AU2008202355B2 - Optimizing glycan processing in plants - Google Patents

Description

AUSTRALIA Patents Act 1990 COMPLETE SPECIFICATION STANDARD PATENT Invention Title: Optimizing glycan processing In plants The following statement is a full description of this Invention, Including the best method of performing it known to us: -1A OPTIMIZING GLYCAN PROCESSING IN PLANTS FIELD OF THE INVENTION The invention is directed to methods for optimizing glycan processing of cell or an organism 5 containing glycoproteins with N-glycans, in particular plants so that a glycoprotein having an N-glycan, high mannose type, hybrid or preferably complex type N-glyoans, including but not limited to bi-antennary N-glycans, and containing a galactose residue on at least one arm of the N glycan and which are devoid of (or reduced in) xylose and fucose residues can be obtained. The invention is father directed to said glycoprotein obtained and In particular a plant host system 10 comprising said protein. BACKGROUND OF THE INVENTION N-linked glycans, specific oligosacoharido structures attached to asparagine residues of glycoproteins,,can contribute significantly to the properties of the protein and, in tnm, to the 15 properties of the organism. Plant proteins can carry N-linked glycans but in marked contrast to mammals only few biological processes are known to which they contribute. Biogenesis ofN-linked glycans begins with the synthesis of a lipid linked oligosaccharide moiety (Glc3Man9GlcNAc2-) which is transferred en bloc to the nascent polypeptide chain in the endoplasmic reticulum (ER). Through a series of trimming reactions by exoglycosidases In the BR 20 and cis-Golgi compartments, the so-called "high mannose" (Man9GlcNAc2 to ManSGloNAc2) glycans are formed. Subsequently, the formation of complex type glycans starts with the transfer of the first GlcNAc onto Man5GQcNAc2 by GnTI and further trimming by mannosidase 11 (Man1) to forn GlcNAcMan3GlcNAc2. Complex glycan biosynthesis continues while the glycoprotein is progressing through the secretory pathway with the transfer in the Golgi apparatus of the second 25 CGlcNAc residue by GnTII as well as other monosaccharide residues onto the OleNAcMan3GlcNAc2 under the action of several other glycosyl transferases. Plants and mammals differ with respect to the formation of complex glycans (see Figure 1, which compares the glycosylation pathway of glycoproteins in plants and mammals). In plants, complex glycans are characterized'by the presence of p(1,2).xylose residues linked to the Man-3 30 and/or an a(1,3)-fucose residue linked to GlcNAc-l, instead of an a(1,6)-fucose residue linked to the GlcNAc-1. Genes encoding the corresponding xylosyl (XylT) and faoosyl (FuoT) transferases have been isolated [Strasser ot al., "Molecular cloning and functional expression of betal,2 xylosyltransferase cDNA from Arabidopsis thaliana," FEBS Lett. 472:105 (2000); Leiter ot al., "Purification, cDNA cloning, and expression of GDP-L-Fuc:Asn-linked GloNAo alpha 1,3 35 fucosyltransferase from mung beans," J Bio. Chem. 274;21830 (1999)]. Plants do not possess p(1,4)-galactosyltransferases nor c(2,6)sialyltransferases and consequently plant glycans lack the ((1,4)-galactose and terminal c,(2,6)NeuA residues often found on mammalian glycans.

The final glycan structures are not only determined by the mere presence of enzymes involved in their biosynthesis and transport but to a large extent by the specific sequence of the various enzymatic reactions. The latter is controlled by discrete sequestering and relative position of these enzymes throughout the ER and Golgi, which is mediated by the interaction of determinants of 5 the transferase and specific characteristics of the sub-Golgi compartment for which the transferase is destined. A number of studies using hybrid molecules have identified that the transmembrane domains of several glycosyltransferases, including that of p(1,4)-galactosyltransferases, play a central role in their sub-Golgi sorting [Grabenhorst et al, J. BIOl. Chem 274:36107 (1999); Colley, X., Glycobtology 7:1 (1997); Munro, S., Trends Cell Biol. 8:11 (1998); Gleeson, PA., istochem. 10 Cell Blol. 109:517 (1998)). Although plants and mammals have diverged a relatively long time ago, N-linked glycosylation seems at least partly conserved. This is evidenced by the similar though not identical glycan structures and by the observation that a mammalian GloNAcTI gene complements a Arabidopsis mutant that is deficient in GleNAcTI activity, and vice versa. The differences in glycan 15 structures can have important consequences. For example, xylose and a(1,3)-fucose epitopes are known to be highly immunogenic and possibly allergenic In some circumstances, which may pose a problem when plants are used for the production of therapeutic glycoproteins, Moreover, blood serum of many allergy patients contains IgE directed against these epitopes but also 50% of non allergic blood donors contains in their sera antibodies specific for core-xylose whereas 25% have 20 antibodies for core-alpha 1,3-fucose (Bardor et al., 2002, in press, Glycobiology) (Advance Access published December 17, 2002) which make these individuals at risk to treatments with recombinant proteins produced in plants containing facose and/or xylose, In addition, this carbohydrate directed IgE in sera might cause false positive reaction in in vitro tests using plant extracts since there is evidence that these carbohydrate specific IgB's are not relevant for the allergenic reaction. In sum, a 25 therapeutic failure with a glycoprotein produced in plants might be the result of accelerated clearance of the recombinant glycoprotein having xylose and/or facose. Accordingly, there is a need to better control glycosylation in plants, and particularly, glycosylation of glycoproteins intended for therapeutic use. 30 DEFINITIONS To facilitate understanding of the invention, a number of terms as used in this specification are defined below. The term "vector" refers to any genetic element, such as a plamid, phage, transposon, cosmid, chromosome, retrovirus, virion, or similar genetic element, which is capable of replication 35 when associated with the proper control elements and which can transfer gene sequences into cells and/or between cells. Thus, this term includes cloning and expression vehicles, as well as viral vectors.

