of glycans enzymatically released from EPO expressed in insect cells (Protein Sciences, Lot # 060302).

Figure 139A depicts the HPLC analysis of the released glycans. Figure 139B
depicts the MALDI analysis of the released glycans. Diamonds represent fucose, and squares represent GIcNAc, circles represent mannose.
Figure 140 depicts the MALDI analysis of glycans released from EPO after the GnT-I/GaIT-1 reaction. The structures of the glycans have been determined by comparison of the peak spectrum with that of standard glycans. The glycan structures are depicted beside the peaks. Diamonds represent fucose, and squares represent GIcNAc, circles represent mannose, stars represent galactose.
Figure 141 depicts the SDS-PAGE analysis of EPO after the GnT-I/GaIT-1 reaction, Superdex 75 purification, ST3Ga13 reaction with SA-PEG (10 kDa) and SA-PEG (20 kDa).
Figure 142 depicts the results of the TF-1 cell in vitro bioassay of PEGylated mono-antennary EPO.
Figure 143, comprising Figures 143A and 143B, depicts the analysis of glycan released from EPO after the GnT-I/GnT-II reaction. Figure 143A depicts the HPLC analysis of the released glycans, where peak 3 represents the bi-antennary GIcNAc glycan. Figure 143B depicts the MALDI analysis of the released glycans. The structures of the glycans have been determined by comparison of the peak spectrum with that of standard glycans. The glycan structures are depicted beside the peaks. Diamonds represent fucose, and squares represent GIcNAc, circles represent mannose.
Figure 144, comprising Figures 144A and 144B, depict the HPLC analysis of glycans released from EPO after the GaIT-1 reaction. Figure 144A depicts the glycans released after the small scale GaIT-1 reaction. Figure 144B depicts the glycans released after the large scale GaIT-1 reaction. In both figures, Peak 1 is the bi-antennary glycan with terminal galactose moieties and Peak 2 is the bi-antennary glycan without terminal galactose moieties.
Figure 145 depicts the Superdex 75 chromatography separation of EPO species after the GaIT-1 reaction. Peak 2 contains EPO with bi-antennary glycans with terminal galactose moieties.
Figure 146 depicts the SDS-PAGE analysis of each of the products of the glycoremodeling process to make bi-antennary glycans with terminal galactose moieties.
Figure 147 depicts the SDS-PAGE analysis of EPO after ST3Ga13 sialylation or PEGylation with SA-PEG (1 kDa) or SA-PEG (10 kDa).

Figure 148 depicts the HPLC analysis of glycams released from EPO after the GnT-I/GnT-II reaction. The structures of the glycans have been determined by comparison of the peak retention with that of standard glycans. The glycan structures are depicted beside the peaks. Diamonds represent fucose, and squares represent GIcNAc, circles represent mannose.
Figure 149 depicts the HPLC analysis of glycans released from EPO after the GnT-V
reaction. The structures of the glycans have been determined by comparison of the peak retention with that of standard glycans. The glycan structures are depicted beside the peaks.
Diamonds represent fucose, and squares represent GIcNAc, circles represent mannose.
Figure 150 depicts the HPLC analysis of glycans released from EPO after the GaIT-1 reaction. The structures of the glycans have been determined by comparison of the peak retention with that of standard glycans. The glycan structures are depicted beside the peaks.
Diamonds represent fucose, and squares represent GIcNAc, circles represent mannose, open circles represent galactose and triangles represent sialic acid.
Figure 151 depicts the HPLC analysis of glycans released from EPO after the ST3Ga13 reaction. The structures of the glycans have been determined by comparison of the peak retention with that of standard glycans. The glycan structures axe depicted beside the peaks. Diamonds represent fucose, and squares represent GIcNAc, circles represent mannose, open circles represent galactose and triangles represent sialic acid.
Figure 152 depicts the HPLC analysis of glycans released from EPO after the ST6Ga11 reaction. The structures of the glycans have been determined by comparison of the peak retention with that of standard glycans. The glycan structures are depicted beside the peaks.
Figure 153 depicts the results of the TF-1 cells ih vit~~o bioassay of EPO
with bi-antennary and triantennary glycans. "Di-SA" refers to EPO with bi-antennary glycans that terminate in sialic acid. "Di-SA 1 OK PEG" refers to EPO with bi-antennary glycans that terminate in sialic acid derivatized with PEG (10 kDa). "Di-SA 1K PEG" refers to EPO with bi-antennary glycans that terminate in sialic acid derivatized with PEG (1 kDa). "Tri-SA ST6 + ST3" refers to EPO with tri-antennary glycans terminating in 2,6-SA capped with 2,3-SA.
"Tri-SA ST3" refers to EPO with tri-antennary glycans terminating in 2,3-SA.

Figure 154 is an image of an IEF gel depicting the pI of the products of the desialylation procedure. Lanes 1 and 5 are IEF standards. Lane 2 is Factor IX
protein. Lane 3 is rFactor IX protein. Lane 4 is the desialylation reaction of rFactor IX
protein at 20 hr.
Figure 155 is an image of an SDS-PAGE gel depicting the molecular weight of Factor IX conjugated with either SA-PEG (1 kDa) or SA-PEG (10 kDa) after reaction with CMP-SA-PEG. Lanes 1 and 6 are SeeBlue +2 molecular weight standards. Lane 2 is rF-IX. Lane 3 is desialylated rF-IX. Lane 4 is rFactor IX conjugated to SA-PEG (1 kDa).
Lane 5 is rFactor IX conjugated to SA-PEG (10 kDa).
Figure 156 is an image of an SDS-PAGE gel depicting the reaction products of direct-sialylation of Factor-IX and sialic acid capping of Factor-IX-SA-PEG. Lane 1 is protein standards, lane 2 is blank; lane 3 is rFactor-IX; lane 4 is SA capped rFactor-IX-SA-PEG (10 kDa); lane 5 is rFactor-IX-SA-PEG (10 kDa); lane 6 is ST3Ga11; lane 7 is ST3Gal3; lanes 8, 9, 10 are rFactor-IX-SA-PEG(10 kDa) with no prior sialidase treatment.
Figure 157 is an image of an isoelectric focusing gel (pH 3-7) of asialo-Factor VIIa.
Lane 1 is rFactor VIIa; lanes 2-5 are asialo-Factor VIIa.
Figure 158 is a graph of a MALDI spectra of Factor VIIa.
Figure 159 is a graph of a MALDI spectra of Factor VIIa-PEG (1 kDa).
Figure 160 is a graph depicting a MALDI spectra of Factor VIIa-PEG (10 kDa).
Figure 161 is an image of an SDS-PAGE gel of PEGylated Factor VIIa. Lane 1 is asialo-Factor VIIa. Lane 2 is the product of the reaction of asialo-Factor VIIa and CMP-SA-PEG(1 kDa) with ST3Gal3 after 48 hr. Lane 3 is the product of the reaction of asialo-Factor VIIa and CMP-SA-PEG (1 kDa) with ST3Gal3 after 48 hr. Lane 4 is the product of the reaction of asialo-Factor VIIa and CMP-SA-PEG (10 kDa) with ST3Ga13 at 96 hr.
Figure 162 is an image of an isoelectric focusing (IEF) gel depicting the products of the desialylation reaction of human pituitary FSH. Lanes 1 and 4 axe isoelectric focusing (IEF) standards. Lane 2 is native FSH. Lane 3 is desialylated FSH.
Figure 163 is an image of an SDS-PAGE gel of the products of the reactions to make PEG-sialylation of rFSH. Lanes 1 and 8 are SeeBlue+2 molecular weight standards. Lane 2 is 15 pig of native FSH. Lane 3 is 15 ~g of asialo-FSH (AS-FSH). Lane 4 is 15 ~.g of the products of the reaction of AS-FSH with CMP-SA. Lane 5 is 15 ~g of the products of the reaction of AS-FSH with CMP-SA-PEG (1 kDa). Lane 6 is 15 ~.g of the products of the reaction of AS-FSH with CMP-SA-PEG (5 kDa). Lane 7 is 15 ~.g of the products of the reaction of AS-FSH with CMP-SA-PEG (10 kDa).
Figure 164 is an image of an isoelectric focusing gel of the products of the reactions to make PEG-sialylation of FSH. Lanes 1 and 8 are IEF standards. Lane 2 is 15 ~g of native FSH. Lane 3 is 15 ~,g of asialo-FSH (AS-FSH). Lane 4 is 15 ~,g of the products of the reaction of AS-FSH with CMP-SA. Lane 5 is 15 ~g of the products of the reaction of AS-FSH with CMP-SA-PEG (1 kDa). Lane 6 is 15 ~,g of the products of the reaction of AS-FSH
with CMP-SA-PEG (5 kDa). Lane 7 is 15 ~,g of the products of the reaction of AS-FSH with CMP-SA-PEG (10 kDa).
Figure 165 is an image of an SDS-PAGE gel of native non-recombinant FSH
produced in human pituitary cells. Lanes 1, 2 and 5 are SeeBlueTM+2 molecular weight standards. Lanes 3 and 4 are native FSH at 5 ~g and 25 ~,g, respectively.
Figure 166 is an image of an isoelectric focusing gel (pH 3-7) depicting the products of the asialylation reaction of rFSH. Lanes 1 and 4 are IEF standards. Lane 2 is native rFSH.
Lane 3 is asialo-rFSH.
Figure 167 is an image of an SDS-PAGE gel depicting the results of the PEG-sialylation of asialo-rFSH. Lane 1 is native rFSH. Lane 2 is asialo-FSH. Lane 3 is the products of the reaction of asialo-FSH and CMP-SA. Lanes 4-7 are the products of the reaction between asialoFSH and 0.5 mM CMP-SA-PEG (10 kDa) at 2 hr, 5 hr, 24 hr, and 48 hr, respectively. Lane 8 is the products of the reaction between asialo-FSH
and 1.0 mM
CMP-SA-PEG (10 kDa) at 48 hr. Lane 9 is the products of the reaction between asialo-FSH
and 1.0 mM CMP-SA-PEG (1 kDa) at 48 hr.
Figure 168 is an image of an isoelectric focusing gel showing the products of PEG-sialylation of asialo-rFSH with a CMP-SA-PEG (1 kDa). Lane 1 is native rFSH.
Lane 2 is asialo-rFSH. Lane 3 is the products of the reaction of asialo-rFSH and CMP-SA
at 24 hr.
Lanes 4-7 are the products of the reaction of asialo-rFSH and 0.5 mM CMP-SA-PEG (1 kDa) at 2 hr, 5 hr, 24 hr, and 48 hr, respectively. Lane 8 is blank. Lanes 9 and 10 axe the products of the reaction at 48 hr of asialo-rFSH and CMP-SA-PEG (10 kDa) at 0.5 mM and 1.0 mM, respectively.
Figure 169 is graph of the pharmacokinetics of rFSH and rFSH-SA-PEG (1 kDa and 10 kDa). This graph illustrates the relationship between the time a rFSH
compound is in the blood stream of the rat, and the mean concentration of the rFSH compound in the blood for glycoPEGylated rFSH as compared to non-PEGylated rFSH.
Figure 170 is a graph of the results of the FSH bioassay using Sertoli cells.
This graph illustrates the relationship between the FSH concentration in the Sertoli cell incubation medium and the amount of 17-(3 estradiol released from the Sertoli cells.
Figure 171 is a graph depicting the results of the Steelman-Pohley bioassay of glycoPEGylated and non-glycoPEGylated FSH. Rats were subcutaneously injected with human chorionic gonadotropin and varying amounts of FSH for three days, and the average ovarian weight of the treatment group determined on day 4. rFSH-SA-PEG refers to recombinant FSH that has been glycoPEGylated with PEG (1 kDa). rFSH refers to non-glycoPEGylated FSH. Each treatment group contains 10 rats.
Figure 172, comprising Figures 172A and 172B, depicts the chromatogram of INF-(3 elution from a Superdex-75 column. Figure 172A depicts the entire chromatogram. Figure 172B depicts the boxed area of Figure 172A containing peaks 4 and 5 in greater detail.
Figure 173, comprising Figures 173A and 173B, depict MALDI analysis of glycans enzymatically released from INF-(3. Figure 173A depicts the MALDI analysis glycans released from native INF-(3. Figure 173B depicts the MALDI analysis of glycans released from desialylated INF-(3. The structures of the glycans have been determined by comparison of the peak spectrum with that of standard glycans. The glycan structures are depicted beside the peaks. Squares represent GIcNAc, triangles represent fucose, circles represent mannose, diamonds represent galactose and stars represent sialic acid.
Figure 174 depicts the lectin blot analysis of the sialylation of the desialylated INF-(3.
The blot on the right side is detected with Maackia amure~csis agglutinin (MAA) labeled with digoxogenin (DIG) (Roche Applied Science, Indianapolis, IL) to detect a,2,3-sialylation. The blot on the left is detected with Erthr~ina cf~istagalli lectin (ECL) labeled with biotin (Vector Laboratories, Burlingame, CA) to detect exposed galactose residues.
Figure 175 depicts the SDS-PAGE analysis of the products of the PEG (10 kDa) PEGylation reaction of INF-(3. "-PEG" refers to INF-(3 before the PEGylation reaction.
"+PEG" refers to INF-(3 after the PEGylation reaction.

Figure 176 depicts the SDS-PAGE analysis of the products of the PEG (20 kDa) PEGylation reaction of INF-[3. "Unmodified" refers to 1NF-(3 before the PEGylation reaction. "Pegylated" refers to INF-(3 after the PEGylation reaction.
Figure 177 depicts the chromatogram of PEG (10 kDa) PEGylated IIlTF-~3 elution from a Superdex-200 column.
Figure 178 depicts the results of a bioassay of peak fractions of PEG (10 kDa) PEGylated INF-(3 shown in the chromatogram depicted Figure INF-PEG 6.
Figure 179 depicts the chromatogram of PEG (20 kDa) PEGylated INF-(3 elution from a Superdex-200 column.
Figure 180, comprising Figures 180A and 180B, is two graphs depicting the MALDI-TOF spectrum of RNaseB (Figure 180A) and the HPLC profile of the oligosaccharides cleaved from RNaseB by N-Glycanase (Figure 180B). The majority of N-glycosylation sites of the peptide are modified with high mannose oligosaccharides consisting of 5 to 9 mannose residues.
Figure 181 is a scheme depicting the conversion of high mannose N-Glycans to hybrid N-Glycans. Enzyme 1 is a,1,2-mannosidase, from Trichodoma reesei or Aspergillus saitoi. Enzyme 2 is GnT-I ((3-1,2-N acetyl glucosaminyl transferase I). Enzyme 3 is GaIT-I
((31,4-galactosyltransfease 1). Enzyme 4 is a2,3-sialyltransferase or a,2,6-sialyltransferase.
Figure 182, comprising Figures 182A and 182B, is two graphs depicting the MALDI-TOF spectrum of RNaseB treated with a recombinant T. reesei a1,2-mannosidase (Figure 182A) and the HPLC profile of the oligosaccharides cleaved by N-Glycanase from the modified RNaseB (Figure 182B).
Figure 183 is a graph depicting the MALDI-TOF spectrum of RNaseB treated with a commercially available a1,2-mannosidase purified from A. saitoi (Glyko &
CalBioChem).
Figure 184 is a graph depicting the MALDI-TOF spectrum of modified RNaseB by treating the product shown in Figure 182 with a recombinant GnT-I (GIcNAc transferase-I).
Figure 185 is a graph depicting the MALDI-TOF spectrum of modified RNaseB by treating the product shown in Figure 184 with a recombinant GaIT 1 (galactosyltransferase 1).

Figure 186 is a graph depicting the MALDI-TOF spectruan of modified RNase~ by treating the product shown in Figure 185 with a recombinant ST3Ga1 III (a2,3-sialyltransferase III) using CMP-SA as the donor for the transferase.
Figure 187 is a graph depicting the MALDI-TOF spectrum of modified RNaseB by treating the product shown in Figure 185 with a recombinant ST3Ga1 III (a2,3-sialyltransferase III) using CMP-SA-PEG (10 kDa) as the donor for the transferase.
Figure 188 is a series of schemes depicting the conversion of high mannose N-glycans to complex N-glycans. Enzyme 1 is a1,2-mannosidase from Tr~ichoderma reesei or Aspe~gillus saitoi. Enzyme 2 is GnT-I. Enzyme 3 is GaIT 1. Enzyme 4 is a2,3 sialyltransferase or a2,6-sialyltransferase. Enzyme 5 is a-mannosidase II.
Enzyme 6 is a-mannosidase. Enzyme 7 is GnT-II. Enzyme 8 is a1,6-mannosidase. Enzyme 9 is a1,3-mannosidase.
Figure 189 is a diagram of the linkage catalyzed by N
acetylglucosaminyltransferase I
to VI (GnT I-VI). R = GIcNAc(31,4G1cNAc-Asn-X.
Figure 190 is an image of an SDS-PAGE gel: standard (Lane 1); native transferrin (Lane 2); asialotransferrin (Lane 3); asialotransferrin and CMP-SA (Lane 4);
Lanes 5 and 6, asialotransferrin and CMP-SA-PEG (1 kDa) at 0.5 mM and 5 mM, respectively;
Lanes 7 and 8, asialotransferrin and CMP-SA-PEG, (5 kDa) at 0.5 xnM and 5 mM, respectively; Lanes 9 and 10, asialotransferrin and CMP-SA-PEG (10 kDa) at 0.5 mM and 5 mM, respectively.
Figure 191 is an image of an IEF gel: native transferrin (Lane 1);
asialotransferrin (Lane 2); asialotransferrin and CMP-SA, 24 hr (Lane 3); asialotransferrin and CMP-SA, 96 hr (Lane 4) Lanes 5 and 6, asialotransferrin and CMP-SA-PEG (1 kDa) at 24 hr and 96 hr, respectively; Lanes 7 and 8, asialotransferrin and CMP-SA-PEG (5 kDa) at 24 hr and 96 hr, respectively; Lanes 9 and 10, asialotransferrin and CMP-SA-PEG (10 kDa) at 24 hr and 96 hr, respectively.
DETAILED DESCRIPTION OF THE INVENTION
The present invention includes methods and compositions for the cell free in vitro addition and/or deletion of sugars to or from a peptide molecule in such a manner as to provide a glycopeptide molecule having a specific customized or desired glycosylation pattern, wherein the glycopeptide is produced at an industrial scale. In a preferred embodiment of the invention, the glycopeptide so produced has attached thereto a modified sugar that has been added to the peptide via an enzymatic reaction. A key feature of the invention is to take a peptide produced by any cell type and generate a core glycan structure on the peptide, following which the glycan structure is then remodeled in vit~°~ to generate a glycopeptide having a glycosylation pattern suitable for therapeutic use in a marmnal. More specifically, it is possible according to the present invention, to prepare a glycopeptide molecule having a modified sugar molecule or other compound conjugated thereto, such that the conjugated molecule confers a beneficial property on the peptide.
According to the present invention, the conjugate molecule is added to the peptide enzymatically because enzyme-based addition of conjugate molecules to peptides has the advantage of regioselectivity and stereoselectivity. The glycoconjugate may be added to the glycan on a peptide before or after glycosylation has been completed. In other words, the order of glycosylation with respect to glycoconjugation may be varied as described elsewhere herein.
It is therefore possible, using the methods and compositions provided herein, to remodel a peptide to confer upon the peptide a desired glycan structure preferably having a modified sugar attached thereto. It is also possible, using the methods and compositions of the invention to generate peptide molecules having desired and or modified glycan structures at an industrial scale, thereby, for the first time, providing the art with a practical solution for the efficient production of improved therapeutic peptides.
Definitions Unless defined otherwise, all technical and scientific terms used herein generally have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Generally, the nomenclature used herein and the laboratory procedures in cell culture, molecular genetics, organic chemistry, and nucleic acid chemistry and hybridization are those well known and commonly employed in the art. Standard techniques are used for nucleic acid and peptide synthesis. The techniques and procedures are generally performed according to conventional methods in the art and various general references (e.g., Sambrook et al., 199, Molecular Cloning: A Laboratory Manual, 2d ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY), which are provided throughout this document.
The nomenclature used herein and the laboratory procedures used in analytical chemistry and organic syntheses described below are those well known and commonly employed in the art.
Standard techniques or modifications thereof, are used for chemical syntheses and chemical analyses.
The articles "a" and "an" are used herein to refer to one or to more than one (i.e. to at least one) of the grammatical object of the article. By way of example, "an element" means one element or more than one element.
The teen "antibody," as used herein, refers to an immunoglobulin molecule which is able to specifically bind to a specific epitope on an antigen. Antibodies can be intact immunoglobulins derived from natural sources or from recombinant sources and can be immunoreactive portions of intact immunoglobulins. Antibodies are typically tetramers of immunoglobulin molecules. The antibodies in the present invention may exist in a variety of forms including, for example, polyclonal antibodies, monoclonal antibodies, Fv, Fab and F(ab)2, as well as single chain antibodies and humanized antibodies (Harlow et al., 1999, Using Antibodies: A Laboratory Manual, Cold Spring Harbor Laboratory Press, NY; Harlow et al., 1989, Antibodies: A Laboratory Manual, Cold Spring Harbor, New York;
Houston et al., 1988, Proc. Natl. Acad. Sci. USA 85:5879-5883; Bird et al., 1988, Science 242:423-426).
By the term "synthetic antibody" as used herein, is meant an antibody which is generated using recombinant DNA technology, such as, for example, an antibody expressed by a bacteriophage as described herein. The term should also be construed to mean an antibody which has been generated by the synthesis of a DNA molecule encoding the antibody and which DNA molecule expresses an antibody protein, or an amino acid sequence specifying the antibody, wherein the DNA or amino acid sequence has been obtained using synthetic DNA or amino acid sequence technology which is available and well known in the art.
As used herein, a "functional" biological molecule is a biological molecule in a form in which it exhibits a property by which it is characterized. A functional enzyme, for example, is one which exhibits the characteristic catalytic activity by which the enzyme is characterized.
As used herein, the structure " ~ " , is the point of connection between an r amino acid or an amino acid sidechain in the peptide chain and the glycan structure.

"N-linked" oligosaccharides are those oligosaccharides that are linked to a peptide backbone through asparagine, by way of an asparagine-N-acetylglucosamine linkage. N-linked oligosaccharides are also called "N-glycans." All N-linked oligosaccharides have a common pentasaccharide core of Te~Ian3GlcNAc2. They differ in the presence of, and in the number of branches (also called antennae) of peripheral sugars such as N-acetylglucosamine, galactose, N-acetylgalactosamine, fucose and sialic acid. Optionally, this structure may also contain a core fucose molecule and/or a xylose molecule.
An "elemental trimannosyl core structure" refers to a glycan moiety comprising solely a trimamiosyl core structure, with no additional sugars attached thereto. When the term "elemental" is not included in the description of the "trimannosyl core structure," then the glycan comprises the trimannosyl core structure with additional sugars attached thereto.
Optionally, this structure may also contain a core fucose molecule and/or a xylose molecule.
The term "elemental trimannosyl core glycopeptide" is used herein to refer to a glycopeptide having glycan structures comprised primarily of an elemental trimannosyl core structure. Optionally, this structure may also contain a core fucose molecule and/or a xylose molecule.
"O-linked" oligosaccharides are those oligosaccharides that are linked to a peptide backbone through threonine, serine, hydroxyproline, tyrosine, or other hydroxy-containing amino acids.
All oligosaccharides described herein are described with the name or abbreviation for the non-reducing saccharide (i.e., Gal), followed by the configuration of the glycosidic bond (a or (3), the ring bond (1 or 2), the ring position of the reducing saccharide involved in the bond (2, 3, 4, 6 or 8), and then the name or abbreviation of the reducing saccharide (i. e., GIcNAc). Each saccharide is preferably a pyranose. For a review of standard glycobiology nomenclature see, Essentials of Glycobiology Varki et al. eds., 1999, CSHL
Press.
The term "sialic acid" refers to any member of a family of nine-carbon carboxylated sugars. The most common member of the sialic acid family is N-acetyl-neuraminic acid (2-keto-5-acetamido-3,5-dideoxy-D-glycero-D-galactononulopyranos-1-onic acid (often abbreviated as NeuSAc, NeuAc, or NANA). A second member of the family is N-glycolyl-neuraminic acid (NeuSGc or NeuGc), in which the N-acetyl group of NeuAc is hydroxylated.
A third sialic acid family member is 2-keto-3-deoxy-nonulosonic acid (KDN) (Nadano et al.

(1986) .J. Bi~l. Chem. 261: 11550-11557; I~anamori et al., J. Biol. Chem. 265:

(1990)). Also included are 9-substituted sialic acids such as a 9-O-C1-C6 acyl-NeuSAc like 9-~-lactyl-NeuSAc or 9-~-acetyl-NeuSAc, 9-deoxy-9-fluoro-NeuSAc and 9-azido-9-deoxy-NeuSAc. For review of the sialic acid family, see, e.~., Varki, (~lycobf~l~~y 2: 25-40 (1992);
Sialic Acids: Chemists y, Metabolism and FuaZCti~rc, R. Schauer, Ed. (Springer-Verlag, New York (1992)). The synthesis and use of sialic acid compounds in a sialylation procedure is disclosed in international application VJ~ 92/16640, published ~ctober 1, 1992.
A peptide having "desired glycosylation", as used herein, is a peptide that comprises one or more oligosaccharide molecules which are required for efficient biological activity of the peptide.
A "disease" is a state of health of an animal wherein the animal cannot maintain homeostasis, and wherein if the disease is not ameliorated then the animal's health continues to deteriorate.
The "area under the curve" or "AUC", as used herein in the context of administering a peptide drug to a patient, is defined as total area under the curve that describes the concentration of drug in systemic circulation in the patient as a function of time from zero to infinity.
The term "half life" or "t'/2", as used herein in the context of administering a peptide drug to a patient, is defined as the time required for plasma concentration of a drug in a patient to be reduced by one half. There may be more than one half life associated with the peptide drug depending on multiple clearance mechanisms, redistribution, and other mechanisms well known in the art. Usually, alpha and beta half lives are defined such that the alpha phase is associated with redistribution, and the beta phase is associated with clearance. However, with protein drugs that are, for the most part, confined to the bloodstream, there can be at least two clearance half lives. For some glycosylated peptides, rapid beta phase clearance may be mediated via receptors on macrophages, or endothelial cells that recognize terminal galactose, N-acetylgalactosamine, N-acetylglucosamine, mannose, or fucose. Slower beta phase clearance may occur via renal glomer-ular filtration for molecules with an effective radius < 2 nm (approximately 68 kD) and/or specific or non-specific uptake and metabolism in tissues. GlycoPEGylation may cap terminal sugars (e.g.
galactose or N-acetylgalactosamine) and thereby block rapid alpha phase clearance via receptors that recognize these sugars. It may also confer a larger effective radius and thereby decrease the volume of distribution and tissue uptake, thereby prolonging the late beta phase.
Thus, the precise impact of glycoPEGylation on alpha phase and beta phase half lives will vary depending upon the size, state of glycosylation, and other parameters, as is well known in the art. Further explanation of "half life" is found in Pharmaceutical Biotechnology (1997, DFA Crommelin and RD Sindelar, eds., Harwood Publishers, Amsterdam, pp 120).
The term "residence time", as used herein in the context of administering a peptide drug to a patient, is defined as the average time that drug stays in the body of the patient after dosing.
An "isolated nucleic acid" refers to a nucleic acid segment or fragment which has been separated from sequences which flank it in a naturally occurring state, e.g., a DNA
fragment which has been removed from the sequences which are normally adjacent to the fragment, e.g., the sequences adjacent to the fragment in a genome in which it naturally occurs. The term also applies to nucleic acids which have been substantially purified from other components which naturally accompany the nucleic acid, e.g., RNA or DNA
or proteins, which naturally accompany it in the cell. The term therefore includes, for example, a recombinant DNA which is incorporated into a vector, into an autonomously replicating plasmid or virus, or into the genomic DNA of a prokaryote or eukaryote, or which exists as a separate molecule (e.g., as a cDNA or a genomic or cDNA fragment produced by PCR or restriction enzyme digestion) independent of other sequences. It also includes a recombinant DNA which is part of a hybrid nucleic acid encoding additional peptide sequence.
A "polynucleotide" means a single strand or parallel and anti-parallel strands of a nucleic acid. Thus, a polynucleotide may be either a single-stranded or a double-stranded nucleic acid.
The term "nucleic acid" typically refers to large polynucleotides. The term "oligonucleotide" typically refers to short polynucleotides, generally no greater than about 50 nucleotides.
Conventional notation is used herein to describe polynucleotide sequences: the left-hand end of a single-stranded polynucleotide sequence is the 5'-end; the left-hand direction of a double-stranded polynucleotide sequence is referred to as the 5'-direction.
The direction of 5' to 3' addition of nucleotides to nascent RNA transcripts is referred to as the transcription direction. The DNA strand having the same sequence as an mRNA is referred to as the "coding strand"; sequences on the DNA strand which are located 5' to a reference point on the DIVA are referred to as "upstream sequences"; sequences on the DNA strand which are 3' to a reference point on the DNA are referred to as "downstream sequences."
"Encoding" refers to the inherent property of specific sequences of nucleotides in a polynucleotide, such as a gene, a cDNA, or an mRNA, to serve as templates for synthesis of other polymers and macromolecules in biological processes having either a defined sequence of nucleotides (i.e., rRNA, tRNA and mRNA) or a defined sequence of amino acids and the biological properties resulting therefrom. Thus, a nucleic acid sequence encodes a protein if transcription and translation of mRNA corresponding to that nucleic acid produces the protein in a cell or other biological system. Both the coding strand, the nucleotide sequence of which is identical to the mRNA sequence and is usually provided in sequence listings, and the non-coding strand, used as the template for transcription of a gene or cDNA, can be referred to as encoding the protein or other product of that nucleic acid or cDNA.
Unless otherwise specified, a "nucleotide sequence encoding an amino acid sequence"
includes all nucleotide sequences that are degenerate versions of each other and that encode the same amino acid sequence. Nucleotide sequences that encode proteins and RNA may include introns.
"Homologous" as used herein, refers to the subunit sequence similarity between two polymeric molecules, e.g., between two nucleic acid molecules, e.g., two DNA
molecules or two RNA molecules, or between two peptide molecules. When a subunit position in both of the two molecules is occupied by the same monomeric subunit, e.g., if a position in each of two DNA molecules is occupied by adenine, then they are homologous at that position. The homology between two sequences is a direct function of the number of matching or homologous positions, e.g., if half (e.g., five positions in a polymer ten subunits in length) of the positions in two compound sequences are homologous then the two sequences are 50%
homologous, if 90% of the positions, e.g., 9 of 10, are matched or homologous, the two sequences share 90% homology. By way of example, the DNA sequences 3'ATTGCCS' and 3'TATGGC share 50% homology.
As used herein, "homology" is used synonymously with "identity."

The deternlination of percent identity between two nucleotide or amino acid sequences can be accomplished using a mathematical algorithm. For example, a mathematical algorithm useful for comparing two sequences is the algorithm of Marlin and Altschul (1990, Proc. Natl. Acad. Sci. USA X7:2264-2260, modified as in I~axlin and Altschul (1993, Proc. Natl. Acad. Sci. USA 90:573-577). This algorithm is incorporated into the NBLAST and XBLAST programs of Altschul, et al. (1990, J. Mol. Biol.
215:403-410), and can be accessed, for example at the National Center for Biotechnology Information (NCBI) world wide web site having the universal resource locator "http://www.ncbi.nlm.nih.gov/BLAST/". BLAST nucleotide searches can be performed with the NBLAST program (designated "blastn" at the NCBI web site), using the following parameters: gap penalty = 5; gap extension penalty = 2; mismatch penalty = 3;
match reward = 1; expectation value 10.0; and word size = 11 to obtain nucleotide sequences homologous to a nucleic acid described herein. BLAST protein searches can be performed with the XBLAST program (designated "blastn" at the NCBI web site) or the NCBI "blastp"
program, using the following parameters: expectation value 10.0, BLOSUM62 scoring matrix to obtain amino acid sequences homologous to a protein molecule described herein.
To obtain gapped alignments for comparison purposes, Gapped BLAST can be utilized as described in Altschul et al. (1997, Nucleic Acids Res. 25:339-3402). Alternatively, PSI-Blast or PHI-Blast can be used to perform an iterated search which detects distant relationships between molecules (Id.) and relationships between molecules which share a common pattern. When utilizing BLAST, Gapped BLAST, PSI-Blast, and PHI-Blast programs, the default parameters of the respective programs (e.g., XBLAST and NBLAST) can be used.
See http://www.ncbi.nlm.nih.gov.
The percent identity between two sequences can be determined using techniques similar to those described above, with or without allowing gaps. In calculating percent identity, typically exact matches axe counted.
A "heterologous nucleic acid expression unit" encoding a peptide is defined as a nucleic acid having a coding sequence for a peptide of interest operably linked to one or more expression control sequences such as promoters and/or repressor sequences wherein at least one of the sequences is heterologous, i. e., not normally found in the host cell.

By describing two polynucleotides as "operably linked" is meant that a single-stranded or double-stranded nucleic acid moiety comprises the two polynucleotides arranged within the nucleic acid moiety in such a mamler that at least one of the two polynucleotides is able to exert a physiological effect by which it is characterised upon the other. By way of example, a promoter operably linked to the coding region of a nucleic acid is able to promote transcription of the coding region.
As used herein, the term "promoter/regulatory sequence" means a nucleic acid sequence which is required for expression of a gene product operably linked to the promoter/regulator sequence. In some instances, this sequence may be the core promoter sequence and in other instances, this sequence may also include an enhancer sequence and other regulatory elements which are required for expression of the gene product. The promoter/regulatory sequence may, for example, be one which expresses the gene product in a tissue specific manner.
A "constitutive promoter is a promoter which drives expression of a gene to which it is operably linked, in a constant manner in a cell. By way of example, promoters which drive expression of cellular housekeeping genes are considered to be constitutive promoters.
An "inducible" promoter is a nucleotide sequence which, when operably linked with a polynucleotide which encodes or specifies a gene product, causes the gene product to be produced in a living cell substantially only when an inducer which corresponds to the promoter is present in the cell.
A "tissue-specific" promoter is a nucleotide sequence which, when operably linked with a polynucleotide which encodes or specifies a gene product, causes the gene product to be produced in a living cell substantially only if the cell is a cell of the tissue type corresponding to the promoter.
A "vector" is a composition of matter which comprises an isolated nucleic acid and which can be used to deliver the isolated nucleic acid to the interior of a cell. Numerous vectors are known in the art including, but not limited to, linear polynucleotides, polynucleotides associated with ionic or amphiphilic compounds, plasmids, and viruses.
Thus, the term "vector" includes an autonomously replicating plasmid or a virus. The term should also be construed to include non-plasmid and non-viral compounds which facilitate transfer of nucleic acid into cells, such as, for example, polylysine compounds, liposomes, and the like. Examples of viral vectors include, but are not limited to, adenoviral vectors, adeno-associated virus vectors, retroviral vectors, and the like.
"Expression vector" refers to a vector comprising a recombinant polynucleotide comprising expression control sequences operatively linked to a nucleotide sequence to be expressed. An expression vector comprises sufficient cis-acting elements for expression9 other elements for expression can be supplied by the host cell or in an in vitr~ expression system. Expression vectors include all those known in the art, such as cosmids, plasmids (e.g., naked or contained in liposomes) and viruses that incorporate the recombinant polynucleotide.
A "genetically engineered" or "recombinant" cell is a cell having one or more modifications to the genetic material of the cell. Such modifications are seen to include, but are not limited to, insertions of genetic material, deletions of genetic material and insertion of genetic material that is extrachromasomal whether such material is stably maintained or not.
A "peptide" is an oligopeptide, polypeptide, peptide, protein or glycoprotein.
The use of the term "peptide" herein includes a peptide having a sugar molecule attached thereto when a sugar molecule is attached thereto.
As used herein, "native form" means the form of the peptide when produced by the cells and/or organisms in which it is found in nature. When the peptide is produced by a plurality of cells and/or organisms, the peptide may have a variety of native forms.
"Peptide" refers to a polymer in which the monomers are amino acids and are joined together through amide bonds, alternatively referred to as a peptide.
Additionally, unnatural amino acids, for example, (3-alanine, phenylglycine and homoarginine are also included.
Amino acids that are not nucleic acid-encoded may also be used in the present inventi~n.
Furthermore, amino acids that have been modified to include reactive groups, glycosylation sites, polymers, therapeutic moieties, biomolecules and the like may also be used in the invention. All of the amino acids used in the present invention may be either the D - or L -isomer thereof. The L -isomer is generally preferred. In addition, other peptidomimetics are also useful in the present invention. As used herein, "peptide" refers to both glycosylated and unglycosylated peptides. Also included are peptides that are incompletely glycosylated by a system that expresses the peptide. For a general review, see, Spatola, A. F., in CHEMISTRY

AND BIOCHEMISTRY ~F AMINO ACIDS, PEPTIDES AND PROTEINS, B. Weinstein, eds., IVIarcel Delclcer, IVew York, p. 267 (193).
The term "peptide conjugate," refers to species of the invention in which a peptide is conjugated with a modified sugar as set forth herein.
The term "amino acid" refers to naturally occurring and synthetic amino acids, as well as amino acid analogs and amino acid mimetics that function in a manner similar to the naturally occurring amino acid. Naturally occurring amino acids are those encoded by the genetic code, as well as those amino acids that axe later modified, e.g., hydroxyproline, ~y-carboxyglutamate, and O-phosphoserine. Amino acid analogs refers to compounds that have the same basic chemical structure as a naturally occurring amino acid, i. e., an a carbon that is linked to a hydrogen, a caxboxyl group, an amino group, and an R group, e.g., homoserine, norleucine, methionine sulfoxide, methionine methyl sulfonium. Such analogs have modified R groups (e.g., norleucine) or modified peptide backbones, but retain the same basic chemical structure as a naturally occurring amino acid. Amino acid mimetics refers to chemical compounds that have a structure that is different from the general chemical structure of an amino acid, but that function in a manner similar to a naturally occurring amino acid.
As used herein, amino acids axe represented by the. full name thereof, by the three letter code corresponding thereto, or by the one-letter code corresponding thereto, as indicated in the following Table 1:
Table 1. Amino acids, and the three letter and one letter codes.

Full Name Three-Letter Code ~ne-Letter Code Aspartic Acid Asp Glutamic Acid Glu E

Lysine Lys Arginine ~'g Histidine His H

Tyrosine TYr Cysteine Cys C

Asparagine Asn Glutamine Gln Serine Ser Threonine Tl~' T

Glycine Gly G

Alanine Ala A

Valine Val V

Leucine Leu L

Isoleucine Ile I

Methionine Met M

Proline Pro P

Phenylalanine Phe F

Tryptophan Trp W

The present invention also provides for analogs of proteins or peptides which comprise a protein as identified above. Analogs may differ from naturally occurring proteins or peptides by conservative amino acid sequence differences or by modifications which do not affect sequence, or by both. For example, conservative amino acid changes may be made, which although they alter the primary sequence of the protein or peptide, do not normally alter its function. Conservative amino acid substitutions typically include substitutions within the following groups:
glycine, alanine;
valine, isoleucine, leucine;
aspartic acid, glutamic acid;
asparagine, glutamine;
serine, threonine;
lysine, arginine;
phenylalanine, tyrosine.

Modifications (which do not normally alter primary sequence) include in vivo, or in vitro, chemical derivatization of peptides, e.g., acetylation, or carboxylation. Also included are modifications of glycosylation, e.g., those made by modifying the glycosylation patterns of a peptide during its synthesis~and processing or in further processing steps9 e.g., by exposing the peptide to enzymes which affect glycosylation, e.g., mammalian glycosylating or deglycosylating enzymes. Also embraced are sequences which have phosphorylated amino acid residues, e.g., phosphotyrosine, phosphoserine, or phosphothreonine.
It will be appreciated, of course, that the peptides may incorporate amino acid residues which are modified without affecting activity. For example, the termini may be derivatized to include blocking groups, i.e. chemical substituents suitable to protect and/or stabilize the N- and C-termini from "undesirable degradation", a term meant to encompass any type of enzymatic, chemical or biochemical breakdown of the compound at its termini which is likely to affect the function of the compound, i.e. sequential degradation of the compound at a terminal end thereof.
Blocking groups include protecting groups conventionally used in the art of peptide chemistry which will not adversely affect the in vivo activities of the peptide. For example, suitable N-terminal blocking groups can be introduced by alkylation or acylation of the N-terminus. Examples of suitable N-terminal blocking groups include C1-CS
branched or unbranched alkyl groups, acyl groups such as formyl and acetyl groups, as well as substituted forms thereof, such as the acetamidomethyl (Acm), Fmoc or Boc groups. Desamino analogs of amino acids are also useful N-terminal blocking groups, and can either be coupled to the N-terminus of the peptide or used in place of the N-terminal reside. Suitable C-terminal blocking groups, in which the carboxyl group of the C-terminus is either incorporated or not, include esters, ketones or amides. Ester or ketone-forming alkyl groups, particularly lower alkyl groups such as methyl, ethyl and propyl, and amide-forming amino groups such as primary amines (-NH2), and mono- and di-alkylamino groups such as methylamino, ethylamino, dimethylamino, diethylamino, methylethylamino and the like are examples of C-terminal blocking groups. Descarboxylated amino acid analogues such as agmatine are also useful C-terminal blocking groups and can be either coupled to the peptide's C-terminal residue or used in place of it. Further, it will be appreciated that the free amino and carboxyl groups at the termini can be removed altogether from the peptide to yield desamino and descarboxylated forms thereof without affect on peptide activity.
~ther modifications can also be incorporated without adversely affecting the activity and these include, but are not limited to, substitution of one or more of the amino acids in the natural L-isomeric form with amino acids in the D-isomeric form. Thus, the peptide may include one or more D-amino acid resides, or may comprise amino acids which are all in the D-form. Retro-inverso forms of peptides in accordance with the present invention are also contemplated, for example, inverted peptides in which all amino acids are substituted with D-amino acid forms.
Acid addition salts of the present invention are also contemplated as functional equivalents. Thus, a peptide in accordance with the present invention treated with an inorganic acid such as hydrochloric, hydrobromic, sulfuric, nitric, phosphoric, and the like, or an organic acid such as an acetic, propionic, glycolic, pyruvic, oxalic, malic, malonic, succinic, malefic, fumaric, tataric, citric, benzoic, cinnamic, mandelic, methanesulfonic, ethanesulfonic, p-toluenesulfonic, salicyclic and the like, to provide a water soluble salt of the peptide is suitable for use in the invention.
Also included are peptides which have been modified using ordinary molecular biological techniques so as to improve their resistance to proteolytic degradation or to optimize solubility properties or to render them more suitable as a therapeutic agent. Analogs of such peptides include those containing residues other than naturally occurring L-amino acids, e.g., D-amino acids or non-naturally occurring synthetic amino acids.
The peptides of the invention are not limited to products of any of the specific exemplary processes listed herein.
As used herein, the term "MALDI" is an abbreviation for Matrix Assisted Laser Desorption Ionization. During ionization, SA-PEG (sialic acid-polyethylene glycol)) can be partially eliminated from the N-glycan structure of the glycoprotein.
As used herein, the terns "glycosyltransferase," refers to any enzyme/protein that has the ability to transfer a donor sugar to an acceptor moiety.
As used herein, the term "modified sugar," refers to a naturally- or non-naturally-occurring carbohydrate that is enzymatically added onto an amino acid or a glycosyl residue of a peptide in a process of the invention. The modified sugar is selected from a number of enzyme substrates including, but not limited to sugar nucleotides (mono-, di-, and tri-phosphates), activated sugaxs (e.g., glycosyl halides, glycosyl mesylates) and sugars that are neither activated nor nucleotides.
The "modified sugar" is covalently functionalized with a "modifying group."
Useful modifying groups include, but are not limited to, water-soluble polymers, therapeutic moieties, diagnostic moieties, biomolecules and the like. The locus of functionalization with the modifying group is selected such that it does not prevent the "modified sugar" from being added enzymatically to a peptide.
The term "water-soluble" refers to moieties that have some detectable degree of solubility in water. Methods to detect and/or quantify water solubility are well known in the art. Exemplary water-soluble polymers include peptides, saccharides, poly(ethers), poly(amines), poly(carboxylic acids) and the like. Peptides can have mixed sequences or be composed of a single amino acid, e.g. poly(lysine). Similarly, sacchaxides can be of mixed sequence or composed of a single saccharide subunit, e.g., dextran, amylose, chitosan, and poly(sialic acid). An exemplary poly(ether) is polyethylene glycol).
Polyethylene imine) is an exemplary polyamine, and poly(aspartic) acid is a representative poly(carboxylic acid).
"Poly(alkylene oxide)" refers to a genus of compounds having a polyether backbone.
Poly(alkylene oxide) species of use in the present invention include, for example, straight-and branched-chain species. Moreover , exemplary poly(alkylene oxide) species can terminate in one or more reactive, activatable, or inert groups. For example, polyethylene glycol) is a poly(alkylene oxide) consisting of repeating ethylene oxide subunits, which may or may not include additional reactive, activatable or inert moieties at either terminus. Useful poly(alkylene oxide) species include those in which one terminus is "capped"
by an inert group, e.g., monomethoxy-poly(alkylene oxide). When the molecule is a branched species, it may include multiple reactive, activatable or inert groups at the termini of the alkylene oxide chains and the reactive groups may be either the same or different.
Derivatives of straight-chain poly(alkylene oxide) species that are heterobifunctional are also known in the art.
The term, "glycosyl linking group," as used herein refers to a glycosyl residue to which an agent (e.g., water-soluble polymer, therapeutic moiety, biomolecule) is covalently attached. In the methods of the invention, the "glycosyl linking group"
becomes covalently attached to a glycosylated or unglycosylated peptide, thereby linking the agent to an amino acid and/or glycosyl residue on the peptide. A "glycosyl linking group" is generally derived from a "modified sugar" by the enzymatic attachment of the "modified sugar" to an amino acid and/or glycosyl residue of the peptide. More specifically, a "glycosyl linking group," as used herein, refers to a moiety that covalently joins a "modifying group," as discussed herein, and an amino acid residue of a peptide. The glycosyl linking group-modifying group adduct has a structure that is a substrate for an enzyme. The enzymes for which the glycosyl linking group-modifying group adduct are substrates are generally those capable of transferring a saccharyl moiety onto an amino acid residue of a peptide, e.g, a glycosyltransferase, amidase, glycosidase, trans-sialidase, etc. The "glycosyl linking group" is interposed between, and covalently joins a "modifying group" and an amino acid residue of a peptide.
An "intact glycosyl linking group" refers to a linking group that is derived from a glycosyl moiety in which the individual saccharide monomer that links the conjugate is not degraded, e.g., oxidized, e.g., by sodium metaperiodate. "Intact glycosyl linking groups" of the invention may be derived from a naturally occurring oligosaccharide by addition of glycosyl units) or removal of one or more glycosyl unit from a parent saccharide structure.
An exemplary "intact glycosyl linking group" includes at least one intact, e.g., non-degraded, saccharyl moiety that is covalently attached to an amino acid residue on a peptide. The remainder of the "linking group" can have substantially any structure. For example, the modifying group is optionally linked directly to the intact saccharyl moiety.
Alternatively, the modifying group is linked to the intact saccharyl moiety via a linker arm.
The linker arm can have substantially any structure determined to be useful in the selected embodiment. In an exemplary embodiment, the linker arm is one or more intact sacchaxyl moieties, i.e. "the intact glycosyl linking group" resembles an oligosaccharide. Another exemplary intact glycosyl linking group is one in which a saccharyl moiety attached, directly or indirectly, to the intact saccharyl moiety is degraded and derivatized (e.g., periodate oxidation followed by reductive amination). Still a further linker arm includes the modifying group attached to the intact saccharyl moiety, directly or indirectly, via a cross-linker, such as those described herein or analogues thereof.
"Degradation," as used herein refers to the removal of one or more carbon atoms from a saccharyl moiety.
The terms "targeting moiety" and "targeting agent", as used herein, refer to species that will selectively localize in a particular tissue or region of the body.
The localization is mediated by specific recognition of molecular determinants, molecular size of the targeting agent or conjugate, ionic interactions, hydrophobic interactions and the like.
Other mechanisms of targeting an agent to a particular tissue or region are known to those of skill in the art.
As used herein, "therapeutic moiety" means any agent useful for therapy including, but not limited to, antibiotics, anti-inflammatory agents, anti-tumor drugs, cytotoxins, and radioactive agents. "Therapeutic moiety" includes prodrugs of bioactive agents, constructs in which more than one therapeutic moiety is linked to a carrier, e.g., multivalent agents.
Therapeutic moiety also includes peptides, and constructs that include peptides. Exemplary peptides include those disclosed in Figure 28 and Tables 6 and 7, herein.
"Therapeutic moiety" thus means any agent useful for therapy including, but not limited to, antibiotics, anti-inflammatory agents, anti-tumor drugs, cytotoxins, and radioactive agents. "Therapeutic moiety" includes prodrugs of bioactive agents, constructs in which more than one therapeutic moiety is linked to a carrier, e.g., multivalent agents.
As used herein, "anti-tumor drug" means any agent useful to combat cancer including, but not limited to, cytotoxins and agents such as antimetabolites, alkylating agents, anthracyclines, antibiotics, antimitotic agents, procarbazine, hydroxyurea, asparaginase, corticosteroids, interferons and radioactive agents. Also encompassed within the scope of the term "anti-tumor drug," are conjugates of peptides with anti-tumor activity, e.g. TNF-a.
Conjugates include, but are not limited to those formed between a therapeutic protein and a glycoprotein of the invention. A representative conjugate is that formed between PSGL-1 and TNF-cc.
As used herein, "a cytotoxin or cytotoxic agent" means any agent that is detrimental to cells. Examples include taxol, cytochalasin B, gramicidin D, ethidium bromide, emetine, mitomycin, etoposide, tenoposide, vincristine, vinblastine, colchicin, doxorubicin, daunorubicin, dihydroxy anthracinedione, mitoxantrone, mithramycin, actinomycin D, 1-dehydrotestosterone, glucocorticoids, procaine, tetracaine, lidocaine, propranolol, and puromycin and analogs or homologs thereof. Other toxins include, for example, ricin, CC-1065 and analogues, the duocannycins. Still other toxins include diphtheria toxin, and snake venom (e.g., cobra venom).
As used herein, "a radioactive agent" includes any radioisotope that is effective in diagnosing or destroying a tumor. Examples include, but are not limited to, indium-111, cobalt-60 and technetium. Additionally, naturally occurring radioactive elements such as uranium, radium, and thorium, which typically represent mixtures of radioisotopes, are suitable examples of a radioactive agent. The metal ions are typically chelated with an organic chelating moiety.
Many useful chelating groups, crown ethers, cryptands and the like are known in the art and can be incorporated into the compounds of the invention (e.g. EDTA, DTPA, DOTA, NTA, HDTA, etc. and their phosphonate analogs such as DTPP, EDTP, HDTP, NTP, etc).
See, for example, Pitt et al., "The Design of Chelating Agents for the Treatment of Iron Overload," In, INORGANIC CHEMISTRY IN BIOLOGY AND MEDICINE; Martell, Ed.;
American Chemical Society, Washington, D.C., 1980, pp. 279-312; Lindoy, THE CHEMISTRY
OF
MACROCYCLIC LIGAND COMPLEXES; Cambridge University Press, Cambridge, 1989;
Dugas, BIOORGANIC CHEMISTRY; Springer-Verlag, New York, 1989, and references contained therein.
Additionally, a manifold of routes allowing the attaclunent of chelating agents, crown ethers and cyclodextrins to other molecules is available to those of skill in the art. See, for example, Meares et al., "Properties of In Vivo Chelate-Tagged Proteins and Polypeptides."
In, MODIFICATION OF PROTEINS: FOOD, NUTRITIONAL, AND PHARMACOLOGICAL ASPECTS;"
Feeney, et al., Eds., American Chemical Society, Washington, D.C., 1982, pp.
370-387;
Kasina et al., Biocohjugate Chem., 9: 108-117 (1998); Song et al., Bioconjugate Chem., 8:
249-255 (1997).
As used herein, "pharmaceutically acceptable carrier" includes any material, which when combined with the conjugate retains the activity of the conjugate activity and is non-reactive with the subject's inunune system. Examples include, but are not limited to, any of the standard pharmaceutical carriers such as a phosphate buffered saline solution, water, emulsions such as oil/water emulsion, and various types of wetting agents.
Other carriers may also include sterile solutions, tablets including coated tablets and capsules. Typically such carriers contain excipients such as starch, milk, sugar, certain types of clay, gelatin, stearic acid or salts thereof, magnesium or calcium stearate, talc, vegetable fats or oils, gums, glycols, or other known excipients. Such carriers may also include flavor and color additives or other ingredients. Compositions comprising such carriers are formulated by well known conventional methods.
As used herein, "administering" means oral administration, administration as a suppository, topical contact, intravenous, intraperitoneal, intramuscular, intralesional, intranasal or subcutaneous administration, intrathecal administration, or the implantation of a slow-release device e.g., a mini-osmotic pump, to the subject.
The term "isolated" refers to a material that is substantially or essentially free from components, which are used to produce the material. For peptide conjugates of the invention, the term "isolated" refers to material that is substantially or essentially free from components, which normally accompany the material in the mixture used to prepare the peptide conjugate.
"Isolated" and "pure" are used interchangeably. Typically, isolated peptide conjugates of the invention have a level of purity preferably expressed as a range. The lower end of the range of purity for the peptide conjugates is about 60%, about 70% or about 80% and the upper end of the range of purity is about 70%, about 80%, about 90% or more than about 90%.
When the peptide conjugates are more than about 90% pure, their purities are also preferably expressed as a range. The lower end of the range of purity is about 90%, about 92%, about 94%, about 96% or about 98%. The upper end of the range of purity is about 92%, about 94%, about 96%, about 98% or about 100% purity.
Purity is determined by any art-recognized method of analysis (e.g., band intensity on a silver stained gel, polyacrylamide gel electrophoresis, HPLC, or a similar means).
"Commercial scale" as used herein means about one or more gram of final product produced in the method.
"Essentially each member of the population," as used herein, describes a characteristic of a population of peptide conjugates of the invention in which a selected percentage of the modified sugars added to a peptide are added to multiple, identical acceptor sites on the peptide. "Essentially each member of the population" speaks to the "homogeneity" of the sites on the peptide conjugated to a modified sugar and refers to conjugates of the invention, which are at least about 80%, preferably at least about 90% and more preferably at least about 95% homogenous.
"Homogeneity," refers to the structural consistency across a population of acceptor moieties to which the modified sugars are conjugated. Thus, in a peptide conjugate of the invention in which each modified sugar moiety is conjugated to an acceptor site having the same structure as the acceptor site to which every other modified sugar is conjugated, the peptide conjugate is said to be about 100% homogeneous. Homogeneity is typically expressed as a range. The lower end of the range of homogeneity for the peptide conjugates is about 60%, about 70% or about 80% and the upper end of the range of purity is about 70%, about 80%, about 90% or more than about 90%.
When the peptide conjugates are more than or equal to about 90% homogeneous, their homogeneity is also preferably expressed as a range. The lower end of the range of homogeneity is about 90%, about 92%, about 94%, about 96% or about 98%. The upper end of the range of purity is about 92%, about 94%, about 96%, about 98% or about 100%
homogeneity. The purity of the peptide conjugates is typically determined by one or more methods known to those of skill in the art, e.g., liquid chromatography-mass spectrometry (LC-MS), matrix assisted laser desorption time of flight mass spectrometry (MALDI-TOF), capillary electrophoresis, and the like.
"Substantially uniform glycoform" or a "substantially uniform glycosylation pattern,"
when referring to a glycopeptide species, refers to the percentage of acceptor moieties that , are glycosylated by the glycosyltransferase of interest (e.g., fucosyltransferase). For example, in the case of a a1,2 fucosyltransferase, a substantially uniform fucosylation pattern exists if substantially all (as defined below) of the Gal(31,4-GIcNAc-R and sialylated analogues thereof are fucosylated in a peptide conjugate of the invention. It will be understood by one of skill in the art, that the starting material may contain glycosylated acceptor moieties (e.g., fucosylated Gal(31,4-GIcNAc-R moieties). Thus, the calculated percent glycosylation will include acceptor moieties that are glycosylated by the methods of the invention, as well as those acceptor moieties already glycosylated in the starting material.
The term "substantially" in the above definitions of "substantially uniform"
generally means at least about 40°/~, at least about 70%, at least about 80%, or more preferably at least about 90%, and still more preferably at least about 95% of the acceptor moieties for a particular glycosyltransferase are glycosylated.
Description of the Invention I Method to Remodel Glycan Chains The present invention includes methods and compositions for the iiz vitr~
addition and/or deletion of sugars to or from a glycopeptide molecule in such a manner as to provide a peptide molecule having a specific customized or desired glycosylation pattern, preferably including the addition of a modified sugar thereto. A key feature of the invention therefore is to take a peptide produced by any cell type and generate a core glycan structure on the peptide, following which the glycan structure is then remodeled in vitro to generate a peptide having a glycosylation pattern suitable for therapeutic use in a mammal.
The importance of the glycosylation pattern of a peptide is well known in the art as are the limitations of present in vivo methods for the production of properly glycosylated peptides, particularly when these peptides are produced using recombinant DNA
methodology. Moreover, until the present invention, it has not been possible to generate glycopeptides having a desired glycan structure thereon, wherein the peptide can be produced at industrial scale.
In the present invention, a peptide produced by a cell is enzymatically treated i~ vitro by the systematic addition of the appropriate enzymes and substrates therefor, such that sugar moieties that should not be present on the peptide are removed, and sugar moieties, optionally including modified sugars, that should be added to the peptide are added in a manner to provide a glycopeptide having "desired glycosylation", as defined elsewhere herein.
A Method to remodel N-linked ~lycans In one aspect, the present invention takes advantage of the fact that most peptides of commercial or pharmaceutical interest comprise a common five sugar structure referred to herein as the trimannosyl core, which is N-linked to asparagine at the sequence Asn-X-Ser/Thr on a peptide chain. The elemental trimannosyl core consists essentially of two N-acetylglucosamine (GIcNAc) residues and three mannose (Man) residues attached to a peptide, i.e., it comprises these five sugar residues and no additional sugars, except that it may optionally include a fucose residue. The first GIcNAc is attached to the amide group of the asparagine and the second GlcNAc is attached to the first via a (31,4 linkage. A mannose residue is attached to the second GIcNAc via a (i1,4 linkage and two mannose residues are attached to this mannose via an a,1,3 and an o,1,6 linkage respectively. A
schematic depiction of a trimannosyl core structure is shown in Figure 1, left side. While it is the case that glycan structures on most peptides comprise other sugars in addition to the trimannosyl core, the trimannosyl core structure represents an essential feature of N-linked glycans on mammalian peptides.
The present invention includes the generation of a peptide having a trimanriosyl core structure as a fundamental element of the structure of the glycan molecules contained thereon. Given the variety of cellular systems used to produce peptides, whether the systems are themselves naturally occurring or whether they involve recombinant DNA
methodology, the present invention provides methods whereby a glycan molecule on a peptide produced in any cell type can be reduced to an elemental trimannosyl core structure. Once the elemental trimannosyl core structure has been generated then it is possible using the methods described herein, to generate in vit~~o, a desired glycan structure on the peptide which confers on the peptide one or more properties that enhances the therapeutic effectiveness of the peptide.
It should be clear from the discussion herein that the term "trimannosyl core"
is used to describe the glycan structure shown in Figure 1, left side. Glycopeptides having a trimannosyl core structure may also have additional sugars added thereto, and for the most part, do have additional structures added thereto irrespective of whether the sugars give rise to a peptide having a desired glycan structure. The term "elemental trimannosyl core structure" is defined elsewhere herein. When the term "elemental" is not included in the description of the "trimannosyl core structure," then the glycan comprises the trimannosyl core structure with additional sugars attached to the mannose sugars.
The term "elemental trimannosyl core glycopeptide" is used herein to refer to a glycopeptide having glycan structures comprised primarily of an elemental trimannosyl core structure. However, it may also optionally contain a fucose residue attached thereto. As discussed herein, elemental trimannosyl core glycopeptides are one optimal, and therefore preferred, starting material for the glycan remodeling processes of the invention.