The term "expression vector" as used herein refers to a recombinant DNA molecule containing a desired coding sequence (or coding sequences) - such as the coding sequence(s) for the hybrid enzyme(s) described In more detail below - and appropriate nucleic acid sequences necessary for the expression of the operably linked coding sequence In a particular host cell or organism. 5 Nucleic acid sequences necessary for expression in prokaryotes usually include a promoter, an operator (optional), and a ribosome binding site, often along with other sequences. Eukaryotic cells are known to utilize promoters, enhancers, and termination and polyadenylation signals. It is not intended that the present invention be limited to particular expression vectors or expression vectors with particular elements. 10 The term "transgenic" when used in reference to a cell retbrs to a cell which contains a transgene, or whose genome has been altered by the introduction of a trasgene. The term "transgenic" when used in reference to a cell, tissue or to a plant refers to a cell, tissue or plant, respectively, which comprises a transgene, where one or more cells of the tissue contain a transgene (such as a gene encoding the hybrid enzyme(s) of the present invention), or a plant whose genome 15 has been altered by the introduction of a transgene. Transgenic cells, tissues and plants may be produced by several methods Including the introduction of a "transgene" comprising nucleic acid (usually DNA) into a target cell or integration of the transgene into a chromosome of a target cell by way of human intervention, such as by the methods described herein. The term "transgene" as used herein refers to any nucleio acid sequence which is Introduced 20 into the genome of a cell by experimental manipulations. A transgene may be an endogenouss DNA sequence," or a "heterologous DNA sequence" (i.e., "foreign DNA"). The term "endogenous DNA sequence" refers to a nucleotide sequence which is naturally found in the cell into which it is introduced so long as it does not contain some modification (e.g., a point mutation, the presence of a selectable marker gene, or other like modifications) relative to the naturally-occurring sequence. 25 The term "heterologous DNA sequence" refers to a nucleotide sequence which is ligated to, or is manipulated to become ligated to, a nucleic acid sequence to which it is not ligated In nature, or to which it is ligated at a different location in nature. Heterologous DNA is not endogenous to the cell into which it is introduced, but has been obtained from another cell. Heterologous DNA also includes an endogenous DNA sequence which contains some modification. Generally, although not 30 necessarily, heterologous DNA encodes RNA and proteins that are not normally produced by the cell into which it is expressed. Examples of heterologous DNA include reporter genes, transcriptional and translational regulatory sequences, selectable marker proteins (e.g., proteins which confer drug resistance), or other similar elements. The term "foreign gene" refers to any nucleio acid (e.g., gene sequence) which is introduced 35 into the genome of a cell by experimental manipulations and may include gene sequences found in that cell so long as the introduced gene contains some modification (e.g., a point mutation, the presence of a selectable marker gone, or other like modifications) relative to the naturally-occurring gene.

-4 The term "fusion protein" refers to a protein wherein at least one part or portion is from a first protein and another part or portion is from a second protein. The term "hybrid enzyme" refers to a fusion protein which is a functional enzyme, wherein at least one part or portion is from a first species and another part or portion is from a second species, Preferred hybrid enzymes of the 5 present invention are functional glycosyltransferases (or portions thereof) wherein at least one part or portion is from a plant and another part or portion Is from a mammal (such as human). The term "introduction into a cell" or "introduction into a host cell" in the context of nucleic acid (e.g., vectors) Is intended to include what the art calls "transformation" or "transfection" or "transduction." Transformation of a cell may be stable or transient - and the present invention 10 contemplates Introduction of vectors under conditions where, on the one hand, there is stable expression, and on the other hand, where there is only transient expression. The term "transient transformation" or "transiently transformed" refers to the introduction of one or more transgenes into a cell in the absence of integration of the transgene into the host cell's genome. Transient transformation may be detected by, fur example, enzyme-linked immunosorbent assay (ELISA) 15 which detects the presence of a polypeptide encoded by one or more of the transgenes. Alternatively, transient transformation may be detected by detecting the activity of the protein (e.g., antigen binding of an antibody) encoded by the tranegene (e.g., the antibody gene). The term "transient transfbrmant" refers to a cell which has transiently incorporated one or more transgenes. In contrast, the term "stable transformation" or "stably transformed" refers to the Introduction and integration of 20 one or more transgenes into the genome of a cell. Stable transformation of a cell may be detected by Southern blot hybridization of genomic DNA ofthe cell with nuclei acid sequences which are capable of binding to one or more ofthe tranagenes. Alternatively, stable transformation of a cell may also be detected by the polymerase chain reaction (PCR) of genonio DNA of the cell to amplify transgene sequences. The term "stable transformant" refers to a cell which has stably integrated one 25 or more transgenes into the genomic DNA. Thus, a stable transformant is distinguished from a transient transformant in that, whereas genomic DNA from the stable transformant contains one or more transgenes, genomie DNA f'om the transient transformant does not contain a transgene. The term "host cell" includes both mammalian (e.g. human B cell clones, Chinese hamster ovary cells, hepatocytes) and non-mammalian cells (e.g. insect cells, bacterial cells, plant cells). In 30 one embodiment, the host cells are mammalian cells and the introduction of a vector expressing a hybrid protein of the present invention (e.g TmGnTII-GaiT) inhibits (or at least reduces) fuoosylation In said mammalian cells. The term "nucleotide sequence of interest" refers to any nucleotide sequence, the manipulation of which may be deemed desirable for any reason (e.g., confer improved qualities, use 35 for production of therapeutic proteins), by one of ordinary skill In the art, Such nucleotide sequences include, but are not limited to, coding sequences of structural genes (e.g., reporter genes, selection marker genes, oncogenes, antibody genes, drug resistance genes, growth factors, and other like genes), and non-coding regulatory sequences which do not encode an mRNA or protein product, -5 (eg., promoter sequence, polyadenylation sequence, termination sequence, enhancer sequence, and other like sequences). The present invention contemplates host cells expressing a heterologous protein encoded by a nucleotide sequence of interest along with one or more hybrid enzymes. The term "isolated" when used in relation to a nucleic acid, as in "an isolated nucleic acid 5 sequence" refers to a nucleic acid sequence that is Identified and separated from one or more other components (e.g,, separated from a cell containing the nucleic acid, or separated from at least one contaminant nucleic acid, or separated from one or more proteins, one or more lipids) with which it is ordinarily associated in its natural source. Isolated nucleic acid is nucleic acid present in a form or setting that is different from that in which It is found in nature. In contrast, non-isolated nucleic 10 acids are nucleic acids such as DNA and ANA which are found in the state they exist in nature. For example, a given DNA sequence (eg., a gene) is found on the host cell chromosome in proximity to neighboring genes; RNA sequences, such as a specific mRNA sequence encoding a specific protein, are found in the cell as a mixture with numerous other mRNAs which encode a multitude of proteins. However, an isolated nucleic acid sequence comprising SEQ ID NO:! includes, by way of 15 example, such nuclei acid sequences In cells which ordinarily contain SEQ ID NO: 1 where the nucleic acid sequence is in a chromosomal or extrachromosomal location different from that of natural cells, or is otherwise flanked by a different nucleic acid sequence than that found in nature. The isolated nucleic acid sequence may be present in single-stranded or double-stranded form. When an isolated nucleic acid sequence is to be utilized to express a protein, the nucleic acid 20 sequence will contain at a minimum at least a portion of the sense or coding strand (i.e., the nucleic acid sequence may be single-stranded). Alternatively, it may contain both the sense and and-sense strands (L.e., the nucleic acid sequence may be double-stranded). As used herein, the term "purified" refers to molecules, either nucleic or amino acid sequences, that are removed from their natural environment, Isolated or separated. An "isolated 25 nucleic acid sequence" is therefore a purified nucleic acid sequence. "Substantially purified" molecules are at least 60% free, preferably at least 75% free, and more preferably at least 90% free, from other components with which they are naturally associated. The present invention contemplates both purified (including substantially purified) and unpurified hybrid enzyme(s) (which are described in more detail below). 30 As used herein, the terms "complementary" or complementarityt" are used in reference to nucleotide sequences related by the base-pairing rules. For example, the sequence 5'-AGT-3' is complementary to the sequence 5'-ACT-3'. Complementarity can be "partial" or "total," "Partial" complementarity is where one or more nucleic acid bases is not matched according to the base pairing rules. "Total" or "complete" complementarity between nucleic acids is where each and every 35 nuclei acid base is matched with another base under the base pairing rules. The degree of complementarity between nucleic acid strands has significant effects on the efficiency and strength of hybridization between nucleic acid strands.