Another optimal starting material for the glycan remodeling process of the invention is a glycan structure having a trimannosyl core wherein one or two additional GIcNAc residues are added to each of the a1,3 and the a1,6 mannose residues (see for example, the structure on the second line of Figure 2, second structure in from the left of the figure). This S structure is referred to herein as "Man3GlcNAc4." ~Jhen the structure is monoantenary, the structure is referred to herein as "Man3GlcNAc3." Optionally, this structure may also contain a core fucose molecule. Once the Man3GlcNAc3 or Man3GlcNAc4 structure has been generated then it is possible using the methods described herein, to generate ire vita~, a desired glycan structure on the glycopeptide which confers on the glycopeptide one or more properties that enhances the therapeutic effectiveness of the peptide.
In their native form, the N-linked glycopeptides of the invention, and particularly the mammalian and human glycopeptides useful in the present invention, are N-linked glycosylated with a trimannosyl core structure and one or more sugars attached thereto.
The terms "glycopeptide" and "glycopolypeptide" are used synonymously herein to refer to peptide chains having sugar moieties attached thereto. No distinction is made herein to differentiate small glycopolypeptides or glycopeptides from large glycopolypeptides or glycopeptides. Thus, hormone molecules having very few amino acids in their peptide chain (e.g., often as few as three amino acids) and other much larger peptides are included in the general terms "glycopolypeptide" and "glycopeptide," provided they have sugar moieties attached thereto. However, the use of the term "peptide" does not preclude that peptide from being a glycopeptide.
An example of an N-linked glycopeptide having desired glycosylation is a peptide having an N-linked glycan having a trimannosyl core with at least one GIcNAc residue attached thereto. This residue is added to the trimannosyl core using N-acetyl glucosaminyltransferase I (GnT-I). If a second GIcNAc residue is added, N-acetyl glucosaminyltransferase II (GnT-II) is used. Optionally, additional GIcNAc residues may be added with GnT-IV and/or GnT-V, and a third bisecting GIcNAc residue may be attached to the X1,4 mannose of the trimannosyl core using N-acetyl glucosaminyltransferase III (GnT-III). Optionally, this structure may be extended by treatment with (31,4 galactosyltransferase to add a galactose residue to each non-bisecting GIcNAc, and even further optionally, using a2,3 or a2,6-sialyltransferase enzymes, sialic acid residues may be added to each galactose residue. The addition of a bisecting GIcNAc to the glycan is not required for the subsequent addition of galactose and sialic acid residues; however, with respect to the substrate affinity of the rat and human GnT-III enzymes, the presence of one or more of the galactose residues on the glycan precludes the addition of the bisecting GIcNAc in that the galactose-containing glycan is not a substrate for these forms of GnT-III. Thus, in instances where the presence of the bisecting GIcNAc is desired and these forms of GnT-III are used, it is important should the glycan contain added galactose and/or sialic residues, that they are removed prior to the addition of the bisecting GIcNAc. ~ther forms of GnT-III rnay not require this specific order of substrates for their activity. In the more preferred reaction, a mixture of GnT-I, GnT-II
and GnT-III is added to the reaction mixture so that the GIcNAc residues can be added in any order.
Examples of glycan structures which represent the various aspects of peptides having "desired glycosylation" are shown in the drawings provided herein. The precise procedures for the ive vitro generation of a peptide having "desired glycosylation" are described elsewhere herein. However, the invention should in no way be construed to be limited solely to any one glycan structure disclosed herein. Rather, the invention should be construed to include any and 'all glycan structures which can be made using the methodology provided herein.
In some cases, an elemental trimannosyl core alone may constitute the desired glycosylation of a peptide. For example, a peptide having only a trimannosyl core has been shown to be a useful component of an enzyme employed to treat Gaucher disease (Mistry et al., 1966, Lancet 348: 1555-1559; Bijsterbosch et al., 1996, Eur. J. Biochem.
237:344-349) .
According to the present invention, the following procedures for the generation of peptides having desired glycosylation become apparent.
a) Beginning with a glycopeptide having one or more glycan molecules which have as a common feature a trimannosyl core structure and at least one or more of a heterogeneous or homogeneous mixture of one or more sugars added thereto, it is possible to increase the proportion of glycopeptides having an elemental trimannosyl core structure as the sole glycan structure or which have Man3GlcNAc3 or Man3GlcNAc4 as the sole glycan structure. This is accomplished in vitro by the systematic addition to the glycopeptide of an appropriate number of enzymes in an appropriate sequence which cleave the heterogeneous or homogeneous mixture of sugars on the glycan structure until it is reduced to an elemental trimannosyl core or Man3GlcNAc3 or Man3GlcNAc4. structure. Specific examples of how this is accomplished will depend on a variety of factors including in large part the type of cell in which the peptide is produced and therefore the degree of complexity of the glycan structures) present on the peptide initially produced by the cell. Examples of how a complex glycan structure can be reduced to an elemental trimannosyl core or a Man3GlcNAc3 or Man3GlcNAc4 structure are presented in Figure 2 or are described in detail elsewhere herein.
b) It is possible, to generate a peptide having an elemental trimannosyl core structure as the sole glycan structure on the peptide by isolating a naturally occurring cell whose glycosylation machinery produces such a peptide. DNA encoding a peptide of choice is then transfected into the cell wherein the DNA is transcribed, translated and glycosylated such that the peptide of choice has an elemental trimannosyl core structure as the sole glycan structure thereon. For example, a cell lacking a functional GnT-I enzyme will produce several types of glycopeptides. In some instances, these will be glycopeptides having no additional sugars attached to the trimannosyl core. However, in other instances, the peptides produced may have two additional mannose residues attached to the trimannosyl core, resulting in a Mans glycan. This is also a desired starting material for the remodeling process of the present invention. Specific examples of the generation of such glycan structures are described herein.
c) Alternatively, it is possible to genetically engineer a cell to confer upon it a specific glycosylation machinery such that a peptide having an elemental trimannosyl core or Man3GlcNAc3 or Man3GlcNAc4 structure as the sole glycan structure on the peptide is produced. DNA encoding a peptide of choice is then transfected into the cell wherein the DNA is transcribed, translated and glycosylated such that the peptide of choice has an increased number of glycans comprising solely an elemental trimannosyl core structure. For example, certain types of cells that are genetically engineered to lack GnT-I, may produce a glycan having an elemental trimannosyl core structure, or, depending on the cell, may produce a glycan having a trimannosyl core plus two additional maimose residues attached thereto (Mans). When the cell produces a Mans glycan structure, the cell may be further genetically engineered to express mannosidase 3 which cleaves off the two additional mannose residues to generate the trimannosyl core. Alternatively, the Mans glycan may be incubated irz vi~~ with mannosidase 3 to have the same effect.
d) When a peptide is expressed in an insect cell, the glycan on the peptide comprises a partially complex chain. Insect cells also express hexosaminidase in the cells which trims the partially complex chain back to a trimannosyl core structure which can then be remodeled as described herein.
e) It is readily apparent from the discussion in b), c) and d) that it is not necessary that the cells produce only peptides having elemental trimannosyl core or Man3GlcNAc3 or Man3GlcNAc4 structures attached thereto. Rather, unless the cells described in b) and c) produce peptides having 100% elemental trimannosyl core structures (i.e., having no additional sugars attached thereto) or 100% of Man3GlcNAc3 or Man3GlcNAc4 structures, the cells in fact produce a heterogeneous mixture of peptides having, in combination, elemental trimannosyl core structures, or Man3GlcNAc3 or Man3GlcNAc4 structures, as the sole glycan structure in addition to these structures having additional sugars attached thereto.
The proportion of peptides having a trimannosyl core or Man3GlcNAc3 or Man3GlcNAc4 structures having additional sugars attached thereto, as opposed to those having one structure, will vary depending on the cell which produces them. The complexity of the glycans (i.e.
which and how many sugars are attached to the trimannosyl core) will also vary depending on the cell which produces them.
f) Once a glycopeptide having an elemental trimannosyl core or a trimannosyl core with one or two GIcNAc residues attached thereto is produced by following a), b) or c) above, according to the present invention, additional sugax molecules are added in vitro to the trimannosyl core structure to generate a peptide having desired glycosylation (i.e., a peptide having an in vitro customized glycan structure).
g) However, when it is the case that a peptide having an elemental trimannosyl core or Man3GlcNAc4 structure with some but not all of the desired sugars attached thereto is produced, then it is only necessary to add any remaining desired sugars without reducing the glycan structure to the elemental trimannosyl core or Man3GlcNAc4 structure.
Therefore, in some cases, a peptide having a glycan structure having a trimannosyl core structure with additional sugars attached thereto, will be a suitable substrate for remodeling.
_77_ Isolation of an elemental trimannosyl core ~lycopet~tide The elemental trimannosyl core or Man3GlcNAc3 or Man3GlcNAc4 glycopeptides of the invention may be isolated and purified, if necessary, using techniques well known in the art of peptide purification. Suitable techniques include chromatographic teclmiques, isoelectric focusing techniques, ultrafiltration techniques and the like.
Using any such techniques, a composition of the invention can be prepared in which the glycopeptides of the invention are isolated from other peptides and from other components normally found within cell culture media. The degree of purification can be, for example, 90% with respect to other peptides or 95%, or even higher, e.g., 98%. See, e.g., Deutscher et al. (ed., 1990, Guide to Protein Purification, Harcourt Brace Jovanovich, San Diego).
The heterogeneity of N-linked glycans present in the glycopeptides produced by the prior art methodology generally only permits the isolation of a small portion of the target glycopeptides which can be modified to produce desired glycopeptides. In the present methods, large quantities of elemental trimannosyl core glycopeptides and other desired glycopeptides, including Man3GlcNAc3 or Man3GlcNAc4 glycans, can be produced which can then be further modified to generate large quantities of peptides having desired glycosylation.
Specific enrichment of any particular type of glycan linked to a peptide may be accomplished using lectins which have an affinity for the desired glycan. Such techniques are well known in the art of glycobiology.
A key feature of the invention which is described in more detail below, is that once a core glycan structure is generated on any peptide, the glycan structure is then remodeled in vitro to generate a peptide having desired glycosylation that has improved therapeutic use in a mammal. The mammal may be any type of suitable mammal, and is preferably a human.
The various scenarios and the precise methods and compositions for generating peptides with desired glycosylation will become evident from the disclosure which follows.
The ultimate objective of the production of peptides for therapeutic use in mammals is that the peptides should comprise glycan structures that facilitate rather than negate the therapeutic benefit of the peptide. As disclosed throughout the present specification, peptides produced in cells may be treated in vitro with a variety of enzymes which catalyze the cleavage of sugars that should not be present on the glycan and the addition of sugars which _78_ should be present on the glycan such that a peptide having desired glycosylation and thus suitable for therapeutic use in mammals is generated. The generation of different glycoforms of peptides in cells is described above. A variety of mechanisms for the generation of peptides having desired glycosylation is now described, where the starting material i.e., the peptide produced by a cell may differ from one cell type to another. As will become apparent from the present disclosure, it is not necessary that the starting material be uniform with respect to its glycan composition. However, it is preferable that the starting material be enriched for certain glycoforms in order that large quantities of end product, i.e., correctly glycosylated peptides are produced.
In a preferred embodiment according to the present invention, the degradation and synthesis events that result in a peptide having desired glycosylation involve at some point, the generation of an elemental trimannosyl core structure or a Man3GlcNAc3 or Man3GlcNAc4 structure on the peptide.
The present invention also provides means of adding one or more selected glycosyl residues to a peptide, after which a modified sugar is conjugated to at least one of the selected glycosyl residues of the peptide. The present embodiment is useful, for example, when it is desired to conjugate the modified sugar to a selected glycosyl residue that is either not present on a peptide or is not present in a desired amount. Thus, prior to coupling a modified sugar to a peptide, the selected glycosyl residue is conjugated to the peptide by enzymatic or chemical coupling. In another embodiment, the glycosylation pattern of a peptide is altered prior to the conjugation of the modified sugar by the removal of a carbohydrate residue from the peptide. See for example WO 98131826.
Addition or removal of any carbohydrate moieties present on the peptide is accomplished either chemically or enzymatically. Chemical deglycosylation is preferably brought about by exposure of the peptide variant to the compound trifluoromethanesulfonic acid, or an equivalent compound. This treatment results in the cleavage of most or all sugars except the linking sugar (N-acetylglucosamine or N-acetylgalactosamine), while leaving the peptide intact. Chemical deglycosylation is described by Hakimuddin et al., 1987, Arch.
Biochem. Biophys. 259: 52 and by Edge et al., 1981, Anal. Eiochem. 118: 131.
Enzymatic cleavage of carbohydrate moieties on peptide variants can be achieved by the use of a variety of endo- and exo-glycosidases as described by Thotakura et al., 1987, Meth.
Enzymol. 138:
350.
Chemical addition of glycosyl moieties is carried out by any art-recognized method.
Enzymatic addition of sugar moieties is preferably achieved using a modification of the methods set forth herein, substituting native glycosyl units for the modified sugars used in the invention. Other methods of adding sugar moieties are disclosed in U.S. Patent No.
5,876,980, 6,030,815, 5,728,554, and 5,922,577.
Exemplary attachment points for selected glycosyl residue include, but are not limited to: (a) sites for N- and O-glycosylation; (b) terminal glycosyl moieties that are acceptors for a glycosyltransferase; (c) arginine, asparagine and histidine; (d) free carboxyl groups; (e) free sulflrydryl groups such as those of cysteine; (f) free hydroxyl groups such as those of serine, threonine, or hydroxyproline; (g) aromatic residues such as those of phenylalanine, tyrosine, or tryptophan; or (h) the amide group of glutamine. Exemplary methods of use in the present invention are described in WO 87/05330 published Sep. 11, 1987, and in Aplin and Wriston, CRC Crit. Rev. Biochem., pp. 259-306 (1981).
Dealing specifically with the examples shown in several of the figures provided herein, a description of the sequence of in vitro enzymatic reactions for the production of desired glycan structures on peptides is now presented. The precise reaction conditions for each of the enzymatic conversions disclosed below are well known to those skilled in the art of glycobiology and are therefore not repeated here. For a review of the reaction conditions for these types of reactions, see Sadler et al., 1982, Methods in Enzymology 83:458-514 and references cited therein.
In Figure 1 there is shown the structure of an elemental trimannosyl core glycan on the left side. It is possible to convert this structure to a complete glycan structure having a bisecting GIcNAc by incubating the elemental trimannosyl core structure in the presence of GnT-I, followed by GnT-II, and further followed by GnT-III, and a sugar donor comprising UDP-GIcNAc, wherein GIcNAc is sequentially added to the elemental trimannosyl core structure to generate a trimannosyl core having a bisecting GIcNAc. In some instances, for example when remodeling Fc glycans as described herein, the order of addition of GnT-I, GnT-II and GnT-III may be contrary to that reported in the literature. The bisecting GIcNAc structure may be produced by adding a mixture of GnT-I, GnT-II and GnT-III and UDP-GIcNAc to the reaction mixture In Figure 3 there is shown the conversion of a bisecting GIcNAc containing trimannosyl core glycan to a complex glycan structure comprising galactose and N-acetyl neuraminic acid. The bisecting GIcNAc containing trimannosyl core glycan is first incubated with galactosyltransferase and UDP-Gal as a donor molecule, wherein two galactose residues are added to the peripheral GIcNAc residues on the molecule. The enzyme NeuAc-transferase is then used to add two NeuAc residues one to each of the galactose residues.
In Figure 4 there is shown the conversion of a high mannose glycan structure to an elemental trimannosyl core glycan. The high mannose glycan (Man9) is incubated sequentially in the presence of the mannosidase 1 to generate a Mans structure and then in the presence of mannosidase 3, wherein all but three mannose residues are removed from the glycan. Alternatively, incubation of the Man9 structure may be trimmed back to the trimannosyl core structure solely by incubation in the presence of mannosidase 3. According to the schemes presented in Figures 1 and 3 above, conversion of this elemental trimannosyl core glycan to a complex glycan molecule is then possible.
In Figure 5 there is shown a typical complex N-linked glycan structure produced in plant cells. It is important to note that when plant cells are deficient in GnT-I enzymatic activity, xylose and fucose cannot be added to the glycan. Thus, the use of GnT-I knock-out cells provides a particular advantage in the present invention in that these cells produce peptides having an elemental trimannosyl core onto which additional sugars can be added without performing any "trimming back" reactions. Similarly, in instances where the structure produced in a plant cell may be of the Mans variety of glycan, if GnT-I is absent in these cells, xylose and fucose cannot be added to the structure. In this case, the Mans structure may be trimmed back to an elemental trimannosyl core (Man3) using mannosidase 3. According to the methods provided herein, it is now possible to add desired sugar moieties to the trimannosyl core to generate a desired glycan structure.
In Figure 6 there is shown a typical complex N-linked glycan structure produced in insect cells. As is evident, additional sugars, such as, for example, fucose may also be present. Further although not shown here, insect cells may produce high mannose glycans having as many as nine mannose residues and may have additional sugars attached thereto. It is also the case in insect cells that GnT-I knock out cells prevent the addition of fucose residues to the glycan. Thus, production of a peptide in insect cells may preferably be accomplished in a GnT-I knock out cell. The glycan thus produced may then be trimmed back in vitt°o if necessary using any of the methods and schemes described herein, and additional sugars may be added irr vitr~ thereto also using the methods and schemes provided herein.
In Figure 2 there is shown glycan structures in various stages of completion.
Specifically, the i~ vitro enzymatic generation of an elemental trimannosyl core structure from a complex carbohydrate glycan structure which does not contain a bisecting GIcNAc residue is shown. Also shown is the generation of a glycan structure therefrom which contains a bisecting GIcNAc. Several intermediate glycan structures which can be produced are shown. These structures can be produced by cells, or can be produced in the in vitro trimming back reactions described herein. Sugar moieties may be added in vitro to the elemental trimannosyl core structure, or to any suitable intermediate structure in order that a desired glycan is produced.
In Figure 7 there is shown a series of possible in vitro reactions which can be performed to trim back and add onto glycans beginning with a high mannose structure. For example, a Man9 glycan may be trimmed using mannosidase 1 to generate a Mans glycan, or it may be trimmed to a trimannosyl core using mannosidase 3 or one or more microbial mannosidases. GnT-I and or GnT-II may then be used to transfer additional GIcNAc residues onto the glycan. Further, there is shown the situation which would not occur when the glycan molecule is produced in a cell that does not have GnT-I (see shaded box). For example, fucose and xylose may be added to a glycan only when GnT-I is active and facilitates the transfer of a GIcNAc to the molecule.
Figure 8 depicts well known strategies for the synthesis of biantennary, triantennaxy and even tetraantennary glycan structures beginning with the trimannosyl core structure.
According to the methods of the invention, it is possible to synthesize each of these structures in vity~o using the appropriate enzymes and reaction conditions well known in the art of glycobiology.
Figure 9 depicts two methods for synthesis of a monoantennary glycan structure beginning from a high mannose (6 to 9 mannose moieties) glycan structures. A
terminal sialic acid-PEG moiety may be added in place of the sialic acid moiety in accordance with glycoPEGylation methodology described herein. In the first method, endo-Ii is used to cleave the glycan structure on the peptide back to the first GIcNAc residue.
Galactose is then added using galactosyltransferase and sialylated-PEG is added as described elsewhere herein.
In the second method, mannosidase I is used to cleave mannose residues from the glycan structure in the peptide. A galactose residue is added to one arm of the remaining mannose residues which were cleaved off the glycan using Sack Eean oc-maimosidase.
Sialylated-PEG
is then added to this structure as directed.
Figure 10 depicts two additional methods for synthesis of a monoantennary glycan structures beginning from high mannose (6 to 9 mannose moieties) glycan structure. As in Figure 9, a terminal sialic acid-PEG moiety may be added in place of the sialic acid moiety in accordance with the glycoPEGylation methodology described herein. In the situation described here, some of the mannose residues from the arm to which sialylated-PEG is not added, are removed.
In Figure 11 there is shown a scheme for the synthesis of yet more complex carbohydrate structures beginning with a trimannosyl core structure. For example, a scheme for the ih vitro production of Lewis x and Lewis a antigen structures, which may or may not be sialylated is shown. Such structures when present on a peptide may confer on the peptide immunological advantages for upregulating or downregulating the immune response. In addition, such structures are useful for targeting the peptide to specific cells, in that these types of structures are involved in binding to cell adhesion peptides and the like.
Figure 12 is an exemplary scheme for preparing an array of O-linked peptides originating with serine or threonine.
Figure 13 is a series of diagrams depicting the four types of O-linked glycan structure termed cores 1 through 4. The core structure is outlined in dotted lines.
Sugars which may also be included in this structure include sialic acid residues added to the galactose residues, and fucose residues added to the GIcNAc residues.
Thus, in preferred embodiments, the present invention provides a method of making an N-linked glycosylated glycopeptide by providing an isolated and purified glycopeptide to which is attached an elemental trimannosyl core or a Man3GlcNAc4 structure, contacting the glycopeptide with a glycosyltransferase enzyme and a donor molecule having a glycosyl moiety under conditions suitable to transfer the glycosyl moiety to the glycopeptide.
Customization of a trimannosyl core glycopeptide or Man3GlcNAc4 glycopeptide to produce a peptide having a desired glycosylation pattern is then accomplished by the sequential addition of the desired sugar moieties, using techniques well lmown in the art.
Determination of Glycan Primary Structure When an N-linked glycopeptide is produced by a cell, as noted elsewhere herein, it may comprise a heterogeneous mixture of glycan structures which must be reduced to a common, generally elemental trima~mosyl core or Man3GlcNAc4 structure, prior to adding other sugar moieties thereto. In order to determine exactly which sugars should be removed from any particular glycan structure, it is sometimes necessary that the primary glycan structure be identified. Techniques for the determination of glycan primary structure are well know in the art and are described in detail, for example, in Montreuil, "Structure and Biosynthesis of Glycopeptides" In Polysaccharides in Medicinal Applications, pp. 273-327, 1996, Eds. Severian Damitriu, Marcel Dekker, NY. It is therefore a simple matter for one skilled in the art of glycobiology to isolate a population of peptides produced by a cell and determine the structures) of the glycans attached thereto. For example, efficient methods are available for (i) the splitting of glycosidic bonds either by chemical cleavage such as hydrolysis, acetolysis, hydrazinolysis, or by nitrous deamination; (ii) complete methylation followed by hydrolysis or methanolysis and by gas-liquid chromatography and mass spectroscopy of the partially methylated monosaccharides; and (iii) the definition of anomeric linkages between monosaccharides using exoglycosidases, which also provide insight into the primary glycan structure by sequential degradation. In particular, the techniques of mass spectroscopy and nuclear magnetic resonance (NMR) spectrometry, especially high field NMR have been successfully used to determine glycan primary structure.
Kits and equipment for carbohydrate analysis are also commercially available.
Fluorophore Assisted Carbohydrate Electrophoresis (FACE~) is available from Glyko, Inc.
(Novato, CA). In FACE analysis, glycoconjugates are released from the peptide with either Endo H or N-glycanase (PNGase F) for N-linked glycans, or hydrazine for Ser/Thr linked glycans. The glycan is then labeled at the reducing end with a fluorophore in a non-structure discriminating manner. The fluorophore labeled glycans are then separated in polyacrylamide gels based on the charge/mass ratio of the saccharide as well as the _84_ hydrodynamic volume. Images are taken of the gel under UV Light and the composition of the glycans are determined by the migration distance as compared with the standards.
~ligosaccharides can be sequenced in this manner by analysing migration shifts due to the sequential removal of saccharides by exoglycosidase digestion.
Exem~lary embodiment The remodeling of N-Linked glycosylation is best illustrated with reference to Formula 1:
(X~ 7) Man-(X3)a ( i 6)a ~-GIcNAc-GIcNAc-Man-(X4)b M ~ n-(X5)c (X')e where X3, X4, X5, X6, X7 and X17 are (independently selected) monosaccharide or oligosaccharide residues; and a, b, c, d, a and x are (independently selected) 0, 1 or 2, with the proviso that at least one member selected from a, b, c, d, a and x are 1 or 2.
Formula 1 describes glycan structure comprising the tri-mannosyl core, which is preferably covalently linked to an asparagine residue on a peptide backbone.
Preferred expression systems will express and secrete exogenous peptides with N-linked glycans comprising the tri-mannosyl core. Using the remodeling method of the invention, the glycan structures on these peptides can be conveniently remodeled to any glycan structure desired.
Exemplary reaction conditions are found throughout the examples and in the literature.
In preferred embodiments, the glycan structures are remodeled so that the structure described in Formula 1 has specific determinates. The structure of the glycan can be chosen to enhance the biological activity of the peptide, give the peptide a new biological activity, remove the biological activity of peptide, or better approximate the glycosylation pattern of the native peptide, among others.

In the first preferred embodiment, the peptide N-linked glycans are remodeled to better approximate the glycosylation pattern of native human proteins. In this embodiment, the glycan structure described in Formula 1 is remodeled to have the following moieties:
X3 and ~5 = ~-GIcNAc-Gal-SA;
aandc=1;
d=Oorl;
b,eandx=0.
This embodiment is particularly advantageous for human peptides expressed in heterologous cellular expression systems. By remodeling the N-linked glycan structures to this configuration, the peptide can be made less immunogenic in a human patient, and/or more stable, among others.
In the second preferred embodiment, the peptide N-linked glycans are remodeled to have a bisecting GIcNAc residue on the tri-mannosyl core. In this embodiment, the glycan structure described in Formula 1 is remodeled to have the following moieties:
X3 and XS are ~-GIcNAc-Gal-SA;
aandc=1;
X4 is GIcNAc;
b=1;
d=Oorl;
eandx=0.
This embodiment is particularly advantageous for recombinant antibody molecules expressed in heterologous cellular systems. When the antibody molecule includes a Fc-mediated cellular cytotoxicity, it is known that the presence of bisected oligosaccharides linked the Fc domain dramatically increased antibody-dependent cellular cytotoxicity.
In a third preferred embodiment, the peptide N-linked glycans are remodeled to have a sialylated Lewis X moiety. In this embodiment, the glycan structure described in Formula 1 is remodeled to have the following moieties:
Fuc X and X are GIcNAc Gal-SA
a,c,d =1; ' b, a and x= 0;
X6= fucose.
This embodiment is particularly advantageous when the peptide which is being remodeling is intended to be targeted to selectin molecules and cells exhibiting the same.
In a fourth preferred embodiment, the peptide N-linked glycans are remodeled to have a conjugated moiety. The conjugated moiety may be a PEG molecule, another peptide, a small molecule such as a drug, among others. In this embodiment, the glycan structure described in Formula 1 is remodeled to have the following moieties:
X3 and XS are ~-GIcNAc-Gal-SA-R;
a and c = 1 or 2;
d=Oorl;
b, d, a and x = 0;
where R = conjugate group.
The conjugated moiety may be a PEG molecule, another peptide, a small molecule such as a drug, among others. This embodiment therefore is useful for conjugating the peptide to PEG
molecules that will slow the clearance of the peptide from the patient's bloodstream, to peptides that will target both peptides to a specific tissue or cell, or to another peptide of complementary therapeutic use.
It will be clear to one of skill in the art that the invention is not limited to the preferred glycan molecules described above. The preferred embodiments are only a few of the many useful glycan molecules that can be made by the remodeling method of the invention. Those skilled in the art will know how to design other useful glycans.
In the first exemplary embodiments, the peptide is expressed in a CHO (Chinese hamster ovarian cell line) according to methods well known in the art. When a peptide with N-linked glycan consensus sites is expressed and secreted from CHO cells, the N-linked glycans will have the structures depicted in top row of Figure 2, but also comprising a core fucose. While all of these structures may be present, by far the most common structures are the two at the right side. In the terms of Formula 1, X3 and XS are ~-GIcNAc-Gal-(SA);
aandc=1;
b, a and x = 0, and _87_ d=Oorl.
Therefore, in one exemplary embodiment, the N-linked glycans of peptides expressed in CFIO cells are remodeled to the preferred humanized glycan by contacting the peptides with a glycosyltransferase that is specific for a galactose acceptor molecule and a sialic acid donor molecule. This process is illustrated in Figure 2 and Example 17. In another exemplary embodiment, the N-linked glycans of a peptide expressed and secreted from CFIO
cells are remodeled to be the preferred PEGylated structures. The peptide is first contacted with a glycosidase specific for sialic acid to remove the terminal SA moiety, and then contacted with a glycosyltransferase specific for a galactose acceptor moiety and an sialic acid acceptor moiety, in the presence of PEG- sialic acid-nucleotide donor molecules.
Optionally, the peptide may then be contacted with a glycosyltransferase specific for a galactose acceptor moiety and an sialic acid acceptor moiety, in the presence of sialic,acid-nucleotide donor molecules to ensure complete the SA capping of all of the glycan molecules.
In other exemplary embodiments, the peptide is expressed in insect cells, such as the sf9 cell line, according to methods well known in the art. When a peptide with N-linked glycan consensus sites is expressed and secreted from sf~3 cells, the N-linked glycans will often have the structures depicted in top row of Figure 6. In the terms of Formula 1:
X3 and XS are ~- GIcNAc;
aandc=Oor 1;
b = 0;
X6 is fucose, d = 0, 1 or 2; and eandx=0.
The trimannose core is present in the vast majority of the N-linked glycans made by insect cells, and sometimes an antennary GIcNAc and/or fucose residues) are also present.
Note that the glycan may have no core fucose, it may have a single core fucose having either linkage, or it may have a single core fucose with a perponderance of a single linkage. In one exemplary embodiment, the N-linked glycans of a peptide expressed and secreted from insect cells is remodeled to the preferred humanized glycan by first contacting the glycans with a glycosidase specific to fucose molecules, then contacting the glycans with a glycosyltransferases specific to the mannose acceptor molecule on each antennary of the _88_ trimannose core, a GIcNAc donor molecule in the presence of nucleotide-GIcNAc molecules;
then contacting the glycans with a glycosyltransferase specific to a GIcNAc acceptor molecule, a Gal donor molecule in the presence of nucleotide-Gal molecules;
and then contacting the glycans with a glyeosyltransferase specific to a galactose acceptor molecule, a sialic acid donor molecule in the presence of nucleotide-SA molecules. One of skill in the art will appreciate that the fucose molecules, if any, can be removed at any time during the procedure, and if the core fucose is of the same alpha 1,6 linkage as found in human glycans, it may be left intact. In another exemplary embodiment, the humanised glycan of the previous example is remodeled further to the sialylated Lewis X glycan by contacting the glycan further with a glycosyltransferase specific to a GIcNAc acceptor molecule, a fucose donor molecule in the presence of nucleotide-fucose molecules. This process is illustrated in Figure 11 and Example 39.
In yet other exemplary embodiments, the peptide is expressed in yeast, such as Saccharomyces cerevisiae, according to methods well known in the art. When a peptide with N-linked glycan consensus sites is expressed and secreted from S. cerevisiae cells, the N-linked glycans will have the structures depicted at the left in Figure 4. The N-linked glycans will always have the trimannosyl core, which will often be elaborated with mannose or related polysaccharides of up to 1000 residues. In the terms of Formula 1:
X3 and XS = I-Man - Man - (Man)o_iooo a and c =1 or 2;
b, d, a and x = 0.
In one exemplary embodiment, the N-linked glycans of a peptide expressed and secreted from yeast cells are remodeled to the elemental trimannose core by first contacting the glycans with a glycosidase specific to a2 mannose molecules, then contacting the glycans with a glycosidase specific to a6 mannose molecules. This process is illustrated in Figure 4 and Example 3 8.
In another exemplary embodiment, the N-linked glycans are further remodeled to make a glycan suitable for an recombinant antibody with Fc-mediated cellular toxicity function by contacting the elemental trimannose core glycans with a glycosyltransferase specific to the mannose acceptor molecule on each antennary of the trimannose core and a GIcNAc donor molecule in the presence of nucleotide-GIcNAc molecules. Then, the glycans are contacted with a glycosyltransferase specific to the acceptor mannose molecule in the middle of the trimannose core, a GIcNAc donor molecule in the presence of nucleotide-GIcNAc molecules and further contacting the glycans with a glycosyltransferase specific to a GIcNAc acceptor molecule, a Gal donor molecule in the presence of nucleotide,-Gal molecules; and then optionally contacting the glycans with a glycosyltransferase specific to a galactose acceptor molecule and further optionally a sialic acid donor molecule in the presence of nucleotide-SA molecules. This process is illustrated in Figures 1, 2 and 3.
In another exemplary embodiment, the peptide is expressed in bacterial cells, in particular E. coli cells, according to methods well known in the art. When a peptide with N-linked glycans consensus sites is expressed in E. coli cells, the N-linked consensus sites will not be glycosylated. In an exemplary embodiment, a humanized glycan molecule is built out from the peptide backbone by contacting the peptides with a glycosyltransferase specific for a N-linked consensus site and a GIcNAc donor molecule in the presence of nucleotide-GIcNAc; and further sequentially contacting the growing glycans with glycosyltransferases specific for the acceptor and donor moieties in the present of the required donor moiety until the desired glycan structure is completed. When a peptide with N-linked glycans is expressed in a eukaryotic cells but without the proper leader sequences that direct the nascent peptide to the golgi apparatus, the mature peptide is likely not to be glycosylated. In this case as well the peptide may be given N-linked glycosylation by building out from the peptide N-linked consensus site as aforementioned. When a protein is chemically modified with a sugar moiety, it can be built out as aforementioned.
These examples are meant to illustrate the invention, and not to limit it. One of skill in the art will appreciate that the steps taken in each example may in some circumstances be able to be performed in a different order to get the same result. One of skill in the art will also understand that a different set of steps may also produce the same resulting glycan. The preferred remodeled glycan is by no means specific to the expression system that the peptide is expressed in. The remodeled glycans are only illustrative and one of skill in the art will know how to take the principles from these examples and apply them to peptides produced in different expression systems to make glycans not specifically described herein.

B Method to remodel O-linked ~lycans O-glycosylation is characterized by the attachment of a variety of monosaccharides in an O-glycosidic linkage to hydroxy amino acids. O-glycosylation is a widespread post-translational modification in the animal and plant kingdoms. The structural complexity of glycans O-linked to proteins vastly exceeds that of N-linked glycans. Serine or threonine residues of a newly translated peptide become modified by virtue of a peptidyl GaINAc transferase in the cis to trans compartments of the Golgi. The site of O-glycosylation is determined not only by the sequence specificity of the glycosyltransferase, but also epigenetic regulation mediated by competition between different substrate sites and competition with other glycosyltransferases responsible for forming the glycan.
The O-linked glycan has been arbitrarily defined as having three regions: the core, the backbone region and the peripheral region. The "core" region of an O-linked glycan is the inner most two or three sugars of the glycan chain proximal to the peptide.
The backbone region mainly contributes to the length of the glycan chain formed by uniform elongation.
The peripheral region exhibits a high degree of structural complexity. The structural complexity of the O-linked glycans begins with the core structure. In most cases, the first sugar residue added at the O-linked glycan consensus site is GaINAc; however the sugar may also be GIcNAc, glucose, mannose, galactose or fucose, among others. Figure 12 is a diagram of some of the known O-linked glycan core structures and the enzymes responsible for their in vivo synthesis.
In mammalian cells, at least eight different O-linked core structures are found, all based on a core-a-GaINAc residue. The four core structures depicted in Figure 13 are the most common. Core 1 and core 2 are the most abundant structures in mammalian cells, and core 3 and core 4 are found in more restricted, organ-characteristic expression systems. O-linked glycans are reviewed in Montreuil, Structure and Synthesis of Glycopeptides, In Polysaccharides in Medicinal Applications, pp. 273-327, 1996, Eds. Severian Damitriu, Marcel Dekker, NY, and in Schachter and Brockhausen, The Biosynthesis of Branched O-Linked Glycans, 199, Society for Experimental Biology, pp. 1-26 (Great Britain).
It will be apparent from the present disclosure that the glycan structure of O-glycosylated peptides can be remodeled using similar techniques to those described for N
linked glycans. O-glycans differ from N-glycans in that they are linked to a serine or threonine residue rather than an asparagine residue. As described herein with respect to N-glycan remodeling, hydrolytic enzymes can be used to cleave unwanted sugar moieties in an O-linked glycan and additional desired sugars can then be added thereto, to build a customized O-glycan structure on the peptide (See Figures 12 and 13).
The initial step in O-glycosylation in mammalian cells is the attachment of N-acetylgalactosamine (GaINAc) using any of a family of at least eleven known a-N-acetylgalactosaminyltransferases, each of which has a restricted acceptor peptide specificity.
Generally, the acceptor peptide recognized by each enzyme constitutes a sequence of at least ten amino acids. Peptides that contain the amino acid sequence recognized by one particular GaINAc-transferase become O-glycosylated at the acceptor site if they are expressed in a cell expressing the enzyme and if they are appropriately localized to the Golgi apparatus where UDP-GaINAc is also present.
However, in the case of recombinant proteins, the initial attachment of the GaINAc may not take place. The a-N-acetylgalactosaminyltransferase enzyme native to the expressing cell may have a consensus sequence specificity which differs from that of the recombinant peptide being expressed.
The desired recombinant peptide may be expressed in a bacterial cell, such as E. coli, that does not synthesize glycan chains. In these cases, it is advantageous to add the initial GaINAc moiety in vitro. The GaINAc moiety can be introduced in vitro onto the peptide once the recombinant peptide has been recovered in a soluble form, by contacting the peptide with the appropriate GaINAc transferase in the presence of UDP-GaINAc.
In one embodiment, an additional sequence of amino acids that constitute an effective acceptor for transfer of an O-linked sugar may be present. Such an amino acid sequence is encoded by a DNA sequence fused in frame to the coding sequence of the peptide, or alternatively, may be introduced by chemical means. The peptide may be otherwise lacking glycan chains. Alternately, the peptide may have N- and/or O-linked glycan chains but require an additional glycosylation site, for example, when an additional glycan substituent is desired.
In an exemplary embodiment, the amino acid sequence PTTTI~-COON, which is the natural GaINAc acceptor sequence in the human mucin MUC-1, is added as a fusion tag. The fusion protein is then expressed in E. coli and purified. The peptide is then contacted with recombinant human GaINAc-transferases T3 or T6 in the presence of UDP-GaINAc to transfer a GaINAc residue onto the peptide in vitro.
This glycan chain on the peptide may then be further elongated using the methods described in reference to the N-linked or O-linked glycans herein.
Alternatively, the GaINAc transferees reaction can be carried out in the presence of UDP-GaINAc to which PEG is covalently substituted in the O-3, 4, or 6 positions or the N-2 position.
Glycoconjugation is described in detail elswhere herein. Any antigenicity introduced into the peptide by the new peptide sequence can be conveniently masked by PEGylation of the associated glycan. The acceptor site fusion technique can be used to introduce not only a PEG moiety, but to introduce other glycan and non-glycan moieties, including, but not limited to, toxins, anti-infectives, cytotoxic agents, chelators for radionucleotides, and glycans with other functionalities, such as tissue targeting.
Exemplary Embodiments The remodeling of O-linked glycosylation is best illustrated with reference to Formula 2:
( ~ 9)m 2 ~AA-GaINAc-(Gal)f-X
(X1 0)n Formula 2 describes a glycan structure comprising a GaINAc which is covalently linked preferably to a serine or threonine residue on a peptide backbone. While this structure is used to illustrate the most common forms of O-linked glycans, it should not be construed to limit the invention solely to these O-linked glycans. Other forms of O-linked glycans are illustrated in Figure 12. Preferred expression systems useful in the present invention express and secrete exogenous peptides having O-linked glycans comprising the GaINAc residue.
Using the remodeling methods of the invention, the glycan structures on these peptides can be conveniently remodeled to generate any desired glycan structure. One of skill in the art will appreciate that O-linked glycans can be remodeled using the same principles, enzymes and reaction conditions as those available in the art once armed with the present disclosure.
Exemplary reaction conditions are found throughout the Examples.

In preferred embodiments, the glycan structures are remodeled so that the structure described in Formula 2 has specific moieties. The structure of the glycan may be chosen to enhance the biological activity of the peptide, confer upon the peptide a new biological activity, remove or alter a biological activity of peptide, or better approximate the glycosylation pattern of the native peptide, among others.
In the first preferred embodiment, the peptide O-linked glycans axe remodeled to better approximate the glycosylation pattern of native human proteins. In this embodiment, the glycan structure described in Formula 2 is remodeled to have the following moieties:
X2 is ~-SA; or ~-SA-SA;
fandn=Oorl;
Xl° is SA;
m = 0.
This embodiment is particularly advantageous for human peptides expressed in heterologous cellular expression systems. By remodeling the O-linked glycan structures to have this configuration, the peptide can be rendered less immunogenic in a human patient and/or more stable.
In the another preferred embodiment, the peptide O-linked glycans axe remodeled to display a sialylated Lewis X antigen. In this embodiment, the glycan structure described in Formula 2 is remodeled to have the following moieties:
X2 is ~-SA;
Xl° is Fuc or ~-GIcNAc(Fuc)-Gal-SA;
fandn= 1;
m = 0.
This embodiment is particularly advantageous when the peptide which is being remodeled is most effective when targeted to a selectin molecule and cells exhibiting the same.
In a yet another preferred embodiment, the peptide O-linked glycans are remodeled to contain a conjugated moiety. The conjugated moiety may be a PEG molecule, another peptide, a small molecule such as a drug, among others. In this embodiment, the glycan structure described in Formula 2 is remodeled to have the following moieties:
XZ is ~-SA-R;
f= 1;

nandm=0;
where R is the conjugate group.
This embodiment is useful for conjugating the peptide to PEG molecules that will slow the clearance of the peptide from the patient's bloodstream, to peptides that will target both peptides to a specific tissue or cell or to another peptide of complementary therapeutic use.
It will be clear to one of skill in the art that the invention is not limited to the preferred glycan molecules described above. The preferred embodiments are only a few of the many useful glycan molecules that can be made using the remodeling methods of the invention.
Those skilled in the art will know how to design other useful glycans once armed with the present invention.
In the first exemplary embodiment, the peptide is expressed in a CHO (Chinese hamster cell line) according to methods well known in the art. When a peptide with O-linked glycan consensus sites is expressed and secreted from CHO cells, the majority of the O-linked glycans will often have the structure, in the terms of Formula 2, X2=~-SA;
f= 1;
mandn=0.
Therefore, most of the glycans in CHO cells do not require remodeling in order to be acceptable for use in a human patient. In an exemplary embodiment, the O-linked glycans of a peptide expressed and secreted from a CHO cell are remodeled to contain a sialylated Lewis X structure by contacting the glycans with a glycosyltransferase specific for the GaINAc acceptor moiety and the fucose donor moiety in the presence of nucleotide-fucose.
This process is illustrated on N-linked glycans in Figure 11 and Example 39.
In other exemplary embodiments, the peptide is expressed in insect cells such as sf9 according to methods well known in the art. When a peptide having O-linked glycan consensus sites is expressed and secreted from most s~ cells, the majority of the O-linked glycans have the structure, in the terms of Formula 2:
XZ = H;
f=Oorl;
nandm=0 See, for example, Marchal et al., (2001, Biol. Chem. 382:151-159). In one exemplary embodiment, the O-linked glycan on a peptide expressed in an insect cell is remodeled to a humanized glycan by contacting the glycans with a glycosyltransferase specific for a GaINAc acceptor molecule and a galactose donor molecule in the presence of nucleotide-Gal; and then contacting the glycans with a glycosyltransferase specific for a Gal aceeptor molecule and a SA donor molecule in the presence of nucleotide-SA. In another exemplary embodiment, the O-linked glycans are remodeled further from the humanized form to the sialylated Lewis ~ form by fiu-ther contacting the glycans with a glycosyltransferase specific for a GaINAc acceptor molecule and a fucose donor molecule in the presence of nucleotide-fucose.
In yet another exemplary embodiment, the peptide is expressed in fungal cells, in particular S. cerevisiae cells, according to methods well known in the art.
When a peptide with O-linked glycans consensus sites is expressed and secreted from S
ce~evisiae cells, the majority of the O-linked glycans have the structure:
~ - AA-Man- Manl_2.
See Gemmill and Trimble (1999, Biochim. Biophys. Acta 1426:227-237). In order to remodel these O-linked glycans for use in human, it is preferable that the glycan be cleaved at the amino acid level and rebuilt from there.
In an exemplary embodiment, the glycan is the O-linked glycan on a peptide expressed in a fungal cell and is remodeled to a humanized glycan by contacting the glycan with an endoglycosylase specific for an amino acid - GaINAc bond; and then contacting the glycan with a glycosyltransferase specific for a O-linked consensus site and a GaINAc donor molecule in the presence of nucleotide-GaINAc; contacting the glycan with a glycosyltransferase specific for a GaINAc acceptor molecule and a galactose donor molecule in the presence of nucleotide-Gal; and then contacting the glycans with a glycosyltransferase specific for a Gal acceptor molecule and a SA donor molecule in the presence of nucleotide-SA.
Alternately, in another exemplary embodiment, the glycan is the O-linked glycan on a peptide expressed in a fungal cell and is remodeled to a humanized glycan by contacting the glycan with an protein O-mannose [3-1,2-N-acetylglucosaminyltransferase (POMGnTI) in the presence of GlcNAc-nucleotide; then contacting the glycan with an galactosyltransferase in the presence of nucleotide-Gal; and then contracting the glycan with an sialyltransferase in the presence of nucleotide-SA.
In another exemplary embodiment, the peptide is expressed in bacterial cells, in particular E. c~li cells, according to methods well known in the art. When a peptide with an O-linked glycan consensus site is expressed in E. a~li cells, the O-linked consensus site will not be glycosylated. In this case, the desired glycan molecule must be built out from the peptide backbone in a manner similar to that describe for S.
cef°evisiae expression above.
Further, when a peptide having an O-linked glycan is expressed in a eukaryotic cell without the proper leader sequences to direct the nascent peptide to the golgi apparatus, the mature peptide is likely not to be glycosylated. In this case as well, an O-linked glycosyl structure may be added to the peptide by building out the glycan directly from the peptide O-linked consensus site. Further, when a protein is chemically modified with a sugar moiety, it can also be remodeled as described herein.
These examples are meant to illustrate the invention, and not to limit it in any way.
One of skill in the art will appreciate that the steps taken in each example may in some circumstances be performed in a different order to achieve the same result.
One of skill in the art will also understand that a different set of steps may also produce the same resulting glycan. Futher, the preferred remodeled glycan is by no means specific to the expression system that the peptide is expressed in. The remodeled glycans are only illustrative and one of skill in the art will know how to take the principles from these examples and apply them to peptides produced in different expression systems to generate glycans not specifically described herein.
C Glycoconiu~ation, in general The invention provides methods of preparing a conjugate of a glycosylated or an unglycosylated peptide. The conjugates of the invention are formed between peptides and diverse species such as water-soluble polymers, therapeutic moieties, diagnostic moieties, targeting moieties and the like. Also provided are conjugates that include two or more peptides linked together through a linker arm, i.e., multifunctional conjugates. The multi-functional conjugates of the invention can include two or more copies of the same peptide or a collection of diverse peptides with different structures, andlor properties.

The conjugates of the invention are formed by the enzymatic attachment of a modified sugar to the glycosylated or unglycosylated peptide. The modif ed sugar, when interposed between the peptide and the modifying group on the sugar becomes what is referred to herein as "an intact glycosyl linking group.'9 Using the exquisite selectivity of enzymes, such as glycosyltransferases, the present method provides peptides that bear a desired group at one or more specific locations. Thus, according to the present invention, a modified sugar is attached directly to a selected locus on the peptide chain or, alternatively, the modified sugar is appended onto a carbohydrate moiety of a peptide.
Peptides in which modified sugars are linked to both a peptide carbohydrate and directly to an amino acid residue of the peptide backbone are also within the scope of the present invention.
In contrast to known chemical and enzymatic peptide elaboration strategies, the methods of the invention make if possible to assemble peptides and glycopeptides that have a substantially homogeneous derivatization pattern; the enzymes used in the invention are generally selective for a particular amino acid residue or combination of amino acid residues of the peptide or particular glycan structure. The methods are also practical for large-scale production of modified peptides and glycopeptides. Thus, the methods of the invention provide a practical means for large-scale preparation of peptides having preselected substantially uniform derivatization patterns. The methods are particularly well suited for modification of therapeutic peptides, including but not limited to, peptides that are incompletely glycosylated during production in cell culture cells (e.g., mammalian cells, insect cells, plant cells, fwgal cells, yeast cells, or prokaryotic cells) or transgenic plants or animals.
The methods of the invention also provide conjugates of glycosylated and unglycosylated peptides with increased therapeutic half life due to, for example, reduced clearance rate, or reduced rate of uptake by the immune or reticuloendothelial system (RES).
Moreover, the methods of the invention provide a means for masking antigenic determinants on peptides, thus reducing or eliminating a host immune response against the peptide.
Selective attachment of targeting agents can also be used to target a peptide to a particular tissue or cell surface receptor that is specific for the particular targeting agent. Moreover, there is provided a class of peptides that are specifically modified with a therapeutic moiety.