-6 A "complement" of a nucleic acid sequence as used herein refers to a nucleotide sequence whose nucleic acids show total complementarity to the nucleic acids of the nucleic acid sequence. For example, the present invention contemplates the complements of SEQ ID NOS: 1, 3, 5, 9, 27, 28, 29, 30, 31, 32, 33, 34, 35, 37, 38, 40, 41 and 43. 5 The term "homology" when used in relation to nucleic acids refrs to a degree of complementarity. There may be partial homology (I.e., partial identity) or complete homology (i.e., complete identity). A partially complementary sequence is one that at least partially inhibits a completely complementary sequence from hybridizing to a target nucleic acid and is referred to using the functional term "substantially homologous." The inhibition of hybridization of the 10 completely complementary sequence to the target sequence may be examined using a hybridization assay (Southern or Northern blot, solution hybridization and the like) under conditions of low stringency. A substantially homologous sequence or probe (i.e, an oligonucleotide which is capable of hybridizing to another oligonucleotide of interest) will compete for and inhibit the binding (i.e., the hybridization) of a completely homologous sequence to a target under conditions of low 15 stringency. This is not to say that conditions of low stringency are such that non-specific binding is permitted; low stringency conditions require that the binding of two sequences to one another be a specific (I.e., selective) Interaction. The absence of non-specizc binding may be tested by the use of a second target which lacks even a partial degree of complementarity (e.g., less than about 30% identity); in the absence of non-specific binding the probe will not hybridize to the second non. 20 complementary target. When used in reference to a double-stranded nucleic acid sequence such as a cDNA or genomic clone, the term "substantially homologous" refers to any probe which can hybridize to either or both strands of the double-stranded nucleic acid sequence under conditions of low stringency as described infra. 25 When used in roferenoe to a single-stranded nucleic acid sequence, the term "substantially homologous" refers to any probe which can hybridize to the single-stranded nuclei acid sequence under conditions of low stringency as described Infra. The term "hybridization" as used herein includes "any process by which a strand ofnucleic acid joins with a complementary strand through base pairing." [Coombs J (1994) DIctionary of 30 Biotechnology, Stockton Press, New York NY]. Hybridization and the strength of hybridization (i.e., the strength of the association between the nucleic acids) is impacted by such factors as the degree of complementarity between the nucleic acids, stringency of the conditions involved, the T. ofthe fbrncd hybrid, and the Q:C ratio within the nucleic acids. As used herein, the term "T." is used in reference to the "melting temperature." The melting 35 temperature is the temperature at which a population of double-stranded nucleic acid molecules becomes half dissociated into single strands. The equation for calculating the Tm ofnucleic acids is well known in the art. As indicated by standard references, a simple estimate of the T. value may be calculated by the equation: Tm = 81.5 +0.41 (% G + C), when a nucleic acid is in aqueous solution -7 at 1 M NaCI (see eg., Anderson and Young, Quantitative filter Hybridization, in: Nucleic Acid Hybridization (1985)]. Other references include more sophisticated computations which take structural as well as sequence characteristics into account for the calculation of Tm. Low stringency conditions when used in reference to nucleic acid hybridization comprise 5 conditions equivalent to binding or hybridization at 68'C in a solution consisting of 5X SSPE (Saline, Sodium Phosphate, EDTA) (43.8 g/I NaCl, 6.9 g/l NaH2POeI2Q and 1.85 g/l EDTA (Ethylenediamninetctracetic Acid), pH adjusted to 7.4 with NaOW), 0.1% SDS (Sodium dodecyl sulfate), 5X Denhardt's reagent [SOX Denhardt's contains the following per 500 ml: 5 g Picoll (Type 400, Pharmacia), 5 g BSA (Bovine Serum Albumin) (Fraction V; Sigma)] and 100 gg/ml denatured 10 salmon sperm DNA followed by washing in a solution comprising between 0.2X and 2.OX SSPE, and 0.1% SDS at room temperature when a DNA probe of about 100 to about 1000 nucleotides in length is employed. High stringency conditions when used in reference to nucleic acid hybridization comprise conditions equivalent to binding or hybridization at 68'C in a solution consisting of 5X SSPE, 1% 15 SDS, 5X Denhardt's reagent and 100 pg/ml denatured salmon sperm DNA followed by washing in a solution comprising 0.JX SSPE, and 0.1% SDS at 68"C when a probe of about 100 to about 1000 nucleotides In length is employed. The term "equivalent" when made in reference to a hybridization condition as it relates to a hybridization condition of interest means that the hybridization condition and the hybridization 20 condition of interest result in hybridization of nucleic acid sequences which have the same range of percent (%) homology. For example, if a hybridization condition of interest results in hybridization of a first nuclelo acid sequence with other nuclei acid sequences that have from 50% to 70% homology to the first nucleic acid sequence, then another hybridization condition is said to be equivalent to the hybridization condition of Interest if this other hybridization condition also results 25 in hybridization of the first nuclei acid sequence with the other nucleic acid sequences that have from 50% to 70% homology to the first nucleic acid sequence. When used in reference to nucleic acid hybridization the art knows well that numerous equivalent conditions may be employed to comprise either low or high stringency conditions; factors such as the length and nature (DNA, RNA, base composition) of the probe and nature of the target 30 (DNA, RNA, base composition, present in solution or immobilized) and the concentration of the salts and other components (c.g., the presence or absence of formamide, dextran sulfate, polyethylene glycol) are considered and the hybridization solution may be varied to generate conditions of either low or high stringency hybridization different from, but equivalent to, the above. listed conditions. 