1. The Coniu~ates In a first aspect, the present invention provides a conjugate between a peptide and a selected moiety. The link between the peptide and the selected moiety includes an intact glycosyl linking group interposed between the peptide and the selected moiety.
As discussed herein, the selected moiety is essentially any species that can be attached to a saccharide unit, resulting in a "modified sugar" that is recognized by an appropriate transferase enzyme, which appends the modified sugar onto the peptide. The saccharide component of the modified sugar, when interposed between the peptide and a selected moiety, becomes an "intact glycosyl linking group." The glycosyl linking group is formed from any mono- or oligo-saccharide that, after modification with a selected moiety, is a substrate for an appropriate transferase.
The conjugates of the invention will typically correspond to the general structure:
Peptide Sugar Linker Sugar t Agent a b c d in which the symbols a, b, c, d and s represent a positive, non-zero integer;
and t is either 0 or a positive integer. The "agent" is a therapeutic agent, a bioactive agent, a detectable label, water-soluble moiety or the like. The "agent" can be a peptide, e.g., enzyme, antibody, antigen, etc. The linker can be any of a wide array of linking groups, infra. Alternatively, the linker may be a single bond or a "zero order linker." The identity of the peptide is without limitation. Exemplary peptides are provided in Figure 2~.
In an exemplary embodiment, the selected moiety is a water-soluble polymer.
The water-soluble polymer is covalently attached to the peptide via an intact glycosyl linking group. The glycosyl linking group is covalently attached to either an amino acid residue or a glycosyl residue of the peptide. Alternatively, the glycosyl linking group is attached to one or more glycosyl units of a glycopeptide. The invention also provides conjugates in which the glycosyl linking group is attached to both an amino acid residue and a glycosyl residue.
In addition to providing conjugates that are formed through an enzymatically added intact glycosyl linking group, the present invention provides conjugates that are highly homogenous in their substitution patterns. Using the methods of the invention, it is possible to form peptide conjugates in which essentially all of the modified sugar moieties across a population of conjugates of the invention are attached to multiple copies of a structurally identical amino acid or glycosyl residue. Thus, in a second aspect, the invention provides a peptide conjugate having a population of water-soluble polymer moieties, which are covalently linked to the peptide through an intact glycosyl linking group. In a preferred conjugate of the invention, essentially each member of the population is linked via the glycosyl linking group to a glycosyl residue of the peptide, and each glycosyl residue of the peptide to which the glycosyl linking group is attached has the same structure.
Also provided is a peptide conjugate having a population of water-soluble polymer moieties covalently linked thereto through an intact glycosyl linking group.
In a preferred embodiment, essentially every member of the population of water soluble polymer moieties is linked to an amino acid residue of the peptide via an intact glycosyl linking group, and each amino acid residue having an intact glycosyl linking group attached thereto has the same structure.
The present invention also provides conjugates analogous to those described above in which the peptide is conjugated to a therapeutic moiety, diagnostic moiety, targeting moiety, toxin moiety or the like via an intact glycosyl linking group. Each of the above-recited moieties can be a small molecule, natural polymer (e.g., peptide) or synthetic polymer.
In an exemplary embodiment, interleukin-2 (IL-2) is conjugated to transferrin via a bifunctional linker that includes an intact glycosyl linking group at each terminus of the PEG
moiety (Scheme 1). For example, one terminus of the PEG linker is functionalized with an intact sialic acid linker that is attached to transferrin and the other is functionalized with an intact GaINAc linker that is attached to IL-2.
In another exemplary embodiment, EPO is conjugated to transferrin. In another exemplary embodiment, EPO is conjugated to glial derived neurotropic growth factor (GDNF). In these embodiments, each conjugation is accomplished via a bifunctional linker that includes an intact glycosyl linking group at each terminus of the PEG
moiety, as aforementioned. Transferrin transfers the protein across the blood brain barrier.
As set forth in the Figures appended hereto, the conjugates of the invention can include intact glycosyl linking groups that are mono- or multi-valent (e.g., antennary structures), see, Figures 14-22. The conjugates of the invention also include glycosyl linking groups that are O-linked glycans originating from serine or threonine (Figure 11). Thus, conjugates of the invention include both species in which a selected moiety is attached to a peptide via a monovalent glycosyl linking group. Also included within the invention are conjugates in which more than one selected moiety is attached to a peptide via a multivalent linking group. One or more proteins can be conjugated together to take advantage of their biophysical and biological properties.
In a still further embodiment, the invention provides conjugates that localize selectively in a particular tissue due to the presence of a targeting agent as a component of the conjugate. In an exemplary embodiment, the targeting agent is a protein.
Exemplary proteins include transferrin (brain, blood pool), human serum (HS)-glycoprotein (bone, brain, blood pool), antibodies (brain, tissue with antibody-specific antigen, blood pool), coagulation Factors V-XII (damaged tissue, clots, cancer, blood pool), serum proteins, e.g., a-acid glycoprotein, fetuin, a-fetal protein (brain, blood pool), (32-glycoprotein (liver, atherosclerosis plaques, brain, blood pool), G-CSF, GM-CSF, M-CSF, and EPO
(immune stimulation, cancers, blood pool, red blood cell overproduction, neuroprotection), and albumin (increase in half life).
In addition to the conjugates discussed above, the present invention provides methods for preparing these and other conjugates. Thus, in a further aspect, the invention provides a method of forming a covalent conjugate between a selected moiety and a peptide.
Additionally, the invention provides methods for targeting conjugates of the invention to a particular tissue or region of the body.
Tn exemplary embodiments, the conjugate is formed between a water-soluble polymer, a therapeutic moiety, targeting moiety or a biomolecule, and a glycosylated or non-glycosylated peptide. The polymer, therapeutic moiety or biomolecule is conjugated to the peptide via an intact glycosyl linking group, which is interposed between, and covalently linked to both the peptide and the modifying group (e.g., water-soluble polymer). The method includes contacting the peptide with a mixture containing a modified sugar and a glycosyltransferase for which the modified sugar is a substrate. The reaction is conducted under conditions sufficient to form a covalent bond between the modified sugar and the peptide. The sugar moiety of the modif ed sugar is preferably selected from nucleotide sugars, activated sugars and sugars, which are neither nucleotides nor activated.

In one embodiment, the invention provides a method for linking two or more peptides through a linking group. The linking group is of any useful structure and may be selected from straight-chain and branched chain struciwres. Preferably, each terminus of the linker, which is attached to a peptide, includes a modified sugar (i.e., a nascent intact glycosyl linking group).
In an exemplary method of the invention, two peptides are linked together via a linker moiety that includes a PEG linker. The construct conforms to the general structure set forth in the cartoon above. As described herein, the construct of the invention includes two intact glycosyl linking groups (i.e., s + t = 1). The focus on a PEG linker that includes two glycosyl groups is for purposes of clarity and should not be interpreted as limiting the identity of linker arms of use in this embodiment of the invention.
Thus, a PEG moiety is functionalized at a first terminus with a first glycosyl unit and at a second terminus with a second glycosyl unit. The first and second glycosyl units are preferably substrates for different transferases, allowing orthogonal attachment of the first and second peptides to the first and second glycosyl units, respectively. In practice, the (glycosyl)1-PEG-(glycosyl)2 linker is contacted with the first peptide and a first transferase for which the first glycosyl unit is a substrate, thereby fornling (peptide)1-(glycosyl)1-PEG-(glycosyl)2. The first transferase and/or unreacted peptide is then optionally removed from the reaction mixture. The second peptide and a second transferase for which the second glycosyl unit is a substrate are added to the (peptide)1-(glycosyl)1-PEG-(glycosyl)2 conjugate, forming (peptide)1-(glycosyl)1-PEG-(glycosyl)2-(peptide)2 . Those of skill in the art will appreciate that the method outlined above is also applicable to forming conjugates between more than two peptides by, for example, the use of a branched PEG, dendrimer, poly(amino acid), polysaccharide or the like.
As noted previously, in an exemplary embodiment, interleukin-2 (IL-2) is conjugated to transferrin via a bifimctional linker that includes an intact glycosyl linking group at each terminus of the PEG moiety (Scheme 1). The IL-2 conjugate has an ivc vivo half life that is increased over that of IL-2 alone by virtue of the greater molecular size of the conjugate.
Moreover, the conjugation of IL-2 to transferrin serves to selectively target the conjugate to the brain. For example, one terminus of the PEG linker is functionalized with a CMP-sialic acid and the other is functionalized with an UDP-GaINAc. The linker is combined with IL-2 in the presence of a GaINAc transferees, resulting in the attachment of the GaINAc of the linker arm to a serine and/or threonine residue on the IL-2.
In another exemplary embodiment, transfemin is conjugated to a nucleic acid for use in gene therapy.
Scheme 1 SA sialidase Gal transferrin transferrin SA ~ ~--Gal 1. sialyltransferase CMP-SA-PEG-GaINAc-UDP
2. GaINAc transferase IL,-2 Gal-SA-PEG-GaINAc-I L-2 transferrin Gal-SA-PEG-GaINAc-IL-2 The processes described above can be carried through as many cycles as desired, and is not limited to forming a conjugate between two peptides with a single linker. Moreover, those of skill in the art will appreciate that the reactions functionalizing the intact glycosyl linking groups at the termini of the PEG (or other) linker with the peptide can occur simultaneously in the same reaction vessel, or they can be carried out in a step-wise fashion.
When the reactions are carried out in a step-wise manner, the conjugate produced at each step is optionally purified from one or more reaction components (e.g., enzymes, peptides).
A still further exemplary embodiment is set forth in Scheme 2. Scheme 2 shows a method of preparing a conjugate that targets a selected protein, e.g., EPO, to bone and increases the circulatory half life of the selected protein.

Scheme 2 /---Gal CMP-SA-PEG-Gal-UDP ~ ~CTaI-SA-PEG-Gal-UDP
HSGP N HSGP N
Gal s~alyltransfernse ~---Gal-SA-PEG-Gal-UDP
EPO
galactosyltransferase Gal-SA-PEG-Gal-EPO
HSGP N
---Gal-SA-PEG-Gal-EPO
The use of reactive derivatives of PEG (or other linkers) to attach one or more peptide moieties to the linker is within the scope of the present invention. The invention is not limited by the identity of the reactive PEG analogue. Many activated derivatives of polyethylene glycol) are available commercially and in the literature. It is well within the abilities of one of skill to choose, and synthesize if necessary, an appropriate activated PEG
derivative with which to prepare a substrate useful in the present invention.
See, Abuchowski et al. Cancer Bioehem. Biophys., 7: 175-186 (1984); Abuchowski et al., J.
Biol. Chem., 252:
3582-3586 (1977); Jackson et al., Anal. BioclZenZ.,165: 114-127 (1987); Koide et al., Biochem Biophys. Res. Commun., 111: 659-667 (1983)), tresylate (Nilsson et al., Methods Enzyrnol., 104: 56-69 (1984); I~elgado et al., Biotechnol. Appl. Biochem.,12:

(1990)); N-hydroxysuccinimide derived active esters (Buckmann et al., Mala°ornol. Chem., 182: 1379-1384 (198I); Joppich et al., Mal~omol. Chena., 180: 1381-1384 (1979);
Abuchowski et al., Cancer Biochem. Biophys., 7: 175-186 (1984); Katreet al.
Proc. Natl.
Acad. Sci. U.S.A., 84: 1487-1491 (1987); Kitarnura et al., Cancer Res., 51:

(1991); Boccu et al., Z Naturfo~sch., 38C: 94-99 (1983), carbonates (Zalipsky et al., POLYETHYLENE GLYCOL) CHEMISTRY: BIOTECHNICAL AND BIOMEDICAL APPLICATIONS, Harris, Ed., Plenum Press, New York, 1992, pp. 347-370; Zalipsky et al., Biotechnol. Appl.
Biochem.,15: 100-I I4 (1992); Veronese et al., Appl. Biochem. Biotech.,11: 141-(1985)), imidazolyl formates (Beauchamp et al., Anal. Biochem.,131: 25-33 (1983); Berger et al., Blood, 7I: 1641-1647 (1988)), 4-dithiopyridines (~oghiren et al., Bioconjugate Chem., 4: 314-318 (1993)), isocyanates (Byun et al., ASAI~ Jourwal, M649-M-653 (1992)) and epoxides (LT.S. Pat. No. 4,806,595, issued to Noishiki et al., (1989).
~ther linking groups include the urethane linkage between amino groups and activated PEG. See, Veronese, et al., Appl. Biochern. Bi~tecl2nol., 11: 141-152 (1985).

In another exemplary embodiment in which a reactive PEG derivative is utilized, the invention provides a method for extending the blood-circulation half life of a selected peptide, in essence targeting the peptide to the blood pool, by conjugating the peptide to a synthetic or natural polymer of a size sufficient to retard the filtration of the protein by the glomerulus (e.g., albumin). This embodiment of the invention is illustrated in scheme 3 in which erythropoietin (EPO) is conjugated to albumin via a PEG linker using a combination of chemical and enzymatic modification.
Scheme 3 CMP-SA-PEG-X
albumin albumin PEG-SA-CMP
X = Activating group ~ EPO
sialyltransferase albumin PEG-SA- EP0 Thus, as shown in Scheme 3, an amino acid residue of albumin is modified with a reactive PEG derivative, such as X-PEG-(CMP-sialic acid), in which X is an activating group (e.g., active ester, isothiocyanate, etc). The PEG derivative and EPO are combined and contacted with a transferase for which CMP-sialic acid is a substrate. In a further illustrative embodiment, an s-amine of lysine is reacted with the N-hydroxysuccinimide ester of the PEG-linker to form the albumin conjugate. The CMP-sialic acid of the linker is enzymatically conjugated to an appropriate residue on EPO, e.g., Gal, thereby forming the conjugate. Those of skill will appreciate that the above-described method is not limited to the reaction partners set forth. Moreover, the method can be practiced to form conjugates that include more than two protein moieties by, for example, utilizing a branched linker having more than two termini.
2. Modified Sugars Modified glycosyl donor species ("modified sugars") are preferably selected from modified sugar nucleotides, activated modified sugars and modified sugars that are simple saccharides that are neither nucleotides nor activated. Any desired carbohydrate structure can be added to a peptide using the methods of the invention. Typically, the structure will be a monosaccharide, but the present invention is not limited to the use of modified monosaccharide sugars; oligosaccharides and polysaccharides are useful as well.
'The modifying group is attached to a sugar moiety by er~ymatic means, chemical means or a combination thereof, thereby producing a modified sugar. The sugars are substituted at any position that allows for the attachment of the modifying moiety, yet which still allows the sugar to function as a substrate for the enzyme used to ligate the modified sugar to the peptide. In a preferred embodiment, when sialic acid is the sugar, the sialic acid is substituted with the modifying group at either the 9-position on the pyruvyl side chain or at the 5-position on the amine moiety that is normally acetylated in sialic acid.
In certain embodiments of the present invention, a modified sugar nucleotide is utilized to add the modified sugar to the peptide. Exemplary sugar nucleotides that are used in the present invention in their modified form include nucleotide mono-, di-or triphosphates or analogs thereof. In a preferred embodiment, the modified sugar nucleotide is selected from a UDP-glycoside, CMP-glycoside, or a GDP-glycoside. Even more preferably, the modified sugar nucleotide is selected from an UDP-galactose, UDP-galactosamine, UDP-glucose, UDP-glucosamine, GDP-mannose, GDP-fucose, CMP-sialic acid, or CMP-NeuAc.
N-acetylamine derivatives of the sugar nucleotides are also of use in the method of the invention.
The invention also provides methods for synthesizing a modified peptide using a modified sugar, e.g., modified-galactose, -fucose, and -sialic acid. When a modified sialic acid is used, either a sialyltransferase or a trans-sialidase (for a2,3-linked sialic acid only) can be used in these methods.
In other embodiments, the modified sugar is an activated sugar. Activated modified sugars, which are useful in the present invention are typically glycosides which have been synthetically altered to include an activated leaving group. As used herein, the term "activated leaving group" refers to those moieties, which are easily displaced in enzyme-regulated nucleophilic substitution reactions. Many activated sugars are known in the art.
See, for example, Vocadlo et al., In CARBOHYDRATE CHEMISTRY AND BIOLOGY, Vol.
2, Ernst et al. Ed., Wiley-VCH Verlag: Weinheim, Germany, 2000; I~odama et al., Tetrahedron Lett.
34: 6419 (1993); Lougheed, et al., ,l. Biol. C'hem. 274: 37717 (1999)).
Examples of activating groups (leaving groups) include fluoro, chloro, bromo, tosylate ester, mesylate ester, triflate ester and the like. Preferred activated leaving groups, for use in the present invention, are those that do not significantly statically encumber the enzymatic transfer of the glycoside to the acceptor. Accordingly, preferred embodiments of activated glycoside derivatives include glycosyl fluorides and glycosyl mesylates, with glycosyl fluorides being particularly preferred. Among the glycosyl fluorides, a-galactosyl fluoride, a-mannosyl fluoride, a-glucosyl fluoride, a-fucosyl fluoride, a-xylosyl fluoride, a-sialyl fluoride, a-N-acetylglucosaminyl fluoride, a-N-acetylgalactosaminyl fluoride, (3-galactosyl fluoride, (3-mannosyl fluoride, [3-glucosyl fluoride, (3-fucosyl fluoride, (3-xylosyl fluoride, (3-sialyl fluoride, (3-N-acetylglucosaminyl fluoride and (3-N-acetylgalactosaminyl fluoride are most preferred.
By way of illustration, glycosyl fluorides can be prepared from the free sugar by first acetylating the sugar and then treating it with HF/pyridine. This generates the thermodynamically most stable anomer of the protected (acetylated) glycosyl fluoride (i. e., the a-glycosyl fluoride). If the less stable anomer (i.e., the (3-glycosyl fluoride) is desired, it can be prepared by converting the peracetylated sugar with HBr/HOAc or with HCl to generate the anomeric bromide or chloride. This intermediate is reacted with a fluoride salt such as silver fluoride to generate the glycosyl fluoride. Acetylated glycosyl fluorides may be deprotected by reaction with mild (catalytic) base in methanol (e.g.
NaOMe/MeOH). In addition, many glycosyl fluorides are commercially available.
Other activated glycosyl derivatives can be prepared using conventional methods known to those of skill in the art. For example, glycosyl mesylates can be prepared by treatment of the fully benzylated hemiacetal form of the sugar with mesyl chloride, followed by catalytic hydrogenation to remove the benzyl groups.
In a further exemplary embodiment, the modified sugar is an oligosaccharide having an antennary structure. In a preferred embodiment, one .or more of the termini of the antennae bear the modifying moiety. When more than one modifying moiety is attached to an oligosaccharide having an antennary structure, the oligosaccharide is useful to "amplify"

the modifying moiety; each oligosaccharide unit conjugated to the peptide attaches multiple copies of the modifying group to the peptide. The general structure of a typical chelate of the invention as set forth in the drawing above, encompasses multivalent species resulting from preparing a conjugate of the invention utilising an antennary structure. Many antemlary saccharide structures are known in the art, and the present method can be practiced with them without limitation.
Exemplary modifying groups are discussed below. The modifying groups can be selected for one or more desirable property. Exemplary properties include, but are not limited to, enhanced pharmacokinetics, enhanced pharmacodynamics, improved biodistribution, providing a polyvalent species, improved water solubility, enhanced or diminished lipophilicity, and tissue targeting.
D. Peptide Com'~ates a) Water-Soluble Polymers The hydrophilicity of a selected peptide is enhanced by conjugation with polar molecules such as amine-, ester-, hydroxyl- and polyhydroxyl-containing molecules.
Representative examples include, but are not limited to, polylysine, polyethyleneimine, polyethylene glycol) and poly(propyleneglycol). Preferred water-soluble polymers are essentially non-fluorescent, or emit such a minimal amount of fluorescence that they are inappropriate for use as a fluorescent marker in an assay. Polymers that are not naturally occurring sugars may be used. In addition, the use of an otherwise naturally occurring sugar that is modified by covalent attachment of another entity (e.g., poly(ethylene glycol), polypropylene glycol), poly(aspartate), biomolecule, therapeutic moiety, diagnostic moiety, etc.) is also contemplated. In another exemplary embodiment, a therapeutic sugar moiety is conjugated to a linker arm and the sugar-linker arm is subsequently conjugated to a peptide via a method of the invention.
Methods and chemistry for activation of water-soluble polymers and saccharides as well as methods for conjugating saccharides and polymers to various species are described in the literature. Commonly used methods for activation of polymers include activation of functional groups with cyanogen bromide, periodate, glutaraldehyde, biepoxides, epichlorohydrin, divinylsulfone, carbodiimide, sulfonyl halides, trichlorotriazine, etc. (see, R.
F. Taylor, (1991), PROTEIN IMMOBILISATION. FUNDAMENTALS AND APPLICATIONS, Marvel Dekker, N.Y.; S. S. Wong, (1992), CHEMISTRY OF PROTEIN CONJUGATION AND
CROSSLINKING, CRC Press, Boca Raton; G. T. FIermanson et al., (1993), IMMOBILIZED
AFFINITY I,IGAND TECHNIQUES, Academic Press, N.Y.; Dunn, R.L., et al., Eds.
POLYMERIC
DRUGS AND DRUG DELIVERY SYSTEMS, ACS Symposium Series Vol. 469, American Chemical Society, Washington, D.C. 1991).
Routes for preparing reactive PEG molecules and forming conjugates using the reactive molecules are kn~wn in the art. For example, U.S. Patent No.
5,672,662 disci~ses a water soluble and isolatable conjugate of an active ester ~f a p~lymer acid selected from linear or branched poly(alkylene oxides), poly(oxyethylated p~lyols), poly(olefinic alc~hols), and poly(acrylomorpholine), wherein the polymer has about 44 or more recurring units.
U.S. Patent No. 6,376,604 sets forth a method for preparing a water-soluble 1-benzotriazolylcarbonate ester of a water-soluble and non-peptidic polymer by reacting a terminal hydroxyl of the polymer with di(1-benzotriazoyl)carbonate in an organic solvent.
The active ester is used t~ form conjugates with a biologically active agent such as a protein or peptide.
WO 99/45964 describes a conjugate comprising a biologically active agent and an activated water soluble polymer comprising a polymer backbone having at least one terminus linked to the polymer backbone through a stable linkage, wherein at least one terminus comprises a branching moiety having proximal reactive groups linked to the branching moiety, in which the biologically active agent is linked to at least one of the proximal reactive groups. Other branched polyethylene glycols) are described in WO 96/21469, U.S. Patent No. 5,932,462 describes a conjugate formed with a branched PEG molecule that includes a branched terminus that includes reactive functional groups. The free reactive groups are available to react with a biologically active species, such as a protein or peptide, forming conjugates between the polyethylene glycol) and the biologically active species. U.S. Patent No. 5,446,090 describes a bifunctional PEG linker and its use in forming conjugates having a peptide at each of the PEG linker termini.
Conjugates that include degradable PEG linkages are described in WO 99/34833;
and WO 99/14259, as well as in U.S. Patent No. 6,348,558. Such degradable linkages are applicable in the present invention.

Although both reactive PEG derivatives and conjugates formed using the derivatives are known in the art, until the present invention, it was not recognized that a conjugate could be formed between PEG (or other polymer) and another species, such as a peptide or glycopeptide, through an intact glycosyl linking group.
Many water-soluble polymers are known to those of skill in the art and are useful in practicing the present invention. The term water-soluble polymer encompasses species such as saccharides (e.g., dextran, amylose, hyalouronic acid, poly(sialic acid), heparans, heparins, etc.); poly (amino acids), e.g., poly(glutamic acid); nucleic acids; synthetic polymers (e.g., poly(acrylic acid), poly(ethers), e.g., poly(ethylene glycol); peptides, proteins, and the like.
The present invention may be practiced with any water-soluble polymer with the sole limitation that the polymer must include a point at which the remainder of the conjugate can be attached.
Methods for activation of polymers can also be found in WO 94/17039, U.S. Pat.
No.
5,324,844, WO 94/18247, WO 94/04193, U.S. Pat. No: 5,219,564, U.S. Pat. No.
5,122,614, WO 90/13540, U.S. Pat. No. 5,281,698, and more WO 93/15189, and for conjugation between activated polymers and peptides, e.g. Coagulation Factor VIII (WO
94/15625), hemoglobin (WO 94/09027), oxygen carrying molecule (U.S. Pat. No. 4,412,989), ribonuclease and superoxide dismutase (Veronese at al., App. Biochem. Biotech.
11: 141-45 (1985)).
Preferred water-soluble polymers are those in which a substantial proportion of the polymer molecules in a sample of the polymer are of approximately the same molecular weight; such polymers are "homodisperse."
The present invention is further illustrated by reference to a polyethylene glycol) conjugate. Several reviews and monographs on the functionalization and conjugation of PEG
are available. See, for example, Harris, Macronol. Chena. Phys. C25: 325-373 (1985);
Scouten, Methods in Enzynaology 135: 30-65 (1987); Wong et al., Enzyme Mic~ob.
Technol.
14: 866-874 (1992); Delgado et al., Critical Reviews in Therapeutic Drug Car~riey~ Systems 9:
249-304 (1992); Zalipsky, Bioco~cjugate Chem. 6: 150-165 (1995); and Bhadra, et al., Pha~r~zazie, 57:5-29 (2002).
Polyethylene glycol) molecules suitable for use in the invention include, but are not limited to, those described by the following Formula 3:

Formula 3.
Z
(~CN2CH2)n~~<~(~H2)m R= H, alkyl, benzyl, aryl, acetal, OHC-, H2N-CH2CH2-, HS-CH2CH2-, Y
w (CH~)q Z , -sugar-nucleotide, protein, methyl, ethyl;
X, Y, W, U (independently selected) = O, S, NH, N-R';
R', R"' (independently selected) = alkyl, benzyl, aryl, alkyl aryl, pyridyl, substituted aryl, axylalkyl, acylaxyl;
n = 1 to 2000;
m, q, p (independently selected) = 0 to 20 o=Oto20;
Z = HO, NH2, halogen, S-R"', activated esters, Y Y
U
(CH2)p V ~ (CH2)p \ (CH2)p V
-sugar-nucleotide, protein, imidazole, HOBT, tetrazole, halide; and V = HO, NH2, halogen, S-R"', activated esters, activated amides, -sugar-nucleotide, protein.
In preferred embodiments, the polyethylene glycol) molecule is selected from the following:

Me-(OCH2CH2)n-O~ Z Me-(OCH2CH2)n-O~ Z
'' TOf O
~ H
Me-(OCH2CH2)n-O~Z Me-(OCH2CH~)n-'P~ ~~~ Z
O O O
Me-(OCH~CH2)n~-~ H O
Z Me-(OCH2CH2)n~N
,Z
O
Me-(OCH2CH2)n-S-Z
Me- OCH CH2)n-N-Z Me-(OCH2CH2)n~HNJ
( 2 I-I O
The polyethylene glycol) useful in forming the conjugate of the invention is either linear or branched. Branched polyethylene glycol) molecules suitable for use in the invention include, but are not limited to, those described by the following Formula:
Formula 4:
R" -V~/~(OCH2CH~)n-X .
(CH2)q m R'-A~(OCH2CH2)P-B ~Z
~''~o IIY
R', R", R"' (independently selected) = H, alkyl, benzyl, aryl, acetal, OHC-, H2N-CH2CHa-, HS-CHZCH2-, -(CH2)qCY-Z, -sugar-nucleotide, protein, methyl, ethyl, heteroaryl, acylalkyl, acylaryl, acylalkylaryl;
X,Y, W, A, B (independently selected) = O, S, NH, N-R', (CH2)i;
n, p (independently selected) = 1 to 2000;
m, q, o (independently selected) = 0 to 20;
Z = HO, NHa, halogen, S-R"', activated esters, Y Y
U
(CH~)p V , (CH~)p '(CH2)p V
-sugar-nucleotide, protein;
V = HO, NH2, halogen, S-R"', activated esters, activated amides, -sugar-nucleotide, protein.

The in vivo half life, area under the curve, and/or residence time of therapeutic peptides can also be enhanced with water-soluble polymers such as polyethylene glycol (PEG) and polypropylene glycol (PPG). Fox example, chemical modification of proteins with PEG (PEGylation) increases their molecular size and decreases their surface-and functional group-accessibility, each of which axe dependent on the size of the PEG
attached to the protein. This results in an improvement of plasma half lives and in proteolytic-stability, and a decrease in immunogenicity and hepatic uptake (Chaffee et al. .I. Clin.
hzvest. 89: 1643-1651 (1992); Pyatak et al. Res. CoyD2f72uYl. Cherrz. PatlZOl Phar~r7zaeol. 29:
113-127 (1980)).
PEGylation of interleukin-2 has been reported to increase its antitumor potency in vivo (Katre et al. Proe. Natl. Acad. Sci. USA. 84: 1487-1491 (1987)) and PEGylation of a F(ab')2 derived from the monoclonal antibody A7 has improved its tumor localization (Kitamura et al.
Biochem. Biophys. Res. Commu~c. 28: 1387-1394 (1990)).
In one preferred embodiment, the ih vivo half life of a peptide derivatized with a water-soluble polymer by a method of the invention is increased relevant to the in vivo half life of the non-derivatized peptide. In another preferred embodiment, the area under the curve of a peptide derivatized with a water-soluble polymer using a method of the invention is increased relevant to the area under the curve of the non-derivatized peptide. In another preferred embodiment, the residence time of a peptide derivatized with a water-soluble polymer using a method of the invention is increased relevant to the residence time of the non-derivatized peptide. ' Techniques to determine the i~c vivo half life, the axes under the curve and the residence time are well known in the art. Descriptions of such techniques can be found in J.G. Wagner, 1993, Pharmacokinetics for the Pharmaceutical Scientist, Technomic Publishing Company, Inc. Lancaster PA.
The increase in peptide in vivo half life is best expressed as a range of percent increase in this quantity. The lower end of the range of percent increase is about 40%, about 60%, about 80%, about 100%, about 150% or about 200%. The upper end of the range is about 60%, about 80°/~, about 100%, about 150%, or more than about 250%.
In an exemplary embodiment, the present invention provides a PEGylated follicle stimulating hormone (Examples 23 and 24). In a further exemplary embodiment, the invention provides a PEGylated transferrin (Example 42).

Other exemplary water-soluble polymers of use in the invention include, but are not limited to linear or branched poly(alkylene oxides), poly(oxyethylated polyols), poly(olefmic alcohols), and poly(acrylomorpholine), dextran, starch, poly(amino acids), etc.
b) Water-insoluble polymers The conjugates of the invention may also include one or more water-insoluble polymers. This embodiment of the invention is illustrated by the use of the conjugate as a vehicle with which to deliver a therapeutic peptide in a controlled manner.
Polymeric drug delivery systems are known in the art. fee, for example, Durm et eel., Eds.
P~LYMERIC
DRUGS AND DRUG DELIVERY SYSTEMS, ACS Symposium Series Vol. 469, American Chemical Society, Washington, D.C. 1991. Those of skill in the art will appreciate that substantially any known drug delivery system is applicable to the conjugates of the present invention.
Representative water-insoluble polymers include, but are not limited to, polyphosphazines, polyvinyl alcohols), polyamides, polycarbonates, polyalkylenes, polyacrylamides, polyalkylene glycols, polyalkylene oxides, polyalkylene terephthalates, polyvinyl ethers, polyvinyl esters, polyvinyl halides, polyvinylpyrrolidone, polyglycolides, polysiloxanes, polyurethanes, poly(methyl methacrylate), poly(ethyl methacrylate), poly(butyl methacrylate), poly(isobutyl methacrylate), poly(hexyl methacrylate), poly(isodecyl methacrylate), poly(lauryl methacrylate), poly(phenyl methacrylate), poly(methyl acrylate), poly(isopropyl acrylate), poly(isobutyl acrylate), poly(octadecyl acrylate) polyethylene, polypropylene, polyethylene glycol), polyethylene oxide), poly (ethylene terephthalate), polyvinyl acetate), polyvinyl chloride, polystyrene, polyvinyl pyrrolidone, pluronics and polyvinylphenol and copolymers thereof.
Synthetically modified natural polymers of use in conjugates of the invention include, but are not limited to, alkyl celluloses, hydroxyalkyl celluloses, cellulose ethers, cellulose esters, and nitrocelluloses. Particularly preferred members of the broad classes of synthetically modified natural polymers include, but are not limited to, methyl cellulose, ethyl cellulose, hydroxypropyl cellulose, hydroxypropyl methyl cellulose, hydroxybutyl methyl cellulose, cellulose acetate, cellulose propionate, cellulose acetate butyrate, cellulose acetate phthalate, carboxymethyl cellulose, cellulose triacetate, cellulose sulfate sodium salt, and polymers of acrylic and methacrylic esters and alginic acid.

These and the other polymers discussed herein can be readily obtained from commercial sources such as Sigma Chemical Co. (St. Louis, MO.), Polysciences (Warrenton, PA.), Aldrich (Milwaukee, WL), Fluka (Ronkonkoma, NY), and BioRad (Richmond, CA), or else synthesized from monomers obtained from these suppliers using standard techniques.
Representative biodegradable polymers of use in the conjugates of the invention include, but are not limited to, polylactides, polyglycolides and copolymers thereof, polyethylene terephthalate), poly(butyric acid), poly(valeric acid), poly(lactide-co-caprolactone), poly(lactide-co-glycolide), polyanhydrides, polyorthoesters, blends and copolymers thereof. Of particular use are compositions that form gels, such as those including collagen, pluronics and the like.
The polymers of use in the invention include "hybrid' polymers that include water-insoluble materials having within at least a portion of their structure, a bioresorbable molecule. An example of such a polymer is one that includes a water-insoluble copolymer, which has a bioresorbable region, a hydrophilic region and a plurality of crosslinkable functional groups per polymer chain.
For purposes of the present invention, "water-insoluble materials" includes materials that are substantially insoluble in water or water-containing environments.
Thus, although certain regions or segments of the copolymer may be hydrophilic or even water-soluble, the polymer molecule, as a whole, does not to any substantial measure dissolve in water.
For purposes of the present invention, the term "bioresorbable molecule"
includes a region that is capable of being metabolized or broken down and resorbed and/or eliminated through normal excretory routes by the body. Such metabolites or break down products are preferably substantially non-toxic to the body.
The bioresorbable region may be either hydrophobic or hydrophilic, so long as the copolymer composition as a whole is not rendered water-soluble. Thus, the bioresorbable region is selected based on the preference that the polymer, as a whole, remains water-insoluble. Accordingly, the relative properties, i. e., the kinds of functional groups contained by, and the relative proportions of the bioresorbable region, and the hydrophilic region are selected to ensure that useful bioresorbable compositions remain water-insoluble.
Exemplary resorbable polymers include, for example, synthetically produced resorbable block copolymers of poly(,-hydroxy-carboxylic acid)/poly(oxyalkylene, (see, Cohn et al., U.S. Patent No. 4,826,945). These copolymers are not crosslinked and are water-soluble so that the body can excrete the degraded block copolymer compositions. See, ~ounes et al., .I ~i~med. lVlate~. Pes. 21: 1301-1316 (1987); and Cohn et al., J~iorned.
Mater'. Res. 22: 993-1009 (1988).
Presently preferred bioresorbable polymers include one or more components selected from poly(esters), poly(hydroxy acids), poly(lactones), poly(amides), polyester-amides), poly (amino acids), poly(anhydrides), poly(orthoesters), poly(carbonates), poly(phosphazines), poly(phosphoesters), poly(thioesters), polysaccharides and mixtures thereof. More preferably still, the biosresorbable polymer includes a poly(hydroxy) acid component. Of the poly(hydroxy) acids, polylactic acid, polyglycolic acid, polycaproic acid, polybutyric acid, polyvaleric acid and copolymers and mixtures thereof are preferred.
In addition to forming fragments that are absorbed i~ vivo ("bioresorbed"), preferred polymeric coatings for use in the methods of the invention can also form an excretable and/or metabolizable fragment.
Higher order copolymers can also be used in the present invention. For example, Casey et al., U.S. Patent No. 4,438,253, which issued on March 20, 1984, discloses tri-block copolymers produced from the transesterification of poly(glycolic acid) and an hydroxyl-ended poly(alkylene glycol). Such compositions are disclosed for use as resorbable monofilament sutures. The flexibility of such compositions is controlled by the incorporation of an aromatic orthocarbonate, such as tetra-p-tolyl orthocarbonate into the copolymer structure.
Other coatings based on lactic and/or glycolic acids can also be utilized. For example, Spinu, U.S. Patent No. 5,202,413, which issued on April 13, 1993, discloses biodegradable mufti-block copolymers having sequentially ordered blocks of polylactide and/or polyglycolide produced by ring-opening polymerization of lactide and/or glycolide onto either an oligomeric diol or a diamine residue followed by chain extension with a di-functional compound, such as, a diisocyanate, diacylchloride or dichlorosilane.
Bioresorbable regions of coatings useful in the present invention can be designed to be hydrolytically and/or enzymatically cleavable. For purposes of the present invention, "hydrolytically cleavable" refers to the susceptibility of the copolymer, especially the bioresorbable region, to hydrolysis in water or a water-containing envirorunent. Similarly, "enzymatically cleavable" as used herein refers to the susceptibility of the copolymer, especially the bioresorbable region, to cleavage by endogenous or exogenous enzymes.
When placed within the body, the hydrophilic region can be processed into excretable and/or metabolizable fragments. Thus, the hydrophilic region can include, for example, polyethers, polyalkylene oxides, polyols, polyvinyl pyrrolidine), polyvinyl alcohol), poly(alkyl oxazolines), polysaccharides, carbohydrates, peptides, proteins and copolymers and mixtures thereof. Furthermore, the hydrophilic region can also be, for example, a poly(alkylene) oxide. Such poly(alkylene) oxides can include, for example, polyethylene) oxide, polypropylene) oxide and mixtures and copolymers thereof.
Polymers that are components of hydrogels are also useful in the present invention.
Hydrogels are polymeric materials that are capable of absorbing relatively large quantities of water. Examples of hydrogel forming compounds include, but are not limited to, polyacrylic acids, sodium carboxymethylcellulose, polyvinyl alcohol, polyvinyl pyrrolidine, gelatin, carrageenan and other polysaccharides, hydroxyethylenemethacrylic acid (HEMA), as well as derivatives thereof, and the like. Hydrogels can be produced that are stable, biodegradable and bioresorbable. Moreover, hydrogel compositions can include subunits that exhibit one or more of these properties.
Bio-compatible hydrogel compositions whose integrity can be controlled through crosslinking are known and are presently preferred for use in the methods of the invention.
For example, Hubbell et al., U.S. Patent Nos. 5,410,016, which issued on April 25, 1995 and 5,529,914, which issued on June 25, 1996, disclose water-soluble systems, which are crosslinked block copolymers having a water-soluble central block segment sandwiched between two hydrolytically labile extensions. Such copolymers are further end-capped with photopolymerizable acrylate functionalities. When crosslinked, these systems become hydrogels. The water soluble central block of such copolymers can include polyethylene glycol); whereas, the hydrolytically labile extensions can be a poly(a-hydroxy acid), such as polyglycolic acid or polylactic acid. See, Sawhney et al., Macromolecules 26:

(1993).
In another preferred embodiment, the gel is a thermoreversible gel.
Thermoreversible gels including components, such as pluronics, collagen, gelatin, hyalouronic acid, polysaccharides, polyurethane hydrogel, polyurethane-urea hydrogel and combinations thereof are presently preferred.
In yet another exemplary embodiment, the conjugate of the invention includes a component of a liposome. Liposomes can be prepared according to methods known to those skilled in the art, for example, as described in Eppstein et al., LT.S. Patent No. 4,522,811, which issued on June 11, 1985. For example, liposome formulations may be prepared by dissolving appropriate lipids) (such as stearoyl phosphatidyl ethanolamine, stearoyl phosphatidyl choline, arachadoyl phosphatidyl choline, and cholesterol) in an inorganic solvent that is then evaporated, leaving behind a thin film of dried lipid on the surface of the container. An aqueous solution of the active compound or its pharmaceutically acceptable salt is then introduced into the container. The container is then swirled by hand to free lipid material from the sides of the container and to disperse lipid aggregates, thereby forming the liposomal suspension.
The above-recited microparticles and methods of preparing the microparticles are offered by way of example and they are not intended to define the scope of microparticles of use in the present invention. It will be apparent to those of skill in the art that an array of microparticles, fabricated by different methods, are of use in the present invention.
c) Biomolecules In another preferred embodiment, the modified sugar bears a biomolecule. In still further preferred embodiments, the biomolecule is a functional protein, enzyme, antigen, antibody, peptide, nucleic acid (e.g., single nucleotides or nucleosides, oligonucleotides, polynucleotides and single- and higher-stranded nucleic acids), lectin, receptor or a combination thereof.
Some preferred biomolecules are essentially non-fluorescent, or emit such a minimal amount of fluorescence that they are inappropriate for use as a fluorescent marker in an assay.
Other biomolecules may be fluorescent. The use of an otherwise naturally occurring sugar that is modified by covalent attachment of another entity (e.g., PEG, biomolecule, therapeutic moiety, diagnostic moiety, etc.) is appropriate. In an exemplary embodiment, a sugar moiety, which is a biomolecule, is conjugated to a linker arm and the sugar-linker arm cassette is subsequently conjugated to a peptide via a method of the invention.

Biomolecules useful in practicing the present invention can be derived from any source. The biomolecules can be isolated from natural sources or they can be produced by synthetic methods. Peptides can be natural peptides or mutated peptides.
1\ilutations can be effected by chemical mutagenesis, site-directed mutagenesis or other means of inducing mutations known to those of skill in the art. Peptides useful in practicing the instant invention include, for example, enzymes, antigens, antibodies and receptors.
Antibodies can be either polyclonal or monoclonal; either intact or fragments. The peptides axe optionally the products of a program of directed evolution.
Both naturally derived and synthetic peptides and nucleic acids axe of use in conjunction with the present invention; these molecules can be attached to a sugar residue component or a crosslinking agent by any available reactive group. For example, peptides' can be attached through a reactive amine, carboxyl, sulfhydryl, or hydroxyl group. The reactive group can reside at a peptide terminus or at a site internal to the peptide chain.
Nucleic acids can be attached through a reactive group on a base (e.g., exocyclic amine) or an available hydroxyl group on a sugar moiety (e.g., 3'- or 5'-hydroxyl). The peptide and nucleic acid chains can be further derivatized at one or more sites to allow for the attachment of appropriate reactive groups onto the chain. See, Chrisey et al. Nucleic Acids Res. 24:
3031-3039 (1996).
In a further preferred embodiment, the biomolecule is selected to direct the peptide modified by the methods of the invention to a specific tissue, thereby enhancing the delivery of the peptide to that tissue relative to the amount of underivatized peptide that is delivered to the tissue. In a still further preferred embodiment, the amount of derivatized peptide delivered to a specific tissue within a selected time period is enhanced by derivatization by at least about 20%, more preferably, at least about 40%, and more preferably still, at least about 100%. Presently, preferred biomolecules for targeting applications include antibodies, hormones and ligands for cell-surface receptors. Exemplary targeting biomolecules include, but are not limited to, an antibody specific for the transferrin receptor for delivery of the molecule to the brain (Penichet et al., 1999, J. Immunol. 163:4421-4426;
Pardridge, 2002, Adv. Exp. Med. Biol. 513:397-430), a peptide that recognizes the vasculature of the prostate (Arap et al., 2002, PNAS 99:1527-1531), and an antibody specific for lung caveolae (lVIcIntosh et al., 2002, PNAS 99:1996-2001).

In a presently preferred embodiment, the modifying group is a protein. In an exemplary embodiment, the protein is an interferon. The interferons are antiviral glycoproteins that, in humans, are secreted by human primary fibroblasts after induction with virus or double-stranded RNA. Interferons are of interest as therapeutics, e.g., antivirals and treatment of multiple sclerosis. For references discussing interferon-Vii, see, e.g., Y~u, et al., J
Neut~oimmun~l., 64(1):91-100 (1996); Schmidt, J.,.I. Neur~~sci. Res., 65(1):59-67 (2001);
blender, et al., F~lia Neur~opathol., 39(2):91-93 (2001); Martin, et al., Spr~inge~ Senain.
Inzrnun~pathol.,18(1):1-24 (1996); Takane, et al., .I: Pharmac~l. Exp. They., 294(2):746-752 (2000); Sburlati, et al., Biotechnol. Prog.,14:189-192 (1998); Dodd, et al., Biochimica et Biophysics Acta, 787:183-187 (1984); Edelbaum, et al., J. Interferon Res.,12:449-453 (1992); Conradt, et al., J. Biol. Chem., 262(30):14600-14605 (1987); Civas, et al., Eur. J.
Biochem.,173:311-316 (1988); Demolder, et al., J. Biotechnol., 32:179-189 (1994); Sedmak, et al., J. Interferon Res., 9(Suppl 1):561-865 (1989); I~agawa, et al., J.
Biol. Chem., 263(33):17508-17515 (1988); Hershenson, et al., U.S. Patent No. 4,894,330;
Jayaram, et al., J. Interfef°on Res., 3(2):177-180 (1983); Menge, et al., Develop. Biol.
Standard., 66:391-401 (1987); Vonk, et al., J. Interferon Res., 3(2):169-175 (1983); and Adolf, et al., J. Interfef on Res.,10:255-267 (1990). For references relevant to interferon-oc, see, Asano, et al., Eu~. J.
Cancer, 27(Suppl 4):521-825 (1991); Nagy, et al., Anticancer Research, 8(3):467-470 (1988); Dron, et al., J. Biol. Regul. Homeost. Agents, 3(1):13-19 (1989);
Habib, et al., Am.
Sing., 67(3):257-260 (3/2001); and Sugyiama, et al., Eur. J. Biochem., 217:921-927 (1993).
In an exemplary interferon conjugate, interferon (3 is conjugated to a second peptide via a linker arm. The linker arm includes an intact glycosyl linking group through which it is attached to the second peptide via a method of the invention. The linker arm also optionally includes a second intact glycosyl linking group, through which it is attached to the interferon.
In another exemplary embodiment, the invention provides a conjugate of follicle stimulating hormone (FSH). FSH is a glycoprotein hormone. See, for example, Saneyoshi, et al., Biol. Repr~od., 65:1686-1690 (2001); Hakola, et al., J.
Endocrinol.,158:441-448 (1998); Stanton, et al., Mol. Cell. Endocrinol., 125:133-141 (1996); Walton, et al., ,I. Clin.
Endocf°inol. Metab., 86(8):3675-3685 (08/2001); Ulloa-Aguirre, et al., End~cr~ine,11(3):205-215 (12/1999); Castro-Fernande~, et a1.1, J. Clin. End~crinol. Matab., 85(12):4603-4610 (2000); Prevost, Rebecca R., Pharmacothe~apy, 18(5):1001-1010 (1998);
Linskens, et al., The FASEB Journal,13:639-645 (04/1999); Butnev, et al., Biol. Reprod., 58:458-469 (1998);
Muyan, et al., Mol. Endo., 12(5):766-772 (1998); Min, et al., Endo. J., 43(5):585-593 (1996);
B~ime, et al., Recent Progress in Ihormone Research, 34:271-289 (1999); and Rafferty, et al., J Erido.,14~:527-533 (1995). The FSH c~njugate can be f~rmed in a manner similar t~ that described for interfer~n.
In yet another exemplary embodiment, the conjugate includes erythropoietin (EPO).
EPO is known to mediate response to hypoxia and to stimulate the producti~n of red blood cells. For pertinent references, see, Cerami, et al., SerrZinars in ~ncology, 28(2)(Supg~~ 8):66-70 (04/2001). An exemplary EPO conjugate is formed analogously to the conjugate of interferon.
In a further exemplary embodiment, the inventi~n provides a conjugate of human granulocyte colony stimulating factor (G-CSF). G-CSF is a glycoprotein that stimulates proliferation, differentiation and activation of neutropoietic progenitor cells into functionally mature neutrophils. Injected G-CSF is known to be rapidly cleared from the body. See, for example, Nohynek, et al., Cancer Chemother. Pharmacol., 39:259-266 (1997);
Lord, et al., Clinical Cancer Research, 7(7):2085-2090 (07/2001); Rotondaro, et al., Molecular Biotechnology, 11(2):117-128 (1999); and Bonig, et al., Bone Marrow Transplantation, 28:259-264 (2001). An exemplary conjugate of G-CSF is prepared as discussed above for the conjugate of the interferons. One of skill in the art will appreciate that many other proteins may be conjugated to interferon using the methods and compositions of the invention, including but not limited to, the peptides listed in Tables 7 and 8 (presented elsewhere herein) and Figure 28, and in Figures 29-57, where individual modification schemes are presented.
In still a further exemplary embodiment, there is provided a conjugate with biotin.
Thus, for example, a selectively biotinylated peptide is elaborated by the attachment of an avidin or streptavidin moiety bearing one or more modifying groups.
In a further preferred embodiment, the biomolecule is selected to direct the peptide modified by the methods of the invention to a specific intracellular compartment, thereby enhancing the delivery of the peptide to that intracellular compartment relative to the amount of underivatized peptide that is delivered to the tissue. In a still further preferred embodiment, the amount of derivatized peptide delivered to a specific intracellular compartment within a selected time period is enhanced by derivatization by at least about 20%, more preferably, at least about 40%, and more preferably still, at least about 100%. In another particularly preferred embodiment, the biomolecule is linked to the peptide by a cleavable linker that can hydrolyze once internalized. Presently, preferred biomolecules for intracellular targeting applications include transferrin, lactotransferrin (lactoferrin), melanotransferrin (p97), ceruloplasmin, and divalent cation transporter, as well as antibodies directed against specific vascular targets. Contemplated linkages include, but are not limited to, protein-sugar-linker-sugar-protein, protein-sugar-linker-protein and multivalent forms thereof, and protein-sugar-linker-drug where the drug includes small molecules, peptides, lipids, among others.
Site-specific and target-oriented delivery of therapeutic agents is desirable for the purpose of treating a wide variety of human diseases, such as different types of malignancies and certain neurological disorders. Such procedures are accompanied by fewer side effects and a higher efficiacy of drug. Various principles have been relied on in designing these delivery systems. For a review, see Gannett, Advanced Drug Delivery Reviews 53:171-216 (2001).
One important consideration in designing a drug delivery system to target tissues specifically. The discovery of tumor surface antigens has made it possible to develop therapeutic approaches where tumor cells displaying definable surface antigens are specifically targeted and killed. There are three main classes of therapeutic monoclonal antibodies (antibody) that have demonstrated effectiveness in human clinical trials in treating malignancies: (1) unconjugated MAb, which either directly induces growth inhibition and/or apoptosis, or indirectly activates host defense mechanisms to mediate antitumor cytotoxicity;
(2) drug-conjugated MAb, which preferentially delivers a potent cytotoxic toxin to the tumor cells and therefore minimizes the systemic cytotoxicity commonly associated with conventional chemotherapy; and (3) radioisotope-conjugated MAb, which delivers a sterilizing dose of radiation to the tumor. See review by Reff et al., Cancer Control 9:152-166 (2002).
In order to arm MAbs with the power to kill malignant cells, the MAbs can be connected to a toxin, which may be obtained from a plant, bacterial, or fungal source, to form chimeric proteins called immunotoxins. Frequently used plant toxins are divided into two classes: (1) holotoxins (or class II ribosome inactivating proteins), such as ricin, abrin, mistletoe lectin, and modeccin, and (2) hemitoxins (class I ribosome inactivating proteins), such as pokeweed antiviral protein (PAP), saporin, ~ryodin 1, bouganin, and gelonin.
Conunonly used bacterial toxins include diphtheria toxin (1)T) and Pseudomonas exotoxin (PE). I~reitman, Cuf~rent Phar~rnaceutical ~i~techaa~l~~y 2:313-325 (2001).
~ther toxins contemplated for use with the present invention include, but are not limited to, those in Table 2.
'Table 2. Toxins.
Chemical Structure Toxin Name/ CAS RN l Indication/ Mechanism Activity (IC50 nM);
Source/ Analogs Toxicity Tumor Type Alternate ID
SW-163E/ 260794-24-9; Cancer and not reported 0.3 P388 Streptomyces 260794-25-0/ Antibacterial/0.2 A2780 sp SNA

15896/ SW-1630; low toxicity 0.4 KB
(mice ip) SW-163E SW-163A; 1.6 colon SW-163B 1.3 HL-60 .~_ Thiocoraline/ 173046-02-1 Breast Cancer; DNA lung, colon, CNS
Micromon~spora marina Melanoma; Non-small Polymerase melanoma (actinomycete) lung cancer / alpha not reported inhibitor (blocks cell progression from G 1 to S) Trunlzamide A'/ 181758-83-8 Cancer/ not reported cell culture (IC50 in LissoclifZUnZ sp (aascidian) not reported micrograms/mL);
0.5 P388;
0.5 A549;
0.5 HT-29;
1.0 MEL-28 NHS
HEN,, Palauaminez/ 148717-58-2 Lung cancer/ not reported cell culture (IC50 in Stylotella agminata LD50 (i.p. in mice) is 13 micrograms/mL);
(sponge) mg/Kg 0.1 P388 0.2 A549 (lung) 2 HT-29 (colon) KB
H
~H
Halichondrin B/ 103614-76-2/ cancer/ antitubulin; NCI tumor panel;
Halichondria ~kadai, isohomohalic myelotoxicity dose sell cycle GI(50) from 50 nM to Axif~ell Carteui arid hondrin B limiting (dogs, rats) inhibitor 0.1 nM;
Phankell carteri (inhibits LC50's from 40 pM to (sponges)/ GTP binding 0.1 nM (many 0.1 to 25 NSC-609385 to tubulin) nM) ~~~I
Isohomo-halichondrin B/ 157078-48-3/ melanoma, lung, CNS, antitubulin; IC50's in 0.1 nM range Halich~aidria Okadai, halichondrin colon, ovary/ sell cycle (NCI tumor panel) Axinell Carteri and B not reported inhibitor Pl~ahkell carteri (inhibits (sponges)/ GTP binding NSC-650467 to tubulin) Halichondrin 253128-15-3/ solid tubulin cell culture B analogs/ tumors/ (not semi-synthetic ER-076349; not reportedbinding reported);
starting from HalichondriaER-086526; agent; animal models active Okadai, AxinellB-1793; disruption Carteri of (tumor regression andPhankell E-7389 mitotic observed) in carteri lymphoma, (sponges)/ spindles colon (mufti-drug ER-076349; ER-086526; resistant).

B-1793; E-7389 NK-130119/ 132707-68-7 antifungal and not reported 25 ng/mL colon S'ta~e~t~~rryces anticancer/ 8.5 nglmL lung bottropensisl not reported H
Tetrocarcin A/ 73666-84-9/ cancer/ inhibits the not reported not reported/ analogs are not reported anti-KF-67544 reported apoptotic functino of Bcl2 Gilvusmycin/ 195052-09-6 cancer/ not reported IC50's in ng/mL:
Streptornyces QM16 not reported 0.08 P388 0.86 K562 (CML) 0.72 A431 (EC) 0.75 MKN28 (GI);
(for all < 1 nM) H
IB-96212/ 220858-11-7/ Cancer and not reported IC50's in ng/xnL:
marine actinomycete/ IB-96212; Antibacterial/ 0.1 P388 IB-96212 IB-98214; notreported BE-563843/ 207570-04-5 cancer/ not reported IC50's in ng/mL:
Streptomyces Sp./ not reported 0.1 P388 BE-56384 0.29 colon 26 0.12 PC-13 0.12 MI~M-45 Palmitoylrhizoxin/ 135819-69-1/ cancer/ tubulin not reported semi-synthetic; Rhizopus Analog of binds LDL; less binding chinensis rhizoxin cytotoxic than rhizoxin agent (cell cycle inhibitor) H~
Rhizoxin/ 95917-95-6; melanoma, lung, CNS, tubulin NCI tumor panel (NSC
Rhizopus chinensisl 90996-54-6 colon, ovary, renal, binding 332598);
WF-1360; NSC-332598; breast, head and neck/ agent (cell log GI50's:
FR-900216 Rapid Drug clearance; cycle 50 nM to 50 fM;
High AUC correlates inhibitor) log LC50's:
with high toxicity 50 pM to 0.5 nM
(several cell lines at 50 fM).
Dolastatin-10/ 110417-88-4/prostate, melanoma,tubulin NCI tumor panel Dolabella auricularia leukemia/ binding (60 cell line;
(sea other GI50);

hare)/ Dolistatins myelotoxicity (tubulin25 nM to 1 (at greater pM (most <

NSC-376128 (ie. 15) and than 0.3 pM) aggregation)1 nM) (three cell lines analogs ~M) soblidotin/ 149606-27-9/ cancer (pancreas, tubulin cell culture: colon, synthetic/ analogs esophageal colon, breast, binding melanoma, M5076 TAT-1027; auristatin PE prepared lung, etc) / agent tumors, P38S with 75-MTD was 1.S mg/Kg 85% inhibition (dose (I~); toxicity not not reported) reported H
O
Dolastatin-15/ not reported/ cancer/Tubulin NCI tumor panel (60 Dolabella auricularia (sea other binding cell line;
not reported GI50); 25 hare) Dolistatins (tubuline nM to 39 pM
(most < 1 (ie. 15) and aggregation)nM) (one cell line 2.5 analogs pM); most active in breast '~3;:
H
. ~ N
~r Cemadotin4/ 1159776-69- melanoma/ tubulin NCI tumor panel (NCS
Synthetic; Parent 9/ hypertension, myocardial binding D-669356); active in Dolastatin-15 was isolated many analogs ischemia and (tubulin breast, ovary, from Dolabella myelosuppression were aggregation) endometrial, sarcomas aurieularia (sea hare)/ dose-limiting toxicities. and drug resistant cell LU-103793; NSC D- lines. Data not public.