35 The term "promoter," "promoter element," or "promoter sequence" as used herein, refers to a DNA sequence which when ligated to a nucleotide sequence of interest is capable of controlling the transcription of the nucleotide sequence of interest into mRNA. A promoter is typically, though not necessarily, located 5' (,e., upstream) of a nucleotide sequence of interest whose transcription into -8 mRiNA it controls, and provides a site for specific binding by RNA polymerase and other transcription factors for initiation of transcription, Promoters may be tissue specific or cell specific. The term "tissue specific" as it applies to a promoter refers to a promoter that is capable of directing selective expression of a nucleotide 5 sequence of interest to a specific type of tissue (e,g., petals) in the relative absence of expression of the same nucleotide sequence of interest in a different type of tissue (e,g., roots). Tissue specificity of a promoter may be evaluated by, for example, operably linking a reporter gene to the promoter sequence to generate a reporter construct, introducing the reporter construct into the genome of a plant such that the reporter construct is integrated into every tissue of the resulting transgenic plant, 10 and detecting the expression of the reporter gene (e.g., detecting mRNA, protein, or the activity of a protein encoded by the reporter gene) in different tissues of the transgenic plant. The detection of a greater level of expression of the reporter gene in one or more tissues relative to the level of expression of the reporter gene in other tissues shows that the promoter is specific for the tissues in which greater levels of expression are detected. The term "cell type specific" as applied to a 15 promoter refers to a promoter which is capable of directing selective expression of a nucleotide sequence of interest in a specific type of cell in the relative absence of expression of the same nucleotide sequence of interest in a different type of cell within the same tissue. The term "cell type specific" when applied to a promoter also means a promoter capable of promoting selective expression of a nucleotide sequence of interest in a region within a single tissue. Cell type 20 specificity of a promoter may be assessed using methods well known in the art, e.g, immuno histochemical staining. Briefly, tissue sections are embedded in paraffin, and paraffin sections are reacted with a primary antibody which is specific for the polypeptide product encoded by the nucleotide sequence of interest whose expression is controlled by the promoter. A labeled (e.g., peroxidase conjugated) secondary antibody which is specific for the primay antibody is allowed to 25 bind to the sectioned tissue and specific binding detected (e.g., with avidin/biotin) by microscopy. Promoters may be constitutive or regulatable. The term "constitutive" when made in reference to a promoter means that the promoter is capable of directing transcription of an operably linked nucleic acid sequence in the absence of a stimulus (e,g,, heat shock, chemicals, light, or similar stimuli). Typically, constitutive promoters are capable of directing expression of a tranegene 30 in substantially any cell and any tissue. In contrast, a "regulatable" promoter Is one which is capable of directing a level of transcription of an operably linked nuclei acid sequence in the presence of a stimulus (e.g., heat shook, chemicals, light, or similar stimuli) which is different from the level of transcription of the operably linked nucleic acid sequence in the absence of the stimulus. The terms "infecting" and "infection" with a bacterium refer to co-incubation of a target 35 biological sample, (e.g., cell, tissue, plant part) with the bacterium under conditions such that nucleic acid sequences contained within the bacterium are introduced into one or more cells of the target biological sample, -9 The term "Agrobacterium" refers to a soil-borne, Gram-negative, rod-shaped phytopathogenic bacterium which causes crown gall. The term "Agrobacterium" includes, but is not limited to, the strains Agrobacterium tumfactens, (which typically causes crown gall in infected plants), and Agobacterum rhtzogens (which causes hairy root disease in infected host plants). 5 Infection of a plant cell with Agrobacterium generally results in the production of opines (e.g., nopalino, agropine, octopine) by the infected cell. Thus, Agrobacterum strains which cause production of nopaline (e.g., strain LBA4301, C58, A208) are referred to as "nopaline-type" Agrobacteria; Agrobacterium strains which cause production of actopine (e.g., strain LBA4404, Ach5, B6) are referred to as "octopine-type" Agrobacteria; and Agrobacterlum strains which cause 10 production of agropine (e.g., strain EHA105, EHAIOI, A281) are referred to as "agropine-typo" Agrobacteria. The terms "bombarding, "bombardment," and "biolistic bombardment" refer to the process of accelerating particles towards a target biological sample (e.g., cell, tissue, plant part - such as a leaf, or intact plant) to effect wounding of the cell membrane of a cell in the target biological sample 15 and/or entry of the particles into the target biological sample. Methods for biolistic bombardment are known in the art (e.g,, U.S. Patent Nos. 5,584,807 and 5,141,131, the contents of both are herein incorporated by reference), and are commercially available (e.g., the helium gas-driven microprojectile accelerator (PDS-1000/Hc) (BioRad). The term "microwounding" when made in reference to plant tissue refers to the introduction 20 of microscopic wounds in that tissue. Microwounding may be achieved by, for example, particle bombardment as described herein. The present invention specifically contemplates schemes for introducing nucleic acid which employ miorowounding. The term "organism" as used herein refers to all organisms and in particular organisms containing glycoproteins with n-linked glycans. 