O., S
/,.... ~-i U ~ U
Epothilone A/ not reported/ cancer/ tubulin IC50's of;

Synthetic or isolated from many analogs not reported binding 1.5 nM MCF-7 (breast) Sorangiuna cellulosum (tubulin 27.1 nM MCF-7/ADR
(myxococcales) strain polymeriza- 2.1 nM KB-31 So ce90) tion) (melanoma) 3.2 nM HCT-116 f,,.. 7H
U ~..~ U

Epothilone B/ 152044054-7/ Solid tumorstubulin IC50's of;
(breast, Synthetic or many analogs ovarian, binding 0.18 nM MCF-7 isolated from etc)/

Sora~gium cellulosumwell tolerated; t1/2 (tubulin(breast) of (myxococcales) 2.5 hrs; partial polymeriza-2.92 nM MCF-7/ADR
strain So ce90) / responses (phase I); tion) 0.19 nM KB-31 EPO-906 diarrhea major side (melanoma) effect. 0.42 nM HCT-116;

broad activity reported H
Epothilone Analog / not reported tubulin IC50's of / cancer/ 0.30 to Synthetic or semi- hundreds of not binding 1.80 nM in reported various synthetic; Original lead, analogs (tubulin tumor cell lines;

Epothilone A, isolated polymeriza- active in drug resistant from Sorangium tion) cell lines cellulosum (myxococcales) strain So ce90)/

ZIP-EPO

-~ ~ ~...,.
N
c~ ~ U
Epothilone I) / 189452-10-9/ Solid tumors (breast, tubulin NCI tumor panel (NSC-Epothilone D, isolated many analogs ovarian, etc)/ binding 703147; IC50);
from S~rangium emesis and anemia; t1/2 (tubulin 0.19 nM ICB-31 cellul~surn of 5-10 hrs. polymeriza- (melanoma) (myxococcales) strain So tion) 0.42 nM HCT-116;
ce90)/ broad activity reported Structure l~Tot Identified Epothilone D analog 5/ 189453-10-9/ tubulin not reported Solid tumors;

Synthetic or semi- hundreds of not binding reported synthetic; Original lead, analogs (tubulin Epothilone D, isolated polymeriza-from Sorangiurrt tion) ~ellulosum (myxococcales) strain So ce90)/

r~nc_~ ~~_~4 Epothilone Analog / not reported/ cancer; tubulin not reported Synthetic; Original lead, hundreds of not reported binding Epothilone A, isolated analogs (tubulin from Sorangium polymeriza-eellulosurn tion) (myxococcales) strain So ce90)/
~C.P-R5715 Epothilone Analog/219989-84-1l non-small tubulin NCI tumor Panel cell Lung, (NSC-Synthetic or hundreds of breast, binding 710428 ~c NSC-semi- stomach tumor synthetic; Originalanalogs (objective responses(tubulin 710468); 8-32 lead, in nM

Epothilone B, breast ovarian and lung)/ (NCI data not isolated polymerizati available) from Sorangiurrzsever toxicity (fatigue,on) cellulosurn anorexia, nauseas, (myxococcales) vomiting, neuropathy strain So ce90)/ myalgia) RMC-?4.7550 to Epothilone Analog / advanced cancers/tubulin broad activity not reported/ with Synthetic or semi- hundredsadverse events binding IC50's of 0.7 of (diarrhea, to 10 nM

synthetic; ~riginal nausea, vomiting(tubulin lead, analogs , Epothilone B, isolated fatigue, neutropenia);polymeriza-from S~rangiunl t1/2 of 3.5 tion) hrs;

cellulosum improved water (myxococcales) strain solubility to So BMS

ce90)/ 247550.

1 ~
~7 r H~ ~ - ~H
HIV ~ H
Discodermolide l 127943-53-7/ solid tubulin Broad activity tumors/ (A549-synthetic; orginally analogs less not stabilizingnsclung, prostate, reported; 100-fold P388, isolated from Discoderrnia potent increaseagent ovarian with in water IC50's dissoluta (deep water solubility over (similarabout 10 nM) taxol to including sponge); rare compound taxol) multi-drug resistant cell (7 mg per 0.5 Kg spongel lines;

H
Chondramide D/ 172430-63-6 cancer/ - tubulin 5 nM A-549 not reported not reported binding (epidermoid carcinoma) agent; actin 15 nM A-498 (kidney) polymeriza- 14 nM A549 (lung) Lion inhibitor 5 nM SIB-~V-3 (ovary) 3 nM U-937 (lymphoma) Cryptophycin 204990-60-3solid tumors, tubulin broad activity analogs colon (lung, (including 52, and 186256-cancer/ polymeriza- breast, colon, 55 and leukemia) others)6/ 67-7/ Phase II studiestion inhibitor with IC50's halted of 2 to 40 Nostoc sp GSV many potentbecause of pM; active against 224 (blue- severe green algae) analogs toxicity with multi-drug resistance isolated one death Cryptophycin prepared resulting fromcell lines (resistant 1./ at drug; to LY-355703; Ly-355702;Lilly MDR pump). NCI

NSC-667642 tumor panel, GI50's from 100 nM to 10 pM;

LC50's from 100 nM to 25 pM.

Cryptophycin 8l 168482-36-8; solid tumors/ tubulin broad spectrum semi-synthetic; starting 168482-40-4; not reported polymeriza- anticancer activity (cell material from Nostoe sp. 18665-94-1; tion inhibitor culture) including 124689-65-2; mufti-drug resistant 125546-14-7/ tumors cryptophycin 5. 15 and 35 ~, Cryptophycin analogs'/ 219660-54-5/ solid tumors/ topoisomer- not reported synthetic; semi-synthetic, LY-404292 not reported ase inhibitors starting material from N~stoc sp./

Arenastatin A analogs8/ not reported/ cancer/ inhibits 8.7 nM (5 pg/mL) I~B
Dysidea arenaria (marine analogs not reported tubulin (nasopharyngeal); NCI
sponge)/ prepared polymerize- tumor panel (GI50's);
Cryptophycin B; NSC- tion 100 pM to 3 pM

ci HiC ,~
4 0 CHl H~0 H C
'~. ~ 0 ' ''CHl 0 N
HZC ~'CH1 ~ OH
~OH

Phomopsin A/ not reported Liver cancer (not as tubulin potent anticancer Diapof~te toxicus or potent in other cancers)/ binding activity especially Phomopsin not reported agent against liver cancer leptostromiforrnis (fungi) Curacin A and analogs/ 155233-30-0/ Cancer/ Tubulin broad activity (cancer Lyng~bya rnajuscula (blue analogs have not reported binding cell lines); 1-29 nM
green cyanobacterium) been prepared agent Hemiasterlins A & B not reported/ Cancer/ Antimitotic broad activity:
and analogs9/ criamide A & not reported agent 0.3-3 nM MCF7 Cymbastela sp. B; (tubulin (breast);
geodiamiolid- binding 0.4 ng/mL P388 C, agent) ~H
Spongistatins (1-9)1/ 149715-96-8; tubulin Most potent compounds cancer/

Spirastrell spinispirulifera 158734-18-0;binding ever tested in not reported NCI panel (sea sponge) 158681-42-6; agent cell line (mean GI50's 158080-65-0; of 0.1 nM;

150642-07-2; Spongistatin-1 GI50's 153698-80-7; of 0.025-0.035 nM with 153745-94-9; extremely potent 150624-44-5; activity against a subset 158734-19-1/ of highly other chemoresistant tumor spongistatins types Maytansine/ 35846-53-8/ cancer! tubulin Broad Activity in NCI

Mezytenus sp./ other related severe toxicity tumor panel binding (NSC-NSC-153858 macrolides agent (causes 153858; NSC-153858);

extensive NCI tumor panel, disassembly GI50's from 3 ~M to of the 0.1 pM; LC50's from microtubule 250 pM to 10 pM. Two and totally different experiments prevents gave very different tubulin potencies.

spiralizaiton) Maytansine-IgG(EGFR not reported/ breast , head and neck, EGFR not reported directed)-conjugate"/ other related Squamous cell binding and semi-synthetic; starting macrolides carcinoma/ tubulin material from Maytenus not reported binding sp n~
0 0~~.

0 ~
w,~

Maytansine-IgG(CD56 not Neuroendocrine,CD56 antigen-specific reported/ small-antigen)-conjugatel2, cell lung, carcinoma/bindingcytotoxicity 3.5 other related and (cell drug molecules per IgG/ mild toxicity tubulinculture; epidermal, macrolides (fatigue, semi-synthetic; startingnausea, headachesbindingbreast, renal and ovarian material from lllexytenz~smild peripheral colon) with IC50's of sp./ neuropathy); 10-40 pM; animal no huN901-DM1 hematological studies (miceSCLC
toxicity;

MTD 60 mg/I~g, tumor--alone LV., and in weekly for 4 combination weeks; only with taxol stable disease or cisplatin reported completely (humans) eliminated tumors).

Maytansine-IgG(CEAnot reported/ non-small-cellCEA binding antigen-specific lung, antigen)-conjugatel3,other related carcinoma and tubulin cytotoxicity 4 pancreas, (cell drug molecules macrolides lung, colon/ binding culture; epidermal, per IgG/

semi-synthetic;mild toxicity (fatigue, breast, renal ovarian starting material from nausea, headaches and colon) with IC50's Maytefzus of sp./ mild peripheral 10-40 pM; animal C424-DM1 neuropathy); pancreatic studies (mice:

lipase elevated; MTD melanoma [COLO-mg/I~g, LV., every 21 205]--alone and in days; only stable diseasecombination with taxol reported (humans); t1/2 or cisplatin completely was 44 hr. eliminated tumors);

Geldanamycin / 30562-34-6/ cancer/ binds Hsp 90 NCI tumor panel (cell Streptonayees natural not reported chaperone culture); 5.3 to 100 hygroscopicus var. derivatives and inhibits nM; most active in Geldanus/ function colon, lung and NSC-212518; Antibiotic leukemia. NCI tumor U 29135; NSC-122750 panel, GI50's from 10 pM to 0.1 nM; LC50's from 100 pM to 100 nM. Two assays with very different potencies.
NHS
Geldanamycin 745747-14-7/solid tumors/ binds Hsp 90 cell culture Analog/ (not semi-synthetic;Kosan, Dose limiting chaperone reported);
/ NCI toxicities animal CP-127374; 17-AAG;and UK (anemia, anorexia,and inhibits models active (tumor NSC-330507 looking diarrhea, nauseafunction regression fox and observed) in - analogs vomiting); breast, ovary, with t1/2 (i.v.) is ' longer about 90 min; melanoma, colon.
t1/2 no and oral objective responses activity;measured at 88 mg/Kg analogs (i.v. daily for 5 days, include: every 21 days);
NSC-255110;

682300;

683661;

683663.

Geldanamycin analog/ not reported/ solid tumors/ binds Hsp 90 not reported semi-synthetic; / analogs not reported chaperone CP-202567 prepared and inhibits function -13 ~-Geldanamycin 345232-4.4.-2/ breast/ binds Hsp 90 cell culture (no conjugates/ analogs not reported chaperone reported); animal semi-synthetic; / prepared and inhibits models performed LY-294002-GM; PI3I~-1- function;
GM binds and inhibits PI-3 kinase Structure Not Reported Geldanamycin Analog/ not reported/ breast, prostate/ binds Hsp 90 not reported not reported/ analogs not reported chaperone CNF-101 prepared and inhibits function Structure Not Reported Geldanamycin- not reported/ prostate/ binds Hsp 90 not reported; conjugate testosterone conjugate/ analogs not reported chaperone has a 15-fold selective semi-synthetic/ prepared and inhibits cytotoxicity for GMT-1 function and androgen positive testosterone prostate cells receptors where it is internalized Podophyllotoxin/ 518-28-5/ Verruca vulgaris, tubulin broad activity (cell Pod~phyllurn sp. many analogs Condyloma/ inhibitor and culture) with IC50's in severe toxicity when topoisomer- pM range given i.v. or s.c. ase inhibitor ..

~~H
an..
H
C-102714/ 120177-69-7 cancer (examined extremely potent DNA (cell Streptornyces hepatoma, breast, lung cleavingculture) IC50's seto~ii C- in pM

1027/ and leukemia/ agent and fM; conjugated to C-1027 not reported antibodies the potency remains the same (ie.

5.5 to 42 pM);

esperamicin-Al/ 99674-26-7 cancer/ . DNA highly potent activity not known/ not reported (suspected cleaving (cell culture); animal BBM-1675A1; BMY- severe toxicity) agent models highly potent 28175: GGM-1675 with optimal dose of HO D
' D
,~ GH9 GHQ ~' NH - 0' Pr~X-S~-t~~ ~ ~'Pd yNH
rNH -. ~ D ~~GG~
D
- Y D HeGDO ~ H9G D NH . D
m-N.5-15 Pr= Ixcteinev~us terrier I~G ~D .~ 9 n ~ D
V~l = ~Iich~rniGin minus, Ms.~-5.~ H~' ,~ ~ D HD D
X = links D D GF~ HD' Y = urtit3orly P 7G.6 HD ~ i ~G -~-D
N~~~rH
HaG - D' pH H9G --~
H,G _ D' Calicheamiein- 113440-58-7; AML/ DNA Kills CD33+ cells (HI,-IgG(CD33 antigen)- 220578-59-6/ mild toxicity cleaving 60, NOMO-1, and conjugate's/ several agent NKM-1) at 100 ng/mL;
semi-synthetic: reported in MDR cell lines are not Micro»zofzospora patents effected by the drug.
eehi~osporal gemtuzumab ozogamicin;
mylotarg; WAY-CMA-676; CMA-676; CDP-HD D
r D
~~ GH9 GHg ~ NH~ D.
Prf-X-S~'V~l~ ~ ,~N~NH~S~S H~. ~~
~NH ~,, II D ~~GG~, HsG~D D H9C D
D NH . D
m =~.5-1~ D
Pr= preteinaceeus r~rrier H H~G,D f I 8 n~~~D r VV = c~liGha~miGin minr,rs Met-~.~ ~; ~, ; D HD o X = linker D D GH9 HD
~' _ ~rtit~.ody P7~,6 HD ~ i ~G -D
N~ ,~~H
H9G - D' pH H9G -/
H9C _ D;
Calicheamicin-IgG- 113440-58-7; cancer/ DNA TBD
conjugateslb/ 220578-59-6 not reported cleaving semi-synthetic: agent Micrornonospora echirZOSpora a ,~.~ CH9 GHs ~ NH 0' Pr~~-S-S-I~d~ ~. NH ~ H'°
N' ~'~ ~~
rNH .w, ~~ D ~4~aa~4 ~,'' H9caa a H~~ ~ a a m=a.5-t~ dp ~ ,NH' D
Pr= pr~teinar~us ~ni~° HgG ~ s n~ a r '~ _ ~lich~miein rninMet-~S Her ,~ ~ D Ha = link~° D a DHa Ho'H~D
~'=~ntlf~ceiyP76.~ HD I
N !-- , , r H9D-0' DH H9C ~! H
H9G _ D;
Calicheamicin- not reported cancer/ DNA all human cancer; data IgG(OBAl antigen) not reported cleaving not reported conjugate/ agent semi-synthetic:
Micromonospora echinosporal ICHa .f NH~ a' P r(-?i-~~-'1~~ ~ ~N ,, NH .. 6 ~, ~ H'7 Y~,NH D ~.. a ~~DD~ ~'' /
HaD .D a HaC d 0 a NH ~ 0 m=a.~-1~ D
Pr= prdeinaceous tamer t~C r s n, 1~a' W = ~licheamicin minus Met-S.~ Her a D ,~ I D~ HD1 D Ha io H = linker .~' Y=~ntifa~yP76.E Ha~ I H°C 1~~ a N~'1 H9C - 0~ OH HaD ~
HaG _ 0;
Calicheamicin- not reported non-Hodgkin lymphoma, DNA all human cancer; data IgG(CD22 antigen) cancer/ cleaving not reported conjugate/ not reported agent semi-synthetic:
Micromonospora echinosporal parially esterified polystyrene malefic acid copolymer (SMA) conjugated to neocarzinostatin (IVCS) Neocarzinostatinl'/ 123760-07-6; liver cancer and brain DNA cell culture data not semi-synthetic; 9014-02-2 cancer/ cleaving reported.
Streptomyces not reported agent carconistaticusl Zinostatin stimalamer;
YM-881: YM-16881 IgG (TES-23)-conjugated to neocaxZinostatin Neocarzinostatin/ not reported solid tumors/ DNA cell culture data not not reported/ toxicity not reported; the cleaving reported.
TES-23-NCS TES-23 antibody agent and (without anticancer immunostim-agent) was as effective at ulator eliminating tumors as the drug conjugated protein Kedarcidin'$/ 128512-40-3; cancer/ DNA cell culture (IC50's in Streptoalloteichus128512-39-0/ not reportedcleaving ng/mL), 0.4 HCT116;
sp NOV strain agent 0.3 HCT116/VP35;
L5856, ATCC
chromophore 53650/ and protein 0.3 HCT116/VM46;

NSC-646276 conjugate 0.2 A2780;

1.3 A2780/DDP.

animal models in P388 and B-16 melanoma.

NCI tumor panel, GI50's from 50 p,M to 5 ~M.

H ~."_.
H~

Eleutherobins/ 174545-76-7/ cancer! tubulin similar potency to taxol;
marine coral sarcodictyins not reported binding not effective against (marine coral) agent MDR cell lines Bryostatin-1/ 83314-O1-6leukemia, melanoma,immunostim-not reported Bugula neritina (marinelung, cancer/ ulant (TNF, bryosoan)/ myalgia; accumulatedGMCSF, GMY-45618; NSC- toxicity; poor etc);
water 339555 solubility; enhances cell dose limiting toxicity kill by current anticancer agents FR-901228/ 128517-07-7 leukemia, T-cell histoneIn vitro cell lines (NCI

Cl~romobacteriurn lymphoma, cancer/ deacetylasetumor panel);

violaceum strain 968/ toxic doses (LD50) 6.4 IC50's of between inhiibitor 0.56 NSC-63-176; FIB-228 and 10 mg/Kg, ip and iv and 4.1 nM (breast, respectively; GI lung, gastric colon, toxicity, lymphoid leukemia) atrophy; dose limiting toxicity (human) 18 mg/Kg; t!/2 of 8 hrs (human) Chlamydocin/ 53342-16-8 cancer/ histone not reported (cell not reported not reported deacetylase culture);
inhiibitor inhibits histone deacetylase at an IC50 of 1.3 nM
9r Phorboxazole A19/ 181377-57-1; leukemia,not reported NCI tumor myeloma/ panel marine sponge 165689-31-6; not reported(induces (details not reported);

180911-82-4; apoptosis) IC50's of 1-10 nM. The 165883-76-1/ inhibition values analogs (clonogenic growth of prepared human cancer cells) at 10 nM ranged from 6.2 to > 99.9% against NALM-6 human B-lineage acute lymophoblastic leukemia cells, BT-20 breast cancer cells and U373 glioblastoma cells, with the specified compound showing inhibition values in the range of 42.4 to >

99.9% against these cell lines.; IC50's are nM

for lViDR cell lines.

N

O H 0, Apicularen A/ 220757-06-2/ cancer/ not reported IC50's of 0.1 to 3 Clzondromyees robustus natural not reported ng/mL (K13-3-A, K13-derivatives Va, K562, HL60, LJ937, A49S, A549, PV3 and SK-OV3) H
H

Taxol/ 33069624/ cancer; breast, prostate, tubulin NCI tumor panel;
Pacific yew and fungi! many analogs ovary, colon, lung, head binding GI50's of 3 nM to 1 Paclitaxel; NSC-125973 & neck, etc./ agent ~,M;
severe toxicity (grade III TGI 50 nM to 25 ~M
and IV) Vitilevuamide/ 191681-63-7 cancer/ tubulin cell culture; IC50's of Didemnurn cuculliferum not reportedbinding 6-311 nM (panel of or Polysyncraton agent tumor cell lines lithostrotum HCT116 cells, A549 cells, SK-MEL-5 cells A498 cells). The increase in lifespan (ILS) for CDF1 mice after ip injection of P388 tumor cells was in the range of -45 to +70% over the dose range of 0.13 to 0.006 mg/kg.

Didemnin B/ 77327-OS-0; non-Hodgkin's inhibits NCI 60-tumor panel Trididemnurn solidurnl lymphoma, breast,protein (GI50's): 100 77327-04-9; nM to 50 NSC-2325319; IND 77327-06-1/carcinoma, CNS,synthesis via IIvI.
colon/

24505 other related Discontinued EF-1 Not potent against due to natural cardiotoxicity;MDI~ cell lines.
nausea, products neuro-muscular toxicity and vomiting MTD 6.3 mg/Kg; toxicity prevented achieving a clinically signif.
effect;

rapidly cleared (t1/2 4.8 hrs O
O OH
Leptomycin B/ 87081-35-4 NCI 60-tumor panel Streptonryces sp. strain (GI50's):
ATS 1287/ 8 p,M to 1 pM; (LC50):
NSC-364372; elactocin 250 ~M to 10 nM
(several cell lines at 0.1 nM). Two testing results with very different potencies.
Cryptopleurin/ NCI 60-tumor panel not known/ (GI50's): 19 nM to 1 NSC-19912 pM; (LC50): 40 ~M to nM (several cell lines at 1 pM).

/°' Silicicolin/ 19186-35-7 NCI 60-tumor panel not known/ (GI50's): 100 nM to 3 NSC-403148, nM; (LC50): 50 ~M to deoxypodophyllotoxin, 10 nM

desoxypodophyllotoxin podophyllotoxin, deox silicicolin c ~ l Ii a ° a ~o a a r.
a ~a Scillaren A/ 124-99-2 NCI 60-tumor panel not known/ (GI50's): 50 nM to 0.1 NSC-7525; Gluco- ~' proscillaridin A; (LC50): 250 ~,M to 0.1 , nM
e,.:nn,.o" a ~1 0 0 0~ 00 ON
~0 ~ ,~ ,"..

I
Cinerubin A-HCl/ not reported NCI 60-tumor panel not known/ (GI50's): 15 nM to 10 NSC-243022; Cinerubin pM; (LC50): 100 ~M
A hydrochloride; to 6 nM
CL 86-F2 HCI;
CL-86-F2-hydrochloride 1 WO-09739025; US-6025466 2 EP-00626383 30 November 1994 4 W~-09705162; W~-09717364 (dolastatui synthesis and analogs) 5 Fosan licensed patent for Epothilone analogs from Sloan-Fettering; US

' WO-09723211 $JP-08092232 '° EP-00608111; EP-00632042; EP-006344.14; WO-09748278 11 EP-00425235; JP-53124692 iaUS-05416064; US-05208020; EP-00425235B
is EP-004252351 JP-53124692; US-063334.1OB1 IaJP-1104183 i6 EP_0068984 5 1'EP-00136791; EP-00087957 1$ US 50001112; US 5143906.
'9 WO-00136048 Conventional immunotoxins contain an MAb chemically conjugated to a toxin that is mutated or chemically modified to minimized binding to normal cells. Examples include anti-B4-blocked ricin, targeting CDS; and RFB4-deglycosylated ricin A chain, targeting CD22. Recombinant immunotoxins developed more recently are chimeric proteins consisting of the variable region of an antibody directed against a tumor antigen fused to a protein toxin using recombinant DNA technology. The toxin is also frequently genetically modified to remove normal tissue binding sites but retain its cytotoxicity. A
large number of differentiation antigens, overexpressed receptors, or cancer-specific antigens have been identified as targets for immunotoxins, e.g., CD19, CD22, CD20, IL-2 receptor (CD25), CD33, IL-4 receptor, EGF receptor and its mutants, ErB2, Lewis carbohydrate, mesothelin, transferrin receptor, GM-CSF receptor, Ras, Bcr-Abl, and c-Kit, for the treatment of a variety of malignancies including hematopoietic cancers, glioma, and breast, colon, ovarian, bladder, and gastrointestinal cancers. See e.g., Brinkmann et al., Expei°t Opin.
Biol. TlZer. 1:693-702 (2001); Perentesis and Sievers, HematologylOncology Clinics of North America 15:677-701 (2001).
MAbs conjugated with radioisotope are used as another means, of treating human malignancies, particularly hematopoietic malignancies, with a high level of specificity and effectiveness. The most commonly used isotopes for therapy are the high-energy -emitters, such as 1311 and 9°Y. Recently, 213Bi-labeled anti-CD33 humanized MAb has also been tested in phase I human clinical trials. Reff et al., supra.
A number of MAbs have been used for therapeutic purposes. For example, the use of rituximab (RituxanTM), a recombinant chimeric anti-CD20 MAb, for treating certain hernatopoietic malignancies was approved by the FDA in 1997. Other MAbs that have since been approved for therapeutic uses in treating human cancers include:
alemtuzumab (Campath-1HTM), a humanized rat antibody against CD52; and gemtuzumab ozogamicin (MylotargTM), a calicheamicin-conjugated humanized mouse antCD33 MAb. The FDA
is also currently examining the safety and efficacy of several other MAbs for the purpose of site-specific delivery of cytotoxic agents or radiation, e.g., radiolabeled ZevalinTM and BexxarTM. Reff et al., supra.
A second important consideration in designing a drug delivery system is the accessibility of a target tissue to a therapeutic agent. This is an issue of particular concern in the case of treating a disease of the central nervous system (CNS), where the blood-brain barrier prevents the diffusion of macromolecules. Several approaches have been developed to bypass the blood-brain barrier for effective delivery of therapeutic agents to the CNS.
The understanding of iron transport mechanism from plasma to brain provides a useful tool in bypassing the blood-brain barrier (BBB). Iron, transported in plasma by transferrin, is an essential component of virtually all types of cells. The brain needs iron for metabolic processes and receives iron through transferrin receptors located on brain capillary endothelial cells via receptor-mediated transcytosis and endocytosis. Moos and Morgan, Cellular and Molecular Neurobiology 20:77-95 (2000). Delivery systems based on transferrin-transferrin receptor interaction have been established for the efficient delivery of peptides, proteins, and liposomes into the brain. For example, peptides can be coupled with a Mab directed against the transferrin receptor to achieve greater uptake by the brain, Moos and Morgan, Supra. Similarly, when coupled with an MAb directed against the transferrin receptor, the transportation of basic fibroblast growth factor (bFGF) across the blood-brain barrier is enhanced. Song et al., The Journal of Pharmacology and Experimental Therapeutics 301:605-610 (2002); Wu et al., Journal of Drug Targeting 10:239-245 (2002).
In addition, a liposomal delivery system for effective transport of the chemotherapy drug, doxorubicin, into C6 glioma has been reported, where transferrin was attached to the distal ends of liposomal PEG chains. Eavarone et al., J. Biomed. Mater. Res. 51:10-14 (2000). A
number of US patents also relate to delivery methods bypassing the blood-brain barrier based on transferrin-transferrin receptor interaction. See e.g., US Patent Nos, 5,154,924;
5,182,107; 5,527,527; 5,833,988; 6,015,555.

There are other suitable conjugation partners for a pharmaceutical agent to bypass the blood=brain barrier. For example, LTS Patent Nos. 5,672,683, 5,977,307 and WO

relate to a method of delivering a neuropharmaceutical agent across the blood-brain barrier, where the agent is administered in the form of a fusion protein with a Iigand that is reactive S with a brain capillary endothelial cell receptor; VoTO 99/001S0 describes a drug delivery system in which the transportation of a drug across the blood-brain barrier is facilitated by conjugation with an MAb directed against human insulin receptor; W~ 89/10134 describes a chimeric peptide, which includes a peptide capable of crossing the blood brain barrier at a relatively high rate and a hydrophilic neuropeptide incapable of transcytosis, as a means of introducing hydrophilic neuropeptides into the brain; WO 01/60411 A1 provides a pharmaceutical composition that can easily transport a pharmaceutically active ingredient into the brain. The active ingredient is bound to a hibernation-specific protein that is used as a conjugate, and administered with a thyroid hormone or a substance promoting thyroid hormone production. In addition, an alternative route of drug delivery for bypassing the 1 S blood-brain barrier has been explored. For instance, intranasal delivery of therapeutic agents without the need for conjugation has been shown to be a promising alternative delivery method (Frey, 2002, Drug Delivery Technology, 2(S):46-49).
In addition to facilitating the transportation of drugs across the blood-brain barrier, transferrin-transferrin receptor interaction is also useful for specific targeting of certain tumor cells, as many tumor cells overexpress transferrin receptor on their surface.
This strategy has been used for delivering bioactive macromolecules into KS62 cells via a transferrin conjugate (Wellhoner et al., The .Iom°nal of Biological Chemistry 266:4309-4314 (1991)), and for delivering insulin into enterocyte-like Caco-2 cells via a transferrin conjugate (Shah and Shen, .Iourual of Pharmaceutical Sciences 85:1306-1311 (1996)).
2S Furthermore, as more becomes known about the functions of various iron transport proteins, such as lactotransferrin receptor, melanotransferrin, ceruloplasmin, and Divalent Cation Transporter and their expression pattern, some of the proteins involved in iron transport mechanism(e.g., melanotransferrin), or their fragments, have been found to be similarly effective in assisting therapeutic agents transport across the blood-brain barrier or targeting specific tissues (WO 02/13843 A2, WO 02/13873 A2). For a review on the use of transferrin and related proteins involved in iron uptake as conjugates in drug delivery, see Li and Qian, Medical Research Reviews 22:225-250 (2002).
The concept of tissue-specific delivery of therapeutic agents goes beyond the interaction between transferrin and transferrin receptor or their related proteins. For example, a bone-specific delivery system has been described in which proteins are conjugated with a bone-seeking aminobisphosphate for improved delivery of proteins to mineralized tissue.
Uludag and Yang, Bioteclz~ol. Pr~g. 18:604-611 (2002). For a review on this topic, see Vyas et al., Critical Reviews i~c Therapeutic Drug Ca~r~ier ~'ystern 18:1-76 (2001).
A variety of linkers may be used in the process of generating bioconjugates for the purpose of specific delivery of therapeutic agents,. Suitable linkers include homo- and heterobifunctional cross-linking reagents, which may be cleavable by, e.g., acid-catalyzed dissociation, or non-cleavable (see, e.g., Srinivasachar and Neville, Biochemistry 28:2501-2509 (1989); Wellhoner et al., The Journal of Biological Chemistry 266:4309-4314 (1991)).
Interaction between many known binding partners, such as biotin and avidin/streptavidin, can also be used as a means to join a therapeutic agent and a conjugate partner that ensures the specific and effective delivery of the therapeutic agent. Using the methods of the invention, proteins may be used to deliver molecules to intracellular compartments as conjugates.
Proteins, peptides, hormones, cytokines, small molecules or the like that bind to specific cell surface receptors that are internalized after ligand binding may be used for intracellular targeting of conjugated therapeutic compounds. Typically, the receptor-ligand complex is internalized into intracellular vesicles that are delivered to specific cell compartments, including, but not limited to, the nucleus, mitochondria, golgi, ER, lysosome, and endosome, depending on the intracellular location targeted by the receptor. By conjugating the receptor ligand with the desired molecule, the drug will be carried with the receptor-ligand complex and be delivered to the intracellular compartments normally targeted by the receptor. The drug can therefore be delivered to a specific intracellular location in the cell where it is needed to treat a disease.
Many proteins may be used to target therapeutic agents to specific tissues and organs.
Targeting proteins include, but are not limited to, growth factors (EP~, HGH, EGF, nerve growth factor, FGF, among others), cytokines (GM-CSF, G-CSF, the interferon family, interleukins, among others), hormones (FSH, LH, the steroid families, estrogen, corticosteroids, insulin, among others), serum proteins (albumin, lipoproteins, fetoprotein, human serum proteins, antibodies and fragments of antibodies, among others), and vitamins (folate, vitamin C, vitamin A, among others). Targeting agents are available that are specific for receptors on most cells types.
Contemplated linkage configurations include, but are not limited to, protein-sugar-linker-sugar-protein and multivalent forms thereof, protein-sugar-linker-protein and multivalent forms thereof, protein-sugar-linker-therapeutic agent, where the therapeutic agent includes, but are not limited to, small molecules, peptides and lipids. In some embodiments, a hydrolysable linker is used that can be hydrolyzed once internalized. An acid labile linker can be used to advantage where the protein conjugate is internalized into the endosomes or lysosomes which have an acidic pH. Once internalized into the endosome or lysosome, the linker is hydrolyzed and the therapeutic agent is released from the targeting agent.
In an exemplary embodiment, transferrin is conjugated via a linker to an enzyme or a nucleic acid vector that encoded the enzyme desired to be targeted to a cell that presents transferrin receptors in a patient. The patient could, for example, require enzyme replacement therapy for that particular enzyme. In particularly preferred embodiments, the enzyme is one that is lacking in a patient with a lysosomal storage disease (see Table 5).
Once in circulation, the transferrin-enzyme conjugate is linked to transferrin receptors and is internalized in early endosomes (Xing et al., 1998, Biochem. J. 336:667; Li et al., 2002, Trends in Pharmcol. Sci. 23:206; Suhaila et al., 1998, J. Biol. Chem.
273:14355). Other contemplated targeting agents that are related to transferrin include, but are not limited to, lactotransferrin (lactoferrin), melanotransferrin (p97), ceruloplasmin, and divalent cation transporter.
In another exemplary embodiment, transferrin-dystrophin conjugates would enter endosomes by the transferrin pathway. Once there, the dystrophin is released due to a hydrolysable linker which can then be taken to the intracellular compartment where it is required. This embodiment may be used to treat a patient with muscular dystrophy by supplementing a genetically defective dystrophin gene and/or protein with the functional dystrophin peptide connected to the transferrin.

E Therapeutic Moieties In another preferred embodiment, the modified sugar includes a therapeutic moiety.
Those of skill in the art will appreciate that there is overlap between the category of therapeutic moieties and biomolecules; many biomolecules have therapeutic properties or potential.
The therapeutic moieties can be agents already accepted for clinical use or they can be drugs whose use is experimental, or whose activity or mechanism of action is under investigation. The therapeutic moieties can have a proven action in a given disease state or can be only hypothesized to show desirable action in a given disease state. In a preferred embodiment, the therapeutic moieties are compounds, which are being screened for their ability to interact with a tissue of choice. Therapeutic moieties, which are useful in practicing the instant invention include drugs from a broad range of drug classes having a variety of pharmacological activities. In some embodiments, it is preferred to use therapeutic moieties that are not sugars. An exception to this preference is the use of a sugar that is modified by covalent attachment of another entity, such as a PEG, biomolecule, therapeutic moiety, diagnostic moiety and the like. In an exemplary embodiment, an antisense nucleic acid moeity is conjugated to a linker arm which is attached to the targeting moiety. In another exemplary embodiment, a therapeutic sugar moiety is conjugated to a linker arm and the sugar-linker arm cassette is subsequently conjugated to a peptide via a method of the invention.
Methods of conjugating therapeutic and diagnostic agents to various other species are well known to those of skill in the art. See, for example Hermanson, BIOCONJUGATE
TECHNIQUES, Academic Press, San Diego, 1996; and Dunn et al., Eds. POLYMERIC
DRUGS
AND DRUG DELIVERY SYSTEMS, ACS Symposium Series Vol. 469, American Chemical Society, Washington, D.C. 1991.
In an exemplary embodiment, the therapeutic moiety is attached to the modified sugar via a linkage that is cleaved under selected conditions. Exemplary conditions include, but are not limited to, a selected pH (e.g., stomach, intestine, endocytotic vacuole), the presence of an active enzyme (e.g., esterase, protease, reductase, oxidase), light, heat and the like. Many cleavable groups are known in the art. See, for example, Jung et al., Biochem.
Biophys. Acta, 761: 152-162 (1983); Joshi et al., .I: Bi~l. CheryZ., 265: 14518-14525 (1990);
~axling et al., ,I.

Immuhol.,124: 913-920 (1980); Bouizar et al., Eu~. J. Biochem.,155: 141-147 (1986); Park et al., J. Biol. Chen2., 261: 205-210 (1986); Browning et al., J.
Imn2unol.,143: 1859-1867 (1989).
Classes of useful therapeutic moieties include, for example, non-steroidal anti-inflammatory drugs (NSAIDS). The IVSAIIW can, for example, be selected from the following categories: (e.g., propionic acid derivatives, acetic acid derivatives, fenamic acid derivatives, biphenylcarboxylic acid derivatives and oxicams); steroidal anti-inflammatory drugs including hydrocortisone and the like; adjuvants; antihistaminic drugs (e.g., chlorpheniramine, triprolidine); antitussive drugs (e.g., dextromethorphan, codeine, caramiphen and carbetapentane); antipruritic drugs (e.g., methdilazine and trimeprazine);
anticholinergic drugs (e.g., scopolamine, atropine, homatropine, levodopa);
anti-emetic and antinauseant drugs (e.g., cyclizine, meclizine, chlorpromazine, buclizine);
anorexic drugs (e.g., benzphetamine, phentermine, chlorphentermine, fenfluramine); central stimulant drugs (e.g., amphetamine, methamphetamine, dextroamphetamine and methylphenidate);
antiarrhythmic drugs (e.g., propanolol, procainamide, disopyramide, quinidine, encainide); (3-adrenergic blocker drugs (e.g., metoprolol, acebutolol, betaxolol, labetalol and timolol);
cardiotonic drugs (e.g., milrinone, amrinone and dobutamine); antihypertensive drugs (e.g., enalapril, clonidine, hydralazine, minoxidil, guanadrel, guanethidine);diuretic drugs (e.g., amiloride and hydrochlorothiazide); vasodilator drugs (e.g., diltiazem, amiodarone, isoxsuprine, nylidrin, tolazoline and verapamil); vasoconstrictor drugs (e.g., dihydroergotamine, ergotamine and methylsergide); antiulcer drugs (e.g., ranitidine and cimetidine); anesthetic drugs (e.g., lidocaine, bupivacaine, chloroprocaine, dibucaine);
antidepressant drugs (e.g., imipramine, desipramine, amitryptiline, nortryptiline); tranquilizer and sedative drugs (e.g., chlordiazepoxide, benacytyzine, benzquinamide, flurazepam, hydroxyzine, loxapine and promazine); antipsychotic drugs (e.g., chlorprothixene, fluphenazine, haloperidol, molindone, thioridazine and trifluoperazine);
antimicrobial drugs (antibacterial, antifungal, antiprotozoal and antiviral drugs).
Classes of useful therapeutic moieties include adjuvants. The adjuvants can, for example, be selected from keyhole lymphet hemocyanin conjugates, monophosphoryl lipid A, mycoplasma-derived lipopeptide MALP-2, cholera toxin B subunit, Eschenichia coli heat-labile toxin, universal T helper epitope from tetanus toxoid, interleukin-12, CpG

oligodeoxynucleotides, dimethyldioctadecylammoniuxn bromide, cyclodextrin, squalene, aluminum salts, meningococcal outer membrane vesicle (OMV), montanide ISA, TiterMaxTM
(available from Sigma, St. Louis MO), nitrocellulose absorption, immune-stimulating complexes such as Quil A, GerbuTM adjuvant (Gerbu Biotechnik, I~irchwald, Germany), threonyl muramyl dipeptide, thymosin alpha, bupivacaine, GM-CSF, Incomplete Freund's Adjuvant, MTP-FE/MF59 (Ciba/Geigy, Basel, Switzerland), polyphosphazene, saponin derived from the soapbark tree Quillaja sap~~zaria, and Syntex adjuvant formulation (Biocine, Emeryville, CA), among others well known to those in the art.
Antimicrobial drugs which are preferred for incorporation into the present composition include, for example, pharmaceutically acceptable salts of (3-lactam drugs, quinolone drugs, ciprofloxacin, norfloxacin, tetracycline, erythromycin, amikacin, triclosan, doxycycline, capreomycin, chlorhexidine, chlortetracycline, oxytetracycline, clindamycin, ethambutol, hexamidine isothionate, metronidazole, pentamidine, gentamycin, kanamycin, lineomycin, methacycline, methenamine, minocycline, neomycin, netilmycin, paromomycin, streptomycin, tobramycin, miconazole and amantadine.
Other drug moieties of use in practicing the present invention include antineoplastic drugs (e.g., antiandrogens (e.g., leuprolide or flutamide), cytocidal agents (e.g., adriamycin, doxorubicin, taxol, cyclophosphamide, busulfan, cisplatin, (3-2-interferon) anti-estrogens (e.g., tamoxifen), antimetabolites (e.g., fluorouracil, methotrexate, mercaptopurine, thioguanine). Also included within this class are radioisotope-based agents for both diagnosis and therapy, and conjugated toxins, such as ricin, geldanamycin, mytansin, CC-1065, C-1027, the duocarmycins, calicheamycin and related structures and analogues thereof, and the toxins listed in Table 2.
The therapeutic moiety can also be a hormone (e.g., medroxyprogesterone, estradiol, leuprolide, megestrol, octreotide or somatostatin); muscle relaxant drugs (e.g., cinnamedrine, cyclobenzaprine, flavoxate, orphenadrine, papaverine, mebeverine, idaverine, ritodrine, diphenoxylate, dantrolene and azumolen); antispasmodic drugs; bone-active drugs (e.g., diphosphonate and phosphonoalkylphosphinate drug compounds); endocrine modulating drugs (e.g., contraceptives (e.g., ethinodiol, ethinyl estradiol, norethindrone, mestranol, desogestrel, medroxyprogesterone), modulators of diabetes (e.g., glyburide or chlorpropamide), anabolics, such as testolactone or stanozolol, androgens (e.g., methyltestosterone, testosterone or fluoxymesterone), antidiuretics (e.g., desmopressin) and calcitonins).
Also of use in the present invention are estrogens (~.g., diethylstilbesterol), glucocorticoids (e.g., triamcinolone, betamethasone, etc.) andprogesterones, such as norethindrone, ethynodiol, norethindrone, levonorgestrel; thyroid agents (e.g., liothyronine or levothyroxine) or anti-thyroid agents (e.g., methimazole);
antihyperprolactinemic drugs (e.g., cabergoline); hormone suppressors (e.g., danazol or goserelin), oxytocics (e.g., methylergonovine or oxytocin) and prostaglandins, such as mioprostol, alprostadil or dinoprostone, can also be employed.
Other useful modifying groups include immunomodulating drugs (e.g., antihistamines, mast cell stabilizers, such as lodoxamide and/or cromolyn, steroids (e.g., triamcinolone, beclomethazone, cortisone, dexamethasone, prednisolone, methylprednisolone, beclomethasone, or clobetasol), histamine H2 antagonists (e.g., famotidine, cimetidine, ranitidine), immunosuppressants (e.g., azathioprine, cyclosporin), etc.
Groups with anti-inflammatory activity, such as sulindac, etodolac, ketoprofen and ketorolac, are also of use. Other drugs of use in conjunction with the present invention will be apparent to those of skill in the art.
Classes of useful therapeutic moieties include, for example, antisense drugs and also naked DNA. The antisense drugs can be selected from for example Affinitak (ISIS, Carlsbad, CA) and Genasense ~ (from Genta, Berkeley Heights, NJ). Naked DNA
can be delivered as a gene therapy therapeutic for example with the DNA encoding for example factors VIII and IX for treatment of hemophilia disorders.
F Prepaxation of Modified Sugars Modified sugars useful in forming the conjugates of the invention are discussed herein. The discussion focuses on preparing a sugar modified with a water-soluble polymer for clarity of illustration. In particular, the discussion focuses on the preparation of modified sugars that include a polyethylene glycol) moiety. Those of skill will appreciate that the methods set forth herein axe broadly applicable to the preparation of modified sugars, therefore, the discussion should not be interpreted as limiting the scope of the invention.

In general, the sugar moiety and the modifying group are linked together through the use of reactive groups, which are typically transformed by the linking process into a new organic functional group or unreactive species. The sugar reactive functional group(s), is located at any position on the sugar moiety. lZeactive groups and classes of reactions useful in practicing the present invention are generally those that are well known in the art of bioconjugate chemistry. Currently favored classes of reactions available with reactive sugar moieties are those, which proceed under relatively mild conditions. These include, but are not limited to nucleophilic substitutions (e.g., reactions of amines and alcohols with aryl halides, active esters), electrophilic substitutions (e.g., enamine reactions) and additions to carbon-carbon and carbon-heteroatom multiple bonds (e.g., Michael reaction, Diels-Alder addition). These and other useful reactions are discussed in, for example, Smith and March, ADVANCED ORGANIC CHEMISTRY, 5th Ed., John Wiley ~ Sons, New York, 2001;
Hermanson, BIOCONJUGATE TECHNIQUES, Academic Press, San Diego, 1996; and Feeney et al., MODIFICATION OF PROTEINS; Advances in Chemistry Series, Vol. 198, American Chemical Society, Washington, D.C., 1982.
Useful reactive functional groups pendent from a sugar nucleus or modifying group include, but are not limited to:
(a) carboxyl groups and various derivatives thereof including, but not limited to, N-hydroxysuccinimide esters, N-hydroxybenzotriazole esters, acid halides, acyl imidazoles, thioesters, p-nitrophenyl esters, alkyl, alkenyl, alkynyl and aromatic esters;
(b) hydroxyl groups, which can be converted to, e.g., esters, ethers, aldehydes, etc.
(c) haloalkyl groups, wherein the halide can be later displaced with a nucleophilic group such as, for example, an amine, a carboxylate anion, thiol anion, caxbanion, or an alkoxide ion, thereby resulting in the covalent attachment of a new group at the functional group of the halogen atom;
(d) dienophile groups, which are capable of participating in Diels-Alder reactions such as, for example, maleimido groups;
(e) aldehyde or ketone groups, such that subsequent derivatization is possible via formation of carbonyl derivatives such as, for example, imines, hydrazones, semicarbazones or oximes, or via such mechanisms as Grignard addition or alkyllithium addition;

(f) sulfonyl halide groups for subsequent reaction with amines, for example, to form sulfonamides;
(g) thiol groups, which can be, for example, converted to disulfides or reacted with alkyl and aryl halides;
(h) amine or sulfhydryl groups, which can be, for example, acylated, alkylated or oxidized;
(i) alkenes, which can undergo, for example, cycloadditions, acylation, Michael addition, etc; and (j) epoxides, which can react with, for example, amines and hydroxyl compounds.
The reactive functional groups can be chosen such that they do not participate in, or interfere with, the reactions necessary to assemble the reactive sugar nucleus or modifying group. Alternatively, a reactive functional group can be protected from participating in the reaction by the presence of a protecting group. Those of skill in the art understand how to protect a particular functional group such that it does not interfere with a chosen set of reaction conditions. For examples of useful protecting groups, see, for example, Greene et al., PROTECTIVE GROUPS IN ORGANIC SYNTHESIS, John Wiley & Sons, New York, 1991.
In the discussion that follows, a number of specific examples of modified sugars that are useful in practicing the present invention are set forth. In the exemplary embodiments, a sialic acid derivative is utilized as the sugar nucleus to which the modifying group is attached. The focus of the discussion on sialic acid derivatives is for clarity of illustration only and should not be construed to limit the scope of the invention. Those of skill in the art will appreciate that a variety of other sugar moieties can be activated and derivatized in a manner analogous to that set forth using sialic acid as an example. For example, numerous methods are available for modifying galactose, glucose, N-acetylgalactosamine and fucose to name a few sugar substrates, which are readily modified by art recognized methods. See, for example, Elhalabi et al., Cur. Med. Chem. 6: 93 (1999); and Schafer et al., J.
O~g. Chem.
65: 24 (2000).
In an exemplary embodiment, the peptide that is modified by a method of the invention is a peptide that is produced in mammalian cells (e.g., CHO cells) or in a transgenic animal and thus, contains N- and/or ~-linked oligosaccharide chains, which are incompletely sialylated. The oligosaccharide chains of the glycopeptide lacking a sialic acid and containing a terminal galactose residue can be PEGylated, PPGylated or otherwise modified with a modified sialic acid.
In Scheme 4, the mannosamine glycoside l, is treated with the active ester of a protected amino acid (e.~., glycine) derivative, converting the sugar amine residue into the corresponding protected amino acid amide adduct. The adduct is treated with an aldolase to form the sialic acid 2. Compound 2 is converted to the corresponding CMP
derivative by the action of CMP-SA synthetase, followed by catalytic hydrogenation of the CMP
derivative to produce compound 3. The amine introduced via formation of the glycine adduct is utilized as a locus of PEG or PPG attachment by reacting compound 3 with an activated PEG
or PPG
derivative (e.g.., PEG-C(O)NHS, PPG-C(O)NHS), producing 4 or 5, respectively.
Scheme 4 OH 1. CMP-SA synthetase, CTP
HO 1. Z-Glycine-NHS HO 2. H2/Pd/C
NHS OH
HO 2. NeuAc Aldolase, pyruvate HO ~ O O-+Na HO O Z~N~NH OH O
OH H

NHS I 'N
O I Nk0 0 0_P_0 o N'~O
O-P_O O +
+ ~ PEG-~-NHS HO OH O- Na HO O Na HO ~ O O-+Na HO OH
II HO ~H 0 O-+Na HO OH ~ O
PEG-C~H~NH OH O H2N ONH OH 3 O
CMP-SA-5-NHCOCH2NH-PEG ppC,-~-NHS CMP-SA-5-NHCOCHzNH2 Table 3 sets forth representative examples of sugar monophosphates that are derivatized with a PEG or PPG moiety. Certain of the compounds of Table 3 are prepared by the method of Scheme 1. ~ther derivatives are prepared by art-recognised methods. S'ee, for example, I~eppler et al., Calyc~bi~l~~y 11: 111 (2001); and Charter et al., Cslye~bi~l~~y 1~:
1049 (2000)). ~ther amine reactive PEG and PPG analogues are commercially available, or they can be prepared by methods readily accessible to those of skill in the art.
Table ~. Examples of sugar monophosphates that are derivati~ed with a PEG or PPG
moiety NHz NHS
I N'~O ~ I N ~O
O-p_O O O.-P_O O
HO OH I ~_+N~ HO H ~_+N~
HOJ ~~ -- ~p ~O'+Na HBO--O(H R-NH ~ O ~-O'+N°a HO~-O(H
R_O-voH o AcN'/,~ o CMP-KDN-5-O-R CMP-NeuAc-9-NH-R NHN
NHa O I ' 'N II N O
.~0 R-NH 0 0' N~a o R-O O P_ ~~ O'~ HO ~H O O'+Na Hp pH
O Na \ /
HO ~~H~ O ~O_+Na HO~IOH ACNH ' OH
AcN\J~~ o OH CMP-NeuAc-8-NH-R
NHz CMP-NeuAc-8-O-R NHS N
O
O I 'k _P_0 O N O
O-P-O~ O N O HO NH-R o-+N~
HO O-R p-+Na N HO O ~--O'+Na HO OH
HO ~~O ~-O-+Na HO OH ACNH ' OH O
AcN\J~~ o CMP-NeuAc-7-NH-R NHS
CMP-NeuAc-7-O-R ~ O I ' N
I Nk0 O-P_O~ O ~N~O
HO OH o P ~~ HO OH p-+Na \ /
O' Na Hp O O'+Na HO~OH
HO ~~O ~-O-+Na HO OH AcNH - O
AcN\/~~ o NH-R
O-R
1 ~ CMP-NeuAc-4-O-R CMI'-NeuAc-4-NH-R

NHZ z 'N ~ 'N
~~O .-P_O ~ N'~O
~ P-O O O
HO OH ~_+i ~ HO H ~.+~
HO',,~-O_+i HO OH R-~~~ ~ ~_+~ HO OH
R-~IH OH O AcNH OH
CMP-SA-5-NH-R CMP-NeuAc-9-O-R
The modified sugar phosphates of use in practicing the present invention can be substituted in other positions as well as those set forth above. "i" may be Na or another salt and "i" may be interchangeable with Na. Presently preferred substitutions of sialic acid are set forth in Formula 5.
Formula 5:
NHS
~~ 'N
O ~'~O
O_P_O O
R2-Y X-R1 o-+i R3-B O O-+i HO OH

~-R5~6 (I) in which X is a linking group, which is preferably selected from -O-, -N(H)-, -S, CH2-, and N(R)2, in which each R is a member independently selected from RI-R5. "i" may be Na or another salt, and Na may be interchangeable with "i:The symbols Y, Z, A and B
each represent a group that is selected from the group set forth above for the identity of X. X, Y, Z, A and B are each independently selected and, therefore, they can be the same or different. The symbols Rl, R2, R3, R4 and RS represent H, polymers, a water-soluble polymer, therapeutic moiety, biomolecule or other moiety. The symbol R6 represents H, OH, or a polymer. Alternatively, these symbols represent a linker that is linked to a polymer, water-soluble polymer, therapeutic moiety, biomolecule or other moiety.
In another exemplary embodiment, a mannosamine is simultaneously acylated and activated for a nucleophilic substitution by the use of chloroacetic anhydride as set forth in Scheme 5. In each of the schemes presented in this section, i+ or Nab can be interchangeable, wherein the salt can be sodium, or can be any other suitable salt.

Scheme 5 0II H3C~0_+Na O
OH
HCI ~H
CI~
~CI
HN~CI
NH

HO. ~ ~ HO
HO Z ~-- H~
HO. s O
Z0 '~ HO ~O COOH
~ CI~' H

O
H~ MeOH
HO

o.,~OH Aldolase O
~, OH 0.1M HEPES

pH 7.5 CTP, CMP-sialic acid Synthetase, Buffer, MgCl2 NHZ
NHz IPI I ~ N
N O~ I~O~N~O
~ ~ ~~ OH
HO O/ I\O~ aO-NI~O ~O~~'n SH HO H HN O 00 Na .O, ~H O- Na HO O COO'+Na CI O HO HO OH
~O~O~S~HN HO HO OH
n O
The resulting chloro-derivatized glycan is contacted with pyruvate in the presence of an aldolase, forming a chloro-derivatized sialic acid. The corresponding nucleotide sugar is prepared by contacted the sialic acid derivative with an appropriate nucleotide triphosphates and a synthetase. The chloro group on the sialic acid moiety is then displaced with a nucleophilic PEG derivative, such as thio-PEG.
In a further exemplary embodiment, as shown is Scheme 6, a mannosamine is acylated with a bis-HOBT dicarboxylate, producing the corresponding amido-alkyl-carboxylic acid, which is subsequently converted to a sialic acid derivative.
The sialic acid derivative is converted to a nucleotide sugar, and the carboxylic acid is activated and reacted with a nucleophilic PEG derivative, such as amino-PEG.
Scheme 6 0 0 0 0 °
NHz HCI H3C~0-+Na HO OH OH
HO. HOBT~HOBT H~NOH O HOCOOH
HO -~H ~N
H~O ~O MeOH/H20 HOO ZO HO
OH OH Aldolase O O
m=0-20 0.1M HEPES
pH 7.5 37 °C CTP, CMP-slalic acid Synthetase, Buffer, MgClp NHZ
I ~N
P
N N ~ _ ' ~ HO OH O/o_\N O N O
P ~~O ~O~O~NHx HO m0 HN O COO-+Na Y
HO OH O'"Na~ BOP, HOBT HO HO OH
~ -O, ~ H COO-+Na 01 v ~HN~HN
HO HO OH
O O

In another exemplary embodiment, set forth in Scheme 7, amine- and carboxyl-protected neuraminic acid is activated by converting the primary hydroxyl group to the corresponding p-toluenesulfonate ester, and the methyl ester is cleaved. The activated neuraminic acid is converted to the corresponding nucleotide sugar, and the activating group is displaced by a nucleophilic PEG species, such as thio-PEG.
Scheme 7 HO OH 1. Tosyl (Ts)-CI O HO OH
HO OH ~ Pyridine Me ~ ~ S-O OH ~ CTP, AcHN ~ COOMe - n ~ O COOH CMP-sialic acid HO Synthetase, 2. H20, base O AcH
Buffer, MgCl2 NHZ NHa O ~N O I ~N
I / P\
P\ ~ ~ ~ , ~ O 01 N O
, ~ HO O ~ + O O N O ~O~O~SH HO OH O-*Na 0O
'O~O~S ~H 0 00' N~ Ts/O AcH'COO *N HO OH
AcHN HO HO OH HO
In yet a further exemplary embodiment, as set forth in Scheme ~, the primary hydroxyl moiety of an amine- and carboxyl-protected neuraminic acid derivative is alkylated using an electrophilic PEG, such as chloro-PEG. The methyl ester is subsequently cleaved and the PEG-sugar is converted to a nucleotide sugar.
Scheme 8 1. Pyridine HO OH ~ H ~O~O~CI w0~0~0 HO OH ~ H CTP, O COOMe ----~ CMP-sialicacid HO AcHN . n AcHN O COOH Synthetase, HO 2. H20, base HO Buffer, MnClz NHZ
c I ~N
\O O N.~O
~O~O~O HO off O O N
AcHN COO'*Na HO HO ~H
Glycans other than sialic acid can be derivatized with PEG using the methods set forth herein. The derivatized glycans, themselves, are also within the scope of the invention.
Thus, Scheme 9 provides an exemplary synthetic route to a PEGylated galactose nucleotide sugar. The primary hydroxyl group of galactose is activated as the corresponding toluenesulfonate ester, which is subsequently converted to a nucleotide sugar.
Scheme 9 ~ Me OH 0~0 ~<~n O
OHC,alactose kinase, ATP _ HO~ O ~ ~NH
HO
HO HO OH UDP-galactose uridyltransferase 0~ i ~O~ i ~O ~NI O
UTP, glucose-1-phosphate, O'*Na O'*Na~
UDP-glucose pyrophosphorylas Ve HO OH
~ j ~ ~n HS~O~O~
~ ~~ ~ ~n ~O~O~
OH S
O O
HO~ 0 O
HO
O~i~O~I~O ~N' 0 O-*Na O°Na~
HO~O/H
Scheme 10 sets forth an exemplary route for preparing a galactose-PEG
derivative that is based upon a galactose-6-amine moiety. Thus, galactosamine is converted to a nucleotide sugar, and the amine moiety of galactosamine is functionalized with an active PEG derivative.
Scheme 10 OH
OH >NH2 t~,alactose kinase,NHz ATP HO~

O
O
II
II

HO HO OH ~p-gaiactose pyrophosphorylaseHO
0 ~
i ~p~
i ~O
I
N
O

UTP 0_+Na O_+Na~
~

HO
OH

BOP n HO~0~0~
~ ~'~ _ , 1f O n m =

HO O
HN

~!
0' O
HO~

O
I ~
NH
HO
I
I
I

P
O~i~0~I~0 N~0 O

-_ I
0_+Na 0_+Na Y

~J

O
H
O
H

Scheme 11 provides another exemplary route to galactose derivatives. The starting point for Scheme 11 is galactose-2-amine, which is converted to a nucleotide sugar. The amine moiety of the nucleotide sugar is the locus for attaching a PEG
derivative, such as Methoxy-PEG (mPEG) carboxylic acid.
Scheme 11 OH OH
HO
O ~ Galactose kinase, ATP _ HOO O NH
HO~~OH UUP-galactosepyrophosphorylase HzNO~P~O~i~O ~'~O
z UTP O-*Na O-*Na~
HO~/OH
, ~O
~p~O~OH BOP
n HO OH
~O~ O
O HO~ P P I NH
'/ ~ HN I
~O Or ~O~ y0 ~O
O O-*Na O-*Na~
HOVOH
Exemplary moieties attached to the conjugates disclosed herein include, but are not limited to, PEG derivatives (e.g., acyl-PEG, acyl-alkyl-PEG, alkyl-acyl-PEG
carbamoyl-PEG, aryl-PEG, alkyl-PEG), PPG derivatives (e.g., acyl-PPG, acyl-alkyl-PPG, alkyl-acyl-PPG carbamoyl-PPG, aryl-PPG), polyapartic acid, polyglutamate, polylysine, therapeutic moieties, diagnostic moieties, mannose-6-phosphate, heparin, heparan, SLe", mannose, mannose-6-phosphate, Sialyl Lewis X, FGF, VFGF, proteins (e.g., transferriri), chondroitin, keratan, dermatan, dextran, modified dextran, amylose, bisphosphate, poly-SA, hyaluronic acid, keritan, albumin, integrins, antennary oligosaccharides, peptides and the like. Methods of conjugating the various modifying groups to a saccharide moiety are readily accessible to those of skill in the art (POLY (ETHYLENE GLYCOL CHEMISTRY : BIOTECHNICAL AND
BIOMEDICAL APPLICATIONS, 1. Milton Harris, Ed., Plenum Pub. Corp., 1992; POLY
(ETHYLENE GLYCOL) CHEMICAL AND BIOLOGICAL APPLICATIONS, J. Milton Harris, Ed., ACS
Symposium Series No. 6~0, American Chemical Society, 1997; Hermanson, BIOCONJUGATE
TECHNIQUES, Academic Press, San Diego, 1996; and Dunn et al., Eds. POLYMERIC
DRUGS

Al~m DRUG DELIVERY SYSTEMS, ACS Symposium Series Vol. 469, American Chemical Society, Washington, D.C. 1991).
Purification of sugars nucleotide sugars and derivatives The nucleotide sugars and derivatives pr~du~ed by the above processes can be used without purification. However, it is usually preferred t~ recover the product.
Standard, well-known techniques for recovery of glycosylated saccharides such as thin or thick layer chromatography, column chromatography, ion exchange chromatography, or membrane filtration can be used. It is preferred t~ use membrane filtration, more preferably utilizing a reverse osmotic membrane, or one or more column chromatographic techniques for the recovery as is discussed hereinafter and in the literature cited herein. For instance, membrane filtration wherein the membranes have molecular weight cutoff of about 3000 to about 10,000 can be used to remove proteins for reagents having a molecular weight of less than 10,000 Da.. Membrane filtration or reverse osmosis can then be used to remove salts and/or purify the product saccharides (see, e.g., WO 98/15581). Nanofilter membranes are a class of reverse osmosis membranes that pass monovalent salts but retain polyvalent salts and uncharged solutes larger than about 100 to about 2,000 Daltons, depending upon the membrane used. Thus, in a typical application, saccharides prepared by the methods of the present invention will be retained in the membrane and contaminating salts will pass through.
G. Cross-linking Groups Preparation of the modified sugar for use in the methods of the present invention includes attachment of a modifying group to a sugar residue and forming a stable adduct, which is a substrate for a glycosyltransferase. Thus, it is often preferred to use a cross-linking agent to conjugate the modifying group and the sugar. Exemplary bifunctional compounds which can be used for attaching modifying groups to carbohydrate moieties include, but are not limited to, bifunctional polyethylene glycols), polyamides, polyethers, polyesters and the like. General approaches for linking carbohydrates t~ other molecules are known in the literature. See, for example, Lee et al., Biochemistry 28: 1856 (1989); Bhatia et al., Anal. Biochenz. 178: 408 (1989); Janda et al., J. Ana. Chew. ~'oc.
112: 8886 (1990) and Bednarski et al., WO 92/18135. In the discussion that follows, the reactive groups are treated as benign on the sugar moiety of the nascent modified sugar. The focus of the discussion is for clarity of illustration. Those of skill in the art will appreciate that the discussion is relevant to reactive groups on the modifying group as well.
An exemplary strategy involves incorporation of a protected sulfhydryl onto the sugar using the heterobifunctional crosslinker SPDP (n-succinimidyl-3-(2-pyridyldithio)propionate and then deprotecting the sulfhydryl for formation of a disulfide bond with another sulfhydryl on the modifying group.
If SPDP detrimentally affects the ability of the modified sugar to act as a glycosyltransferase substrate, one of an array of other crosslinkers such as 2-iminothiolane or N-succinimidyl S-acetylthioacetate (SATA) is used to form a disulfide bond. 2-iminothiolane reacts with primary amines, instantly incorporating an unprotected sulfhydryl onto the amine-containing molecule. SATA also reacts with primary amines, but incorporates a protected sulfhydryl, which is later deacetylated using hydroxylamine to produce a free sulfhydryl. In each case, the incorporated sulfhydryl is free to react with other sulfliydryls or protected sulfllydryl, like SPDP, forming the required disulfide bond.
The above-described strategy is exemplary, and not limiting, of linkers of use in the invention. Other crosslinkers are available that can be used in different strategies for crosslinking the modifying group to the peptide. For example, TPCH(S-(2-thiopyridyl)-L-cysteine hydrazide and TPMPH ((S-(2-thiopyridyl) mercapto-propionohydrazide) react with carbohydrate moieties that have been previously oxidized by mild periodate treatment, thus forming a hydrazone bond between the hydrazide portion of the crosslinker and the periodate generated aldehydes. TPCH and TPMPH introduce a 2-pyridylthione protected sulfhydryl group onto the sugar, which can be deprotected with DTT and then subsequently used for conjugation, such as forming disulfide bonds between components.
If disulfide bonding is found unsuitable for producing stable modified sugars, other crosslinkers may be used that incorporate more stable bonds between components. The heterobifunctional crosslinkers GMBS (N-gama-malimidobutyryloxy)succinimide) and SMCC (succinimidyl 4-(N-maleimido-methyl)cyclohexane) react with primary amines, thus introducing a maleimide group onto the component. The maleimide group can subsequently react with sulfhydryls on the other component, which can be introduced by previously mentioned crosslinkers, thus forming a stable thioether bond between the components. If steric hindrance between components interferes with either component's activity or the ability of the modified sugar to act as a glycosyltransferase substrate, crosslinkers can be used which introduce long spacer arms between components and include derivatives of some of the previously mentioned crosslinkers (i.e., SPDP). Thus, there is an abundance of suitable crosslinkers, which are useful; each of which is selected depending on the effects it has on optimal peptide conjugate and modified sugar production.
A variety of reagents are used to modify the components of the modified sugar with intramolecular chemical crosslinks (for reviews of crosslinking reagents and crosslinking procedures see: Wold, F., Meth. E~zymol. 25: 623-651, 1972; Weetall, H. H., and Cooney, D.
A., In: ENZYMES AS DRUGS. (Holcenberg, and Roberts, eds.) pp. 395-442, Wiley, New York, 1981; Ji, T. H., Meth. EhzynZOl. 91: 580-609, 1983; Mattson et al., Mol. Biol.
Rep. 17: 167-183, 1993, all of which are incorporated herein by reference). Preferred crosslinking reagents are derived from various zero-length, homo-bifunctional, and hetero-bifunctional crosslinking reagents. Zero-length crosslinking reagents include direct conjugation of two intrinsic chemical groups with no introduction of extrinsic material. Agents that catalyze formation of a disulfide bond belong to this category. Another example is reagents that induce condensation of a carboxyl and a primary amino group to form an amide bond such as carbodiimides, ethylchloroformate, Woodward's reagent I~ (2-ethyl-5-phenylisoxazolium-3'-sulfonate), and carbonyldiimidazole. In addition to these chemical reagents, the enzyme transglutaminase (glutamyl-peptide ~y-glutamyltransferase; EC 2.3.2.13) may be used as zero-length crosslinking reagent. This enzyme catalyzes acyl transfer reactions at caxboxamide groups of protein-linked glutaminyl residues, usually with a primary amino group as substrate. Preferred homo- and hetero-bifunctional reagents contain two identical or two dissimilar sites, respectively, which may be reactive for amino, sulfhydryl, guanidino, indole, or nonspecific groups.
2 Preferred Specific Sites in Crosslinking Reagents a. Amino-Reactive Groups In one preferred embodiment, the sites on the cross-linker are amino-reactive groups.
Useful non-limiting examples of amino-reactive groups include N-hydroxysuccinimide (NHS) esters, imidoesters, isocyanates, acylhalides, arylazides, p-nitrophenyl esters, aldehydes, and sulfonyl chlorides.

NHS esters react preferentially with the primary (including aromatic) amino groups of a modified sugar component. The imidazole groups of histidines are known to compete with primaxy amines for reaction, but the reaction products are unstable and readily hydrolyzed.
The reaction involves the nucleophilic attack of an amine on the acid carboxyl of an NHS
ester to form an amide, releasing the N-hydroxysuccinimide. Thus, the positive chaxge of the original amino group is lost.
Imidoesters are the most specific acylating reagents for reaction with the amine groups of the modified sugar components. At a pH between 7 and 10, imidoesters react only with primary amines. Primary amines attack imidates nucleophilically to produce an intermediate that breaks down to amidine at high pH or to a new imidate at low pH. The new imidate can react with another primary amine, thus crosslinking two amino groups, a case of a putatively monofunctional imidate reacting bifunctionally. The principal product of reaction with primary amines is an amidine that is a stronger base than the original amine.
The positive charge of the original amino group is therefore retained.
Isocyanates (and isothiocyanates) react with the primary amines of the modified sugar components to form stable bonds. Their reactions with sulfliydryl, imidazole, and tyrosyl groups give relatively unstable products.
Acylazides are also used as amino-specific reagents in which nucleophilic amines of the affinity component attack acidic caxboxyl groups under slightly alkaline conditions, e.g.
pH 8.5.
Arylhalides such as 1,5-difluoro-2,4-dinitrobenzene react preferentially with the amino groups and tyrosine phenolic groups of modified sugar components, but also with sulfhydryl and imidazole groups.
p-Nitrophenyl esters of mono- and dicarboxylic acids are also useful amino-reactive groups. Although the reagent specificity is not very high, a- and s-amino groups appear to react most rapidly.
Aldehydes such as glutaraldehyde react with primary amines of modified sugar.
Although unstable Schiff bases are formed upon reaction of the amino groups with the aldehydes of the aldehydes, glutaraldehyde is capable of modifying the modified sugax with stable crosslinks. At pH 6-8, the pH of typical crosslinking conditions, the cyclic polymers undergo a dehydration to form a,-(3 unsaturated aldehyde polymers. Schiff bases, however, are stable, when conjugated to another double bond. The resonant interaction of both double bonds prevents hydrolysis of the Schiff linkage. Furthermore, amines at high local concentrations can attack the ethylenic double bond to forni a stable Michael addition product.
Aromatic sulfonyl chlorides react with a variety of sites of the modified sugar components, but reaction with the amino groups is the most important, resulting in a stable sulfonamide linkage.
b Sulfliydryl-Reactive Grouts In another preferred embodiment, the sites are sulfhydryl-reactive groups.
Useful, non-limiting examples of sulfliydryl-reactive groups include maleimides, alkyl halides, pyridyl disulfides, and thiophthalimides.
Maleimides react preferentially with the sulfllydryl group of the modified sugar components to form stable thioether bonds. They also react at a much slower rate with primary amino groups and the imidazole groups of histidines. However, at pH 7 the maleimide group can be considered a sulfhydryl-specific group, since at this pH the reaction rate of simple thiols is 1000-fold greater than that of the corresponding amine.
Alkyl halides react with sulfliydryl groups, sulfides, imidazoles, and amino groups.
At neutral to slightly alkaline pH, however, alkyl halides react primarily with sulfliydryl groups to form stable thioether bonds. At higher pH, reaction with amino groups is favored.
Pyridyl disulfides react with free sulfliydryls via disulfide exchange to give mixed disulfides. As a result, pyridyl disulfides are the most specific sulfhydryl-reactive groups.
Thiophthalimides react with free sulfhydryl groups to form disulfides.
c Carboxyl-Reactive Residue In another embodiment, carbodiimides soluble in both water and organic solvent, are used as carboxyl-reactive reagents. These compounds react with free carboxyl groups forming a pseudourea that can then coupled to available amines yielding an amide linkage.
Procedures to modify a carboxyl group with carbodiimide is well know in the art (see, Yamada et al., Biochemistry 20: 4836-4842, 1981).

3 Preferred Nonspecific Sites in Crosslinkina Reagents In addition to the use of site-specific reactive moieties, the present invention contemplates the use of non-specific reactive groups to link the sugar to the modifying group.
Exemplary non-specific cross-linkers include photoactivatable groups, completely inert in the dark, which are converted to reactive species upon absorption of a photon of appropriate energy. In one preferred embodiment, photoactivatable groups are selected from precursors of nitrenes generated upon heating or photolysis of azides.
Electron-deficient nitrenes are extremely reactive and can react with a variety of chemical bonds including N-H, ~-H, C-H, and C=C. Although three types of azides (aryl, alkyl, and acyl derivatives) may be employed, arylazides are presently preferred. The reactivity of arylazides upon photolysis is better with N-H and O-H than C-H bonds. Electron-deficient arylnitrenes rapidly ring-expand to form dehydroazepines, which tend to react with nucleophiles, rather than form C-H
insertion products. The reactivity of arylazides can be increased by the presence of electron-withdrawing substituents such as vitro or hydroxyl groups in the ring. Such substituents push the absorption maximum of arylazides to longer wavelength. Unsubstituted arylazides have an absorption maximum in the range of 260-280 nm, while hydroxy and nitroarylazides absorb significant light beyond 305 nm. Therefore, hydroxy and nitroarylazides are most preferable since they allow to employ less harmful photolysis conditions for the affinity component than unsubstituted arylazides.
In another preferred embodiment, photoactivatable groups are selected from fluorinated arylazides. The photolysis products of fluorinated arylazides are arylnitrenes, all of which iuldergo the characteristic reactions of this group, including C-H
bond insertion, with high efficiency (Keana et al., J. Oy~g. Chem. 55: 3640-3647, 1990).
In another embodiment, photoactivatable groups are selected from benzophenone residues. Benzophenone reagents generally give higher crosslinking yields than arylazide reagents.
In another embodiment, photoactivatable groups are selected from diazo compounds, which form an electron-deficient carbene upon photolysis. These carbenes undergo a variety of reactions including insertion into C-H bonds, addition to double bonds (including aromatic systems), hydrogen attraction and coordination to nucleophilic centers to give carbon ions.

In still another embodiment, photoactivatable groups are selected from diazopyruvates. For example, the p-nitrophenyl ester of p-nitrophenyl diazopyruvate reacts with aliphatic amines to give diazopyruvic acid amides that undergo ultraviolet photolysis to form aldehydes. The photoly~ed dia~opyruvate-modified affinity component will react like formaldehyde or glutaraldehyde forming crosslinks.
4. Homobifunctional Reagents a. Homobifitnctional crosslinkers reactive with primaxy amines Synthesis, properties, and applications of amine-reactive cross-linkers are commercially described in the literature (for reviews of crosslinking procedures and reagents, see above). Many reagents are available (e.g., Pierce Chemical Company, Rockford, Ill.;
Sigma Chemical Company, St. Louis, Mo.; Moleculax Probes, Inc., Eugene, OR.).
Preferred, non-limiting examples of homobifunctional NHS esters include disuccinimidyl glutarate (DSG), disuccinimidyl suberate (DSS), bis(sulfosuccinimidyl) suberate (BS), disuccinimidyl tartarate (DST), disulfosuccinimidyl tartarate (sulfo-DST), bis-2-(succinimidooxycarbonyloxy)ethylsulfone (BSOCOES), bis-2-(sulfosuccinimidooxy-carbonyloxy)ethylsulfone (sulfo-BSOCOES), ethylene glycolbis(succinimidylsuccinate) (EGS), ethylene glycolbis(sulfosuccinimidylsuccinate) (sulfo-EGS), dithiobis(succinimidyl-propionate (DSP), and dithiobis(sulfosuccinimidylpropionate (sulfo-DSP).
Preferred, non-limiting examples of homobifunctional imidoesters include dimethyl malonimidate (DMM), dimethyl succinimidate (DMSC), dimethyl adipimidate (DMA), dimethyl pimelimidate (DMP), dimethyl suberimidate (DMS), dimethyl-3,3'-oxydipropionimidate (DODP), dimethyl-3,3'-(methylenedioxy)dipropionimidate (DMDP), dimethyl-,3'-(dimethylenedioxy)dipropionimidate (DDDP), dimethyl-3,3'-(tetramethylenedioxy)-dipropionimidate (DTDP), and dimethyl-3,3'-dithiobispropionimidate (DTBP).
Preferred, non-limiting examples of homobifunctional isothiocyanates include:
p-phenylenediisothiocyanate (DITC), and 4,4'-diisothiocyano-2,2'-disulfonic acid stilbene (DIDS).
Preferred, non-limiting examples of homobifunctional isocyanates include xylene-diisocyanate, toluene-2,4-diisocyanate, toluene-2-isocyanate-4-isothiocyanate, methoxydiphenylmethane-4,4'-diisocyanate, 2,2'-dicarboxy-4,4'-azophenyldiisocyanate, and hexamethylenediisocyanate.
Preferred, non-limiting examples of homobifunctional arylhalides include 1,5-difluoro-2,4-dinitrobenzene (DFDNB), and 4,4~'-difluoro-3,3'-dinitrophenyl-sulfone.
Preferred, non-limiting examples of hornobifunctional aliphatic aldehyde reagents include glyoxal, malondialdehyde, and glutaraldehyde.
Preferred, non-limiting examples of homobifunctional acylating reagents include nitrophenyl esters of dicarboxylic acids.
Preferred, non-limiting examples of homobifunctional aromatic sulfonyl chlorides include phenol-2,4-disulfonyl chloride, and a-naphthol-2,4-disulfonyl chloride.
Preferred, non-limiting examples of additional amino-reactive homobifunctional reagents include erythritolbiscarbonate which reacts with amines to give biscarbamates.
b Homobifunctional Crosslinkers Reactive with Free Sulfliydryl Grou s Synthesis, properties, and applications of such reagents are described in the literature (for reviews of crosslinking procedures and reagents, see above). Many of the reagents are commercially available (e.g., Pierce Chemical Company, Rockford, Ill.; Sigma Chemical Company, St. Louis, Mo.; Molecular Probes, Inc., Eugene, OR).
Preferred, non-limiting examples of homobifunctional maleimides include bismaleimidohexane (BMH), N,N'-(1,3-phenylene) bismaleimide, N,N'-(1,2-phenylene)bismaleimide, azophenyldimaleimide, and bis(N-maleimidomethyl)ether.
Preferred, non-limiting examples of homobifunctional pyridyl disulfides include 1,4-di-3'-(2'-pyridyldithio)propionamidobutane (DPDPB).
Preferred, non-limiting examples of homobifunctional alkyl halides include 2,2'-dicarboxy-4,4'-diiodoacetamidoazobenzene, a,a'-diiodo-p-xylenesulfonic acid, a, a'-dibromo-p-xylenesulfonic acid, N,N'-bis(b-bromoethyl)benzylamine, N,N'-di(bromoacetyl)phenylthydrazine, and 1,2-di(bromoacetyl)amino-3-phenylpropane.

c. Homobifunctional Photoactivatable Crosslinkers Synthesis, properties, and applications of such reagents are described in the literature (for reviews of crosslinking procedures and reagents, see ab~ve). Some of the reagents are commercially available (~.g., Pierce Chemical Company, Rockford, Ill.; Sigma Chemical Company, St. Louis, Mo.; Molecular Probes, Inc., Eugene, OR).
Preferred, non-limiting examples of homobifunctional photoactivatable crosslinker include bis-~3-(4-azidosalicylamido)ethyldisulfide (BASED), di-N-(2-vitro-4-azidophenyl)-cystamine-S,S-dioxide (DNCO), and 4,4'-dithiobisphenylazide.
5. HeteroBifunctional Reagents a Amino-Reactive HeteroBifunctional Reagents with a Pyridyl Disulfide Moiety Synthesis, properties, and applications of such reagents are described in the literature (for reviews of crosslinking procedures and reagents, see above). Many of the reagents are conunercially available (e.g., Pierce Chemical Company, Rockford, Ill.; Sigma Chemical Company, St. Louis, Mo.; Molecular Probes, Inc., Eugene, OR).
Preferred, non-limiting examples of hetero-bifunctional reagents with a pyridyl disulfide moiety and an amino-reactive NHS ester include N-succinimidyl-3-(2-pyridyldithio)propionate (SPDP), succinimidyl 6-3-(2-pyridyldithio)propionamidohexanoate (LC-SPDP), sulfosuccinimidyl 6-3-(2-pyridyldithio)propionamidohexanoate (sulfo-LCSPDP), 4-succinimidyloxycarbonyl-a-methyl-a-(2-pyridyldithio)toluene (SMPT), and sulfosuccinimidyl 6-a-methyl-a-(2-pyridyldithio)toluamidohexanoate (sulfo-LC-SMPT).
b Amino-Reactive HeteroBifunctional Reagents with a Maleimide Moiet Synthesis, properties, and applications of such reagents are described in the literature.
Preferred, non-limiting examples of hetero-bifunctional reagents with a maleimide moiety and an amino-reactive NHS ester include succinimidyl maleimidylacetate CAMAS), succinimidyl 3-maleimidylpropionate (BMPS), N- ~y-maleimidobutyryloxysuccinimide ester (GMBS)N-y-maleimidobutyryloxysulfo succinimide ester (sulfo-GMBS) succinimidyl maleimidylhexanoate (FMCS), succinimidyl 3-maleimidylbenzoate (SMB), m-maleimidobenzoyl-N-hydroxysuccinimide ester (MBS), m-maleimidobenzoyl-N-hydroxysulfosuccinimide ester (sulfo-MBS), succinimidyl 4-(N-maleimidomethyl)-cyclohexane-1-carboxylate (SMCC), sulfosuccinimidyl 4-(IV-maleimidomethyl)cyclohexane-1-carboxylate (sulfo-SMCC), succinimidyl 4-(p-maleimidophenyl)butyrate (SMPB), and sulfosuccinimidyl 4-(p-maleimidophenyl)butyrate (sulfo-SMPB).
c Amino-Reactive HeteroBifunctional I~ea~ents with an Alkyl Halide Moiet Synthesis, properties, and applications of such reagents are described in the literature.
Preferred, non-limiting examples of hetero-liifunctional reagents with an alkyl halide moiety and an amino-reactive NHS ester include N-succinimidyl-(4-iodoacetyl)aminobenzoate (SIAB), sulfosuccinimidyl-(4-iodoacetyl)aminobenzoate (sulfo-SIAB), succinimidyl-6-(iodoacetyl)aminohexanoate (SIAX), succinimidyl-6-(6-((iodoacetyl)-amino)hexanoylamino)hexanoate (SIAXX), succinimidyl-6-(((4-(iodoacetyl)-amino)-methyl)-cyclohexane-1-carbonyl)aminohexanoate (SIACX), and succinimidyl-4((iodoacetyl)-amino)methylcyclohexane-1-carboxylate (SIAC).
A preferred example of a hetero-bifunctional reagent with an amino-reactive NHS
ester and an alkyl dihalide moiety is N-hydroxysuccinimidyl 2,3-dibromopropionate (SDBP).
SDBP introduces intramolecular crosslinks to the affinity component by conjugating its amino groups. The reactivity of the dibromopropionyl moiety towards primary amine groups is controlled by the reaction temperature (McKenzie et al., Protein Chem. 7:

(1988)).
Preferred, non-limiting examples of hetero-bifunctional reagents with an alkyl halide moiety and an amino-reactive p-nitrophenyl ester moiety include p-nitrophenyl iodoacetate (NPIA).
Other cross-linking agents are known to those of skill in the art. See, for example, Pomato et al., LT.S. Patent No. 5,965,106. It is within the abilities of one of skill in the art to choose an appropriate cross-linking agent for a particular application.

d. Cleavable Linker Crroups In yet a further embodiment, the linker group is provided with a group that can be cleaved to release the modifying group from the sugar residue. Many cleavable groups are known in the art. fee, for example, Jung et al., Bi~chern. Bi~phys. Acta 761:
152-162 (1983);
Joshi et al., .7. Bi~l. ChenZ. 265: 14518-14525 (1990); Zarling et al., .l.
Immun~l. 124: 913-920 (1980); Bouizar et al., Eur. ,I. BioclZem. 155: 141-147 (1986); Park et al., .I. Biol. Chem. 261:
205-210 (1986); Brov~nning et al., .I. Inamun~l. 143: 1859-1867 (1989).
Moreover a broad range of cleavable, bifunctional (both homo- and hetero-bifunctional) linker groups is commercially available from suppliers such as Pierce.
Exemplary cleavable moieties can be cleaved using light, heat or reagents such as thiols, hydroxylamine, bases, periodate and the like. Moreover, certain preferred groups are cleaved in vivo in response to being endocytosed (e.g., cis-aconityl; see, Shen et al., Biochem.
Biophys. Res. Conzmun. 102: 1048 (1991)). Preferred cleavable groups comprise a cleavable moiety which is a member selected from the group consisting of disulfide, ester, imide, carbonate, nitrobenzyl, phenacyl and benzoin groups.
e. Coniu~ation of Modified Sugars to Peptides The modified sugars are conjugated to a glycosylated or non-glycosylated peptide using an appropriate enzyme to mediate the conjugation. Preferably, the concentrations of the modified donor sugar(s), enzymes) and acceptor peptides) are selected such that glycosylation proceeds until the acceptor is consumed. The considerations discussed below, while set forth in the context of a sialyltransferase, are generally applicable to other glycosyltransferase reactions.
A number of methods of using glycosyltransferases to synthesize desired oligosaccharide structures are known and are generally applicable to the instant invention.
Exemplary methods are described, for instance, WO 96/32491, Ito et al., Pure Appl. Chem.
65: 753 (1993), and U.S. Pat. Nos. 5,352,670, 5,374,541, and 5,545,553.
The present invention is practiced using a single glycosyltransferase or a combination of glycosyltransferases. for example, one can use a combination of a sialyltransferase and a galactosyltransferase. In those embodiments using more than one enzyme, the enzymes and substrates axe preferably combined in an initial reaction mixture, or the enzymes and reagents for a second enzymatic reaction are added to the reaction medium once the first enzymatic reaction is complete or nearly complete. By conducting two enzymatic reactions in sequence in a single vessel, overall yields are improved over procedures in which an intermediate species is isolated. Moreover, cleanup and disposal of extra solvents and by-products is reduced.
In a preferred embodiment, each of the first and second enzyme is a glycosyltransferase. In another preferred embodiment, one enzyme is an endoglycosidase. In another preferred embodiment, one enzyme is an exoglycosidase. In an additional preferred embodiment, more than two enzymes are used to assemble the modified glycoprotein of the invention. The enzymes are used to alter a saccharide structure on the peptide at any point either before or after the addition of the modified sugar to the peptide.
In another embodiment, at least two of the enzymes are glycosyltransferases and the last sugar added to the saccharide structure of the peptide is a non-modified sugar. Instead, the modified sugar is internal to the glycan structure and therefore need not be the ultimate sugar on the glycan. In an exemplary embodiment, galactosyltransferase may catalyze the transfer of Gal-PEG from UDP-Gal-PEG onto the glycan, followed by incubation in the presence of ST3Gal3 and CMP-SA, which serves to add a "capping" unmodified sialic acid onto the glycan (Figure 23A).
In another embodiment, at least two of the enzymes used are glycosyltransferases, and at least two modified sugars are added to the glycan structures on the peptide. In this manner, two or more different glycoconjugates may be added to one or more glycans on a peptide.
This process generates glycan structures having two or more functionally different modified sugars. In an exemplary embodiment, incubation of the peptide with GnT-I, II
and UDP-GIcNAc-PEG serves to add a GIcNAc-PEG molecule to the glycan; incubation with galactosyltransferase and UDP-Gal then serves to add a Gal residue thereto;
and, incubation with ST3Ga13 and CMP-SA-Man-6-Phosphate serves to add a SA-mannose-6-phosphate molecule to the glycan. This series of reactions results in a glycan chain having the functional characteristics of a PEGylated glycan as well as mannose-6-phosphate targeting activity (Figure 23B). ' In another embodiment, at least two of the enzymes used in the reaction are glycosyltransferases, and again, different modified sugars are added to N-linked and ~-linked glycans on the peptide. This embodiment is useful when two different modified sugars are to be added to the glycans of a peptide, but when it is important to spatially separate the modified sugars on the peptide from each other. For example, if the modified sugars comprise bulky molecules, including but not limited to, PEG and other molecules such as a linker molecule, this method may be preferable. The modified sugars may be added simultaneously to the glycan structures on a peptide, or they may be added sequentially. In an exemplary embodiment, incubation with ST3Ga13 and CMP-SA-PEG serves to add sialic acid-PEG to the N-linked glycans, while incubation with ST3Ga11 and CMP-SA-bisPhosphonate serves to add sialic acid-BisPhosphonate to the ~-linked glycans (Figure 23C).
In another embodiment, the method makes use of one or more exo- or endoglycosidase. The glycosidase is typically a mutant, which is engineered to form glycosyl bonds rather than rupture them. The mutant glycanase, sometimes called a glycosynthase, typically includes a substitution of an amino acid residue for an active site acidic amino acid residue. For example, when the endoglycanase is endo-H, the substituted active site residues will typically be Asp at position 130, Glu at position 132 or a combination thereof. The amino acids are generally replaced with serine, alanine, asparagine, or glutamine.
Exoglycosidases such as transialylidase are also useful.
The mutant enzyme catalyzes the reaction, usually by a synthesis step that is analogous to the reverse reaction of the endoglycanase hydrolysis step. In these embodiments, the glycosyl donor molecule (e.g., a desired oligo- or mono-saccharide structure) contains a leaving group and the reaction proceeds with the addition of the donor molecule to a GIcNAc residue on the protein. For example, the leaving group can be a halogen, such as fluoride. In other embodiments, the leaving group is a Asn, or a Asn-peptide moiety. In yet further embodiments, the GIcNAc residue on the glycosyl donor molecule is modified. For example, the GIcNAc residue may comprise a 1,2 oxazoline moiety.
In a preferred embodiment, each of the enzymes utilized to produce a conjugate of the invention axe present in a catalytic amount. The catalytic amount of a particular enzyme vaxies according to the concentration of that enzyme's substrate as well as to reaction conditions such as temperature, time and pH value. Means for determining the catalytic amount for a given enzyme under preselected substrate concentrations and reaction conditions are well known to those of skill in the art.
The temperature at which an above-described process is carried out can range from just above freezing to the temperature at which the most sensitive enzyme denatures.
Preferred temperature ranges are about 0 °C to about 55 °C, and more preferably about 20 ° C
to about 37 °C. In another exemplary embodiment, one or more components of the present method are conducted at an elevated temperature using a thermophilic enzyme.
The reaction mixture is maintained for a period of time sufficient for the acceptor to be glycosylated, thereby forming the desired conjugate. Some of the conjugate can often be detected after a few hours, with recoverable amounts usually being obtained within 24 hours or less. Those of skill in the art understand that'the rate of reaction is dependent on a number of variable factors (e.g, enzyme concentration, donor concentration, acceptor concentration, temperature, solvent volume), which are optimized for a selected system.
The present invention also provides for the industrial-scale production of modified peptides. As used herein, an industrial scale generally produces at least one gram of finished, purified conjugate.
In the discussion that follows, the invention is exemplified by the conjugation of modified sialic acid moieties to a glycosylated peptide. The exemplary modified sialic acid is labeled with PEG. The focus of the following discussion on the use of PEG-modified sialic acid and glycosylated peptides is for clarity of illustration and is not intended to imply that the invention is limited to the conjugation of these two partners. One of skill understands that the discussion is generally applicable to the additions of modified glycosyl moieties other than sialic acid. Moreover, the discussion is equally applicable to the modification of a glycosyl unit with agents other than PEG including other water-soluble polymers, therapeutic moieties, and biomolecules.
An enzymatic approach can be used for the selective introduction of PEGylated or PPGylated carbohydrates onto a peptide or glycopeptide. The method utilizes modified sugars containing PEG, PPG, or a masked reactive functional group, and is combined with the appropriate glycosyltransferase or glycosynthase. By selecting the glycosyltransferase that will make the desired carbohydrate linkage and utilizing the modified sugar as the donor substrate, the PEG or PPG can be introduced directly onto the peptide backbone, onto existing sugar residues of a glycopeptide or onto sugar residues that have been added to a peptide.
An acceptor for the sialyltransferase is present on the peptide to be modified by the methods of the present invention either as a naturally occurring structure or one placed there recombinantly, enzymatically or chemically. Suitable acceptors, include, for example, galactosyl acceptors such as Gal(31,4G1cNAc, Gal(31,4GalNAc, Gal(31,3GalNAc, facto-N-tetraose, Gal(31,3G1cNAc, Gal(31,3Ara, Gal(31,6G1cNAc, Gal(31,4G1c (lactose), and other acceptors known to those of skill in the art (see, e.g., Paulson et al., .I::
~i~l. C'dzern. 2~~: 5617-5624 (1978)).
In one embodiment, an acceptor for the sialyltransferase is present on the peptide to be modified upon in vivo synthesis of the peptide. Such peptides can be sialylated using the claimed methods without prior modification of the glycosylation pattern of the peptide.
Alternatively, the methods of the invention can be used to sialylate a peptide that does not include a suitable acceptor; one first modifies the peptide to include an acceptor by methods known to those of skill in the art. In an exemplary embodiment, a GaINAc residue is added by the action of a GaINAc transferase.
In an exemplary embodiment, the galactosyl acceptor is assembled by attaching a galactose residue to an appropriate acceptor linked to the peptide, e.g., a GIcNAc. The method includes incubating the peptide to be modified with a reaction mixture that contains a suitable amount of a galactosyltransferase (e.g., gal(31,3 or gal/31,4), and a suitable galactosyl donor (e.g., LTDP-galactose). The reaction is allowed to proceed substantially to completion or, alternatively, the reaction is terminated when a preselected amount of the galactose residue is added. Other methods of assembling a selected saccharide acceptor will be apparent to those of skill in the art.
In yet another embodiment, peptide-linked oligosacchaxides are first "trimmed,"
either in whole or in part, to expose either an acceptor for the sialyltransferase or a moiety to which one or more appropriate residues can be added to obtain a suitable acceptor. Enzymes such as glycosyltransferases and endoglycosidases (see, for example U.S.
Patent No.
5,716,812) are useful for the attaching and trimming reactions. A detailed discussion of "trimming" and remodeling N-linked and O-linked glycans is provided elsewhere herein.

In the discussion that follows, the method of the invention is exemplified by the use of modified sugars having a water-soluble polymer attached thereto. The focus of the discussion is for clarity of illustration. Those of skill will appreciate that the discussion is equally relevant to those embodiments in which the modified sugar bears a therapeutic moiety, biomolecule or the like.
An exemplary embodiment of the invention in which a carbohydrate residue is "trimmed" prior to the addition of the modified sugar is set forth in Figure 14~, which sets forth a scheme in which high mannose is trimmed back to the first generation biantennary structure. A modified sugar bearing a water-soluble polymer is conjugated to one or more of the sugar residues exposed by the "trimming back." In one example, a water-soluble polymer is added via a GIcNAc moiety conjugated to the water-soluble polymer. The modified GIcNAc is attached to one or both of the terminal mannose residues of the biantennaxy structure. Alternatively, an unmodified GIcNAc can be added to one or both of the termini of the branched species.
In another exemplary embodiment, a water-soluble polymer is added to one or both of the terminal mannose residues of the biantennary structure via a modified sugar having a galactose residue, which is conjugated to a GIcNAc residue added onto the terminal mannose residues. Alternatively, an unmodified Gal can be added to one or both terminal GIcNAc residues.
In yet a further example, a water-soluble polymer is added onto a Gal residue using a modified sialic acid.
Another exemplary embodiment is set forth in Figure 15, which displays a scheme similar to that shown in Figure 14, in which the high mannose structure is "trimmed back" to the mannose from which the biantennary structure branches. In one example, a water-soluble polymer is added via a GIcNAc modified with the polymer. Alternatively, an unmodified GIcNAc is added to the mannose, followed by a Gal with an attached water-soluble polymer.
In yet another embodiment, unmodified GIcNAc and Gal residues are sequentially added to the mannose, followed by a sialic acid moiety modified with a water-soluble polymer.
Figure 16 sets forth a further exemplary embodiment using a scheme similar to that shown in Figure 14, in which high mannose is "trimmed back" to the GIcNAc to which the first mannose is attached. The GIcNAc is conjugated to a Gal residue bearing a water-soluble polymer. Alternatively, an unmodified Gal is added to the GIcNAc, followed by the addition of a sialic acid modified with a water-soluble sugar. In yet a further example, the terminal GIcNAc is conjugated with Gal and the GIcNAc is subsequently fucosylated with a modified fucose bearing a water-soluble polymer.
Figure 17 is a scheme similar to that shown in Figure 14, in which high mannose is trimmed back to the first GlcNAc attached to the Asn of the peptide. In one example, the GIcNAc of the GIcNAc-(Fuc)a residue is conjugated with a GIcNAc bearing a water soluble polymer. In another example, the GIcNAc of the GIcNAc-(Fuc)~ residue is modified with Gal, which bears a water soluble polymer. In a still further embodiment, the GIcNAc is modified with Gal, followed by conjugation to the Gal of a sialic acid modified with a water-soluble polymer.
Other exemplary embodiments are set forth in Figures 18-22. An illustration of the array of reaction types with which the present invention may be practiced is provided in each of the aforementioned figures.
The Examples set forth above provide an,illustration of the power of the methods set forth herein. Using the methods of the invention, it is possible to "trim back" and build up a carbohydrate residue of substantially any desired structure. The modified sugar can be added to the termini of the carbohydrate moiety as set forth above, or it can be intermediate between the peptide core and the terminus of the carbohydrate.
In an exemplary embodiment, an existing sialic acid is removed from a glycopeptide using a sialidase, thereby unmasking all or most of the underlying galactosyl residues.
Alternatively, a peptide or glycopeptide is labeled with galactose residues, or an oligosaccharide residue that terminates in a galactose unit. Following the exposure of or addition of the galactose residues, an appropriate sialyltransferase is used to add a modified sialic acid. The approach is summarized in Scheme 12.

Scheme 12 NHz ~ I \ N G81 Glycoprotein n Nko Gal O-p_O O
HO off ~ O N ~al HO~J~:~~-~ Na HO OH
PEG or PPG~N.~NH OH
H ~ Sialyltransferase CMP-SA-5-NHC~CH2NH-PEG(PPG) .

Glycoprotein Gal Gal-SA-5-NHCOCH2NH-PEG
Gal In yet a further approach, summarized in Scheme 13, a masked reactive functionality is present on the sialic acid. The masked reactive group is preferably unaffected by the conditions used to attach the modified sialic acid to the peptide. After the covalent attachment of the modified sialic acid to the peptide, the .mask is removed and the peptide is conjugated with an agent such as PEG, PPG, a therapeutic moiety, biomolecule or other agent. The agent is conjugated to the peptide in a specific manner by its reaction with the unmasked reactive group on the modified sugar residue.

Scheme 13 G21 Glycoprotein NH2 Gad ;.
SA-5-NHCOCH~S-SEt -~~ Gal Gal o-P-o o N
H~ H~ ~~H~,, ~~ ~~_+N~ Ho off Sialyltransferase Gal-SA-5-NHCOCH2S-SEt EtS~S~N\/~~ ° Gal SA-5-NHCOCH2S-SEt Glycoprotein Gad 1. dithiothreitol Gal-SA-5-NHCOCH2S-PEG 2. PEG-halide or PPG halide Gal Any modified sugar can be used with its appropriate glycosyltransferase, depending on the terminal sugars of the oligosaccharide side chains of the glycopeptide (Table 4). As discussed above, the terminal sugar of the glycopeptide required for introduction of the PEGylated or PPGylated structure can be introduced naturally during expression or it can be produced post expression using the appropriate glycosidase(s), glycosyltransferase(s) or mix of glycosidase(s) and glycosyltransferase(s).

Table 4. Modified sugars.
Q
R~-~f ~ X_R1 R3-Y I X-R~

o ~ p~ _Z :.
R~_~ .R =A ~ ~ ~ ~ R4 A ~ ~ °~o 4 ~-P~ -p ON O ~ P O P~
o_,~N~ o_ N~ ~ Na ~-+Na HO OH
HO OH Upp_g~actosamine-derivatives UDP-galactose-derivatives (when A = NH, R4 may be acetyl) ~\ X-R1 C~\ ~~-!~~
R3_Y ~ o R3_lf o _ R2'~ NH
O
R2 Z R4-A II o ~O R4 A O-p~0-P_O O N~O
O 0 +Na p- N~ O +Na O-+Na ° HO~-O( H
UDP-Glucose-derivatives UDP-Glucosamine-derivatives (when A = NH, Rq may be acetyl) Q X_R~ o N NH

R -Y O II II N N NHS
~N I NH O-P~O-P-O O
O O \ ~~ O-+Na p- Na II II N N NHS
0 0 +NO o Na o R~ X o A-R4 Ho OH

HO OH RZ-Y GDP-fucose-derivatives GDP-Mannose-derivatives X = O, NH, S, CH2, N-(R~-5)2~
Y = X; Z = X; A = X; B = X. Ligand of interest = acyl-PEG, acyl-PPG, alkyl-PEG, acyl-alkyl-PEG, acyl-alkyl-PEG, carbamoyl-PEG, carbamoyl-PPG, PEG, PPG, Q = H2, O, S, NH, N-R. acyl-aryl-PEG, acyl-aryl-PPG, aryl-PEG, aryl-PPG, Mannose-6-phosphate, heparin, heparan, SLex, Mannose, FGF, VFGF, R, Rl-4 = H, Linker-M, M. protein, chondroitin, keratan, dermatan, albumin, integrins, peptides, etc.
M = Ligand of interest In a further exemplary embodiment, IJDP-galactose-PEG is reacted with bovine milk [31,4-galactosyltransferase, thereby transferring the modified galactose to the appropriate terminal N-acetylglucosamine structure. The terminal GIcNAc residues on the glycopeptide may be produced during expression, as may occur in such expression systems as mammalian, insect, plant or fungus, but also can be produced by treating the glycopeptide with a sialidase and/or glycosidase and/or glycosyltransferase, as required.
-1~7-In another exemplary embodiment, a GIcNAc transferase, such as GnT-I-IV, is utilized to transfer PEGylated-GlcNc to a mannose residue on a glycopeptide.
In a still further exemplary embodiment, the N- and/or ~-linked glycan structures are enzymatically removed from a glycopeptide to expose an amino acid or a terminal glycosyl residue that is subsequently conjugated with the modified sugar. for example, an endoglycanase is used to remove the N-linked structures of a glycopeptide to expose a terminal GIcNAc as a GIcNAc-linked-Asn on the glycopeptide. UDP-Gal-PEG and the appropriate galactosyltransferase is used to introduce the PEG- or PPG-galactose functionality onto the exposed GIcNAc.
In an alternative embodiment, the modified sugar is added directly to the peptide backbone using a glycosyltransferase known to transfer sugar residues to the peptide backbone. This exemplary embodiment is set forth in Scheme 14. Exemplary glycosyltransferases useful in practicing the present invention include, but are not limited to, GaINAc transferases (GaINAc Tl-14), GIcNAc transferases, fucosyltransferases, glucosyltransferases, xylosyltransferases, mannosyltransferases and the like.
Use of this approach allows the direct addition of modified sugars onto peptides that lack any carbohydrates or, alternatively, onto existing glycopeptides. In both cases, the addition of the modified sugar occurs at specific positions on the peptide backbone as defined by the substrate specificity of the glycosyltransferase and not in a random manner as occurs during modification of a protein's peptide backbone using chemical methods. An array of agents can be introduced into proteins or glycopeptides that lack the glycosyltransferase substrate peptide sequence by engineering the appropriate amino acid sequence into the peptide chain.
Scheme 14 HO OH
O O Protein or Glycoprotein Ho ~ ~NH ~ GaiNH-CO(CH2)4NH-PEG
O-P-_O-P_O O
o N H o °u I N'~o Q-+Na ~_+N 1'-I
HO off GalNAc Transferase (GaINAc T3) GaINH-CO(CH2)4NH-PEG
NH
s PEG
In each of the exemplary embodiments set forth above, one or more additional chemical or enzymatic modification steps can be utilized following the conjugation of the modified sugar to the peptide. In an exemplary embodiment, an enzyme (e.g., fucosyltransferase) is used to append a glycosyl unit (e.g., fucose) onto the terminal modified sugar attached to the peptide. In another example, an enzymatic reaction is utilized to "cap"
sites to which the modified sugar failed to conjugate. Alternatively, a chemical reaction is utilized to alter the structure of the conjugated modified sugar. For example, the conjugated modified sugar is reacted with agents that stabilize or destabilize its linkage with the peptide component to which the modified sugar is attached. In another example, a component of the modified sugar is deprotected following its conjugation to the peptide. ~ne of skill will appreciate that there is an array of enzymatic and chemical procedures that are useful in the methods of the invention at a stage after the modified sugar is conjugated to the peptide.
Further elaboration of the modified sugar-peptide conjugate is within the scope of the invention.
Peptide Targeting With Mannose-6-Phosphate In an exemplary embodiment the peptide is derivatized with at least one mannose-6-phosphate moiety. The mannose-6-phosphate moiety targets the peptide to a lysosome of a cell, and is useful, for example, to taxget therapeutic proteins to lysosomes for therapy of lysosomal storage diseases.
Lysosomal storage diseases are a group of over 40 disorders which are the result of defects in genes encoding enzymes that break down glycolipid or polysaccharide waste products within the lysosomes of cells. The enzymatic products, e.g., sugars and lipids, are then recycled into new products. Each of these disorders results from an inherited autosomal or X-linked recessive trait which affects the levels of enzymes in the lysosome. Generally, there is no biological or functional activity of the affected enzymes in the cells and tissues of affected individuals. Table 5 provides a list of representative storage diseases and the enzymatic defect associated with the diseases. In such diseases the deficiency in enzyme function creates a progressive systemic deposition of lipid or carbohydrate substrate in lysosomes in cells in the body, eventually causing loss of organ function and death. The genetic etiology, clinical manifestations, molecular biology and possibility of the lysosomal storage diseases axe detailed in Scriver et al., eds., THE METABOLIC AND
MOLECULAR BASIS
of INHERITED DISEASE, 7^th Ed., Vol. II, McGraw Hill, (1995).
Table 5. Lysosomal storage diseases and associated enzymatic defects -1~9-Disease Enzymatic Defect Pompe disease acid a-glucosidase (acid maltase) MPSI* (Hurler disease)a-L-iduronidase MPSII (Hunter disease)iduronate sulfatase MPSIII (Sanfilippo) heparan N-sulfatase MPS IV (Morquio A) galactose-6-sulfatase MPS IV (Morquio B) acid (3-galactosidase MPS VII (Sly disease)(3-glucoronidase I-cell disease N-acetylglucosamine-1-phosphotransferase Schindler disease a-N-acetylgalactosaminidase (a-galactosidase B) Wolinan disease acid lipase Cholesterol ester acid lipase storage disease Farber disease lysosomal acid ceramidase Niemann-Pick diseaseacid sphingomyelinase Gaucher disease glucocerebrosidase Krabbe disease galactosylceramidase Fabry disease a-galactosidase A

GMl gangliosidosis acid ~3-galactosidase Galactosialidosis (3-galactosidase and neuraminidase Tay-Sach's disease hexosaminidase A

Magakaryotic leukodystrophyarylsulphatase a Sandhoff disease hexosaminidase A
and B

*MPS = mucopolysaccaridosis De Duve first suggested that replacement of the missing lysosomal enzyme with exogenous biologically active enzyme might be a viable approach to treatment of lysosomal storage diseases (De Duve, Fed. Proc. 23: 1045 (1964). Since that time, various studies have suggested that enzyme replacement therapy may be beneficial for treating various lysosomal storage diseases. The best success has been shown with individuals with type I
Craucher disease, who have been treated with exogenous enzyme ([3-glucocerebrosidase), prepared from placenta (CeredaseTM) or, more recently, recombinantly (CerezymeTM). It has been suggested that enzyme replacement may also be beneficial for treating Fabry's disease, as well as other lysosomal storage diseases. See, for example, Dawson et al., Ped. Res. 7(8):
684-690 (1973) (i~ vitro) and Mapes et al., Science 169: 987 (1970) (in vivo).
Clinical trials of enzyme replacement therapy have been reported for Fabry patients using infusions of normal plasma (Mapes et al., Science 169: 987-989 (1970)), ~,-galactosidase A
purified from placenta (Brady et al., N. End. ,I. pled. 279: 1163 (1973)); or cc-galactosidase A purified from spleen or plasma (Desnick et al., Proc. Natl. Acad. Sci., ZISA 76: 5326-5330 (1979)) and have demonstrated the biochemical effectiveness of direct enzyme replacement for Fabry disease.

These studies indicate the potential for eliminating, or significantly reducing, the pathological glycolipid storage by repeated enzyme replacement. For example, in one study (Desnick et al., supra), intravenous injection of purified enzyme resulted in a transient reduction in the plasma levels of the stored lipid substrate, globotriasylceramide.
Accordingly, there exists a need in the art for methods for providing sufficient quantities of biologically active lysosomal enzymes, such as human cc-galactosidase A, to deficient cells. Recently, recombinant approaches have attempted to address these needs, see, e.g., U.S. Fat. No. 5,658,567; 5,580,757; Bishop et al., Pt°~c. Natl.
Acad. Sci., LISA. ~3: 4859-4863 (1986); Medin et al., Proc. Natl. Acad. Sci., USA. 93: 7917-7922 (1996);
Novo, F. J., Gene Therapy. 4: 488-492 (1997); Ohshima et al., Proc. Natl. Acad. Sci., USA.
94: 2540-2544 (1997); and Sugimoto et al., Human Gehe Therapy 6: 905-915, (1995).
Through the mannose-6-phosphate mediated targeting of therapeutic peptides to lysosomes, the present invention provides compositions and methods for delivering sufficient quantities of biologically active lysosomal peptides to deficient cells.
Thus, in an exemplary embodiment, the present invention provides a peptide according to Table 7 that is derivatized with mannose-6-phosphate (Figure 24 and Figure 25).
The peptide may be recombinantly or chemically prepared. Moreover, the peptide can be the full, natural sequence, or it may be modified by, for example, truncation, extension, or it may include substitutions or deletions. Exemplary proteins that are remodeled using a method of the present invention include glucocerebrosidase, (3-glucosidase, a-galactosidase A, acid-a-glucosidase (acid maltase). Representative modified peptides that are in clinical use include, but are not limited to, CeredaseTM, CerezymeTM, and FabryzymeTM. A glycosyl group on modified and clinically relevant peptides may also be altered utilizing a method of the invention. The mannose-6-phosphate is attached to the peptide via a glycosyl linking group.
In an exemplary embodiment, the glycosyl linking group is derived from sialic acid.
Exemplary sialic acid-derived glycosyl linking groups are set forth in Table 3, in which one or more of the "R" moieties is mannose-6-phosphate or a spacer group having one or more mannose-6-phosphate moieties attached thereto. The modified sialic acid moiety is preferably the terminal residue of an oligosacchaxide linked to the surface of the peptide (Figure 26) In addition to the mannose-6-phosphate, the peptides of the invention may be further derivatized with a moiety such as a water-soluble polymer, a therapeutic moiety, or an additional targeting moiety. Methods for attaching these and other groups are set forth herein. In an exemplary embodiment, the group other than maneose-6-phosphate is attached to the peptide via a derivatized sialic acid derivative according to Table 3, in which one or more of the "1~" moieties is a group other than mannose-6-phosphate.
In an exemplary embodiment, a sialic acid moiety modified with a Cbz-protected glycine-based linker arm is prepared. The corresponding nucleotide sugar is prepared and the Cbz group is removed by catalytic hydrogenation. The resulting nucleotide sugar has an available, reactive amine that is contacted with an activated mannose-6-phosphate derivative, providing a mannose-6-phosphate derivatized nucleotide sugar that is useful in practicing the methods of the invention.
As shown in the scheme below (scheme 15), an exemplary activated mannose-6-phosphate derivative is formed by converting a 2-bromo-benzyl-protected phosphotriester into the corresponding triflate, in situ, and reacting the triflate with a linker having a reactive oxygen-containing moiety, forming an ether linkage between the sugar and the linker. The benzyl protecting groups are removed by catalytic hydrogenation, and the methyl ester of the linker is hydrolyzed, providing the corresponding carboxylic acid. The carboxylic acid is activated by any method known in the art. An exemplary activation procedure relies upon the conversion of the carboxylic acid to the N-hydroxysuccinimide ester.