25 The term "plant" as used herein refers to a plurality of plant cells which are largely differentiated into a structure that is present at any stage of a plant's development. Such structures include, but are not limited to, a ftit, shoot, stem, root, leaf, seed, flower petal, or similar structure. The term "plant tissue" includes differentiated and undifferentiated tissues of plants including, but not limited to, roots, shoots, leaves, pollen, seeds, tumor tissue and various types of cells in culture 30 (e.g., single cells, protoplasts, embryos, callus, protocorm-like bodies, and other types of cells). Plant tissue may be inplanta, in organ culture, tissue culture, or cell culture. Similarly, "plant cells" may be cells in culture or may be part of a plant. Glycosyltransferases are enzymes that catalyze the processing reactions that determine the structures of cellular oligosaccharides, including the oligosaccharides on glycoproteins. As used 35 herein, "glycosyltransferase" is meant to Include mannosidases, even though these enzymes trim glycans and do not "transfer" a monosaccharide. Glyoosyltransferases share the feature of a type II membrane orientation, Each glycosyltransferase is comprised of an amino terminal cytoplasmic tail (shown for Illustration purposes below as a made up of a string of amino acids arbitrarily labeled "X" -10 - without Intending to suggest the actual size of the region), a signal anchor domain (shown below as made up of a string of amino acids labeled "'" for hydrophobic - without Intending to suggest the actual size of the domain and without intending to suggest that the domain is only made up of hydrophobic anino acids) that spans the membrane (referred to herein as a "transmembrane 5 domain"), followed by a luminal stem (shown below as made up of a string of amino acids arbitrarily labeled "S"- without intended to suggest the actual size of the region) or stalk region, and a carboxyetrminal catalytic domain (shown below as made up of a string of amino acids arbitrarily labeled "C" - without intending to suggest the actual size of the domain M2rXXXXXXHHHHHHHHSSSCCCCCCCC 10 Collectively, The Cytoplasmic Tail-Transmembrane-Stem Region or "CTS" (which has been underlined in the above schematic for clarity) can be used (or portions thereof) in embodiments contemplated by the present invention wherein the catalytic domain is exchanged or "swapped" with a corresponding catalytic domain from another molecule (or portions of such regions/domains) to create a hybrid protein. 15 For example, in a preferred embodiment, the present invention contemplates nucleic acid encoding a hybrid enzyme (as well as vectors containing such nucleio acid, host cells containing such vectors, and the hybrid enzyme Itself), said hybrid enzyme comprising at least a portion of a CTS region [e.g., the cytoplasmic tail ("C"), the transmembrane domain ('T'), the cytoplasmic tail 20 together with the transmembrane domain ("CT"), the transmombrane domain together with the stem ("TS"), or the complete CTS region] of a first glyoosyltransferase (e.g. plant glycosyltransferase) and at least a portion of a catalytic region of a second glycosyltransferase (e.g. mammalian glycosyltransferase). To create such an embodiment, the coding sequence for the entire CTS region (or portion thereof) may be deleted from nucleic acid coding for the mammalian glycosyltransfrrase 25 and replaced with the coding sequence for the entire CTS region (or portion thereof) of a plant glycosyltransferase. On the other hand, a different approach might be taken to create this embodiment; for example, the coding sequence for the entire catalytic domain (or portion thereof) may be deleted from the coding sequence for the plant glycosyltransferase and replaced with the coding sequence for the entire catalytic domain (or portion thereof) of the mammalian 30 glycosyltransferase. In such a case, the resulting hybrid enzyme would have the amino-terminal cytoplasmic tail of the plant glycosyltransferase linked to the plant glycosyltransferase transmembrane domain linked to the stem region of the plant glycosyltransferase in the normal manner of the wild-type plant enzyme - but the stem region would be linked to the catalytic domain of the mammalian glycosyltransferase (or portion thereof). 