Scheme 15 off 1. CMP-SA synthetase, CTP
HO NH 1. Z-Glycine-NHS Ho off 2. HZ/Pd/C
H~ z 2. NeuAc Aldolase, pyruvate Ho'~~-o-+Na Ho ~ ~ ~ yN~,NH OH
OH H o NHz 03PO off ~ I 'N
HO -o o-P-O o N
O
HOII HO o-+N~
OH _ o-Linker -C-activated HO O ~ +Na Ho OH
m NH~/I O
HaN~( OH
O
CMP-SA-5-NHCOCHzNHz N H~
~'N
O ~~O
03P0 OH O~P-O O
HO -O HO OH O_+N I-1 HO O HO O' o O-+Na HO OH
o-Linker IC~N~.NH off H O
m o ~ / o II II
O pP\o OAc ~ o 0P\0 OAc Ac0 -o HO-Linker-C-OMe -o / I Ac0 , Ac0 Ac0 O
\ gr AgOTf, sieves, CHZC12 ~ ~ II
\ O-Linker-C-OMe 1. HZ/Pd/C
2. NaOMe, MeOH, H20 II II
O_pP~o off Activating agent -~ i~0 OH
HO -O ~-_ o Ho -O
HO O HO O
II II
O-Linker-C-activated O-Linker-C-OH
In another exemplary embodiment, as shown in the scheme below (scheme 16), a N-acetylated sialic acid is converted to an amine by manipulation of the pyruvyl moiety. Thus, the primary hydroxyl is converted to a sulfonate ester and reacted with sodium azide. The azide is catalytically reduced to the corresponding amine. The sugar is subsequently converted to its nucleotide analogue and coupled, through the amine group, to the linker arm-derivatized mannose-6-phosphate prepared as discussed above.
~cherr~e 1~
1. MeOH, Dowex (H+) OH 2. Ts-CI, pyridine OH
HO ~H 3. NaN3, DMF HO ~H +
HO O O-+Na HzN O O- Na AcNH O
AcNH OH ~ 4. HZ/Pd/C OH
5. NaOMe, MeOH, H20 CMP-SA synthetase, CTP
NHz 03P0 OH ~' N
~'~O
HO -O O P-O
HO ~~ HO O-+Na O-Linker -C-activated HzN ~H O O'+Na HO OH
AcHN O
OH

NHz 03P0 OH O I ' N
HO 'O II O N~O
HO O O~P-O
O_+Na O-Linker ~~NH HO off O O- Na + HO OH
AcNH OH O
Peptides useful to treat lysosomal storage disease can be derivatized with other targeting moieties including, but not limited to, transferrin (to deliver the peptide across the blood-brain barrier, and to endosomes), carnitine (to deliver the peptide to muscle cells), and phosphonates, e.g, bisphosphonate (to target the peptide to bone and other calciferous tissues). The targeting moiety and therapeutic peptide are conjugated by any method discussed herein or otherwise known in the art.
In an exemplary embodiment, the targeting agent and the therapeutic peptide are coupled via a linker moiety. In this embodiment, at least one of the therapeutic peptide or the targeting agent is coupled to the linker moiety via an intact glycosyl linking group according to a method of the invention. In an exemplary embodiment, the linker moiety includes a poly(ether) such as polyethylene glycol). In another exemplary embodiment, the linker moiety includes at least one bond that is degraded in vivo, releasing the therapeutic peptide from the targeting agent, following delivery of the conjugate to the targeted tissue or region of the body.
In yet another exemplary embodiment, the i~a vivo distribution of the therapeutic moiety is altered via altering a glycoform on the therapeutic moiety without conjugating the therapeutic peptide to a targeting moiety. For example, the therapeutic peptide can be shunted away from uptake by the reticuloendothelial system by capping a terminal galactose moiety of a glycosyl group with sialic acid (or a derivative thereof) (Figures 24 and 27).
Sialylation to cover terminal Gal avoids uptake of the peptide by hepatic asialoglycoprotein (ASGP) receptors, and may extend the half life of the peptide as compared with peptides having only complex glycan chains, in the absence of sialylation.
II. Peptide/Glycopeptides of the Invention In one embodiment, the present invention provides a composition comprising multiple copies of a single peptide having an elemental trimannosyl core as the primary glycan structure attached thereto. In preferred embodiments, the peptide may be a therapeutic molecule. The natural form of the peptide may comprise complex N-linked glycans or may be a high mannose glycan. The peptide may be a mammalian peptide, and is preferably a human peptide. In some embodiments the peptide is selected from the group consisting of an immunoglobulin, erythropoietin, tissue-type activator peptide, and others (See Figure 28).
Exemplary peptides whose glycans can be remodeled using the methods of the invention are set forth in Figure 28.
Table 6. Preferred peptides for glycan remodeling Hormones and Growth FactorsReceptors and Chimeric Receptors GM-CSF Tumor Necrosis Factor receptor (TNF-R) TPO TNF-R:IgG Fc fusion EPO Alpha-CD~O

EPO variants PSGL-1 F SH Complement HGH GIyCAM or its chimera insulin N-CAM or its chimera alpha-TNF Monoclonal Antibodies (Immuno~lobulins) Leptin MAb-anti-RSV

human chorionic gonadotropinMAb-anti-IL-2 receptor Enzymes and Inhibitors MAb-anti-CEA

TPA MAb-anti-glycoprotein IIb/IIIa TPA variants MAb-anti-EGF

Urokinase MAb-anti-Her2 Factors VII, VIII, IX, MAb-CD20 X

DNase MAb-alpha-CD3 Glucocerebrosidase MAb-TNFa Hirudin MAb-CD4 al antitrypsin (al proteaseMAb-PSGL-1 inhibitor) Mab-anti F protein of Respiratory Antithrombin III Syncytial Virus Acid a-glucosidase (acid Anti-thrombin-III
maltase) a galactosidase A Cells a-L-iduronidase Red blood cells Urokinase White blood cells (e.g., T
cells, B cells, Cytokines and Chimeric dendritic cells, macrophages, Cytokines NIA cells, _ neutrophils, monocytes and Interleukin-1 (IL-1), the like) 1B, 2, 3, 4 Interferon-alpha (IFN-alpha)Stem cells IFN-alpha-2b Others IFN-beta Hepatits B surface antigen (HbsAg) IFN-gamma IFN-omega Chimeric diphtheria toxin-IL-2 Table 7. Most preferred peptides for glycan remodeling Alpha-galactosidase A Interleukin-2 (IL-2) Alpha-L-iduronidase Factor VIII

Anti-thrombin-III hrDNase Granulocyte colony Insulin stimulating factor (G-CSF)Hepatitis B surface protein (HbsAg) Interferon a Human Growth Hormone (HGH) Interferon ~i Human chorionic gonadotropin Interferon omega Urokinase Factor VII clotting factor TNF receptor-IgG Fc fusion (EnbrelTM) Factor IX clotting factor MAb-Her-2 (HerceptinTM) Follicle Stimulating Hormone MAb-F protein of Respiratory (FSH) Erythropoietin (EPO) Syncytial Virus (SynagisTM) Granulocyte-macrophage colony MAb-CI~20 (RituxanT'~) stimulating factor (GM-CSF) MAb-TNFa (RemicadeTM) Interferon y MAb-Glycoprotein IIb/IIIa (ReoproTM) al protease inhibitor (al antitrypsin) Tissue-type plasminogen activator (TPA) Glucocerebrosidase (Cere~ymeT~) A more detailed list of peptides useful in the invention and their source is provided in Figure 2~.
Other exemplary peptides that are modified by the methods of the invention include members of the immunoglobulin family (e.g., antibodies, MHC molecules, T cell receptors, and the like), intercellular receptors (e.g., integrins, receptors for hormones or growth factors and the like) lectins, and cytokines (e.g., interleukins). Additional examples include tissue-type plasminogen activator (TPA), renin, clotting factors such as Factor VIII and Factor IX, bombesin, thrombin, hematopoietic growth factor, colony stimulating factors, viral antigens, complement peptides, al-antitrypsin, erythropoietin, P-selectin glycopeptide ligand-1 (PSGL-1), granulocyte-macrophage colony stimulating factor, anti-thrombin III, interleukins, interferons, peptides A and C, fibrinogen, herceptinTM, leptin, glycosidases, among many others. This list of peptides is exemplary and should not be considered to be exclusive. Rather, as is apparent from the disclosure provided herein, the methods of the invention are applicable to any peptide in which a desired glycan structure can be fashioned.
The methods of the invention are also useful for modifying chimeric peptides, including, but not limited to, chimeric peptides that include a moiety derived from an immunoglobulin, such as IgG.
Peptides modified by the methods of the invention can be synthetic or wild-type peptides or they can be mutated peptides, produced by methods known in the art, such as site-directed mutagenesis. Glycosylation of peptides is typically either N-linked or O-linked. An exemplary N-linkage is the attachment of the modified sugar to the side chain of an asparagine residue. The tripeptide sequences asparagine-X-serine and asparagine-X
threonine, where X is any amino acid except proline, are the recognition sequences for enzymatic attachment of a carbohydrate moiety to the asparagine side chain.
Thus, the presence of either of these tripeptide sequences in a peptide creates a potential glycosylation site. As described elsewhere herein, O-linked glycosylation refers to the attachment of one sugar (e.~., N-acetylgalactosamine, galactose, mannose, GIcNAc, glucose, fitcose or xylose) to a hydroxy side chain of a hydroxyamino acid, preferably serine or threonine, although 5-hydroxyproline or 5-hydroxylysine may also be used.
Several exemplary embodiments of the invention are discussed below. While several of these embodiments use peptides having names having trademarks, and other specific peptides as the exemplary peptide, these examples are not confined to any specific peptide.
The following exemplary embodiments are contemplated to include all peptide equivalents and variants of any peptide. Such variants include, but are not limited to, adding and deleting N-linked and O-linked glycosylation sites, and fusion proteins with added glycosylation sites.
One of skill in the art will appreciate that the following embodiments and the basic methods disclosed therein can be applied to many peptides with equal success.
In one exemplary embodiment, the present invention provides methods for modifying Granulocyte Colony Stimulating Factor (G-CSF). Figures 29A to 29G set forth some examples of how this is accomplished using the methodology disclosed herein.
In Figure 29B, a G-CSF peptide that is expressed in a mammalian cell system is trimmed back using a sialidase. The residues thus exposed are modified by the addition of a sialic acid-polyethylene glycol) moiety (PEG moiety), using an appropriate donor therefor and ST3Ga11. Figure 29C sets forth an exemplary scheme for modifying a G-CSF
peptide that is expressed in an insect cell. The peptide is modified by adding a galactose moiety using an appropriate donor thereof and a galactosyltransferase. The galactose residues are functionalized with PEG via a sialic acid-PEG derivative, through the action of ST3Ga11. In Figure 29D, bacterially expressed G-CSF is contacted with an N-acetylgalactosamine donor and N-acetylgalactosamine transferase. The peptide is functionalized with PEG, using a PEGylated sialic acid donor and a sialyltransferase. In Figure 29E, mammalian cell expressed G-CSF is contacted with a sialic acid donor that is modified with levulinic acid, adding a reactive ketone to the sialic acid donor. After addition to a glycosyl residue on the glycan on the peptide, the ketone is derivatized with a moiety such as a hydrazine- or amine-PEG. In Figure 29F, bacterially expressed G-CSF is remodeled by contacting the peptide with an endo-GaINAc enzyme under conditions where it functions in a synthetic, rather than a hydrolytic manner, thereby adding a PEG-Gal-GaINAc molecule from an activated derivative thereof. Figure 29G provides another route for remodeling bacterially expressed G-CSF. The polypeptide is derivatized with a PEGylated N-acetylgalactosamine residue by contacting the polypeptide with an N-acetylgalactosamine transferase and an appropriate donor of PEGylated N-acetylgalactosamine.
In another exemplary embodiment, the invention provides methods for modifying Interferon a-14~C (IFNaI4C), as shown in Figures 30A to 30N. The various forms of IFNa are disclosed elsewhere herein. In Figure 30B, IFNaI4C expressed in mammalian cells is first treated with sialidase to trim back the sialic acid units thereon, and then the molecule is PEGylated using ST3Ga13 and a PEGylated sialic acid donor. In Figure 30C, N-acetylglucosamine is first added to IFNaI4C which has been expressed in insect or fungal cells, where the reaction is conducted via the action of GnT-I and/or II using an N-acetylglucosamine donor. The polypeptide is then PEGylated using a galactosyltransferase and a donor of PEG-galactose. In Figure 30D, IFNaI4C expressed in yeast is first treated with Endo-H to trim back the glycosyl units thereon. The molecules is galactosylated using a galactosyltransferase and a galactose donor, and it is then PEGylated using ST3Gal3 and a donor of PEG-sialic acid. In Figure 30F, IFNaI4C produced by mammalian cells is modified to inched a PEG moiety using ST3Ga13 and a donor of PEG-sialic acid. In Figure 30G, IFNaI4C expressed in insect of fungal cells first has N-acetylglucosamine added using one or more of GnT-I, II, IV, and V, and an N-acetylglucosamine donor. The protein is subsequently galactosylated using an appropriate donor and a galactosyltransferase. Then, IFNaI4C is PEGylated using ST3Gal3 and a donor of PEG-sialic acid. In Figure 30H, yeast produced IFNaI4C is first treated with mannosidases to trim back the mannosyl groups. N-acetylglucosamine is then added using a donor of N-acetylglucosamine and one or more of GnT-I, II, IV, and V. IFNaI4C is further galactosylated using an appropriate donor and a galactosyltransferase. Then, the polypeptide is PEGylated using ST3Gal3 and a donor of PEG-sialic acid. In Figure 30I, NS~ cell expressed IFNaI4C is modified by capping appropriate terminal residues with a sialic acid donor that is modified with levulinic acid, thereby adding a reactive ketone to the sialic acid donor. After addition to a glycosyl residue of the peptide, the ketone is derivatized with a moiety such as a hydrazine-or amine- PEG.

In Figure 30J, IFNaI4C expressed by mammalian cells is PEGylated using a donor of PEG-sialic acid and a 2,8-sialyltransferase. In Figure 30K, IFNaI4C produced by mammalian cells is first treated with sialidase to trim back the terminal sialic acid residues, and then the molecule is PEGylated using trans-sialidase and PEGylated sialic acid-lactose complex. In Figure 30L, IFNaI4C expressed in a mammalian system is sialylated using a donor of sialic acid and a 2,8-sialyltransferase. In Figure 30M, IFNaI4C expressed in insect or fungal cells first has N-acetylglucosamine added using an appropriate donor and GnT-I
and/or II. The molecule is then contacted with a galactosyltransferase and a galactose donor that is derivatized with a reactive sialic acid via a linker, so that the polypeptide is attached to the reactive sialic acid via the linker and the galactose residue. The polypeptide is then contacted with ST3Gal3 and transferrin, and thus becomes connected with transferrin via the sialic acid residue. In Figure 30N, IFNaI4C expressed in either insect or fungal cells is first treated with endoglycanase to trim back the glycosyl groups, and is then contacted with a galactosyltransferase and a galactose donor that is derivatized with a reactive sialic acid via a linker, so that the polypeptide is attached to the reactive sialic acid via the linker and the galactose residue. The molecule is then contacted with ST3Gal3 and transferrin, and thus becomes connected with transferrin via the sialic acid residue.
In another exemplary embodiment, the invention provides methods for modifying Interferon a-2a or 2b (IFNa), as shown in Figures 300 to 30EE. In Figure 30P, IFNa produced in mammalian cells is first treated with sialidase to trim back the glycosyl units, and is then PEGylated using ST3Gal3 and a PEGylated sialic acid donor. In Figure 30Q, IFNa expressed in insect cells is first galactosylated using an appropriate donor and a galactosyltransferase, and is then PEGylated using ST3Ga11 and a PEGylated sialic acid donor. Figure 30R offers another method for remodeling IFNa expressed in bacteria:
PEGylated N-acetylgalactosamine is added to the protein using an appropriate donor and N-acetylgalactosamine transferase. In Figure 305, IFNa expressed in mammalian cells is modified by capping appropriate terminal residues with a sialic acid donor that is modified with levulinic acid, adding a reactive ketone to the sialic acid donor. After addition to a glycosyl residue of the peptide, the ketone is derivatized with a moiety such as a hydrazine-or amine- PEG. In Figure 30T, IFNa expressed in bacteria is PEGylated using a modified enzyme Endo-N-acetylgalactosamidase, which functions in a synthetic instead of a hydrolytic manner, and using a N-acetylgalactosamine donor derivatized with a PEG moiety.
In Figure 30U, N-acetylgalactosamine is first added IFNa using an appropriate donor and N-acetylgalactosamine transferase, and then is PEGylated using a sialyltransferase and a PEGylated sialic acid donor. In Figure 30~, IFNa expressed in a mammalian system is first treated with sialidase to trim back the sialic acid residues, and is then PEGylated using a suitable donor and ST3Ga11 and/or ST3Ga13. In Figure 30W, IFNa expressed in mammalian cells is first treated with sialidase to trim back the sialic acid residues.
The polypeptide is then contacted with ST3Gall and two reactive sialic acid residues that are connect via a linker, so that the polypeptide is attached to one reactive sialic acid via the linker and the second sialic acid residue. The polypeptide is subsequently contacted with ST3Gal3 and transferrin, and thus becomes connected with transferrin via the sialic acid residue. In Figure 30Y, IFNa expressed in mammalian cells is first treated with sialidase to trim back the sialic acid residues, and is then PEGylated using ST3Ga11 and a donor of PEG-sialic acid. In Figure 30Z, IFNa produced by insect cells is PEGylated using a galactosyltransferase and a donor of PEGylated galactose. In Figure 30AA, bacterially expressed IFNa first has N-acetylgalactosamine added using a suitable donor and N-acetylgalactosamine transferase.
The protein is then PEGylated using a sialyltransferase and a donor of PEG-sialic acid. In Figure 30CC, IFNa expressed in bacteria is modified in another procedure:
PEGylated N-acetylgalactosamine is added to the protein by N-acetylgalactosamine transferase using a donor of PEGylated N-acetylgalactosamine. In Figure 30DD, IFNa expressed in bacteria is remodeled in yet another scheme. The polypeptide is first contacted with N-acetylgalactosamine transferase and a donor of N-acetylgalactosamine that is derivatized with a reactive sialic acid via a linker, so that IFNa is attached to the reactive sialic acid via the linker and the N-acetylgalactosamine. IFNa is then contacted with ST3Ga13 and asialo-transferrin so that it becomes connected with transferrin via the sialic acid residue. Then, IFNa is capped with sialic acid residues using ST3Gal3 and a sialic acid donor. An additional method for modifying bacterially expressed IFNa is disclosed in Figure 30EE, where IFNa is first exposed to NHS-CO-linker-SA-CMP and is then connected to a reactive sialic acid via the linker. It is subsequently conjugated with transferrin using ST3Ga13 and transferrin.

The methods for remodeling INN omega are essentially identical to those presented here for IFN alpha except that the attachment of the glycan to the IFN omega peptide occurs at amino acid residue 101 in SEQ ID N~:75. The nucleotide and amino acid sequences for IFN omega are presented herein as SEQ ID N~S:74a and 75. Methods of making and using IFN omega are found in LT.S. Patent No. 4,917,887 and 5,317,089, and in EP
Patent No.
0170204-A.
In another exemplary embodiment, the invention provides methods for modifying Interferon (3 (IFN-[3), as shown in Figures 31A to 31 S. In Figure 318, IFN-(3 expressed in a mammalian system is first treated with sialidase to trim back the terminal sialic acid residues.
The protein is then PEGylated using ST3Ga13 and a donor of PEGylated sialic acid. Figure 31 C is a scheme for modifying IFN-~3 produced by insect cells. First, N-acetylglucosamine is added to IFN-(3 using an appropriate donor and GnT-I and/or -II. The protein is then galactosylated using a galactose donor and a galactosyltransferase. Finally, IFN-[3 is PEGylated using ST3Gal3 and a donor of PEG-sialic acid. In Figure 31D, IFN-(3 expressed in yeast is first treated with Endo-H to trim back its glycosyl chains, and is then galactosylated using a galactose donor and a galactosyltransferase, and is then PEGylated using ST3Ga13 and a donor of PEGylated sialic acid. In Figure 31E, IFN-(3 produced by mammalian cells is modified by PEGylation using ST3Gal3 and a donor of sialic acid already derivatized with a PEG moiety. In Figure 31F, IFN-(3 expressed in insect cells first has N-acetylglucosamine added by one or more of GnT-I, II, IV, and V using a N-acetylglucosamine donor, and then is galactosylated using a galactose donor and a galactosyltransferase, and is then PEGylated using ST3Ga13 and a donor of PEG-sialic acid.
In Figure 31 G, IFN-[3 expressed in yeast is first treated with mannosidases to trim back the mannosyl units, then has N-acetylglucosamine added using a N-acetylglucosamine donor and one or more of GnT-I, II, IV, and V. The protein is further galactosylated using a galactose donor and a galactosyltransferase, and then PEGylated using ST3Ga13 and a PEG-sialic acid donor. In Figure 31H, mammalian cell expressed IFN-(3 is modified by capping appropriate terminal residues with a sialic acid donor that is modified with levulinic acid, adding a reactive ketone to the sialic acid donor. After addition to a glycosyl residue of the peptide, the ketone is derivatized with a moiety such as a hydrazine- or amine- PEG. In Figure 31I, IFN-~ expressed in a mammalian system is PEGylated using a donor of PEG-sialic acid and u, 2,8-sialyltransferase. In Figure 31J, IFN-(3 expressed by mammalian cells is first treated with sialidase to trim back its terminal sialic acid residues, and then PEGylated using trans-sialidase and a donor of PEGylated sialic acid. In Figure 31I~, IFN-(3 expressed in mammalian cells is first treated with sialidase to trim back terminal sialic acid residues, then PEGylated using ST3Ga13 and a donor of PEG-sialic acid, and then sialylated using ST3Ga13 and a sialic acid donor. In Figure 31L, IFN-[3 expressed in mammalian cells is first treated with sialidase and galactosidase to trim back the glycosyl chains, then galactosylated using a galactose donor and an a-galactosyltransferase, and then PEGylated using ST3Ga13 or a sialyltransferase and a donor of PEG-sialic acid. In Figure 31M, IFN-(3 expressed in mammalian cells is first treated with sialidase to trim back the glycosyl units: It is then PEGylated using ST3Gal3 and a donor of PEG-sialic acid, and is then sialylated using ST3Ga13 and a sialic acid donor. In Figure 31N, IFN-(3 expressed in mammalian cells is modified by capping appropriate terminal residues with a sialic acid donor that is modified with levulinic acid, adding a reactive ketone to the sialic acid donor. After addition to a glycosyl residue of the peptide, the ketone is derivatized with a moiety such as a hydrazine-or amine- PEG. In Figure 31 ~, IFN-(3 expressed in mammalian cells is sialylated using a sialic acid donor and a 2,8-sialyltransferase. In Figure 31Q, IFN-(3 produced by insect cells first has N-acetylglucosamine added using a donor of N-acetylglucosamine and one or more of GnT-I, II, IV, and V, and is further PEGylated using a donor of PEG-galactose and a galactosyltransferase. In Figure 31R, IFN-(3 expressed in yeast is first treated with endoglycanase to trim back the glycosyl groups, then galactosylated using a galactose donor and a galactosyltransferase, and then PEGylated using ST3Gal3 and a donor of PEG-sialic acid. In Figure 31 S, IFN-(3 expressed in a mammalian system is first contacted with ST3Ga13 and two reactive sialic acid residues connected via a linker, so that the polypeptide is attached to one reactive sialic acid via the linker and the second sialic acid residue.
The polypeptide is then contacted with ST3Ga13 and desialylated transferrin, and thus becomes connected with transferrin via the sialic acid residue. Then, IFN-(3 is further sialylated using a sialic acid donor and ST3Gal3.
In another exemplary embodiment, the invention provides methods for modifying Factor VII or VIIa, as shown in Figures 32 A to 32D. In Figure 32B, Factor VII
or VIIa produced by a mammalian system is first treated with sialidase to trim back the terminal sialic acid residues, and then PEGylated using ST3Ga13 and a donor of PEGylated sialic acid.
Figure 32C, Factor VII or VIIa expressed by mammalian cells is first treated with sialidase to trim back the terminal sialic acid residues, and then PEGylated using ST3Gal3 and a donor of PEGylated sialic acid. Further, the polypeptide is sialylated with ST3Ga13 and a sialic acid donor. Figure 32D offers another modification scheme for Factor VII or VIIa produced by mammalian cells: the polypeptide is first treated with sialidase and galactosidase to trim back its sialic acid and galactose residues, then galactosylated using a galactosyltransferase and a galactose donor, and then PEGylated using ST3Gal3 and a donor of PEGylated sialic acid.
In another exemplaxy embodiment, the invention provides methods for modifying Factor IX, some examples of which are included in Figures 33A to 33G. In Figure 33B, Factor IX produced by mammalian cells is first treated with sialidase to trim back the terminal sialic acid residues, and is then PEGylated with ST3Ga13 using a PEG-sialic acid donor. In Figure 33C, Factor IX expressed by mammalian cells is first treated with sialidase to trim back the terminal sialic acid residues, it is then PEGylated using ST3Ga13 and a PEG-sialic acid donor, and further sialylated using ST3Ga11 and a sialic acid donor. Another scheme for remodeling mammalian cell produced Factor IX can be found in Figure 33D. The polypeptide is first treated with sialidase to trim back the terminal sialic acid residues, then galactosylated using a galactose donor and a galactosyltransferase, further sialylated using a sialic acid donor and ST3Ga13, and then PEGylated using a donor of PEGylated sialic acid and ST3Gall. In Figure 33E, Factor IX that is expressed in a mammalian system is PEGylated through the process of sialylation catalyzed by ST3 Gala using a donor of PEG-sialic acid. In Figure 33F, Factor IX expressed in mammalian cells is modified by capping appropriate terminal residues with a sialic acid donor that is modified with levulinic acid, adding a reactive ketone to the sialic acid donor. After addition to a glycosyl residue of the peptide, the ketone is derivatized with a moiety such as a hydrazine- or amine-PEG. Figure 33G provides an additional method of modifying Factor IX. The polypeptide, produced by mammalian cells, is PEGylated using a donor of PEG-sialic acid and a 2,8-sialyltransferase.
In another exemplary embodiment, the invention provides methods for modification of Follicle Stimulating Hormone (FSH). Figures 34A to 34J present some examples. In Figure 34B, FSH is expressed in a mannnalian system and modified by treatment of sialidase to trim back terminal sialic acid residues, followed by PEGylation using ST3Ga13 and a donor of PEG-sialic acid. In Figure 34C, FSH expressed in mammalian cells is first treated with sialidase to trim back terminal sialic acid residues, then PEGylated using ST3Gal3 and a donor of PEG-sialic acid, and then sialylated using ST3Gal3 and a sialic acid donor. Figure 34I~ provides a scheme for modifying FSH expressed 111 a manlnlallall System.
The polypeptide is treated with sialidase and galactosidase to trim back its sialic acid and galactose residues, then galactosylated using a galactose donor and a galactosyltransferase, and then PEGylated using ST3Gal3 and a donor of PEG-sialic acid. In Figure 34E, FSH
expressed in mammalian cells is modified in the following procedure: FSH is first treated with sialidase to trim back the sialic acid residues, then PEGylated using ST3Gal3 and a donor of PEG-sialic acid, and is then sialylated using ST3Gal3 and a sialic acid donor.
Figure 34F offers another example of modifying FSH produced by mammalian cells: The polypeptide is modified by capping appropriate terminal residues with a sialic acid donor that is modified with levulinic acid, adding a reactive ketone to the sialic acid donor. After addition to a glycosyl residue of the peptide, the ketone is derivatized with a moiety such as a hydrazine- or amine- PEG. In Figure 34G, FSH expressed in a mammalian system is modified in another procedure: the polypeptide is remodeled with addition of sialic acid using a sialic acid donor and an a 2,8-sialyltransferase. In Figure 34H, FSH is expressed in insect cells and modified in the following procedure: N-acetylglucosamine is first added to FSH
using an appropriate N-acetylglucosamine donor and one or more of GnT-I, II, IV, and V;
FSH is then PEGylated using a donor of PEG-galactose and a galactosyltransferase. Figure 34I depicts a scheme of modifying FSH produced by yeast. According to this scheme, FSH
is first treated with endoglycanase to trim back the glycosyl groups, galactosylated using a galactose donor and a galactosyltransferase, and is then PEGylated with ST3Gal3 and a donor of PEG-sialic acid. In Figure 34J, FSH expressed by mammalian cells is first contacted with ST3Ga13 and two reactive sialic acid residues via a linker, so that the polypeptide is attached to a reactive sialic acid via the linker and a second sialic acid residue. The polypeptide is then contacted with ST3Gall and desialylated chorionic gonadotrophin (CG) produced in CHO, and thus becomes connected with CG via the second sialic acid residue.
Then, FSH is sialylated using a sialic acid donor and ST3Gal3 and/or ST3Gall.
In another exemplary embodiment, the invention provides methods for modifying erythropoietin (EPO), Figures 35A to 35AA set forth some examples which are relevant to the remodeling of both wild-type and mutant EPO peptides. In Figure 35B, EPO
expressed in various mammalian systems is remodeled by contacting the expressed protein with a sialidase to remove terminal sialic acid residues. The resulting peptide is contacted with a sialyltransferase and a CMP-sialic acid that is derivatized with a PEG moiety.
In Figure 35C, EPO that is expressed in insect cells is remodeled with N-acetylglucosamine, using GnT-I
and/or GnT-II. Galactose is then added to the peptide, using galactosyltransferase. PEG
group is added to the remodeled peptide by contacting it with a sialyltransferase and a CIVIP-sialic acid that is derivatized with a PEG moiety. In Figure 35I~, EPO that is expressed in a mammalian cell system is remodeled by removing terminal sialic acid moieties via the action of a sialidase. The terminal galactose residues of the N-linked glycosyl units are "capped"
with sialic acid, using ST3Gal3 and a sialic acid donor. The terminal galactose residues on the O-linked glycan are functionalized with a sialic acid bearing a PEG
moiety, using an appropriate sialic acid donor and ST3Ga11. In Figure 35E, EPO that is expressed in a mammalian cell system is remodeled by functionalizing the N-linked glycosyl residues with a PEG-derivatized sialic acid moiety. The peptide is contacted with ST3Gal3 and an appropriately modified sialic acid donor. In Figure 35F, EPO that is expressed in an insect cell system, yeast or fungi, is remodeled by adding at least one N-acetylglucosamine residues by contacting the peptide with a N-acetylglucosamine donor and one or more of GnT-I, GnT-II, and GnT-V. The peptide is then PEGylated by contacting it with a PEGylated galactose donor and a galactosyltransferase. In Figure 35G, EPO that is expressed in an insect cell system, yeast or fungi, is remodeled by the addition of at least one N-acetylglucosamine residues, using an appropriate N-acetylglucosamine donor and one or more of GnT-I, GnT-II, and GnT-V. A galactosidase that is altered to operate in a synthetic, rather than a hydrolytic manner is used to add an activated PEGylated galactose donor to the N-acetylglucosamine residues. In Figure 35H, EPO that is expressed in an insect cell system, yeast or fungi, is remodeled by the addition of at least one terminal N-acetylglucosamine-PEG
residue. The peptide is contacted with GnT-I and an appropriate N-acetlyglucosamine donor that is derivatized with a PEG moiety. In Figure 35I, EPO that is expressed in an insect cell system, yeast or fungi, is remodeled by adding one or more terminal galactose-PEG
residues. The peptide is contacted with GnT-I and an appropriate N-acetylglucosamine donor that is derivatized with a PEG moiety. The peptide is then contacted with galactosyltransferase and an appropriate galactose donor that is modified with a PEG moiety. In Figure 35J, EPO
expressed in an insect cell system, yeast or fungi, is remodeled by the addition of one more terminal sialic acid-PEG residues. The peptide is contacted with an appropriate N-acetylglucosamine donor and GnT-I. The peptide is further contacted with galactosyltransferase and an appropriate galactose donor. The peptide is then contacted with ST3Gal3 and an appropriate sialic acid donor that is derivatized with a PEG
moiety. In Figure 35I~, EPO expressed in an insect cell system, yeast or fungi, is remodeled by the addition of terminal sialic acid-PEG residues. The peptide is contacted with an appropriate N-acetylglucosamine donor and one or more of GnT-I, GnT-II, and GnT-V. The peptide is then contacted with galactosyltransferase and an appropriate galactose donor.
The peptide is further contacted with ST3Gal3 and an appropriate sialic acid donor that is derivatized with a PEG moiety. In Figure 35L, EPO expressed in an insect cell system, yeast or fungi, is remodeled by the addition of one or more terminal a2,6-sialic acid-PEG
residues. The peptide is contacted with an appropriate N-acetylglucosamine donor and one or more of GnT-I, GnT-II, and GnT-V. The peptide is further contacted with galactosyltransferase and an appropriate galactose donor. The peptide is then contacted with a2,6-sialyltransferase and an appropriately modified sialic acid donor. In Figure 35M, EPO expressed in a mammalian cell system is remodeled by addition of one or more terminal sialic acid-PEG
residues. The peptide is contacted with a sialidase to remove terminal sialic acid residues.
The peptide is further contacted with a sialyltransferase and an appropriate sialic acid donor. The peptide is further contacted with a sialyltransferase and an appropriate sialic acid donor that is derivatized with a PEG moiety. In Figure 35N, EPO expressed in a mammalian cell system is remodeled by the addition of one or more terminal sialic acid-PEG residues.
The peptide is contacted with a sialyltransferase and an appropriate sialic acid donor that is derivatized with a PEG moiety. In Figure 350, EPO expressed in a mammalian cell system is remodeled by the addition of one or more terminal a,2,8-sialic acid-PEG residues to primarily O-linked glycans. The peptide is contacted with cc2,8-sialyltransferase and an appropriate sialic acid donor that is derivatized with a PEG moiety. In Figure 35P, EPO expressed in a manunalian cell is remodeled by the addition of one or more terminal a,2,8-sialic acid-PEG residues to O-linked and N-linked glycans. The peptide is contacted with x,2,8-sialyltransferase and an appropriate sialic acid donor that is derivatized with a PEG moiety. In Figure 35Q, EPO
expressed in yeast or fungi is remodeled by the addition of one or more terminal sialic acid-PEG residues. 'The peptide is contacted with mannosidases to remove terminal mannose residues. Next, the peptide is contacted with GnT-I and an appropriate N-acetylglucosamine donor. The peptide is further contacted with galaetosyltransferase and an appropriate galactose donor. The peptide is then contacted with a sialyltransferase and an appropriate sialic acid donor that is derivatized with a PEG moiety. In Figure 35R, EPO
expressed in yeast or fungi is remodeled by the addition of at least one terminal N-acetylglucosamine-PEG
residues. The peptide is contacted with mannosidases to remove terminal mannose residue.
The peptide is then contacted with GnT-I and an appropriate N-acetylglucosamine donor that is derivatized with a PEG moiety. In Figure 355, EPO expressed in yeast or fungi is remodeled by the additon of one mor more terminal sialic acid-PEG residues.
The peptide is contacted with mannosidase-I to remove oc2 mannose residues. The peptide is further contacted with GnT-I and an appropriate N-acetylglucosamine donor. The peptide is then contacted with galactosyltransferase and an appropriate galacose donor. The peptide is then contacted with a sialyltransferase and an appropriate sialic acid donor that is derivatized with a PEG moiety. In Figure 35U, EPO expressed in yeast or fungi is remodeled by addition of one or more galactose-PEG residues. The peptide is contacted with endo-H to trim back glycosyl groups. The peptide is then contacted with galactosyltransferase and an appropriate galactose donor that is derivatized with a PEG moiety. In Figure 35V, EPO
expressed in yeast or fungi is remodeled by the addition of one or more terminal sialic acid-PEG residues.
The peptide is contacted with endo-H to trim back glycosyl groups. The peptide is further contacted with galactosyltransferase and an appropriate galactose donor. The peptide is then contacted with a sialyltransferase and an appropriate sialic acid donor that is derivatized with a PEG moiety. In Figure 35W, EPO expressed in an insect cell system is remodeled by the addition of terminal galactose-PEG residues. The peptide is contacted with mannosidases to remove terminal mannose residues. The peptide is then contacted with galactosyltransferase and an appropriate galactose donor that is derivatized with a PEG moeity. In Figure 35Y, a mutant EPO called "novel erythropoiesis-stimulating protein" or NESP, expressed in NSO
marine myeloma cells is remodeled by capping appropriate terminal residues with a sialic acid donor that is modified with levulinic acid, adding a reactive ketone to the sialic acid donor. After addition to a glycosyl residue of the peptide, the ketone is derivatized with a moiety such as a hydrazine- or amine-PEG. In Figure 35Z, mutant EP~, i.e.
NESP, expressed in a mammalian cell system is remodeled by addition of one or more terminal sialic acid-PEG residues. PEG is added to the glycosyl residue on the glycan using a PEG-modified sialic acid and an ce 2,8-sialyltransferase. In Figure 35AA, NESP
expressed in a mammalian cell system is remodeled by the addition of terminal sialic acid residues. The sialic acid is added to the glycosyl residue using a sialic acid donor and an a2,8-sialyltransferase.
In another exemplary embodiment, the invention provides methods for modifying granulocyte-macrophage colony-stimulating factor (GM-CSF), as shown in Figures 36A to 36K. In Figure 36B, GM-CSF expressed in mammalian cells is first treated with sialidase to trim back the sialic acid residues, and then PEGylated using ST3Ga13 and a donor of PEG-sialic acid. In Figure 36C, GM-CSF expressed in mammalian cells is first treated with sialidase to trim back the sialic acid residues, then PEGylated using ST3Ga13 and a donor of PEG-sialic acid, and then is further sialylated using a sialic acid donor and ST3Ga11 and/or ST3Ga13. In Figure 36D, GM-CSF expressed in NSO cells is first treated with sialidase and a-galactosidase to trim back the glycosyl groups, then sialylated using a sialic acid donor and ST3Ga13, and is then PEGylated using ST3Ga11 and a donor of PEG-sialic acid.
In Figure 36E, GM-CSF expressed in mammalian cells is first treated with sialidase to trim back sialic acid residues, then PEGylated using ST3Ga13 and a donor of PEG-sialic acid, and then is further sialylated using ST3Gal3 and a sialic acid donor. In Figure 36F, GM-CSF expressed in mammalian cells is modified by capping appropriate terminal residues with a sialic acid donor that is modified with levulinic acid, adding a reactive ketone to the sialic acid donor.
After addition to a glycosyl residue of the peptide, the ketone is derivatized with a moiety such as a hydrazine- or amine- PEG. In Figure 36G, GM-CSF expressed in mammalian cells is sialylated using a sialic acid donor and a 2,8-sialyltransferase. In Figure 36I, GM-CSF
expressed in insect cells is modified by addition of N-acetylglucosamine using a suitable donor and one or more of GnT-I, II, IV, and V, followed by addition of PEGylated galactose using a suitable donor and a galactosyltransferase. In Figure 36J, yeast expressed GM-CSF is first treated with endoglycanase and/or mannosidase to trim back the glycosyl units, and subsequently PEGylated using a galactosyltransferase and a donor of PEG-galactose. In Figure 36K, GM-CSF expressed in mammalian cells is first treated with sialidase to trim back sialic acid residues, and is subsequently sialylated using ST3Ga13 and a sialic acid donor. The polypeptide is then contacted with ST3Gal1 and two reactive sialic acid residues comzected via a linker, so that the polypeptide is attached to one reactive sialic acid via the linker and second sialic acid residue. The polypeptide is further contacted with ST3Gal3 and transferrin, and thus becomes connected with transferrin.
In another exemplary embodiment, the invention provides methods for modification of Interferon gamma (IFNy). Figures 37A to 37N contain some examples. In Figure 37E, IFNy expressed in a variety of mammalian cells is first treated with sialidase to trim back terminal sialic acid residues, and is subsequently PEGylated using ST3Gal3 and a donor of PEG-sialic acid. In Figure 37C, IFNY expressed in a mammalian system is first treated with sialidase to trim back terminal sialic acid residues. The polypeptide is then PEGylated using ST3Gal3 and a donor of PEG-sialic acid, and is further sialylated with ST3Gal3 and a donor of sialic acid. In Figure 37D, mammalian cell expressed IFNy is first treated with sialidase and a-galactosidase to trim back sialic acid and galactose residues. The polypeptide is then galactosylated using a galactose donor and a galactosyltransferase. Then, IFNy is PEGylated using a donor of PEG-sialic acid and ST3Gal3. In Figure 37E, IFNy that is expressed in a mammalian system is first treated with sialidase to trim back terminal sialic acid residues.
The polypeptide is then PEGylated using ST3Ga13 and a donor of PEG-sialic acid, and is further sialylated with ST3Gal3 and a sialic acid donor. Figure 37F describes another method for modifying IFNy expressed in a mammalian system. The protein is modified by capping appropriate terminal residues with a sialic acid donor that is modified with levulinic acid, adding a reactive ketone to the sialic acid donor. After addition to a glycosyl residue of the peptide, the ketone is derivatized with a moiety such as a hydrazine- or amine-PEG. In Figure 37G, IFN~y expressed in mammalian cells is remodeled by addition of sialic acid using a sialic acid donor and an a 2,8-sialyltransferase. In Figure 37I, IFNy expressed in insect or fungal cells is modified by addition of N-acetylglucosamine using an appropriate donor and one or more of GnT-I, II, IV, and V. The protein is further modified by addition of PEG
moieties using a donor of PEGylated galactose and a galactosyltransferase.
Figure 37J offers a method for modifying IFNy expressed in yeast. The polypeptide is first treated with endoglycanase to trim back the saccharide chains, and then galactosylated using a galactose donor and a galactosyltransferase. Then, IFNy is PEGylated using a donor of PEGylated sialic acid and ST3Gal3. In Figure 37I~, IFN7 produced by mammalian cells is modified as follows: the polypeptide is first contacted with ST3Ga13 and a donor of sialic acid that is derivatized with a reactive galactose via a linker, so that the polypeptide is attached to the reactive galactose via the linker and sialic acid residue. The polypeptide is then contacted with a galactosyltransferase and transferrin pre-treated with endoglycanase, and thus becomes connected with transferrin via the galactose residue. In the scheme illustrated by Figure 37L, IFN~y, which is expressed in a mammalian system, is modified via the action of ST3Ga13:
PEGylated sialic acid is transferred from a suitable donor to IFNy. Figure 37M
is an example of modifying IFNy expressed in insect or fungal cells, where PEGylation of the polypeptide is achieved by transferring PEGylated N-acetylglucosamine from a donor to IFNy using GnT-I and/or II. In Figure 37N, IFNy expressed in a mammalian system is remodeled with addition of PEGylated sialic acid using a suitable donor and an a 2,8-sialyltransferase.
In another exemplary embodiment, the invention provides methods for modifying al anti-trypsin (al-protease inhibitor). Some such examples can be found in Figures 38A to 38N. In Figure 38B, a1 anti-trypsin expressed in a variety of mammalian cells is first treated with sialidase to trim back sialic acid residues. PEGylated sialic acid residues are then added using an appropriate donor, such as CMP-SA-PEG, and a sialyltransferase, such as ST3Ga13.
Figure 38C demonstrates another scheme of al anti-trypsin modification. al anti-trypsin expressed in a mammalian system is first treated with sialidase to trim back sialic acid residues. Sialic acid residues derivatized with PEG are then added using an appropriate donor and a sialyltransferase, such as ST3Ga13. Subsequently, the molecule is further modified by the addition of sialic acid residues using a sialic acid donor and ST3Gal3.
Optionally, mammalian cell expressed al anti-trypsin is first treated with sialidase and a-galactosidase to trim back terminal sialic acid and a-linkage galactose residues. The polypeptide is then galactosylated using galactosyltransferase and a suitable galactose donor.
Further, sialic acid derivatized with PEG is added by the action of ST3Gal3 using a PEGylated sialic acid donor. In Figure 38D, al anti-trypsin expressed in a mammalian system first has the terminal sialic acid residues trimmed back using sialidase. PEG is then added to N-linked glycosyl residues via the action of ST3Gal3, which mediates the transfer of PEGylated sialic acid from a donor, such as CMP-SA-PEG, to al anti-trypsin.
More sialic acid residues are subsequently attached using a sialic acid donor and ST3Gal3.
Figure 38E
illustrates another process through which al anti-trypsin is remodeled. al anti-trypsin expressed in mammalian cells is modified by capping appropriate terminal residues with a sialic acid donor that is modified with levulinic acid, adding a reactive ketone to the sialic acid donor. After addition to a glycosyl residue of the peptide, the ketone is derivatized with a moiety such as a hydrazine- or amine- PEG. In Figure 38F, yet another method of al anti-trypsin modification is disclosed. al anti-trypsin obtained from a mammalian expression system is remodeled with addition of sialic acid using a sialic acid donor and an a 2,8-sialyltransferase. In Figure 38H, al anti-trypsin is expressed in insect or yeast cells, and remodeled by the addition of terminal N-acetylglucosamine residues by way of contacting the polypeptide with UDP-N-acetylglucosamine and one or more of GnT-I, II, IV, or V. Then, the polypeptide is modified with PEG moieties using a donor of PEGylated galactose and a galactosyltransferase. In Figure 38I, al anti-trypsin expressed in yeast cells is treated first with endoglycanase to trim back glycosyl chains. It is then galactosylated with a galactosyltransferase and a galactose donor. Then, the polypeptide is PEGylated using ST3Ga13 and a donor of PEG-sialic acid. In Figure 38J, al anti-trypsin is expressed in a mammalian system. The polypeptide is first contacted with ST3Gal3 and a donor of sialic acid that is derivatized with a reactive galactose via a linker, so that the polypeptide is attached to the reactive galactose via the linker and sialic acid residue. The polypeptide is then contacted with a galactosyltransferase and transferrin pre-treated with endoglycanase, and thus becomes connected with transferrin via the galactose residue. In Figure 38L, al anti-trypsin expressed in yeast is first treated with endoglycanase to trim back its glycosyl groups. The protein is then PEGylated using a galactosyltransferase and a donor of galactose with a PEG moiety. In Figure 38M, al anti-trypsin expressed in plant cells is treated with hexosaminidase, mannosidase, and xylosidase to trim back its glycosyl chains, and subsequently modified with N-acetylglucosamine derivatized with a PEG moiety, using N-acetylglucosamine transferase and a suitable donor. In Figure 38N, al anti-trypsin expressed in mammalian cells is modified by adding PEGylated sialic acid residues using ST3Ga13 and a donor of sialic acid derivatized with PEG.
In another exemplary embodiment, the invention provides methods for modifying glucocerebrosidase ((3-glucosidase, CerezymeTM or CeredaseTM), as shown in Figures 39A to 39K. In Figure 39B, CerezymeTM expressed in a mammalian system is first treated with sialidase to trim back terminal sialic acid residues, and is then PEGylated using ST3Ga13 and a donor of PEG-sialic acid. In Figure 39C, CerezymeTM expressed in mammalian cells is first treated with sialidase to trim back the sialic acid residues, then has mannose-6-phosphate group attached using ST3Gal3 and a reactive sialic acid derivatized with mannose-6-phosphate, and then is sialylated using ST3Gal3 and a sialic acid donor.
Optionally, NSO
cell expressed CerezymeTM is first treated with sialidase and galactosidase to trim back the glycosyl groups, and is then galactosylated using a galactose donor and an a-galactosyltransferase. Then, mannose-6-phosphate moiety is added to the molecule using ST3Gal3 and a reactive sialic acid derivatized with mannose-6-phosphate. In Figure 39D, CerezymeTM expressed in mammalian cells is first treated with sialidase to trim back the sialic acid residues, it is then PEGylated using ST3Ga13 and a donor of PEG-sialic acid, and is then sialylated using ST3Ga13 and a sialic acid donor. In Figure 39E, CerezymeTM
expressed in mammalian cells is modified by capping appropriate terminal residues with a sialic acid donor that is modified with levulinic acid, adding a reactive ketone to the sialic acid donor. After addition to a glycosyl residue of the peptide, the ketone is derivatized with a moiety such as one or more mannose-6-phosphate groups. In Figure 39F, CerezymeTM
expressed in mammalian cells is sialylated using a sialic acid donor and a 2,8-sialyltransferase. In Figure 39H, CerezymeTM expressed in insect cells first has N-acetylglucosamine added using a suitable donor and one or more of GnT-I, II, IV, and V, and then is PEGylated using a galactosyltransferase and a donor of PEG-galactose.
Tn Figure 39I, CerezymeTM expressed in yeast is first treated with endoglycanase to trim back the glycosyl groups, then galactosylated using a galactose donor and a galactosyltransferase, and then PEGylated using ST3Ga13 and a donor of PEG-sialic acid. Tn Figure 39JK, CerezymeTM
expressed in mammalian cells is first contacted with ST3Ga13 and two reactive sialic acid residues connected via a linker, so that the polypeptide is attached to one reactive sialic acid via the linker and the second sialic acid residue. The polypeptide is then contacted with ST3Ga13 and desialylated transferrin, and thus becomes connected with transferrin. Then, the polypeptide is sialylated using a sialic acid donor and ST3Gal3.
In another exemplary embodiment, the invention provides methods for modifying Tissue-Type Plasminogen Activator (TPA) and its mutant. Several specific modification schemes are presented in Figures 40A to 40W. Figure 40B illustrates one modification procedure: after TPA is expressed by mammalian cells, it is treated with one or more of mannosidase(s) and sialidase to trim back mannosyl and/or sialic acid residues. Terminal N-acetylglucosamine is then added by contacting the polypeptide with a suitable donor ofN-acetylglucosamine and one or more of GnT-I, II, IV, and V. TPA is further galactosylated using a galactose donor and a galactosyltransferase. Then, PEG is attached to the molecule by way of sialylation catalyzed by ST3Ga13 and using a donor of sialic acid derivatized with a PEG moiety. In Figure 40C, TPA is expressed in insect or fungal cells. The modification includes the steps of addition of N-acetylglucosamine using an appropriate donor of N-acetylglucosamine and GnT-I and/or II; galactosylation using a galactose donor and a galactosyltransferase; and attachment of PEG by way of sialylation using ST3Gal3 and a donor of sialic acid derivatized with PEG. In Figure 40D, TPA is expressed in yeast and subsequently treated with endoglycanase to trim back the sacchaxide chains.
The polypeptide is further PEGylated via the action of a galactosyltransferase, which catalyzes the transfer of a PEG-galactose from a donor to TPA. In Figure 40E, TPA is expressed in insect or yeast cells. The polypeptide is then treated with a- and (3- mannosidases to trim back terminal mannosyl residues. Further, PEG moieties are attached to the molecule via transfer of PEG-galactose from a suitable donor to TPA, which is mediated by a galactosyltransferase. Figure 40F provides a different method for modification of TPA obtained from an insect or yeast system: the polypeptide is remodeled by addition of N-acetylglucosamine using a donor of N-acetylglucosamine and GnT-I and/or II, followed by PEGylation using a galactosyltransferase and a donor of PEGylated galactose. Figure 40G offers another scheme for remodeling TPA
expressed in insect or yeast cells. Terminal N-acetylglucosamine is added using a donor of N-acetylglucosamine and GnT-I and/or II. A galactosidase that is modified to operate in a synthetic, rather than a hydrolytic manner, is utilized to add PEGylated galactose from a proper donor to the N-acetylglucosamine residues. In Figure 40I, TPA expressed in a mammalian system is first treated with sialidase and galactosidase to trim back sialic acid and galactose residues. The polypeptide is further modified by capping appropriate terminal residues with a sialic acid donor that is modified with levulinic acid, adding a reactive ketone to the sialic acid donor. After addition to a glycosyl residue of the peptide, the ketone is derivatized with a moiety such as a hydrazine- or amine- PEG. In Figure 4.0J, TPA, which is expressed in a mammalian system, is remodeled following this scheme: first, the polypeptide is treated with a- and (3- mannosidases to trim back the terminal mannosyl residues; sialic acid residues are then attached to terminal galactosyl residues using a sialic acid donor and ST3Gal3; further, TPA is PEGylated via the transfer of PEGylated galactose from a donor to a N-acetylglucosaminyl residue catalyzed by a galactosyltransferase. In Figure 40K, TPA is expressed in a plant system. The modification procedure in this example is as follows: TPA
is first treated with hexosaminidase, mannosidase, and xylosidase to trim back its glycosyl groups; PEGylated N-acetylglucosamine is then added to TPA using a proper donor and N-acetylglucosamine transferase. In Figure 40M, a TPA mutant (TNK TPA), expressed in mammalian cells, is remodeled. Terminal sialic acid residues are first trimmed back using sialidase; ST3Ga13 is then used to transfer PEGylated sialic acid from a donor to TNK TPA, such that the polypeptide is PEGylated. In Figure 40N, TNK TPA expressed in a mammalian system is first treated with sialidase to trim back terminal sialic acid residues. The protein is then PEGylated using CMP-SA-PEG as a donor and ST3Ga13, and further sialylated using a sialic acid donor and ST3Gal3. In Figure 400, NSO cell expressed TNK TPA is first treated with sialidase and a-galactosidase to trim back terminal sialic acid and galactose residues.
TNK TPA is then galactosylated using a galactose donor and a galactosyltransferase. The last step in this remodeling scheme is transfer of sialic acid derivatized with PEG moiety from a donor to TNK TPA using a sialyltransferase such as ST3Gal3. In Figure 40Q, TNK
TPA is expressed in a mammalian system and is first treated with sialidase to trim back terminal sialic acid residues. The protein is then PEGylated using ST3Gal3 and a donor of PEGylated sialic acid. Then, the protein is sialylated using a sialic acid donor and ST3Gal3.
In Figure 40R, TNK TPA expressed in a mammalian system is modified by capping appropriate terminal residues with a sialic acid donor that is modified with levulinic acid, adding a reactive ketone to the sialic acid donor. After addition to a glycosyl residue of the peptide, the ketone is derivatized with a moiety such as a hydrazine= or amine-PEG. In Figure 405, TNK TPA expressed in mammalian cells is modified via a different method: the polypeptide is remodeled with addition of sialic acid using a sialic acid donor and a 2,8-sialyltransferase. In Figure 40U, TNK TPA expressed in insect cells is remodeled by addition of N-acetylglucosamine using an appropriate donor and one or more of GnT-I, II, IV, and V. The protein is furfher modified by addition of PEG moieties using a donor of PEGylated galactose and a galactosyltransferase. In Figure 40V, TNI~ TPA is expressed in yeast. The polypeptide is first treated with endoglycanase to trim back its glycosyl chains and then PEGylated using a galactose donor derivatized with PEG and a galactosyltransferase. In Figure 40VJ, T1~TI~ TPA is produced in a mammalian system. The polypeptide is first contacted with ST3Gal3 and a donor of sialic acid that is derivatized with a reactive galactose via a linker, so that the polypeptide is attached to the reactive galactose via the linker and sialic acid residue. The polypeptide is then contacted with a galactosyltransferase and anti-TNF IG chimera produced in CFI~, and thus becomes connected with the chimera via the galactose residue.
In another exemplary embodiment, the invention provides methods for modifying Interleukin-2 (IL-2). Figures 41A to 41G provide some examples. Figure 41B
provides a two-step modification scheme: IL-2 produced by mammalian cells is first treated with sialidase to trim back its terminal sialic acid residues, and is then PEGylated using ST3Gal3 and a donor of PEGylated sialic acid. In Figure 41 C, insect cell expressed IL-2 is modified first by galactosylation using a galactose donor and a galactosyltransferase.
Subsequently, IL-2 is PEGylated using ST3Ga13 and a donor of PEGylated sialic acid. In Figure 41D, IL-2 expressed in bacteria is modified with N-acetylgalactosamine using a proper donor and N-acetylgalactosamine transferase, followed by a step of PEGylation with a PEG-sialic acid donor and a sialyltransferase. Figure 41E offers another scheme of modifying IL-2 produced by a mammalian system. The polypeptide is modified by capping appropriate terminal residues with a sialic acid donor that is modified with levulinic acid, adding a reactive ketone to the sialic acid donor. After addition to a glycosyl residue of the peptide, the ketone is derivatized with a moiety such as a hydrazine- or amine- PEG. Figure 41F
illustrates an example of remodeling IL-2 expressed by E. coli. The polypeptide is PEGylated using a reactive N-acetylgalactosamine complex derivatized with a PEG group and an enzyme that is modified so that it functions as a synthetic enzyme rather than a hydrolytic one. In Figure 41 G, IL-2 expressed by bacteria is modified by addition of PEGylated N-acetylgalactosamine using a proper donor and N-acetylgalactosamine transferase.
In another exemplary embodiment, the invention provides methods for modifying Factor VIII, as shown in Figures 42A to 42N. In Figure 42B, Factor VIII
expressed in mammalian cells is first treated with sialidase to trim back the sialic acid residues, and is then PEGylated using ST3Gal3 and a donor of PEG-sialic acid. In Figure 42C, Factor VIII
expressed in mammalian cells is first treated with sialidase to trim back the sialic acid residues, then PEGylated using ST3Ga13 and a proper donor, and is then further sialylated using ST3Ga11 and a sialic acid donor.
In Figure 42E, mammalian cell produced Factor VIII is modified by the single step of PEGylation, using ST3Gal3 and a donor of PEGylated sialic acid. Figure 42F
offers another example of modification of Factor VIII that is expressed by mammalian cells.
The protein is PEGylated using ST3Gal1 and a donor of PEGylated sialic acid. In Figure 42G, mammalian cell expressed Factor VIII is remodeled following another scheme: it is PEGylated using a 2,8-sialyltransferase and a donor of PEG-sialic acid. In Figure 42 I, Factor VIII produce by mammalian cells is modified by capping appropriate terminal residues with a sialic acid donor that is modified with levulinic acid, adding a reactive ketone to the sialic acid donor.
After addition to a glycosyl residue of the peptide, the ketone is derivatized with a moiety such as a hydrazine- or amine- PEG. In Figure 42J, Factor VIII expressed by mammalian cells is first treated with Endo-H to trim back glycosyl groups. It is then PEGylated using a galactosyltransferase and a donor of PEG-galactose. In Figure 42K, Factor VIII
expressed in a mammalian system is first sialylated using ST3Ga13 and a sialic acid donor, then treated with Endo-H to trim back the glycosyl groups, and then PEGylated with a galactosyltransferase and a donor of PEG-galactose. In Figure 42L, Factor VIII
expressed in a mammalian system is first treated with mannosidases to trim back terminal mannosyl residues, then has an N-acetylglucosamine group added using a suitable donor and GnT-I
and/or II, and then is PEGylated using a galactosyltransferase and a donor of PEG-galactose.
In Figure 42M, Factor VIII expressed in mammalian cells is first treated with mannosidases to trim back mannosyl units, then has N-acetylglucosamine group added using N-acetylglucosamine transferase and a suitable donor. It is further galactosylated using a galactosyltransferase and a galactose donor, and then sialylated using ST3Gal3 and a sialic acid donor. In Figure 42N, Factor VIII is produced by mammalian cells and modified as follows: it is first treated with mannosidases to trim back the terminal mannosyl groups. A
PEGylated N-acetylglucosamine group is then added using GnT-I and a suitable donor of PEGylated N-acetylglucosamine.

In another exemplary embodiment, the invention provides methods for modifying urokinase, as shown in Figures 43A to 43M. In Figure 43B, urokinase expressed in mammalian cells is first treated with sialidase to trim back sialic acid residues, and is then PEGylated using ST3Ga13 and a donor of PEGylated sialic acid. In Figure 43C, urokinase expressed in mammalian cells is first treated with sialidase to trim back sialic acid residues, then PEGylated using ST3Gal3 and a donor of PEGylated sialic acid, and then sialylated using ST3Ga13 and a sialic acid donor. ~ptionally, urokinase expressed in a mammalian system is first treated with sialidase and galactosidase to trim back glycosyl chains, then galactosylated using a galactose donor and an a-galactosyltransferase, and then PEGylated using ST3Ga13 or sialyltransferase and a donor of PEG-sialic acid. In Figure 43D, urokinase expressed in mammalian cells is first treated with sialidase to trim back sialic acid residues, then PEGylated using ST3Ga13 and a donor of PEG-sialic acid, and then further sialylated using ST3Ga13 and a sialic acid donor. In Figure 43E, urokinase expressed in mammalian cells is modified by capping appropriate terminal residues with a sialic acid donor that is modified with levulinic acid, adding a reactive ketone to the sialic acid donor. After addition to a glycosyl residue of the peptide, the ketone is derivatized with a moiety such as a hydrazine- or amine- PEG. In Figure 43F, urokinase expressed in mammalian cells is sialylated using a sialic acid donor and a 2,8-sialyltransferase. In Figure 43H, urokinase expressed in insect cells is modified in the following steps: first, N-acetylglucosamine is added to the polypeptide using a suitable donor of N-acetylglucosamine and one or more of GnT-I, II, IV, and V; then PEGylated galactose is added, using a galactosyltransferase and a donor of PEG-galactose. In Figure 43I, urokinase expressed in yeast is first treated with endoglycanase to trim back glycosyl groups, then galactosylated using a galactose donor and a galactosyltransferase, and then PEGylated using ST3Ga13 and a donor of PEG-sialic acid.
In Figure 43J, urokinase expressed in mammalian cells is first contacted with ST3Gal3 and two reactive sialic acid residues that are connected via a linker, so that the polypeptide is attached to one reactive sialic acid via the linker and second sialic acid residue. The polypeptide is then contacted with ST3Ga11 and desialylated urokinase produced in mammalian cells, and thus becomes connected with a second molecule of urokinase. Then, the whole molecule is further sialylated using a sialic donor and ST3Ga11 and/or ST3Ga13.
In Figure 43K, isolated urokinase is first treated with sulfohydrolase to remove sulfate groups, and is then PEGylated using a sialyltransferase and a donor of PEG-sialic acid. In Figure 43LM, isolated urokinase is first treated with sulfohydrolase and hexosaminidase to remove sulfate groups and hexosamine groups, and then PEGylated using a galactosyliransferase and a donor of PEG-galactose.
In another exemplary embodiment, the invention provides methods for modifying DNase I, as shown in Figures 44A to 44J. In Figure 44B, DNase I is expressed in a mammalian system and modified in the following steps: first, the protein is treated with sialidase to trim back the sialic acid residues; then the protein is PEGylated with ST3Ga13 using a donor of PEG-sialic acid. In Figure 44C, DNase I expressed in mammalian cells is first treated with sialidase to trim back the sialic acid residues, then PEGylated with ST3Gal3 using a PEG-sialic acid donor, and is then sialylated using ST3Ga13 and a sialic acid donor.
Optionally, DNase I expressed in a mammalian system is first exposed to sialidase and galactosidase to trim back the glycosyl groups, then galactosylated using a galactose donor and an a-galactosyltransferase, and then PEGylated using ST3Gal3 or sialyltransferase and a donor of PEG-sialic acid. In Figure 44D, DNase I expressed in mammalian cells is first treated with sialidase to trim back the sialic acid residues, then PEGylated using ST3Gal3 and a PEG-sialic acid donor, and then sialylated with ST3Ga13 using a sialic acid donor. In Figure 44E, DNase I expressed in mammalian cells is modified by capping appropriate terminal residues with a sialic acid donor that is modified with levulinic acid, adding a reactive ketone to the sialic acid donor. After addition to a glycosyl residue of the peptide, the ketone is derivatized with a moiety such as a hydrazine- or amine- PEG. In Figure 44F, DNase I expressed in mannnalian cells is sialylated using a sialic acid donor and a 2,8-sialyltransferase. In Figure 44H, DNase I expressed in insect cells first has N-acetylglucosamine added using a suitable donor and one or more of GnT-I, II, IV, and V.
The protein is then PEGylated using a galactosyltransferase and a donor of PEG-galactose.
In Figure 44I, DNase I expressed in yeast is first treated with endoglycanase to trim back the glycosyl units, then galactosylated using a galactose donor and a galactosyltransferase, and then PEGylated using ST3Ga13 and a donor of PEG-sialic acid. In Figure 44JI~, DNase I
expressed in mammalian cells is first contacted with ST3Ga13 and two reactive sialic acid residues connected via a linker, so that the polypeptide is attached to one reactive sialic acid via the linker and the second sialic acid residue. The polypeptide is then contacted with ST3Ga11 and desialylated a-1-protease inhibitor, and thus becomes connected with the inhibitor via the sialic acid residue. Then, the polypeptide is further sialylated using a suitable donor and ST3Ga11 and/or ST3Gal3.
In another exemplary embodiment, the invention provides methods for modifying S insulin that is mutated to contain an N-glycosylation site, as shown in Figures 45A to 45L. In Figure 45B, insulin expressed in a mammalian system is first treated with sialidase to trim back the sialic acid residues, and then PEGylated using ST3Ga13 and a PEG-sialic acid donor. In Figure 45C, insulin expressed in insect cells is modified by addition of PEGylated N-acetylglucosamine using an appropriate donor and GnT-I and/or II. In Figure 45D, insulin expressed in yeast is first treated with Endo-H to trim back the glycosyl groups, and then PEGylated using a galactosyltransferase and a donor of PEG-galactose. In Figure 45F, insulin expressed in mammalian cells is first treated with sialidase to trim back the sialic acid residues and then PEGylated using ST3Gal1 and a donor of PEG-sialic acid. In Figure 45G, insulin expressed in insect cells is modified by means of addition of PEGylated galactose using a suitable donor and a galactosyltransferase. In Figure 45H, insulin expressed in bacteria first has N-acetylgalactosamine added using a proper donor and N-acetylgalactosamine transferase. The polypeptide is then PEGylated using a sialyltransferase and a donor of PEG-sialic acid. In Figure 45J, insulin expressed in bacteria is modified through a different method: PEGylated N-acetylgalactosamine is added to the protein using a suitable donor and N-acetylgalactosamine transferase. In Figure 45K, insulin expressed in bacteria is modified following another scheme: the polypeptide is first contacted with N-acetylgalactosamine transferase and a reactive N-acetylgalactosamine that is derivatized with a reactive sialic acid via a linker, so that the polypeptide is attached to the reactive sialic acid via the linker and N-acetylgalactosamine. The polypeptide is then contacted with ST3Gal3 and asialo-transferrin, and therefore becomes connected with transferrin.
Then, the polypeptide is sialylated using ST3Ga13 and a sialic acid donor. In Figure 45L, insulin expressed in bacteria is modified using yet another method: the polypeptide is first exposed to NHS-C~-linker-SA-CMP and becomes connected to the reactive sialic acid residue via the linker. The polypeptide is then conjugated to transferrin using ST3Gal3 and asialo-transferrin. Then, the polypeptide is further sialylated using ST3Ga13 and a sialic acid donor.

In another exemplary embodiment, the invention provides methods for modifying Hepatitis B antigen (M antigen-preS2 and S), as shown in Figures 46A to 46K.
In Figure 468, M-antigen is expressed in a mammalian system and modified by initial treatment of sialidase to trim back the sialic acid residues and subsequent conjugation with lipid A, using ST3Ga13 and a reactive sialic acid linked to lipid A via a linker. In Figure 46C, M-antigen expressed in mammalian cells is first treated with sialidase to trim back the terminal sialic acid residues, then conjugated with tetanus toxin via a linker using ST3Ga11 and a reactive sialic acid residue linked to the toxin via the linker, and then sialylated using ST3Ga13 and a sialic acid donor. In Figure 46I~, M-antigen expressed in a mammalian system is first treated with a galactosidase to trim back galactosyl residues, and then sialylated using ST3Ga13 and a sialic acid donor. The polypeptide then has sialic acid derivatized with KLH
added using ST3Gal1 and a suitable donor. In Figure 46E, yeast expressed M-antigen is first treated with a mannosidase to trim back the mannosyl residues, and then conjugated to a diphtheria toxin using GnT-I and a donor of N-acetylglucosamine linked to the diphtheria toxin.
In Figure 46F, mammalian cell expressed M-antigen is modified by capping appropriate terminal residues with a sialic acid donor that is modified with levulinic acid, adding a reactive ketone to the sialic acid donor. After addition to a glycosyl residue of the peptide, the ketone is derivatized with a moiety such as a hydrazine- or amine- PEG. In Figure 46G, M-antigen obtained from a mammalian system is remodeled by sialylation using a sialic acid donor and poly a 2,8-sialyltransferase. In Figure 46I, M-antigen expressed in insect cells is conjugated to a Neisseria protein by using GnT-II and a suitable donor of N-acetylglucosamine linked to the Neisseria protein. In Figure 46J, yeast expressed M-antigen is first treated with endoglycanase to trim back its glycosyl chains, and then conjugated to a Neisseria protein using a galactosyltransferase and a proper donor of galactose linked to the Neisseria protein.
Figure 46K is another example of modification of M-antigen expressed in yeast.
The polypeptide is first treated with mannosidases to trim back terminal mannosyl residues, and then has N-acetylglucosamine added using GnT-I andlor II. Subsequently, the polypeptide is galactosylated using a galactose donor and a galactosyltransferase, and then capped with sialic acid residues using a sialyltransferase and a sialic acid donor.
In another exemplary embodiment, the invention provides methods for modifying human growth hormone (N, V, and variants thereof), as shown in Figures 47A to 47K. In Figure 47B, human growth hormone either mutated to contain a N-linked site, or a naturally occurring isoform that has an N-linked side (i.e., the placental enzyme) produced by mammalian cells is first treated with sialidase to trim back terminal sialic acid residues and subsequently PEGylated with ST3Gal3 and using a donor of PEGylated sialic acid. In Figure 47C, human growth hormone expressed in insect cells is modified by addition of PEGylated N-acetylglucosamine using GnT-I and/or II and a proper donor of PEGylated N-acetylglucosamine. In Figure 47D, human growth hormone is expressed in yeast, treated with Endo-H to trim back glycosyl groups, and further PEGylated with a galactosyltransferase using a donor of PEGylated galactose. In Figure 47F, human growth hormone-mucin fusion protein expressed in a mammalian system is modified by initial treatment of sialidase to trim back sialic acid residues and subsequent PEGylation using a donor of PEG-sialic acid and ST3Ga11. In Figure 47G, human growth hormone-mucin fusion protein expressed in insect cells is remodeled by PEGylation with a galactosyltransferase and using a donor of PEGylated galactose. In Figure 47H, human growth hormone-mucin fusion protein is produced in bacteria. N-acetylgalactosamine is first added to the fusion protein by the action of N-acetylgalactosamine transferase using a donor of N-acetylgalactosamine, followed by PEGylation of the fusion protein using a donor of PEG-sialic acid and a sialyltransferase. Figure 47I describes another scheme of modifying bacterially expressed human growth hormone-mucin fusion protein: the fusion protein is PEGylated through the action of N-acetylgalactosamine transferase using a donor of PEGylated N-acetylgalactosamine. Figure 47J provides a further remodeling scheme for human growth hormone-mucin fusion protein. The fusion protein is first contacted with N-acetylgalactosamine transferase and a donor of N-acetylgalactosamine that is derivatized with a reactive sialic acid via a linker, so that the fusion protein is attached to the reactive sialic acid via the linker and N-acetylgalactosamine. The fusion protein is then contacted with a sialyltransferase and asialo-transferrin, and thus becomes connected with transferrin via the sialic acid residue. Then, the fusion protein is capped with sialic acid residues using ST3Ga13 and a sialic acid donor. In Figure 47K, yet another scheme is given for modification of human growth hormone(N) produced in bacteria. The polypeptide is first contacted with NHS-CO-linker-SA-CMP and becomes coupled with the reactive sialic acid through the linker. The polypeptide is then contacted with ST3Ga13 and asialo-transferrin and becomes linked to transferrin via the sialic acid residue. Then, the polypeptide is sialylated using ST3Gal3 and a sialic acid donor.
In another exemplary embodiment, the invention provides methods for remodeling T1VF receptor IgG fusion protein (TNFR-IgG, or EnbrelTM), as shown in Figures 4~8A to 4~8G.
Figure 48B illustrates a modification procedure in which TNFR-IgG, expressed in a mammalian system is first sialylated with a sialic acid donor and a sialyltransferase, ST3Gall; the fusion protein is then galactosylated with a galactose donor and a galactosyltransferase; then, the fusion protein is PEGylated via the action of ST3Ga13 and a donor of sialic acid derivatized with PEG. In Figure 48C, TNFR-IgG expressed in mammalian cells is initially treated with sialidase to trim back sialic acid residues. PEG
moieties are subsequently attached to TNFR-IgG by way of transferring PEGylated sialic acid from a donor to the fusion protein in a reaction catalyzed by ST3Gall. In Figure 48D, TNFR-IgG is expressed in a mammalian system and modified by addition of PEG
through the galactosylation process, which is mediated by a galactosyltransferase using a PEG-galactose donor. In Figure 48E, TNFR-IgG is expressed in a mammalian system.
The first step in remodeling of the fusion protein is adding O-linked sialic acid residues using a sialic acid donor and a sialyltransferase, ST3Gall. Subsequently, PEGylated galactose is added to the fusion protein using a galactosyltransferase and a suitable donor of galactose with a PEG
moiety. In Figure 48F, TNFR-IgG expressed in mammalian cells is modified first by capping appropriate terminal residues with a sialic acid donor that is modified with levulinic acid, adding a reactive ketone to the sialic acid donor. After addition to a glycosyl residue of the fusion protein, the ketone is derivatized with a moiety such as a hydrazine-or amine- PEG.
In Figure 48G, TNFR-IgG expressed in mammalian cells is remodeled by 2,8-sialyltransferase, which catalyzes the reaction in which PEGylated sialic acid is transferred to the fusion protein from a donor of sialic acid with a PEG moiety.
In another exemplary embodiment, the invention provides methods for generating HerceptinTM conjugates, as shown in Figures 49A to 49D. In Figure 49B, HerceptinTM is expressed in a mammalian system and is first galactosylated using a galactose donor and a galactosyltransferase. HerceptinTM is then conjugated with a toxin via a sialic acid through the action of ST3Gal3 using a reactive sialic acid-toxin complex. In Figure 49C, HerceptinTM
produced in either mammalian cells or fungi is conjugated to a toxin through the process of galactosylation, using a galactosyltransferase and a reactive galactose-toxin complex. Figure 49D contains another scheme of making HerceptinTM conjugates: HerceptinTM
produced in fungi is first treated with Endo-H to trim back glycosyl groups, then galactosylated using a galactose donor and a galactosyltransferase, and then conjugated with a radioisotope by way of sialylation, by using ST3Ga13 and a reactive sialic acid-radioisotope complex.
Alternatively, the reactive sialic acid moiety may have attached only the chelating moiety can then be loaded with radioisotope at a subsequent stage.
In another exemplary embodiment, the invention provides methods for making SynagisTM conjugates, as shown in Figures SOA to SOD. In Figure SOB, SynagisTM
expressed in marninalian cells is first galactosylated using a galactose donor and a galactosyltransferase, and then PEGylated using ST3Ga13 and a donor of PEG-sialic acid. In Figure SOC, SynagisTM expressed in mammalian or fungal cells is PEGylated using a galactosyltransferase and a donor of PEG-galactose. In Figure SOD, SynagisTM expressed in first treated with Endo-H to trim back the glycosyl groups, then galactosylated using a galactose donor and a galactosyltransferase, and is then PEGylated using ST3Gal3 and a donor of PEG-sialic acid.
In another exemplary embodiment, the invention provides methods for generating RemicadeTM conjugates, as shown in Figures SlA to S1D. In Figure S1B, RemicadeTM
expressed in a mammalian system is first galactosylated using a galactose donor and a galactosyltransferase, and then PEGylated using ST3Gal3 and a donor of PEG-sialic acid. In Figure 51 C, RemicadeTM expressed in a mammalian system is modified by addition of PEGylated galactose using a suitable donor and a galactosyltransferase. In Figure S 1D, RemicadeTM expressed in fungi is first treated with Endo-H to trim back the glycosyl chains, then galactosylated using a galactose donor and a galactosyltransferase, and then conjugated to a radioisotope using ST3Ga13 and a reactive sialic acid derivatized with the radioisotope.
In another exemplary embodiment, the invention provides methods for modifying Reopro, which is mutated to contain an N glycosylation site. Figures 52A to 52L contain such examples. In Figure 52B, Reopro expressed in a mammalian system is first treated with sialidase to trim back the sialic acid residues, and then PEGylated using ST3Ga13 and a donor of PEG-sialic acid. In Figure 52C, Reopro expressed in insect cells is modified by addition of PEGylated N-acetylglucosamine using an appropriate donor and GnT-I and/or II. In Figure 52D, Reopro expressed in yeast is first treated with Endo-H to trim back the glycosyl groups. Subsequently, the protein is PEGylated using a galactosyltransferase and a donor of PEG-galactose. In Figure 52F, Reopro expressed in mammalian cells is first treated with sialidase to trim back the sialic acid residues and then PEGylated with ST3Ga11 using a donor of PEGylated sialic acid. In Figure 52G, Reopro expressed in insect cells is modified by PEGylation using ~ galactosyltransferase and a donor of PEG-galactose. In Figure 52H, Reopro expressed in bacterial first has N-acetylgalactosamine added using N-acetylgalactosamine transferase and a suitable donor. The protein is then PEGylated using a sialyltransferase and a donor of PEG-sialic acid. In Figure 52J, Reopro expressed in bacteria is modified in a different scheme: it is PEGylated via the action of N-acetylgalactosamine transferase, using a donor of PEGylated N-acetylgalactosamine. In Figure 52K, bacterially expressed Reopro is modified in yet another method: first, the polypeptide is contacted with N-acetylgalactosamine transferase and a donor of N-acetylgalactosamine that is derivatized with a reactive sialic acid via a linker, so that the polypeptide is attached to the reactive sialic acid via the linker and N-acetylgalactosamine. The polypeptide is then contacted with ST3Ga13 and asialo-transferrin and thus becomes connected with transferrin via the sialic acid residue. Then, the polypeptide is capped with sialic acid residues using a proper donor and ST3Ga13. Figure 52L offers an additional scheme of modifying bacterially expressed Reopro. The polypeptide is first exposed to NHS-CO-linker-SA-CMP and becomes connected with the reactive sialic acid through the linker. The polypeptide is then contacted with ST3Ga13 and asialo-transferrin and thus becomes connected with transferrin via the sialic acid residue. Then, the polypeptide is capped with sialic acid residues using a proper donor and ST3Ga13.
In another exemplary embodiment, the invention provides methods for producing RituxanTM conjugates. Figures 53A to 53G presents some examples. In Figure 53B, RituxanTM expressed in various mammalian systems is first galactosylated using a proper galactose donor and a galactosyltransferase. The peptide is then functionalized with a sialic acid derivatized with a toxin moiety, using a sialic acid donor and ST3Ga13.
In Figure 53C, RituxanTM expressed in mammalian cells or fungal cells is galactosylated using a galactosyltransferase and a galactose donor, which provides the peptide galactose containing a drug moiety. Figure 53D provides another example of remodeling RituxanTM
expressed in a fungal system. The polypeptide's glycosyl groups are first trimmed back using Endo-H.

Galactose is then added using a galactosyltransferase and a galactose donor.
Subsequently, a radioisotope is conjugated to the molecule through a radioisotope-complexed sialic acid donor and a sialyltransferase, ST3Gal3. In Figure 53F, RituxanTM is expressed in a mammalian system and first galactosylatcd using a galactosyltransferase and a proper galactose donor; sialic acid with a PEG moiety is then attached to the molecule using ST3Gal3 and a PEGylated sialic acid donor. As shown in Figure 53G, RituxanTM
expressed in fungi, yeast, ox mammalian cells can also be modified in the following process: first, the polypeptide is treated with o,- and ~3- mannosidases to remove terminal mannosyl residues;
GIcNAc is then attached to the molecule using GnT-I, II and a GIcNAc donor, radioisotope is then attached by way of galactosylation using a galactosyltransferase and a donor of galactose that is coupled to a chelating moiety capable of binding a radioisotope.
In another exemplary embodiment, the invention provides methods for modifying anti-thrombin III (AT III). Figures 54A to 540 present some examples. In Figure 54B, anti-thrombin III expressed in various mammalian systems is remodeled by the addition of one or more terminal sialic acid-PEG moieties. The AT III molecule is first contacted with sialidase to remove terminal sialic acid moieties. Then, the molecule is contacted with a sialyltransferase and an appropriate sialic acid donor that has been derivatized with a PEG
moiety. In Figure 54C, AT III expressed in various mammalian systems is remodeled by the addition of sialic acid-PEG moieties. The AT III molecule is contacted with sialidase to remove terminal sialic acid moieties. The molecule is then contacted with a ST3Ga13 and an appropriate sialic acid donor that has been derivatized with a PEG moiety at 1.2 mol eq. The molecule is then contacted with a ST3Ga13 and an appropriate sialic acid donor to cap remaining terminal galactose moieties. In Figure 54D, AT III is expressed in NSO marine rnyeloma cells is remodeled to have complex glycan molecules with terminal sialic acid-PEG
moieties. The AT III molecule is contacted with sialidase and a-galactosidase to remove terminal sialic acid and galactose moieties. The molecule is then contacted with galactosyltransferase and an appropriated galactose donor. The molecule is then contacted with a ST3Ga13 and an appropriate sialic acid donor that has been derivatized with a PEG
moiety. In Figure 54E, AT III expressed in various mammalian systems is remodeled to have nearly complete terminal sialic acid-PEG moieties. The AT III molecule is contacted with sialidase to remove terminal sialic acid moieties. The molecule is then contacted with a ST3Gal3 and an appropriate sialic acid donor that has been derivatized with a PEG moiety at 16 mol eq. The molecule is then contacted with ST3Ga13 and an appropriate sialic acid donor to cap remaining terminal galactose moieties. In Figure 54F, AT III
expressed in various mammalian systems is remodeled by the addition of one or more terminal sialic acid PEG moieties. The AT III molecule is contacted with ST3Gal3 and an appropriate sialic acid donor that has been derivatized with a levulinate moiety. The molecule is then contacted with hydrazine-PEG. In Figure 54~G, AT III expressed in various mammalian systems is remodeled by the addition of one or more terminal poly-oc2,~-linked sialic acid moieties. The AT III molecule is contacted with poly-a,2,8-sialyltransferase and an appropriate sialic acid donor. In Figure 54I, AT III expressed in insect, yeast or fungi cells is remodeled by the addition of branching N- N-acetylglucosamine -PEG moieties. The AT III
molecule is contacted with GnT-I and an appropriate N-acetylglucosamine donor that has been derivatized with PEG. In Figure 54J, AT III expressed in yeast is remodeled by removing high maimose glycan structures and the addition of terminal sialic acid-PEG
moieties. The AT III molecule is contacted with endoglycanase to trim back glycosyl groups.
The molecule is then contacted with galactosyltransferase and an appropriate galactose donor. The molecule is then contacted with ST3Ga13 and an appropriate sialic acid donor that has been derivatized with a PEG moiety. In Figure 54K, AT III expressed in various mammalian systems is remodeled by the addition of glycoconjugated transferrin. The AT
III molecule is contacted with ST3Ga13 and an appropriate sialic acid donor that has been derivatized with a linker-galactose donor moiety. The molecule is then contacted with galactosyltransferase and endoglycanase-treated transferrin. In Figure 54M, AT III expressed in yeast is remodeled by the removal of mannose glycan structures and the addition of terminal galactose-PEG
moieties. The molecule is contacted with endoglycanase to trim back glycosyl groups. The molecule is further contacted with galactosyltransferase and an appropriate galactose donor that has been derivatized with a PEG moiety. In Figure 54N, AT III expressed in plant cells is remodeled by converting the glycan structures into mammalian-type complex glycans and then adding one or more terminal galactose-PEG moieties. The AT III molecule is contacted with xylosidase to remove xylose residues. The molecule is then contacted with galactosyltransferase and an appropriate galactose donor that has been derivatized with a PEG moiety. In Figure 540, AT III expressed in various mammalian systems is remodeled by the addition of one or more terminal sialic acid-PEG moieties to terminal galactose moieties. The AT III molecule is contacted with ST3Gal3 and an appropriate sialic acid PEG
donor that has been derivatized with PEG.
In another exemplary embodiment, the invention provides methods for modifying the oc and ~3 subunits of human Chorionic Gonadoiropin (hCG). Figures SSA to SSJ
present some examples. W Figure SSB, hCG expressed in various mammalian and insect systems is remodeled by the addition of terminal sialic acid-PEG moieties. The hCG
molecule is contacted with sialidase to remove terminal sialic acid moieties. The molecule is then contacted with ST3Ga13 and an appropriate sialic acid donor molecule that has been derivatized with a PEG moiety. In Figure SSC, hCG expressed in insect cell, yeast or fungi systems is remodeled by building out the N-linked glycans and the addition of terminal sialic acid-PEG moieties. The hCG molecule is contacted with GnT-I and GnT-II, and an appropriated N-acetylglucosamine donor. The molecule is then contacted with galactosyltransferase and an appropriate galactose donor. The molecule is further contacted with ST3 Gal3 and an appropriate sialic acid donor that has been derivatized with a PEG
moiety. In Figure SSD, hCG expressed in vaxious mammalian and insect systems is remodeled by the addition of one or more terminal sialic acid-PEG moieties on O-linked glycan structures. The hCG molecule is contacted with sialidase to remove terminal sialic acid moieties. The molecule is then contacted with ST3Gal3 and an appropriate sialic acid donor to cap the glycan structures with sialic acid moieties. The molecule is then contacted with ST3Gal1 and an appropriate sialic acid donor that has been derivatized with PEG. In Figure SSE, hCG expressed in various mammalian and insect systems is remodeled by the addition of sialic acid-PEG moieties to N-linked glycan structures. The hCG
molecule is contacted with ST3Ga13 and an appropriate sialic acid donor that has been derivatized with PEG. In Figure SSF, hCG expressed in insect cells, yeast or fungi, is remodeled by the addition of terminal N-acetylglucosamine-PEG molecules. The hCG molecule is contacted with GnT-I and GnT-II, and an appropriate N-acetylglucosamine donor that has been derivatized with PEG. In Figure SSG, hCG expressed in insect cells, yeast or fungi, is remodeled by the addition of not more than one N-acetylglucosamine-PEG moiety per N-linked glycan structure. The hCG molecule is contacted with GnT-I and an appropriate N-acetylglucosamine donor that has been derivatized with a PEG moiety. In Figure SSH, hCG

expressed in various mammalian systems is remodeled by the addition of one or more terminal sialic acid-PEG moiety to O-linked glycan structures. The hCG
molecule is contacted with ST3Gal3 and an appropriate sialic acid donor that has been derivatized with PEG. In Figure SSI, hCG expressed in various mamnxalian systems is remodeled by the addition of terminal sialic acid-PEG moieties. The hCG molecule is contacted with oc2,8-SA
and an appropriate sialic acid donor that has been derivatized with a PEG
moiety. In Figure SSJ, hCG expressed in various mammalian systems is remodeled by the addition of terminal sialic acid moieties. The hCG molecule is contacted with poly-alpha2,8-ST and an appropriate sialic acid donor that has been derivatized with a PEG moiety.
In another exemplary embodiment, the invention provides methods for modifying alpha-galactosidase A (FabrazymeTM). Figures 56A to 56J present some examples.
In Figure 56B, alpha-galactosidase A expressed in and secreted from various mammalian and insect systems is remodeled by the addition of one or more terminal galactose-PEG-transferrin moieties. The alpha-galactosidase A molecule is contacted with Endo-H to trim back glycosyl groups. The molecule is then contacted with galactosyltransferase and an appropriate galactose donor that has been derivatized with PEG and transferrin. In Figure 56C, alpha-galactosidase A expressed in and secreted from various mammal and insect cell systems is remodeled by the addition of one or more terminal sialic acid-linker-mannose-6-phosphate moieties. The alpha-galactosidase A molecule is contacted with sialidase to remove terminal sialic acid moieties. The molecule is further contacted with ST3Gal3 and an appropriate sialic acid donor that has been conjugated via a linker to mannose-6-phosphate.
In Figure 56D, alpha-galactosidase A expressed in NSO marine myeloma cells is remodeled by the addition of terminal sialic acid-linker-mannose-6-phosphate moieties.
The alpha-galactosidase A molecule is contacted with sialidase and a-galactosidase to remove terminal sialic acid and galactose moieties. The molecule is then contacted with galactosyltransferase and an appropriate galactose donor. The molecule is then contacted with sialyltransferase and an appropriate sialic acid donor that has been conjugated via a linker to mannose-6-phosphate. In Figure 56E, alpha-galactosidase A expressed in and secreted from various mammalian and insect cell systems is remodeled by the addition of one or more terminal sialic acid-PEG moieties. The alpha-galactosidase A molecule is contacted with sialidase to remove terminal sialic acid moieties. The molecule is then contacted with sialyltransferase and an appropriate sialic acid donor that has been derivatized with a PEG
moiety. In Figure 56F, alpha-galactosidase A expressed in marmnalian, insect, yeast or fungi systems, is remodeled by the addition of one or more terminal mannose-linker-ApoE
moieties. The alpha-galactosidase A molecule is contacted with mannosyltransferase and an appropriate mannose donor that has been conjugated via a linker to ApoE. In Figure 56G, alpha-galactosidase A expressed in mammalian, insect, yeast or fungal systems is remodeled by the addition of galactose-linker-alpha2-macroglobulin moieties. The alpha-galactosidase A
molecule is contacted with Endo-H to trim back glycosyl groups. The molecule is then contacted with galactosyltransferase and an appropriate galactose donor that has been conjugated via a linker to alpha2-macroglobulin. In Figure 56H, alpha-galactosidase A
expressed in insect, yeast and fungal systems, is remodeled by the addition of one or more N-acetylglucosamine-PEG-mannose-6-phosphate moieties. The alpha-galactosidase molecule is contacted with GnT-I and an appropriate N-acetyl-glucosamine donor that has been derivatized with PEG and mannose-6-phosphate. In Figure 56I, alpha-galactosidase A
expressed in insect, yeast or fungal systems, is remodeled by the addition of one or more terminal galactose-PEG-transferrin moieties. The alpha-galactosidase A
molecule is contacted with GnT-I and an appropriate N-acetyl-glucosamine donor. The molecule is then contacted with galactosyltransferase and an appropriate galactose donor that has been derivatized with PEG and transferrin. In Figure 56J, alpha-galactosidase A
expressed in insect, yeast or fungi systems is remodeled by the addition of one or more terminal sialic acid-PEG-melanotransferrin moieties. The alpha-galactosidase A molecule is contacted with GnT-I and GnT-II and an appropriate N-acetyl-glucosamine donor. The molecule is then contacted with galactosyltransferase and an appropriate galactose donor. The molecule is then contacted with sialyltransferase and an appropriate sialic acid donor that has been derivatized with PEG and rnelanotransferrin.
In another exemplary embodiment, the invention provides methods for modifying alpha-iduronidase (AldurazymeTM). Figures 57A to 57J present some examples. In Figure 57B, alpha-iduronidase expressed in and secreted from various mammalian and insect systems is remodeled by the addition of one or more terminal galactose-PEG-transferrin moieties. The alpha-iduronidase molecule is contacted with Endo-H to trim back glycosyl groups. The molecule is then contacted with galactosyltransferase and an appropriate galactose donor that has been derivatized with PEG and transferrin. In Figure 57C, alpha-iduronidase expressed in and secreted from various mammal and insect cell systems is remodeled by the addition of terminal sialic acid-linker-mannose-6-phosphate moieties. The alpha-iduronidase molecule is contacted with sialidase to remove terminal sialic acid moieties. The molecule is then contacted with ST3Gal3 and an appropriate sialic acid donor that has been conjugated via a linker to mannose-6-phosphate. In Figure 57D, alpha-iduronidase expressed in NS~ marine myeloma cells is remodeled by the addition of one or more terminal sialic acid-linker-mannose-6-phosphate moieties. The alpha-iduronidase molecule is contacted with sialidase and o~-galactosidase to remove terminal sialic acid and galactose moieties. The molecule is then contacted with galactosyltransferase and an appropriate galactose donor. The molecule is further contacted with sialyltransferase and an appropriate sialic acid donor that has been conjugated via a linker to mannose-6-phosphate.
In Figure 57E, alpha-iduronidase expressed in and secreted from various mammalian and insect cell systems is remodeled by the addition of one or more terminal sialic acid-PEG
moieties. The alpha-iduronidase molecule is contacted with sialidase to remove terminal sialic acid moieties. The molecule is further contacted with sialyltransferase and an appropriate sialic acid donor that has been derivatized with a PEG moiety. In Figure 57F, alpha-iduronidase expressed in mammalian, insect, yeast or fungi systems is remodeled by the addition of one or more terminal mannose-linker-ApoE moieties. The alpha-iduronidase molecule is contacted with mannosyltransferase and an appropriate mannose donor that has been conjugated via a linker to ApoE. In Figure 57G, alpha-iduronidase expressed in mammalian, insect, yeast or fungal systems is remodeled by the addition of one or more galactose-linker-alpha2-macroglobulin moieties. The alpha-iduronidase molecule is contacted with Endo-H to trim back glycosyl groups. The molecule is then contacted with galactosyltransferase and an appropriate galactose donor that has been conjugated via a linker to alpha2-macroglobulin. In Figure 57H, alpha-iduxonidase expressed in insect, yeast and fungal systems, is remodeled by the addition of one or more N-acetylglucosamine-PEG-mannose-6-phosphate moieties. The alpha-galactosidase molecule is contacted with GnT-I
and an appropriate N-acetyl-glucosamine donor that has been derivatized with PEG and mannose-6-phosphate. In Figure 57I, alpha-iduronidase expressed in insect, yeast or fungal systems, is remodeled by the addition of one or more terminal galactose-PEG-transferrin moieties. The alpha-iduronidase molecule is contacted with GnT-I and an appropriate N-acetyl-glucosamine donor. The molecule is then contacted with galactosyltransferase and an appropriate galactose donor that has been derivatized with PEG and transferrin. In Figure 57J, alpha-iduronidase expressed in insect, yeast or fungi systems, is remodeled by the addition of one or more terminal sialic acid-PEG-melanotransferrin moieties.
The alpha-iduronidase molecule is contacted with GnT-I and GnT-II and an appropriate N-acetyl-glucosamine donor. The molecule is then contacted with galactosyltransferase and an appropriate galactose donor. The molecule is further contacted with sialyltransferase and an appropriate sialic acid donor that has been derivatized with PEG and melanotransferrin.
A Creation or elimination of N-linked ~lycosylation sites The present invention contemplates the use of peptides in which the site of the glycan chains) on the peptide have been altered from that of the native peptide.
Typically, N-linked glycan chains are linked to the primary peptide structure at asparagine residues where the asparagine residue is within an amino acid sequence that is recognized by a membrane-bound glycosyltransferase in the endoplasmic reticulum (ER). Typically, the recognition site on the primary peptide structure is the sequence asparagine-X-serine/threonine where X can be any amino acid except proline and aspartic acid. While this recognition site is typical, the invention further encompasses peptides that have N-linked glycan chains at other recognition sites where the N-linked chains are added using natural or recombinant glycosyltransferases.
Since the recognition site for N-linked glycosylation of a peptide is known, it is within the skill of persons in the art to create mutated primary peptide sequences wherein a native N-linked glycosylation recognition site is removed, or alternatively or in addition, one or more additional N-glycosylation recognition sites are created. Most simply, an asparagine residue can be removed from the primary sequence of the peptide thereby removing the attachment site for a glycan, thus removing one glycan from the mature peptide. For example, a native recognition site with the sequence of aspar~,ine-serine-serine can be genetically engineered to have the sequence leucine-serine-serine, thus eliminating a N-linked glycosylation site at this position.
Further, an N-linked glycosylation site can be removed by altering the residues in the recognition site so that even though the asparagine residue is present, one or more of the additional recognition residues are absent. For example, a native sequence of asparagine-serine-serine can be mutated to asparagine-serine-lysine, thus eliminating an N-glycosylation site at that position. In the case of N-linked glycosylation sites comprising residues other than the typical recognition sites described above, the skilled artisan can determine the sequence and residues required for recognition by the appropriate glycosyltransferase, and then mutate at least one residue so the appropriate glycosyltransferase no longer recognizes that site. In other words, it is well within the skill of the artisan to manipulate the primary sequence of a peptide such that glycosylation sites are either created or are removed, or both, thereby generating a peptide having an altered glycosylation pattern. The invention should therefore not be construed to be limited to any primary peptide sequence provided herein as the sole sequence for glycan remodeling, but rather should be construed to include any and all peptide sequences suitable for glycan remodeling.
To create a mutant peptide, the nucleic acid sequence encoding the primary sequence of the peptide is altered so that native codons encoding native amino acid residues are mutated to generate a codon encoding another amino acid residue. Techniques for altering nucleic acid sequence are common in the art and are described for example in any well-known molecular biology manual.
In addition, the nucleic acid encoding a primary peptide structure can be synthesized i~ vitro, using standard techniques. For example, a nucleic acid molecule can be synthesized in a "gene machine" using protocols such as the phosphoramidite method. If chemically-synthesized double stranded DNA is required for an application such as the synthesis of a nucleic acid or a fragment thereof, then each complementary strand is synthesized separately. The production of short nucleic acids (60 to 80 base pairs) is technically straightforward and can be accomplished by synthesizing the complementary strands and then annealing them. For the production of longer nucleic acids (>300 base pairs), special strategies may be required, because the coupling efficiency of each cycle during chemical DNA synthesis is seldom 100%. To overcome this problem, synthetic genes (double-stranded) are assembled in modular form from single-stranded fragments that are from 20 to 100 nucleotides in length. For reviews on polynucleotide synthesis, see, for example, Glick and Pasternak (Molecular Biotechnology, Principles and Applications of Recombinant DNA, 1994, ASM Press), Itakura et al. (1984, Annu. Rev. Biochem. 53:323), and Climie et al.
(1990, Proc. Nat'1 Acad. Sci. USA 87:633).
Additionally, changes in the nucleic acid sequence encoding the peptide can be made by site-directed mutagenesis. As will be appreciated, this technique typically employs a phage vector which exists in both a single stranded and double stranded form.
Typical vectors useful in site-directed mutagenesis include vectors such as the M13 phage. These phage are readily available and their use is generally well known to those skilled in the art.
Double stranded plasmids are also routinely employed in site-directed mutagenesis which eliminates the step of transferring the nucleic acid of interest from a plasmid to a phage.
In general, site-directed mutagenesis is performed by first obtaining a single-stranded vector or melting the two strands of a double stranded vector which includes within its sequence a DNA sequence which encodes the desired peptide. An oligonucleotide primer bearing the desired mutated sequence is prepared generally synthetically. This primer is then annealed with the single-stranded vector, and subjected to DNA polymerizing enzymes such as E. coli polymerase I Klenow fragment, in order to complete the synthesis of the mutation-bearing strand. Thus, a heteroduplex is formed wherein one strand encodes the original non-mutated sequence and the second strand bears the desired mutation. This heteroduplex vector is then used to transform or transfect appropriate cells, such as E. coli cells, and clones are selected which include recombinant vectors bearing the mutated sequence arrangement. A
genetic selection scheme was devised by Kunkel et al. (1987, Kunkel et al., Methods Enzymol. 154:367-382) to enrich for clones incorporating the mutagenic oligonucleotide.
Alternatively, the use of PCRTM with commercially available thermostable enzymes such as Taq polymerase may be used to incorporate a mutagenic oligonucleotide primer into an amplified DNA fragment that can then be cloned into an appropriate cloning or expression vector. The PCRTM-mediated mutagenesis procedures of Tomic et al. (1990, Nucl.
Acids Res., 12:1656) and Upender et al. (1995, Biotechniques, 18:29-31) provide two examples of such protocols. A PCRTM employing a thermostable ligase in addition to a thermostable polymerase may also be used to incorporate a phosphorylated mutagenic oligonucleotide into an amplified DNA fragment that may then be cloned into an appropriate cloning or expression vector. The mutagenesis procedure described by Michael (1994, Biotechniques 16:410-412) provides an example of one such protocol.

Not all Asn-X-Ser/Thr sequences are N-glycosylated suggesting the context in which the motif is presented is important. In another approach, libraries of mutant peptides having novel N-linked consensus sites are created in order to identify novel N-linked sites that are glycosylated iy~ viv~ and are beneficial to the activity, stability or other characteristics of the peptide.
As noted previously, the consensus sequence for the addition of N-linked glycan chains in glycoproteins is Asn-X-Ser/Thr where X can be any amino acid. The nucleotide sequence encoding the amino acid two positions to the carboxyl terminal side of the Asn may be mutated to encode a Ser and/or Thr residue using standard procedures known to those of ordinary skill in the art. As stated above not all Asn-X-Ser/Thr sites are modified by the addition of glycans. Therefore, each recombinant mutated glycoprotein must be expressed in a fungal, yeast or animal or mammalian expression system and analyzed for the addition of an N-linked glycan chain. The techniques for the characterization of glycosylation sites are well known to one skilled in the art. Further, the biological function of the mutated recombinant glycoprotein can be determined using assays standard for the particular protein being examined. Thus, it becomes a simple matter to manipulate the primary sequence of a peptide and identify novel glycosylation sites contained therein, and further determine the effect of the novel site on the biological activity of the peptide.
In an alternative embodiment, the nucleotide sequence encoding the amino acid two positions to the amino terminal side of Ser/Thr residues may be mutated to encode an Asn using standard procedures known to those of ordinary skill in the art. The procedures to determine whether a novel glycosylation site has been created and the effect of this site on the biological activity of the peptide are described above.
B Creation or elimination of O-linked ~lycosylation sites The addition of an O-linked glycosylation site to a peptide is conveniently accomplished by altering the primary amino acid sequence of the peptide such that it contains one or more additional O-linked glycosylation sites compared with the beginning primary amino acid sequence of the peptide. The addition of an O-linked glycosylation site to the peptide may also be accomplished by incorporation of one or more amino acid species into the peptide which comprises an -OH group, preferably serine or threonine residues, within the sequence of the peptide, such that the OH group is accessible and available for O-linked glycosylation. Similar to the discussion of alteration of N-linked glycosylation sites in a peptide, the primary amino acid sequence of the peptide is preferably altered at the nucleotide level. Specific nucleotides in the DNA sequence encoding the peptide may be altered such that a desired amino acid is encoded by the sequence. IVIutation(s) in DNA are preferably made using methods known in the art, such as the techniques of phosphoramidite method DNA synthesis and site-directed mutagenesis described above.
Alternatively, the nucleotide sequence encoding a putative site for O-linked glycan addition can be added to the DNA molecule in one or several copies to either 5' or the 3' end of the molecule. The altered DNA sequence is then expressed in any one of a fungal, yeast, or animal or mammalian expression system and analyzed for the addition of the sequence to the peptide and whether or not this sequence is a functional O-linked glycosylation site.
Briefly, a synthetic peptide acceptor sequence is introduced at either the 5' or 3' end of the nucleotide molecule. In principle, the addition of this type of sequence is less disruptive to the resulting glycoprotein when expressed in a suitable expression system. The altered DNA
is then expressed in CHO cells or other suitable expression system and the proteins expressed thereby are examined for the presence of an O-linked glycosylation site. In addition, the presence or absence of glycan chains can be determined.
In yet another approach, advantageous sites for new O-linked sites may be found in a peptide by creating libraries of the peptide containing various new O-linked sites. For example, the consensus amino acid sequence for N-acetylgalactosamine addition by an N-acetylgalactosaminyltransferase depends on the specific transferase used. The amino acid sequence of a peptide may be scanned to identify contiguous groups of amino acids that can be mutated to generate potential sites for addition of O-linked glycan chains.
These mutations can be generated using standard procedures known to those of ordinary skill in the art as described previously. In order to determine if any discovered glycosylation site is actually glycosylated, each recombinant mutated peptide is then expressed in a suitable expression system and is subsequently analyzed for the addition of the site and/or the presence of an O-linked glycan chain.
C. Chemical synthesis of peptides While the primary structure ~f peptides useful in the invention can be generated most efficiently in a cell-based expression system, it is within the scope of the present invention that the peptides may be generated synthetically. Chemical synthesis of peptides is well known in the art and include, without limitation, stepwise solid phase synthesis, and fragment condensation either in solution or on solid phase. A classic stepwise solid phase synthesis of involves covalently linking an amino acid corresponding to the carboxy-terminal amino acid of the desired peptide chain to a solid support and extending the peptide chain toward the amino end by stepwise coupling of activated amino acid derivatives having activated carboxyl groups. After completion of the assembly of the fully protected solid phase bound peptide chain, the peptide-solid phase covalent attachment is cleaved by suitable chemistry and the protecting groups are removed to yield the product peptide. See, R.
Merrifield, Solid Phase Peptide Synthesis: The Synthesis of a Tetrapeptide, J. Am. Chem. Soc., 85:2149-2154 (1963). The longer the peptide chain, the more challenging it is to obtain high-purity well-defmed products. Due to the production of complex mixtures, the stepwise solid phase synthesis approach has size limitations. In general, well-defined peptides of 100 contiguous amino acid residues or more are not routinely prepared via stepwise solid phase synthesis.
The segment condensation method involves preparation of several peptide segments by the solid phase stepwise method, followed by cleavage from the solid phase and purification of these maximally protected segments. The protected segments are condensed one-by-one to the first segment, which is bound to the solid phase.
The peptides useful in the present invention may be synthesized by exclusive solid phase synthesis, partial solid phase methods, fragment condensation or classical solution synthesis. These synthesis methods are well-known to those of skill in the art (see, for example, Merrifield, J. Am. Chem. Soc. 85:2149 (1963), Stewart et al., "Solid Phase Peptide Synthesis" (2nd Edition), (Pierce Chemical Co. 1984), Bayer and Rapp, Chem.
Pept. Prot. 3:3 (1986), Atherton et al., Solid Phase Peptide Synthesis: A Practical Approach (IRL Press 1989), Fields and Colowick, "Solid-Phase Peptide Synthesis," Methods in Enzymology Volume 289 (Academic Press 1997), and Lloyd-Williams et al., Chemical Approaches to the Synthesis of Peptides and Peptides (CRC Press, Inc. 1997)). Variations in total chemical synthesis strategies, such as "native chemical ligation" and "expressed peptide ligation" are also standard (see, for example, Dawson et al., Science 266:776 (1994), Hackeng et al., Proc.
Nat'1 Acad. Sci. USA 94:7845 (1997), Dawson, Methods Enzymol. 287: 34 (1997), Muir et al, Proc. Nat'1 Acad. Sci. USA 95:6705 (1998), and Severinov and Muir, J.
Biol. Chem.

273:16205 (1998)). Also useful are the solid phase peptide synthesis methods developed by Gryphon Sciences, South San Francisco, CA. See, U.S. Patent Nos. 6,326,468, 6,217,873, 6,174,530, and 6,001,364, all of which are incorporated in their entirety by reference herein.
S D. Post-translational modifications It will be appreciated to one of ordinary skill in the art that peptides may undergo post-translational modification besides the addition of N-linked and/or ~-linked glycans thereto. It is contemplated that peptides having post-translational modifications other than glycosylation can be used as peptides in the invention, as long as the desired biological activity or function of the peptide is maintained or improved. Such post-translational modifications may be natural modifications usually carried out in vivo, or engineered modifications of the peptide carried out in vitro. Contemplated known modifications include, but are not limited to, acetylation, acylation, ADP-ribosylation, amidation, covalent attaclunent of flavin, covalent attachment of a heme moiety, covalent attachment of a nucleotide or nucleotide derivative, covalent attachment of a lipid or lipid derivative, covalent attachment of phosphotidylinositol, cross-linking, cyclization, disulfide bond formation, demethylation, formation of covalent crosslinks, formation of cysteine, formation of pyroglutamate, formylation, gamma carboxylation, glycosylation, GPI anchor formation, hydroxylation, iodination, methylation, myristoylation, oxidation, proteolytic processing, phosphorylation, prenylation, racemization, selenoylation, sulfation, transfer-RNA mediated addition of amino acids to peptides such as arginylation, and ubiquitination.
Enzymes that may be used to carry out many of these modifications are well known in the art, and available commercially from companies such as Boehringer Mannheim (Indianapolis, IN) and Sigma Chemical Company (St. Louis, MO), among others.
Such modifications are well known to those of skill in the art and have been described in great detail in the scientific literature. Several particularly common modifications, glycosylation, lipid attachment, sulfation, gamma-carboxylation of glutamic acid residues, hydroxylation and ADP-ribosylation, for instance, are described in most basic texts, such as Peptides--Structure and Molecular Properties, 2nd Ed., T. E. Creighton, W. H.
Freeman and Company, New York (1993). Many detailed reviews are available on this subject, such as by Wold, F., Post-translational Covalent Modification of Peptides, B. C. Johnson, Ed., Academic Press, New York 1-12 (1983); Seifter et al. (Meth. Enzymol. 182: 626-646 (1990)) and Rattan et al. (Ann. N.Y. Acad. Sci. 663:48-62 (1992)).
Covalent modifications of a peptide may also be introduced into the molecule iva vita°~
by reacting targeted amino-acid residues of the peptide with an organic derivatizing agent S that is capable of reacting with selected side chains or terminal amino-acid residues. Most commonly derivatized residues are cysteinyl, histidyl, lysinyl, arginyl, tyrosyl, glutaminyl, asparaginyl and amino terminal residues. Hydroxylation of proline and lysine, phosphorylation of hydroxyl groups of seryl and threonyl residues, methylation of the alpha-amino groups of lysine, histidine, and histidine side chains, acetylation of the N-terminal amine and amidation of the C-terminal carboxylic groups. Such derivatized moieties may improve the solubility, absorption, biological half life and the like. The moieties may also eliminate or attenuate any undesirable side effect of the peptide and the like.
In addition, derivatization with bifunctional agents is useful for cross-linking the peptide to water insoluble support matrices or to other macromolecular carriers. Commonly used cross-linking agents include glutaraldehyde, N-hydroxysuccinimide esters, homobifimctional imidoesters, 1,1-bis(-diazoloacetyl)-2-phenylethane, and bifunctional maleimides. Derivatizing agents such as methyl-3-[9p-azidophenyl)]dithiopropioimidate yield photoactivatable intermediates that are capable of forming crosslinks in the presence of light. Alternatively, reactive water-insoluble matrices such as cyanogen bromide activated carbohydrates and the reactive substrates described in U.S. Pat. Nos.
3,969,287 and 3,691,016 may be employed for peptide immobilization.
E. Fusion peptides/peptides Peptides useful in the present invention may comprise fusion peptides. Fusion peptides are particularly advantageous where biological and/or functional characteristics of two peptides are desired to be combined in one peptide molecule. Such fusion peptides can present combinations of biological activity and function that are not found in nature to create novel and useful molecules of therapeutic and industrial applications.
Biological activities of interest include, but are not limited to, enzymatic activity, receptor and/or ligand activity, immunogenic motifs, and structural domains.

Such fusion peptides are well known in the art, and the methods of creation will be well-known to those in the art. For example, a human a,-interferon human albumin fusion peptide has been made wherein the resulting peptide has the therapeutic benefits of o,-interferon combined with the long circulating life of albumin, thereby creating a therapeutic composition that allows reduced dosing frequency and potentially reduced side effects,in patients. See, AlbuferonTM from Human Genome Sciences, Inc. and U.S. Patent No.
5,766,~~3. ~ther fusion peptides include antibody molecules that are described elsewhere herein.
F Generation of smaller "biologically active" molecules The peptides used in the invention may be variants of native peptides, wherein a fragment of the native peptide is used in place of the full length native peptide. In addition, pre-pro-, and pre-peptides are contemplated. Variant peptides may be smaller in size that the native peptide, and may comprise one or more domains of a larger peptide.
Selection of specific peptide domains can be advantageous when the biological activity of certain domains in the peptide is desired, but the biological activity of other domains in the peptide is not desired. Also included are truncations of the peptide and internal deletions which may enhance the desired therapeutic effect of the peptide. Any such forms of a peptide is contemplated to be useful in the present invention provided that the desired biological activity of the peptide is preserved.
Shorter versions of peptides may have unique advantages not found in the native peptide. In the case of human albumin, it has been found that a truncated form comprising as little as 63% of the native albumin peptide is advantageous as a plasma volume expander.
The truncated albumin peptide is considered to be better than the native peptide for this therapeutic purpose because an individual peptide dose of only one-half to two-thirds that of natural-human serum albumin, or recombinant human serum albumin is required for the equivalent colloid osmotic effect. See U.S. Patent No. 5,30,712, the entirety of which is incorporated by reference herein.
Smaller "biologically active" peptides have also been found to have enhanced therapeutic activity as compared to the native peptide. The therapeutic potential of IL-2 is limited by various side effects dominated by the vascular leak syndrome. A
shorter chemically synthesized version of the peptide consisting of residues 1-30 corresponding to the entire a-helix was found to fold properly and contain the natural IL-2 biological activity with out the attending side effects.
G. C"aeneration of novel peptides The peptide of the invention may be derived from a primary sequence of a native peptide, or may be engineered using any of the many means known to those of skill in the art. Such engineered peptides can be designed and/or selected because of enhanced or novel properties as compared with the native peptide. For example, peptides may be engineered to have increased enzyme reaction rates, increased or decreased binding affinity to a substrate or ligand, increased or decreased binding affinity to a receptor, altered specificity for a substrate, ligand, receptor or other binding partner, increased or decreased stability ih vitro and/or ire vivo, or increased or decreased immunogenicity in an animal.
H. Mutations 1. Rational design mutation The peptides useful in the methods of the invention may be mutated to enhance a desired biological activity or function, to diminish an undesirable property of the peptide, and/or to add novel activities or functions to the peptide. "Rational peptide design" may be used to generate such altered peptides. Once the amino acid sequence and structure of the peptide is known and a desired mutation planned, the mutations can be made most conveniently to the corresponding nucleic acid codon which encodes the amino acid residue that is desired to be mutated. One of skill in the art can easily determine how the nucleic acid sequence should be altered based on the universal genetic code, and knowledge of codon preferences in the expression system of choice. A mutation in a codon may be made to change the amino acid residue that will be polymerized into the peptide during translation.
Alternatively, a codon may be mutated so that the corresponding encoded amino acid residue is the same, but the codon choice is better suited to the desired peptide expression system.
For example, cys-residues may be replaced with other amino acids to remove disulfide bonds from the mature peptide, catalytic domains may be mutated to alter biological activity, and in general, isoforms of the peptide can be engineered. Such mutations can be point mutations, deletions, insertions and truncations, among others.

Techniques to mutate specific amino acids in a peptide are well known in the art. The technique of site-directed mutagenesis, discussed above, is well suited for the directed mutation of codons. The oligonucleotide-mediated mutagenesis method is also discussed in detail in Sambrook et al. (2001, Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, New York, starting at page 15.51). Systematic deletions, insertions and truncations can be made using linker insertion mutagenesis, digestion with nuclease Ba131, and linker-scanning mutagenesis, among other method well known to those in the art (Sambrook et al., 2001, Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, New York).
Rational peptide design has been successfully used to increase the stability of enzymes with respect to thermoinactivation and oxidation. For example, the stability of an enzyme was improved by removal of asparagine residues in a-amylase (Declerck et al., 2000, J. Mol. Biol. 301:1041-1057), the introduction of more rigid structural elements such as proline into a-amylase (Igarashi et al., 1999, Biosci. Biotechnol. Biochem.
63:1535-1540) and D-xylose isomerase (Zhu et al., 1999, Peptide Eng. 12:635-638). Further, the introduction of additional hydrophobic contacts stabilized 3-isopropylmalate dehydrogenase (Akanuma et al., 1999, Eur. J. Biochem. 260:499-504) and formate dehydrogenase obtained from Pseudomonas sp. (Rojkova et al., 1999, FEBS Lett. 445:183-188). The mechanisms behind the stabilizing effect of these mutations is generally applicable to many peptides.
These and similar mutations are contemplated to be useful with respect to the peptides remodeled in the methods of the present invention.
2. Random muta~enesis techniques Novel peptides useful in the methods of the invention may be generated using techniques that introduce random mutations in the coding sequence of the nucleic acid. The nucleic acid is then expressed in a desired expression system, and the resulting peptide is assessed for properties of interest. Techniques to introduce random mutations into DNA
sequences axe well known in the art, and include PCR mutagenesis, saturation mutagenesis, and degenerate oligonucleotide approaches. See Sambrook and Russell (2001, Molecular Cloning, A Laboratory Approach, Cold Spring Harbor Press, Cold Spring Harbor, NY) and Ausubel et al. (2002, Current Protocols in Molecular Biology, John Wiley &
Sons, NY).

In PCR mutagenesis, reduced Taq polymerase fidelity is used to introduce random mutations into a cloned fragment of DNA (Leung et al., 1989, Technique 1:11-15). This is a very powerful and relatively rapid method of introducing random mutations into a DNA
sequence. The DNA region to be mutagenized is amplified using the polymerase chain reaction (PCR) under conditions that reduce the fidelity of DNA synthesis by Taq DNA
polymerase, e.g., by using an altered dGTP/dATP ratio and by adding Mn2+ to the PCR
reaction. The pool of amplified DNA fragments are inserted into appropriate cloning vectors to provide random mutant libraries.
Saturation mutagenesis allows for the rapid introduction of a large number of single base substitutions into cloned DNA fragments (Mayers et al., 1985, Science 229:242). This technique includes generation of mutations, e.g., by chemical treatment or irradiation of single-stranded DNA in vitro, and synthesis of a complementary DNA strand. The mutation frequency can be modulated by modulating the severity of the treatment, and essentially all possible base substitutions can be obtained. Because this procedure does not involve a genetic selection for mutant fragments, both neutral substitutions as well as those that alter function, are obtained. The distribution of point mutations is not biased toward conserved sequence elements.
A library of nucleic acid homologs can also be generated from a set of degenerate oligonucleotide sequences. Chemical synthesis of a degenerate oligonucleotide sequences can be carried out in an automatic DNA synthesizer, and the synthetic genes may then be ligated into an appropriate expression vector. The synthesis of degenerate oligonucleotides is known in the art (see for example, Narang, SA (1983) Tetrahedron 39:3; Itakura et al. (1981) Recombinant DNA, Proc 3rd Cleveland Sympos. Macromolecules, ed. AG Walton, Amsterdam: Elsevier pp. 273-289; Itakura et al. (1984) Annu. Rev. Biochem.
53:323; Itakura et al. (1984) Science 198:1056; Ike et al. (1983) Nucleic Acid Res. 11:477.
Such techniques have been employed in the directed evolution of other peptides (see, for example, Scott et al.
(1990) Science 249:386-390; Roberts et al. (1992) PNAS 89:2429-2433; Devlin et al. (1990) Science 249: 404-406; Cwirla et al. (1990) PNAS 87: 6378-6382; as well as U.S.
Pat. Nos.
5,223,409, 5,198,346, and 5,096,815).
a. Directed evolution Peptides useful in the methods of the invention may also be generated using "directed evolution" techniques. In contrast to site directed mutagenesis techniques where knowledge of the structure of the peptide is required, there now exist strategies to generate libraries of mutations from which to obtain peptides with improved properties without knowledge of the structural features of the peptide. These strategies are generally known as "directed evolution" technologies and are different from traditional random mutagenesis procedures in that they involve subjecting the nucleic acid sequence encoding the peptide of interest to recursive rounds of mutation, screening and amplification.
In some "directed evolution" techniques, the diversity in the nucleic acids obtained is generated by mutation methods that randomly create point mutations in the nucleic acid sequence. The point mutation techniques include, but are not limited to, "error-prone PCRTM" (Caldwell and Joyce, 1994; PCR Methods Appl. 2: 28-33; and Ke and Madison, 1997, Nucleic Acids Res. 25: 3371-3372), repeated oligonucleotide-directed mutagenesis (Reidhaar-Olson et al., 1991, Methods Enzymol. 208:564-586), and any of the aforementioned methods of random mutagenesis.
Another method of creating diversity upon which directed evolution can act is the use of mutator genes. The nucleic acid of interest is cultured in a mutator cell strain the genome of which typically encodes defective DNA repair genes (U.S. Patent No.
6,365,410;
Selifonova et al., 2001, Appl. Environ. Microbiol. 67:3645-3649; Long-McGie et al., 2000, Biotech. Bioeng. 68:121-125; see, Genencor International Inc, Palo Alto CA).
Achieving diversity using directed evolution techniques may also be accomplished using saturation mutagenesis along with degenerate primers (Gene Site Saturation MutagenesisTM, Diversa Corp., San Diego, CA). In this type of saturation mutagenesis, degenerate primers designed to cover the length of the nucleic acid sequence to be diversified are used to prime the polymerase in PCR reactions. In this manner, each codon of a coding sequence for an amino acid may be mutated to encode each of the remaining common nineteen amino acids. This technique may also be used to introduce mutations, deletions and insertions to specific regions of a nucleic acid coding sequence while leaving the rest of the nucleic acid molecule untouched. Procedures for the gene saturation technique are well known in the art, and can be found in U.S. Patent 6,171,820.
b. DNA shuffling Novel peptides useful in the methods of the invention may also be generated using the techniques of gene-shuffling, motif shuffling, axon-shuffling, and/or codon-shuffling (collectively referred to as "DNA shuffling"). DNA shuffling techniques are may be employed to modulate the activities of peptides useful in the invention and may be used to generate peptides having altered activity. See, generally, U.S. Pat. Nos.
5,605,793;
5,811,238; 5,830,721; 5,834,252; and 5,837,458, and Stammer et al. (1994, Nature 370(6488):389-391); Crameri et al. (1998, Nature 391 (6664):288-291); Zhang et al. (1997, Proc. Natl. Acad. Sci. USA 94(9):4504-4509); Stammer et al. (1994, Proc. Natl.
Acad. Sci USA 91(22):10747-10751), Patten et al. (1997, Curr. ~pinion Biotechnol. 8:724-33);
Harayama, (1998, Trends Biotechnol. 16(2):76-82); Hansson, et al., (1999, J.
Mol. Biol.
287:265-76); and Lorenzo and Blasco (1998, Biotechniques 24(2):308-13) (each of these patents are hereby incorporated by reference in its entirety).
DNA shuffling involves the assembly of two or more DNA segments by homologous or site-specific recombination to generate variation in the polynucleotide sequence. DNA
shuffling has been used to generate novel variations of human immunodeficiency virus type 1 proteins (Pekrun et al., 2002, J. Virol. 76(6):2924-35), triazine hydrolases (Raillard et al.
2001, Chem Biol 8(9):891-898), marine leukemia virus (MLV) proteins (Powell et al. 2000, Nat Biotechnol 18(12):1279-1282), and indoleglycerol phosphate synthase (Merz et al. 2000, Biochemistry 39(5):880-889).
The technique of DNA shuffling was developed to generate biomolecular diversity by mimicking natural recombination by allowing i~ vitro homologous recombination of DNA
(Stemmler, 1994, Nature 370: 389-391; and Stemmler, 1994, PNAS 91: 10747-10751).
Generally, in this method a population of related genes is fragmented and subjected to recursive cycles of denaturation, rehybridization, followed by the extension of the 5' overhangs by Taq polymerase. With each cycle, the length of the fragments increases, and DNA recombination occurs when fragments originating from different genes hybridize to each other. The initial fragmentation of the DNA is usually accomplished by nuclease digestion, typically using DNase (see Stemmler references, above), but may also be accomplished by interrupted PCR synthesis (LJ.S. Patent 5,965,408, incorporated herein by reference in its entirety; see, Diversa Corp., San Diego, CA). DNA shuffling methods have advantages over random point mutation methods in that direct recombination of beneficial mutations generated by each round of shuffling is achieved and there is therefore a self selection for improved phenotypes of peptides.
The techniques of DNA shuffling are well known to those in art. Detailed explanations of such technology is found in Stemmler, 1994, Nature 370: 389-391 and S Stemmler, 1994, PNAS 91: 10747-10751. The DNA shuffling technique is also described in U.S. Patents 6,180,406, 6,165,793, 6,132,970, 6,117,679, 6,096,548, 5,837,458, 5,834,252, 5,830,721, 5,811,238, and 5,605,793 (all of which are incorporated by reference herein in their entirety) .
The art also provides even more recent modifications of the basic technique of DNA
shuffling. In one example, exon shuffling, exons or combinations of exons that encode specific domains of peptides are amplified using chimeric oligonucleotides.
The amplified molecules are then recombined by self priming PCR assembly (I~olkman and Stemmler, 2001, Nat. Biotech. 19:423-428). In another example, using the technique of random chimeragenesis on transient templates (RACHITT) library construction, single stranded parental DNA fragments are annealed onto a full-length single-stranded template (Coco et al., 2001, Nat. Biotechnol. 19:354-359). In yet another example, staggered extension process (StEP), thermocycling with very abbreviated annealing/extension cycles is employed to repeatedly interrupt DNA polymerization from flanking primers (Zhao et al., 1998, Nat.
Biotechnol. 16: 258-261). In the technique known as CLERY, in vitro family shuffling is combined with in vivo homologous recombination in yeast (Abecassis et al., 2000, Nucleic Acids Res. 28:E88; ). To maximize intergenic recombination, single stranded DNA from complementary strands of each of the nucleic acids are digested with DNase and annealed (Kikuchi et al., 2000, Gene 243:133-137). The blunt ends of two truncated nucleic acids of variable lengths that are linked by a cleavable sequence are then ligated to generate gene fusion without homologous recombination (Sieber et al., 2001, Nat Biotechnol.
19:456-460;
Lutz et al., 2001, Nucleic Acids Res. 29:E16; Ostermeier et al., 1999, Nat.
Biotechnol.
17:1205-1209; Lutz and Benkovic, 2000, Curr. Opin. Biotechnol. 11:319-324).
Recombination between nucleic acids with little sequence homology in common has also been enhanced using exonuclease-mediated blunt-ending of DNA fragments and ligating the fragments together to recombine them (U.S. Patent No. 6,361,974, incorporated herein by reference in its entirety). The invention contemplates the use of each and every variation described above as a means of enhancing the biological properties of any of the peptides and/or enzymes useful in the methods of the invention.
In addition to published protocols detailing directed evolution and gene shuffling techniques, commercial services are now available that will undertake the gene shuffling and selection procedures on peptides of choice. Ie~Iaxygen (Redwood City, CA) offers commercial services to generate custom DNA shuffled libraries. In addition, this company will perform customized directed evolution procedures including gene shuffling and selection on a peptide family of choice.
Optigenix, Inc. (Newark, DE) offers the related service of plasmid shuffling.
Optigenix uses families of genes to obtain mutants therein having new properties. The nucleic acid of interest is cloned into a plasmid in an Aspergillus expression system. The DNA of the related family is then introduced into the expression system and recombination in conserved regions of the family occurs in the host. Resulting mutant DNAs are then expressed and the peptide produced therefrom are screened for the presence of desired properties and the absence of undesired properties.
c. Screening procedures Following each recursive round of "evolution," the desired peptides expressed by mutated genes are screened for characteristics of interest. The "candidate"
genes are then amplified and pooled for the next round of DNA shuffling. The screening procedure used is highly dependant on the peptide that is being "evolved" and the characteristic of interest.
Characteristics such as peptide stability, biological activity, antigenicity, among others can be selected using procedures that are well known in the art. Individual assays for the biological activity of preferred peptides useful in the methods of the invention are described elsewhere herein.
d. Combinations of techniques It will be appreciated by the skilled artisan that the above techniques of mutation and selection can be combined with each other and with additional procedures to generate the best possible peptide molecule useful in the methods of the invention. Thus, the invention is not limited to any one method for the generation of peptides, and should be construed to encompass any and all of the methodology described herein. For example, a procedure for introducing point mutations into a nucleic acid sequence may be performed initially, followed by recursive rounds of DNA shuffling, selection and amplification. The initial introduction of point mutations may be used to introduce diversity into a gene population where it is lacking, and the following round of DNA shuffling and screening will select and recombine advantageous point mutations.
III. Glycosidases and Glycotransferases A. Gl~cosidases Glycosidases are glycosyltransferases that use water as an acceptor molecule, and as such, are typically glycoside-hydrolytic enzymes. Glycosidases can be used for the formation of glycosidic bonds in vitro by controlling the thermodynamics or kinetics of the reaction mixture. Even with modified reaction conditions, though, glycosidase reactions can be difficult to work with, and glycosidases tend to give low synthetic yields as a result of the reversible transglycosylase reaction and the competing hydrolytic reaction.
A glycosidase can function by retaining the stereochemistry at the bond being broken during hydrolysis or by inverting the stereochemistry at the bond being broken during hydrolysis, classifying the glycosidase as either a "retaining" glycosidase or an "inverting"
glycosidase, respectively. Retaining glycosidases have two critical carboxylic acid moieties present in the active site, with one carboxylate acting as an acid/base catalyst and the other as a nucleophile, whereas with the inverting glycosidases, one carboxylic acid functions as an acid and the other functions as a base.
Methods to determine the activity and linkage specificity of any glycosidase are well known in the art, including a simplified HPLC protocol (Jacob and Scudder, 1994, Methods in Enzymol. 230: 280-300). A general discussion of glycosidases and glycosidase treatment is found in Glycobiology, A Practical Approach, (1993, Fukuda and Kobata eds., Oxford University Press Inc., New York).
Glycosidases useful in the invention include, but are not limited to, sialidase, galactosidase, endoglycanase, mannosidase (i.e., a and (3, ManI, ManII and ManIII,) xylosidase, fucosidase, Agrobacterium sp. (3-glucosidase, C'ellulonaonas fifni mannosidase 2A, Humicola insolens glycosidase, Sulf~l~bus s~lfataricus glycosidase and ~aeillus licheni~'~f~fnis glycosidase.
The choice of fucosidases for use in the invention depends on the linkage of the fucose to other molecules. The specificities of many a-fucosidases useful in the methods of the invention are well known to those in the art, and many varieties of fucosidase are also commercially available (Glyko, Novato, CA; PR~zyme, San Leandro, CA;
Calbiochem-Novabiochem Corp., San Diego, CA; among others). a-Fucosidases of interest include, but are not limited to, a-fucosidases from Tur°b~ c~r~rzuta~s, C'har~~nia lanapas, Bacillus fulrrzinans, ~lsracr~~illus niter', Clostridium per~f~iragens, Bovine kidney (Glyko), chicken liver (Tyagarajan et al., 1996, Glycobiology 6:83-93) and a-fucosidase II from Xanthonronas rnanih~tis (Glyko, PROzyme). Chicken liver fucosidase is particularly useful for removal of core fucose from N-linked glycans.
B. Glycosyltransferases Glycosyltransferases catalyze the addition of activated sugars (donor NDP-sugars), in a step-wise fashion, to a protein, glycopeptide, lipid or glycolipid or to the non-reducing end of a growing oligosaccharide. N-linked glycopeptides are synthesized via a transferase and a lipid-linked oligosaccharide donor Dol-PP-NAG2G1c3Man9 in an en block transfer followed by trimming of the core. In this case the nature of the "core" saccharide is somewhat different from subsequent attachments. A very large number of glycosyltransferases are known in the art.
The glycosyltransferase to be used in the present invention may be any as long as it can utilize the modified sugar as a sugar donor. Examples of such enzymes include Leloir pathway glycosyltransferases, such as galactosyltransferase, N-acetylglucosaminyltransferase, N-acetylgalactosaminyltransferase, fucosyltransferase, sialyltransferase, mannosyltransferase, xylosyltransferase, glucurononyltransferase and the like.
For enzymatic saccharide syntheses that involve glycosyltransferase reactions, glycosyltransferases can be cloned, or isolated from any source. Many cloned glycosyltransferases are known, as are their polynucleotide sequences. See, e.g., Taniguchi et al., 2002, Handbook of glycosyltransferases and related genes, Springer, Tokyo.
Glycosyltransferase amino acid sequences and nucleotide sequences encoding glycosyltransferases from which the amino acid sequences can be deduced are also found in various publicly available databases, including GenBank, Swiss-Prot, EMBL, and others.

Glycosyltransferases that can be employed in the methods of the invention include, but are not limited to, galactosyltransferases, fucosyltransferases, glucosyltransferases, N-acetylgalactosaminyltransferases, N-acetylglucosaminyltransferases, glucuronyltransferases, sialyltransferases, mannosyltransferases, glucuronic acid transferases, galacturonic acid transferases, and oligosaccharyltransferases. suitable glycosyltransferases include those obtained from eukaryotes, as well as from prokaryotes.
DNA encoding glycosyltransferases may be obtained by chemical synthesis, by screening reverse transcripts of mRNA from appropriate cells or cell line cultures, by screening genomic libraries from appropriate cells, or by combinations of these procedures.
Screening of mRNA or genomic DNA may be carried out using oligonucleotide probes generated from the glycosyltransferases nucleic acid sequence. Probes may be labeled with a detectable label, such as, but not limited to, a fluorescent group, a radioactive atom or a chemiluminescent group in accordance with known procedures and used in conventional hybridization assays. In the alternative, glycosyltransferases nucleic acid sequences may be obtained by use of the polymerase chain reaction (PCR) procedure, with the PCR
oligonucleotide primers being produced from the glycosyltransferases nucleic acid sequence.
See, U.S. Pat. No. 4,683,195 to Mullis et al. and U.S. Pat. No. 4,683,202 to Mullis.
A glycosyltransferases enzyme may be synthesized in a host cell transformed with a vector containing DNA encoding the glycosyltransferases enzyme. A vector is a replicable DNA construct. Vectors are used either to amplify DNA encoding the glycosyltransferases enzyme and/or to express DNA which encodes the glycosyltransferases enzyme. An expression vector is a replicable DNA construct in which a DNA sequence encoding the glycosyltransferases enzyme is operably linked to suitable control sequences capable of effecting the expression of the glycosyltransferases enzyme in a suitable host. The need for such control sequences will vary depending upon the host selected and the transformation method chosen. Generally, control sequences include a transcriptional promoter, an optional operator sequence to control transcription, a sequence encoding suitable mRNA
ribosomal binding sites, and sequences which control the termination of transcription and translation.
Amplification vectors do not require expression control domains. All that is needed is the ability to replicate in a host, usually conferred by an origin of replication, and a selection gene to facilitate recognition of transformants.

1. Fucos~transferases In some embodiments, a glycosyltransferase used in the method of the invention is a fucosyltransferase. Fucosyltransferases are known to those of skill in the art. Exemplary fucosyltransferases include enzymes, which transfer L-fucose from GDP-fucose to a hydroxy position of an acceptor sugar. Fucosyltransferases that transfer from non-nucleotide sugars to an acceptor are also of use in the present invention.
In some embodiments, the acceptor sugar is, for example, the GIcNAc in a Gal(3(1 a3,4)GlcNAc(3- group in an oligosaccharide glycoside. Suitable fucosyltransferases for this reaction include the Gal(3(1~3,4)GIcNAc(31-a,(1-~3,4)fucosyltransferase (FTIII E.C.
No. 2.4.1.65), which was first characterized from human milk (see, Palcic, et al., Carbohydrate Res. 190: 1-11 (1989); Prieels, et al., J. Biol. Chem. 256: 10456-10463 (1981);
and Nunez, et al., Can. J. Chem. 59: 2086-2095 (1981)) and the Gal(3(1-~4)GIcNAc(3-afucosyltransferases (FTIV, FTV, FTVI) which are found in human serum. FTVII
(E.C. No.
2.4.1.65), a sialyl oc(2-~3)Gal(3((1-~3)GIcNAc[3 fucosyltransferase, has also been characterized. A recombinant form of the Gal/3(1 X3,4) GlcNAc(3-a(1~3,4)fucosyltransferase has also been characterized (see, Dumas, et al., Bioorg. Med.
Letters l: 425-428 (1991) and Kukowska-Latallo, et al., Genes and Development 4: 1288-1303 (1990)). Other exemplary fucosyltransferases include, for example, a,1,2 fucosyltransferase (E.C. No. 2.4.1.69). Enzymatic fucosylation can be carried out by the methods described in Mollicone, et al., Eur. J. Biochem. 191: 169-176 (1990) or U.S. Patent No. 5,374,655.
2. GalactosYltransferases In another group of embodiments, the glycosyltransferase is a galactosyltransferase.
Exemplary galactosyltransferases include a(1,3) galactosyltransferases (E.C.
No. 2.4.1.151, see, e.g., Dabkowski et al., Transplant Proc. 25:2921 (1993) and Yamamoto et al. Nature 345:
229-233 (1990), bovine (GenBank j04989, Joziasse et al., J. Biol. Chem. 264:

(1989)), marine (GenBank m26925; Larsen et al., Proc. Nat'1. Acad. Sci. USA
86: 8227-8231 (1989)), porcine (GenBank L36152; Strahan et al., Immunogenetics 41: 101-105 (1995)).
Another suitable ccl,3 galactosyltransferase is that which is involved in synthesis of the blood group B antigen (EC 2.4.1.37, Yamamoto et al., J. Biol. Chem. 265: 1146-1151 (1990) (human)).
Also suitable for use in the methods ofthe invention are (3(1,4) galactosyltransferases, which include, for example, EC 2.4.1.90 (LacNAc synthetase) and EC 2.4.1.22 (lactose synthetase) (bovine (D'Agostaro et al., Eur. J. Biochem. 183: 211-217 (1989)), human (Masri et al., Biochem. Biophys. Res. Commun. 157: 657-663 (1988)), murine (Nakazawa et al., J.
Biochem. 104: 165-168 (1988)), as well as E.C. 2.4.1.38 and the ceramide galactosyltransferase (EC 2.4.1.45, Stahl et al., J. Neurosci. Res. 38: 234-242 (1994)). Other suitable galactosyltransferases include, for example, a1,2 galactosyltransferases (from e.g., Schizosaccharomyces pombe, Chapell et al., Mol. Biol. Cell 5: 519-528 (1994)).
For further suitable galactosyltransferases, see Taniguchi et al. (2002, Handbook of Glycosyltransferases and Related Genes, Springer, Tokyo), Guo et al. (2001, Glycobiology, 11(10):813-820), and Breton et al. (1998, J Biochem. 123:1000-1009).
The production of proteins such as the enzyme GaINAc TI_~v from cloned genes by genetic engineering is well known. See, e.g., U.S. Pat. No. 4,761,371. One method involves collection of sufficient samples, then the amino acid sequence of the enzyme is determined by N-terminal sequencing. This information is then used to isolate a cDNA
clone encoding a full-length (membrane bound) transferase which upon expression in the insect cell line S~
resulted in the synthesis of a fully active enzyme. The acceptor specificity of the enzyme is then determined using a semiquantitative analysis of the amino acids surrounding known glycosylation sites in 16 different proteins followed by i~c vitro glycosylation studies of synthetic peptides. This work has demonstrated that certain amino acid residues are overrepresented in glycosylated peptide segments and that residues in specific positions surrounding glycosylated serine and threonine residues may have a more marked influence on acceptor efficiency than other amino acid moieties.
3. Sialyltransferases Sialyltransferases are another type of glycosyltransferase that is useful in the recombinant cells and reaction mixtures of the invention. Examples of sialyltransferases that are suitable for use in the present invention include ST3Ga1 III (e.g., a rat or human ST3Ga1 III), ST3Ga1 IV, ST3Ga1 I, ST6Gal I, ST3Ga1 V, ST6Ga1 II, ST6GalNAc I, ST6GaINAc II, and ST6GalNAc III (the sialyltransferase nomenclature used herein is as described in Tsuji et al., Glycobiology 6: v=xiv (1996)). An exemplary oc(2,3)sialyltransferase referred to as oc(2,3)sialyltransferase (EC 2.4.99.6) transfers sialic acid to the non-reducing terminal Gal of a Gal[31 ~3Glc disaccharide or glycoside. See9 Van den Eijnden et a1.9 J.
Biol. Chem. 256:
3159 (1981), Weinstein et al., J. Biol. Chem. 257: 13845 (1982) and Wen et al., J. Biol.
Chem. 267: 21011 (1992). Another exemplary cc2,3-sialyltransferase (EC
2.4.99.4) transfers sialic acid to the non-reducing terminal Gal of the disaccharide or glycoside.
see, Rearick et al., J. Biol. Chem. 254: 4444 (1979) and Gillespie et al., J. Biol. Chem. 267:
21004 (1992).
Further exemplary enzymes include Gal-(3-1,4-GIcNAc oc-2,6 sialyltransferase (See, I~urosawa et al. Eur. J. Biochem. 219: 375-381 (1994)).
Preferably, for glycosylation of carbohydrates of glycopeptides the sialyltransferase will be able to transfer sialic acid to the sequence Gal[31,4G1cNAc-, Gal(31,3G1cNAc-, or Gal[31,3GalNAc-, the most common penultimate sequences underlying the terminal sialic acid on fully sialylated carbohydrate structures (see, Table 8). a2,8-Sialyltransferases capable of transfering sialic acid to a2,3Ga1(31,4G1cNAc are also useful in the methods of the invention.
Table 8. Sialyltransferases which use the Gal(31,4G1cNAc sequence as an acceptor substrate SialyltransferaseSource Sequences) formed Ref.

ST6GalI Mammalian NeuAca2,6Ga1(31,4G1cNAc-1 ST3GalIII Mammalian NeuAca2,3Ga1(31,4G1cNAc-1 NeuAca2,3Ga1(31,3G1cNAc-ST3GalIV Mammalian NeuAca2,3Ga1(31,4G1cNAc-1 NeuAca2,3Ga1(31,3G1cNAc-ST6GalII Mammalian NeuAca2,6Ga1(31,4G1cNAc-ST6GalII PhotobacteriumNeuAca2,6Ga1(31,4G1cNAc-2 ST3Ga1 V N.~ meningitidesNeuAca2,3Ga1(31,4G1cNAc-3 N. goyaorrlaoeae 1) Goochee et al., Bio/Technology 9: 1347-1355 (1991) 2) Yamamoto et al., J. Biochem. 120: 104-110 (1996) 3) Gilbert et al., J. Biol. Chem. 271: 28271-28276 (1996) An example of a sialyltransferase that is useful in the claimed methods is ST3Ga1 III, which is also referred to as a(2,3)sialyltransferase (EC 2.4.99.6). This enzyme catalyzes the transfer of sialic acid to the Gal of a Gal(31,3G1cNAc or Gal~i1,4G1cNAc glycoside (see, e.g., Wen et al., J. Biol. Chem. 267: 21011 (1992); Van den Eijnden et al., J. Biol.
Chem. 256:
3159 (1991)) and is responsible for sialylation of asparagine-linked oligosaccharides in glycopeptides. The sialic acid is linked to a Gal with the formation of an a-linkage between the two saccharides. Bonding (linkage) between the saccharides is between the 2-position of NeuAc and the 3-position of Gal. This particular enzyme can be isolated from rat liver (Weinstein et al., J. Biol. Chem. 257: 13845 (1982)); the human cDNA (Sasaki et al. (1993) J. Biol. Chem. 268: 22782-22787; I~itagawa & Paulson (1994) J. Biol. Chem.
269: 1394-1401) and genomic (I~itagawa et al. (1996) J. Biol. Chem. 271: 931-938) DNA
sequences are known, facilitating production of this enzyme by recombinant expression. In a preferred embodiment, the claimed sialylation methods use a rat ST3Ga1 III.
An example of a sialyltransferase that is useful in the claimed methods is CST-I from Campylobacter (see ,for example, U.S. Pat. No. 6,503744, 6,096,529, and 6,210933 and W099/49051, and published U.S. Pat. Application 2002/2,042,369). This enzyme catalyzes the transfer of sialic acid to the Gal of a Gal(31,4G1c or Gal(31,3Ga1NAc.
Other exemplary sialyltransferases of use in the present invention include those isolated from Campylobacter jejuni, including the a(2,3) sialyltransferase. See, e.g, W099/49051.
Other sialyltransferases, including those listed in Table 8, are also useful in an economic and efficient large-scale process for sialylation of commercially important glycopeptides. As a simple test to find out the utility of these other enzymes, various amounts of each enzyme (1-100 mU/mg protein) are reacted with asialo-al AGP
(at 1-10 mg/ml) to compare the ability of the sialyltransferase of interest to sialylate glycopeptides relative to either bovine ST6Ga1 I, ST3Ga1 III or both sialyltransferases.
Alternatively, other glycopeptides or glycopeptides, or N-linked oligosaccharides enzymatically released from the peptide backbone can be used in place of asialo-al AGP for this evaluation.
Sialyltransferases with the ability to sialylate N-linked oligosaccharides of glycopeptides more efficiently than ST6Ga1 I are useful in a practical large-scale process for peptide sialylation (as illustrated for ST3Ga1 III in this disclosure).

4. Other ~lycosyltransferases One of skill in the art will understand that other glycosyltransferases can be substituted into similar transferase cycles as have been described in detail for the sialyltransferase. In particular, the glycosyltransferase can also be, for instance, glucosyltransferases, e.g., AlgB (Stagljov et al., Proc. Natl. Acad. Sci. USA
91: 5977 (1994)) or AlgS (Heesen et al., Eur. J. Biochem. 224: 71 (1994)).
N-acetylgalactosaminyltransferases are also of use in practicing the present invention.
Suitable N-acetylgalactosaminyltransferases include, but are not limited to, a(1,3) N-acetylgalactosaminyltransferase, (3(1,4) N-acetylgalactosaminyltransferases (Nagata et al., J.
Biol. Chem. 267: 12082-12089 (1992) and Smith et al., J. Biol Chem. 269: 15162 (1994)) and peptide N-acetylgalactosaminyltransferase (Homa et al., J. Biol. Chem.
268: 12609 (1993)). Suitable N-acetylglucosaminyltransferases include GnT-I (2.4.1.101, Hull et al., BBRC 176: 608 (1991)), GnT-II, GnT-III (Ihara et al., J. Biochem. 113: 692 (1993)), GnT-IV, GnT-V (Shoreibah et al., J. Biol. Chem. 268: 15381 (1993)) and GnT-VI, O-linked N-acetylglucosaminyltransferase (Bierhuizen et al., Proc. Natl. Acad. Sci. USA
89: 9326 (1992)), N-acetylglucosamine-1-phosphate transferase (Rajput et al., Biochem J. 285: 985 (1992), and hyaluronan synthase.
Mannosyltransferases are of use to transfer modified mannose moieties.
Suitable mannosyltransferases include a(1,2) mannosyltransferase, a(1,3) mannosyltransferase, a(1,6) mannosyltransferase, (3(1,4) mannosyltransferase, Dol-P-Man synthase, OChl, and Pmtl (see, Kornfeld et al., Annu. Rev. Biochem. 54: 631-664 (1985)).
Xylosyltransferases are also useful in the present invention. See, for example, Rodgers, et al., Biochem. J., 288:817-822 (1992); and Elbain, et al., U.S.
Patent No., 6,168,937.
Other suitable glycosyltransferase cycles are described in Ichikawa et al., JACS 114:
9283 (1992), Wong et al., J. Org. Chem. 57: 4343 (1992), and Ichikawa et al.
in CARBOHYDRATES AND CARBOHYDRATE POLYMERS. Yaltami, ed. (ATL Press, 1993).
Prokaryotic glycosyltransferases are also useful in practicing the invention.
Such glycosyltransferases include enzymes involved in synthesis of lipooligosaccharides (LOS), which are produced by many gram negative bacteria. The LOS typically have terminal glycan sequences that mimic glycoconjugates found on the surface of human epithelial cells or in host secretions (Preston et al., Critical Reviews in Microbiology 23(3):
139-180 (1996)).
such enzymes include, but are not limited to, the proteins of the rfa operons of species such as E. c~li and Salm~nella typhimut~iurfa, which include a (31,6 galactosyltransferase and ~. X31,3 galactosyltransferase (see, e.g., EMBL Accession Nos. M80599 and M86935 (E.
a~li);
EMBL Accession No. 556361 (S. typlziraauriurrz)), a glucosyltransferase (Swiss-Prot Accession No. P25740 (E. c~li), ~an (31,2-glucosyltransferase (rfaJ)(Swiss-Prot Accession No.
P27129 (E. c~li) and Swiss-Prot Accession No. P19817 (S typhimuriurn)), and an (31,2-N-acetylglucosaminyltransferase (~faI~)(EMBL Accession No. U00039 (E. coli).
Other glycosyltransferases for which amino acid sequences are known include those that are encoded by operons such as ~faB, which have been characterized in organisms such as Klebsiella pneumoniae, E. coli, Salmonella typhimur~ium, Salmonella euterica, Ye~si~cia enterocolitica, Mycobacterium leprosum, and the rhl operon of Pseudomonas aerwgi~cosa.
Also suitable for use in the present invention are glycosyltransferases that are involved in producing structures containing lacto-N-neotetraose, D-galactosyl-[3-1,4-N-acetyl-D-glucosaminyl-(3-1,3-D-galactosyl-(3-1,4-D-glucose, and the Pk blood group trisaccharide sequence, D-galactosyl-a-1,4-D-galactosyl-(3-1,4-D-glucose, which have been identified in the LOS of the mucosal pathogens Neisse~ia gonno~hoeae and N.
meniv~gitidis (Scholten et al., J. Med. Microbiol. 41: 236-243 (1994)). The genes from N.
meningitidis and N. go~corrhoeae that encode the glycosyltransferases involved in the biosynthesis of these structures have been identified from N. meningitidis immunotypes L3 and L1 (Jennings et al., Mol. Microbiol. 18: 729-740 (1995)) and the N. gonorrhoeae mutant F62 (Gotshlich, J. Exp.
Med. 180: 2181-2190 (1994)). In N. meningitidis, a locus consisting of three genes, lgtA, lgtB and lg E, encodes the glycosyltransferase enzymes required for addition of the last three of the sugars in the lacto-N neotetraose chain (Wakarchuk et al., J. Biol.
Chem. 271: 19166-73 (1996)). Recently the enzymatic activity of the lgtB and ZgtA gene product was demonstrated, providing the first direct evidence for their proposed glycosyltransferase function (Wakarchuk et al., J. Biol. Chem. 271(45): 28271-276 (1996)). In N.
gofzort~hoeae, there are two additional genes, lgtD which adds (3-D-GaINAc to the 3 position of the terminal galactose of the lacto-N neotetraose structure and lgtC which adds a terminal a-D-Gal to the lactose element of a truncated LOS, thus creating the Pk blood group antigen structure (Gotshlich (1994), supra.). In N. mehingitidis, a separate immunotype L1 also expresses the Pk blood group antigen and has been shown to carry an lgtC gene (Jennings et al., (1995), supf~a.). Neisser°ia glycosyltransferases and associated genes are also described in USPN
5,545,553 (Gotschlich). Genes for oel,2-fucosyltransferase and X1,3-fucosyltransferase from ~lelic~bacte~ pyl~r°i has also been characterized (Martin et al., J.
Biol. Chem. 272: 21349-21356 (1997)). Also of use in the present invention are the glycosyltransferases of Campyl~bactef° jejuni (see, Taniguchi et al., 2002, Handbook of glycosyltransferases and related genes, Springer, Tokyo).
B. Sulfotransferases The invention also provides methods for producing peptides that include sulfated molecules, including, for example sulfated polysaccharides such as heparin, hepaxan sulfate, carragenen, and related compounds. Suitable sulfotransferases include, for example, chondroitin-6-sulphotransferase (chicken cDNA described by Fukuta et al., J.
Biol. Chem.
270: 18575-18580 (1995); GenBank Accession No. D49915), glycosaminoglycan N-acetylglucosamine N-deacetylase/N-sulphotransferase 1 (Dixon et al., Genomics 26: 239-241 (1995); UL18918), and glycosaminoglycan N-acetylglucosamine N-deacetylase/N-sulphotransferase 2 (marine cDNA described in Orellana et al., J. Biol. Chem.
269: 2270-2276 (1994) and Eriksson et al., J. Biol. Chem. 269: 10438-10443 (1994); human cDNA
described in GenBank Accession No. U2304).
C. Cell-Bound Glycosyltransferases In another embodiment, the enzymes utilized in the method of the invention are cell-bound glycosyltransferases. Although many soluble glycosyltransferases are known (see, for example, U.S. Pat. No. 5,032,519), glycosyltransferases are generally in membrane-bound form when associated with cells. Many of the membrane-bound enzymes studied thus far are considered to be intrinsic proteins; that is, they are not released from the membranes by sonication and require detergents for solubilization. Surface glycosyltransferases have been identified on the surfaces of vertebrate and invertebrate cells, and it has also been recognized that these surface transferases maintain catalytic activity under physiological conditions.
However, the more recognized function of cell surface glycosyltransferases is for intercellular recognition (Roth, 1990, Molecular Approaches to Supracellular Phenomena,).

Methods have been developed to alter the glycosyltransferases expressed by cells.
For example, Larsen et al., Proc. Natl. Acad. Sci. USA 86: 8227-8231 (1989), report a genetic approach to isolate cloned cDNA sequences that determine expression of cell surface oligosaccharide structures and their cognate glycosyltransferases. A cDNA
library generated from mI~NA isolated from a marine cell line known to express UDP-galactose:.(3.-D-galactosyl-1,4-N-acetyl-D-glucosaminide oc-1,3-galactosyltransferase was transfected into COS-1 cells. The transfected cells were then cultured and assayed for ce 1-3 galactosyltransferase activity.
Francisco et al., Proc. Natl. Acad. Sci. USA 89: 2713-2717 (1992), disclose a method of anchoring (3-lactamase to the external surface of Esche~ichia coli. A
tripartite fusion consisting of (i) a signal sequence of an outer membrane protein, (ii) a membrane-spanning section of an outer membrane protein, and (iii) a complete mature (3-lactamase sequence is produced resulting in an active surface bound [3-lactamase molecule. However, the Francisco method is limited only to prokaryotic cell systems and as recognized by the authors, requires the complete tripartite fusion for proper functioning.
D. Fusion Enzymes In other exemplary embodiments, the methods of the invention utilize fusion peptides that have more than one enzymatic activity that is involved in synthesis of a desired glycopeptide conjugate. The fusion peptides can be composed of, for example, a catalytically active domain of a glycosyltransferase that is joined to a catalytically active domain of an accessory enzyme. The accessory enzyme catalytic domain can, for example, catalyze a step in the formation of a nucleotide sugar that is a donor for the glycosyltransferase, or catalyze a reaction involved in a glycosyltransferase cycle. For example, a polynucleotide that encodes a glycosyltransferase can be joined, in-frame, to a polynucleotide that encodes an enzyme involved in nucleotide sugar synthesis. The resulting fusion peptide can then catalyze not only the synthesis of the nucleotide sugar, but also the transfer of the sugar moiety to the acceptor molecule. The fusion peptide can be two or more cycle enzymes linked into one expressible nucleotide sequence. In other embodiments the fusion ppeptide includes the catalytically active domains of two or more glycosyltransferases. See, for example, U.S.
Patent No. 5,641,668. The modified glycopeptides of the present invention can be readily designed and manufactured utilizing various suitable fusion peptides (see, for example, PCT
Patent Application PCT/CA98/O11 ~0, which was published as WO 99/31224 on June 24, 1999.) E. Immobilized Enzymes In addition to cell-bound enzymes, the present invention also provides for the use of enzymes that are immobilized on a solid and/or soluble support. In an exemplary embodiment, there is provided a glycosyltransferase that is conjugated to a PEG via an intact glycosyl linker according to the methods of the invention. The PEG-linker-enzyme conjugate is optionally attached to solid support. The use of solid supported enzymes in the methods of the invention simplifies the work up of the reaction mixture and purification of the reaction product, and also enables the facile recovery of the enzyme. The glycosyltransferase conjugate is utilized in the methods of the invention. Other combinations of enzymes and supports will be apparent to those of skill in the art.
F. Muta~enesis of Gl~yltransferases The novel forms of the glycosyltransferases, sialyltransferases, sulfotransferases, and any other enzymes used in the method of the invention can be created using any of the methods described previously, as well as others well known to those in the art. Of particular interest are transferases with altered acceptor specificity and/or donor specificity. Also of interest are enzymes with higher conversion rates and higher stability among others.
The techniques of rational design mutagenesis can be used when the sequence of the peptide is known. Since the sequences as well as many of the tertiary structures of the transferases and glucosidases used in the invention are known, these enzymes are ideal for rational design of mutants. For example, the catalytic site of the enzyme can be mutated to alter the donor and/or acceptor specificity of the enzyme.
The extensive tertiary structural data on the glycosyltransferases and glycosidase hydrolases also make these enzyme idea for mutations involving domain exchanges.
Glycosyltransferases and glycosidase hydrolases are modular enzymes (see, Bourne and Henrissat, 2001, Current Opinion in Structural Biology 11:593-600).
Glycosyltransferases are divided into two families bases on their structure: GT-A and GT-B. The glycosyltransferases of the GT-A family comprise two dissimilar domains, one involved in nucleotide binding and the other in acceptor binding. Thus, one could conveniently fuse the DNA sequence encoding the domain from one gene in frame with a domain from a second gene to create a new gene that encodes a protein with a new acceptor/donor specificity. Such exchanges of domains could additionally include the carbohydrate modules and other accessory domains.
The techniques of random mutation and/or directed evolution, as described above, may also be used to create novel forms of the glycosyltransferases and glycosidases used in the invention.
IV. In vitro and in vivo expression systems A. Cells for the t~roduction of ~lycopeptides The action of glycosyltransferases is key to the glycosylation of peptides, thus, the difference in the expression of a set of glycosyltransferases in any given cell type affects the pattern of glycosylation on any given peptide produced in that cell. For a review of host cell dependent glycosylation of peptides, see Kabata and Takasaki, "Structure and Biosynthesis of Cell Surface Carbohydrates," in Cell Surface Carbohydrates and Cell Development, 1991, pp.
1-24, Eds. Minoru Fukuda, CRC Press, Boca Raton, FL.
According to the present disclosure, the type of cell in which the peptide is produced is relevant only with respect to the degree of remodeling required to generate a peptide having desired glycosylation. For example, the number and sequence of enzymatic digestion reactions and the number and sequence of enzymatic synthetic reactions that are required in vitro to generate a peptide having desired glycosylation will vary depending on the structure of the glycan on the peptide produced by a particular cell type. While the invention should in no way be construed to be limited to the production of peptides from any one particular cell type including any cell type disclosed herein, a discussion of several cell systems is now presented which establishes the power of the present invention and its independence of the cell type in which the peptides are generated.
In general, and to express a peptide from a nucleic acid encoding it, the nucleic acid must be incorporated into an expression cassette, comprising a promoter element, a terminator element, and the coding sequence of the peptide operably linked between the two.
The expression cassette is then operably linked into a vector. Toward this end, adapters or linkers may be employed to join the nucleotide fragments or other manipulations may be involved to provide for convenient restriction sites, removal of superfluous nucleotides, removal of restriction sites, or the like. For this purpose, in vitro mutagenesis, primer repair, restriction, annealing, resubstitutions, e.g., transitions and transversions, may be involved. A
shuttle vector has the genetic elements necessary for replication in a cell.
Some vectors may be replicated only in prokaryotes, or may be replicated in both prokaryotes and eukaryotes.
Such a plasmid expression vector will be maintained in one or more replication systems, preferably two replication systems, that allow for stable maintenance within a yeast host cell for expression purposes, and within a prokaryotic host for cloning purposes.
Many vectors with diverse characteristics are now available commercially. Vectors are usually plasmids or phages, but may also be cosmids or mini-chromosomes. Conveniently, many commercially available vectors will have the promoter and terminator of the expression cassette already present, and a multi-linker site where the coding sequence for the peptide of interest can be inserted. The shuttle vector containing the expression cassette is then transformed in E. coli where it is replicated during cell division to generate a preparation of vector that is sufficient to transform the host cells of the chosen expression system. The above methodology is well know to those in the art, and protocols by which to accomplish can be found Sambrook et al.
(2001, Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, New York).
The vector, once purified from the cells in which it is amplified, is then transformed into the cells of the expression system. The protocol for transformation depended on the kind of the cell and the nature of the vector. Transformants are grown in an appropriate nutrient medium, and, where appropriate, maintained under selective pressure to insure retention of endogenous DNA. Where expression is inducible, growth can be permitted of the yeast host to yield a high density of cells, and then expression is induced. The secreted, mature heterologous peptide can be harvested by any conventional means, and purified by chromatography, electrophoresis, dialysis, solvent-solvent extraction, and the like.
The techniques of molecular cloning are well-known in the art. Further, techniques for the procedures of molecular cloning can be found in Sambrook et al. (2001, Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.); Glover et al., (1985, DNA Cloning: A Practical Approach, Volumes I and II); Gait et al., (1985, Oligonucleotide Synthesis); Hames and Higgins (1985, Nucleic Acid Hybridization ); Hames and Higgins (1984, Transcription And Translation);
Freshney et al., (1986, Animal Cell Culture); Perbal, (1986, Immobilized Cells And Enzymes, IRL
Press);
Perbal,(1984, A Practical Guide To lblolecular Cloning); Ausubel et al. (2002, Current Protocols in Molecular Biology, John Wiley ~ Sons, Inc.).
B. Fungi and yeast Peptides produced in yeast are glycosylated and the glycan structures present thereon are primarily high mannose structures. In the case of N-glycans, the glycan structures produced in yeast may contain as many as nine or more mannose residues which may or may not contain additional sugars added thereto. An example of the type of glycan on peptides produced by yeast cells is shown in Figure 4, left side. Irrespective of the number of mannose residues and the type and complexity of additional sugars added thereto, N-glycans as components of peptides produced in yeast cells comprise a trimannosyl core structure as shown in Figure 4. When the glycan structure on a peptide produced by a yeast cell is a high mannose structure, it is a simple matter for the ordinary skilled artisan to remove, in vitro using available mannosidase enzymes, all of the mannose residues from the molecule except for those that comprise the trimannosyl core of the glycan, thereby generating a peptide having an elemental trimannosyl core structure attached thereto. Now, using the techniques available in the art and armed with the present disclosure, it is a simple matter to enzymatically add, in vitro, additional sugar moieties to the elemental trimannosyl core structure to generate a peptide having a desired glycan structure attached thereto. Similarly, when the peptide produced by the yeast cell comprises a high mannose structure in addition to other complex sugars attached thereto, it is a simple matter to enzymatically cleave off all of the additional sugars, including extra mannose residues, to arrive at the elemental trimannosyl core structure. Once the elemental trimannosyl core structure is produced, generation of a peptide having desired glycosylation is possible following the directions provided herein.
By "yeast" is intended ascosporogenous yeasts (Endomycetales), basidiosporogenous yeasts, and yeast belonging to the Fungi Imperfecti (Blastomycetes). The ascosporogenous yeasts are divided into two families, Spermophthoraceae and Saccharomycetaceae. The later is comprised of four subfamilies, Schizosaccharvrnycoideae (e.g., genus Schizosaccha~~myces), Nadso~ioideae, Lip~mycvideae, and Sacchar~ornycoideae (e.g., genera DEMANDE OU BREVET VOLUMINEUX
LA PRESENTE PARTIE DE CETTE DEMANDE OU CE BREVET COMPREND
PLUS D'UN TOME.

NOTE : Pour les tomes additionels, veuillez contacter 1e Bureau canadien des brevets JUMBO APPLICATIONS/PATENTS
THIS SECTION OF THE APPLICATION/PATENT CONTAINS MORE THAN ONE
VOLUME

NOTE: For additional volumes, please contact the Canadian Patent Office NOM DU FICHIER / FILE NAME
NOTE POUR LE TOME / VOLUME NOTE:

Claims

1. A cell-free, in vitro method of remodeling a peptide comprising poly(ethylene glycol), the peptide having the formula:

wherein AA is a terminal or internal amino acid residue of the peptide;
X1-X2 is a saccharide covalently linked to the AA, wherein X1 is a first glycosyl residue; and X2 is a second glycosyl residue covalently linked to X1, wherein X1 and X2 are selected from monosaccharyl and oligosaccharyl residues;
the method comprising:
(a) removing X2 or a saccharyl subunit thereof from the peptide, thereby forming a truncated glycan.

2. The method according to claim 1 wherein said truncated glycan is formed by removing a Sia residue.

3. The method according to claim 1 wherein said peptide has the formula:

wherein X3, X4, X5, X6, X7, and X17, are independently selected monosaccharyl or oligosaccharyl residues; and a, b, c, d, e, and x are independently selected from the integers 0, 1 and 2.

4. The method according to claim 3 wherein said oligosaccharyl residue is a member selected from GlcNAc-Gal-Sia and GlcNAc-Gal.

5. The method according to claim 3 wherein at least one member selected from a, b, c, d, a and x is 1 or 2.

6. The method of claim 3, wherein said removing of step (a) produces a truncated glycan in which at least one of a, b, c, a and x are 0.

7. The method of claim 6, wherein X3, X5 and X7 are members independently selected from (mannose)z and (mannose)z-(X8) wherein X8 is a glycosyl moiety selected from mono- and oligo-saccharides; and z is an integer between 1 and 20, wherein when z is 3 or greater, each (mannose)z is independently selected from linear and branched structures.

8. The method of claim 6 wherein X4 is selected from the group consisting of GlcNAc and xylose.

9. The method of claim 6, wherein X3, X5 and X7 are (mannose)u wherein a is selected from the integers between 1 and 20, and when a is 3 or greater, each (mannose)u is independently selected from linear and branched structures.

10. The method according to claim 3 wherein said peptide has the formula:

wherein r, s, and t are integers independently selected from 0 and 1.

11. The method of claim 1, wherein said peptide has the formula:

wherein X9 and X10 are independently selected monosaccharyl or oligosaccharyl residues; and m, n and f are integers independently selected from 0 and 1.

12. The method of claim 11, wherein said peptide has the formula:

wherein X16 is a member selected from:

wherein s and i are integers independently selected from 0 and 1.

13. The method of claim 12, wherein said peptide has the formula:

wherein X13, X14, and X15 are independently selected glycosyl residues; and g, h, i, j, k, and p are independently selected from the integers 0 and 1

14. The method according to claim 13 wherein at least one of g, h, i, j, k and p is 1.

15. The method of claim 13, wherein X14 and X15 are members independently selected from GlcNAc and Sia; and i and k are independently selected from the integers 0 and 1.

16. The method according to claim 15 wherein at least one of i and k is 1, and if k is 1, g, h, and j are 0.

17. The method according to claim 1, further comprising:
(b) contacting the truncated glycan with at least one glycosyltransferase and at least one glycosyl donor under conditions suitable to transfer the at least one glycosyl donor to the truncated glycan, thereby remodeling said peptide comprising poly(ethylene glycol).

18. The method according to claim 17 wherein said glycosyl donor comprises a modifying group covalently linked thereto.

19. The method of claim 1, further comprising:
(c) removing X1, thereby exposing AA.

20. The method according to claim 19, further comprising:
(d) contacting AA with at least one glycosyltransferase and at least one glycosyl donor under conditions suitable to transfer said at least one glycosyl donor to AA, thereby remodeling said peptide comprising poly(ethylene glycol).

21. The method according to claim 20 wherein said at least one glycosyl donor comprises a modifying group covalently linked thereto.

22. The method according to claim 21 wherein said modifying group is poly(ethylene glycol).

23. The method according to claim 22 wherein said poly(ethylene glycol) has a molecular weight distribution that is essentially homodisperse.

24. The method of claim 17, further comprising:
(e) prior to step (b), removing a group added to said saccharide during post-translational modification.

25. The method of claim 24 wherein said group is a member selected from phosphate, sulfate, carboxylate and esters thereof.

26. The method of claim 1 wherein said peptide has the formula:

wherein Z is a member selected from O, S, NH and a cross-linker.

27. The method of claim 1, wherein said peptide has the formula:

wherein X11 and X12 are independently selected glycosyl moieties; and r and x are integers independently selected from 0 and 1.

28. The method of claim 27, wherein X11 and X12 axe (mannose)q, wherein q is selected from the integers between 1 and 20, and when q is three or greater, (mannose)q is selected from linear and branched structures.

29. A pharmaceutical composition comprising a pharmaceutically acceptable diluent and a remodeled peptide according to claim 1.

30. A cell-free, in vitro method of remodeling a peptide comprising poly(ethylene glycol), said peptide having the formula:

wherein AA is a terminal or internal amino acid residue of said peptide;
X1 is a glycosyl residue covalently linked to said AA, selected from monosaccharyl and oligosaccharyl residues; and a is an integer selected from 0 and 1, said method comprising:
contacting said peptide with at least one glycosyltransferase and at least one glycosyl donor under conditions suitable to transfer said at least one glycosyl donor to said truncated glycan, thereby remodeling said peptide.

31. The method according to claim 30 wherein said at least one glycosyl donor comprises a modifying group covalently linked thereto.

32. The method according to claim 30 wherein said modifying group is poly(ethylene glycol).

33. The method according to claim 32 wherein said poly(ethylene glycol) has a molecular weight distribution that is essentially homodisperse.

34. A pharmaceutical composition comprising a pharmaceutically acceptable diluent and a remodeled peptide according to claim 30.