35 It is not intended that the present invention be limited only to the two approaches outlined above. Other variations in the approach are contemplated. For example, to create nucleic acid encoding a hybrid enzyme, said hybrid enzyme comprising at least a portion of a transmembrane region of a plant glycosyltransforase and at least a portion of a catalytic region of a mammalian -11 glycosyltransferase, one might use less than the entire coding sequence for the CTS region (e.g, only the transmembrane domain of the plant glycosytransferase, or the complete cytoplasmic tail together with all or a portion ofthe transmembrane domain, or the complete cytoplasmic tail together with all of the transmembrane domain together with a portion of the stem region). One might delete the 5 mammalian coding sequence for the entire cytoplasmic tail together with the coding sequence for the transmembrane domain (or portion thereof) - followed by replacement with the corresponding coding sequence for the cytoplasmic tail and transmembrane domain (or portion thereof) of the plant glycosyltransferase. In such a case, the resulting hybrid enzyme would have the stem region of the mammalian glycosyltransfbrase linked to the plant glycosyltransferase transmembrane domain (or 10 portion thereof) which In turn would be linked to the amino-terminal cytoplasmic tail of the plant glycosyltransferase, with the stem region being linked to the catalytic domain of the mammalian glycosyltransferase (i.e. two of the four regions/domains would be of plant origin and two would be of mammalian origin). Tn other embodiments, the present invention contemplates nucleic acid encoding a hybrid 15 enzyme (along with vectors, host cells containing the vectors, plants - or plant parts - containing the host cells), said hybrid enzyme comprising at least a portion of an amino-terminal cytoplasmio tail of a plant glycosyltransferase and at least a portion of a catalytic region of a mammalian glycosyltransferase. In this embodiment, the hybrid enzyme encoded by the nucleic acid might or might not contain other plant sequences (e.g., the transmembrane domain or portion thereof, the stem 20 region or portion thereof). For example, to create such an embodiment, the coding sequence for the entire cytoplasmic tail (or portion thereof) may be deleted from nucleic acid coding for the mammalian glycosyltransferase and replaced with the coding sequence for the entire cytoplasmic domain (or portion thereof) of a plant glycosyltransferase. In such a case, the resulting hybrid enzyme would have the amino-terminal cytoplasmic tail (or portion thereof) of the plant 25 glycosyltransferase linked to the mammalian glycosyltransferase transmembrane domain, which in turn Is linked to stem region of the mammalian glycosyltransferase, the stem region being linked to the catalytic domain of the mammalian glycosyltransferase. On the other hand, a different approach might be taken to create this embodiment; for example, the coding sequence for the entire catalytic domain (or portion thereof) may be deleted from the coding sequence for the plant 30 glycosyltransferase and replaced with the coding sequence for the entire catalytic domain (or portion thereof) of the mammalian glycosyltransferase. In such a case, the resulting hybrid enzyme would have the amino-terminal cytoplasmic tail of the plant glycosyltransferase linked to the plant glycosyltransferase transmembrane domain linked to the stern region of the plant glycosyltransferase in the normal manner of the wild-type plant enzyme - but the stem region would be linked to the 35 catalytic domain of the mammalian glycosyltransferase (or portion thereof). In the above discussion, the use of the phrase "or portion thereof' was used to expressly indicate that less than the entire region/domain might be employed in the particular case (e.g., a fragment might be used), For example, the eytoplasmic tail of glycosyltransferases ranges from -12 approximately 5 to 50 amino acids In length, and more typically 15 to 30 amino acids, depending on the particular transferase. A "portion" of the cytoplasmic tail region is herein defined as no fewer than four amino acids and can be as large as up to the full length of the region/domain less one amino acid. It is desired that the portion function in a manner analogous to the full length 5 region/domain - but need not function to the same degree. For example, to the extent the full-length cytoplasmic tail functions as a Golgi retention region or ER retention signal, it is desired that the portion employed in the above-named embodiments also function as a Golgi or ER retention region, albeit perhaps not as efficiently as the full-length region. Similarly, the transmembrane domain is typically 15-25 amino acids in length and made up 10 of primarily hydrophobic amino acids. A "portion" of the transmembrane domain is herein defined as no fewer than ten amino acids and can be as large as up to the full length of the region/domain (for the particular type of transferase) less one amino acid It is desired that the portion function in a manner analogous to the full length region/domain - but need not function to the same degree. For example, to the extent the full-length transmembrane domain functions as the primary Golgi 15 retention region or ER retention signal, it is desired that the portion employed in the above-named embodiments also function as a Golgi or ER retention region, albeit perhaps not as efficiently as the full-length region. The present invention specifically contemplates conservative substitutions to create variants of the wild-type tranamembrane domain or portions thereof. For example, the present invention contemplates replacing one or more hydrophobic amino acids (shown as "H" in the 20 schematic above) of the wild-type sequence with one or more different amino acids, preferably also hydrophobic amino acids. A portion of the catalytic domain can be as large as the full length of the domain less on amino acid. Where the catalytic domain Is from a betal,4-galactosyltransferase, it is preferred that the portion include at a minimum residues 345-365 which are believed to be involved in the 25 conformation conferring an oligosaccharide acceptor binding site (it is preferred that the portion include this region at a minimum and five to ten amino acids on either side to permit the proper conformation). The present Invention also Includes synthetic CTS regions and portions thereof. A "portion" of a CTS region must Include at least one (and may include more than one) entire domain (e,g., the 30 entire transmembrane domain) but less than the entire CTS region. Importantly, by using the term "CTS region" or "transmembrane domain" it is not intended that only wild type sequences be encompassed. Indeed, this invention is not limited to natural glycosyltransferases and enzymes involved in glycosylation, but also includes the use of synthetic enzymes exhibit the same or similar function. In one embodiment, wild type domains are changed 35 (e.g. by deletion, insertion, replacement and the like). Finally, by using the indicator "Tm" when referring to a particular hybrid (e.g, "TmXyl-), entire transmembrane/CTS domains (with or without changes to the wild-type sequence) as well as portions (with or without changes to the wild-type sequence) are intended to be encompassed; -13 SUMMARY OF THE INVENTION The present invention contemplates nucleic acid (whether DNA or RNA) encoding hybrid enzymes (or "fusion proteins"), vectors containing such nucleic acid, host cells (including but not 5 limited to cells in plant tissue and whole plants) containing such vectors an expressing the hybrid enzymes, and the isolated hybrid enzyme(s) themselves, In one embodiment, expression of said hybrid enzymes (or "fusion proteins") results in changes in glycosylation, such as, but not limited to, reduction of sugar moieties such as xylose, fucose, Lewis41' or other sugar structures that interfere with desired glycoform accumulation. In one embodiment, the present invention contemplates 10 nucleic acid encoding a hybrid enzyme, said hybrid enzyme comprising a CTS region (or portion thereof) of a glycosyltransferase (including but not limited to a plant glycosyltransferase) and a catalytic region (or portion thereof) of a non-plant glycosyltransfbrase (e.g., mammalian, fish, amphibian, fungal). It is preferred that, when expressed, the CTS region (or portion thereof) is linked (directly or indirectly) in operable combination to said catalytic region (or portion thereof). 15 The linking is preferably covalent and the combination is operable in that the catalytic region exhibits catalytic function (even if said catalytic function is reduced as compared to the wild-type enzyme). The linking can be direct in the sense that there are no intervening amino acids or other regions/domains. On the other hand, the linking can be indirect in that there are intervening amino acids (or other chemical groups) and/or other regions/domains between them. Of course, the nucleic 20 acid used to make the nucleic acid encoding the above-described hybrid enzyme(s) can be obtained enzymatically from a physical sequence (e.g. genomic DNA, a cDNA, and the like) or alternatively, made synthetically using a reference sequence (e.g. electronic or hardcopy sequence) as a guide. In a particular embodiment, the present invention contemplates nuclei acid encoding a hybrid enzyme, said hybrid enzyme comprising a transmemnbrane region (e.g., at least a 25 transmembrane region and optionally more of the CTS region) of a plant glycosyltransferase and a catalytic region (or portion thereof) of a non-plant (such as a mammalian) glycosyltransferase. Again, it is preferred that, when expressed, these regions are linked (directly or indirectly) in operable combination. In yet another embodiment, the present invention contemplates nucleic acid encoding a hybrid enzyme, said hybrid enzyme comprising a transmembrane domain (or portion 30 thereof) of a plant glycosyltransferase and a catalytic region (or portion thereof) of a mammalian glycosyltransferasc. Again, it is preferred that, when expressed, these regions are linked (directly or indirectly) in operable combination, It is not intended that the present invention be limited to particular transferases. In one embodiment, the plant glycosyltransferase is a xylosyltransferase. In another embodiment, the plant 35 glycosyltransferasc is a N-acetylglucosaminyltransferase. In another embodiment, the plant glycosyltransferase is a fucosyltransferase. In a preferred embodiment, the mammalian glycosyltnsferase is a human galaotosyltransferase (such as the human beta 1,4- -14 galactosyltransferase encoded by SEQ ID NO: 1 wherein the nucleotides encoding the transmembrane domain are deleted and replaced). In one embodiment, the invention contemplates a hybrid glycosyltransferase comprising a first fragment that comprises the transmembrane domain of a plant glycosyltransferase and a 5 second fragment that comprises the catalytic domain of a mammalian glycosyltransferase, wherein the hybrid glycosyltransferase is selected from the group consisting of (i) a hybrid glycosyltransferase in which the first fragment is from an N-acetylglucosaminyltransferase I (GnTI) and the second fragment is from a mannosidase II (ManII) or N acetylglucosaminyltransferase II (GnTII), (ii) a hybrid glycosyltransferase in which the first 10 fragment is from a mannosidase I (ManI) and the second fragment is from a GnTI, ManII, or GnTII, (iii) a hybrid glycosyltransferase in which the first fragment is from a fucosyltransferase (FucT) and the second fragment is from galactosyltransferase (GalT), (iv) a hybrid glycosyltransferase having the amino acid sequence of SEQ ID NO:10, and (v) a hybrid glycosyltransferase having the amino acid sequence of SEQ ID NO:4. 15 It is not intended that the present invention is limited to the use of a plant-derived glycosyltransferase CTS-domain and a human glycosyltransferase catalytic domain but also vice versa and the use of any CTS-domain of a glycosyltransferase in combination with the catalytic fragment of at least one other glycosyltransferase. Indeed, the present invention broadly contemplates, in one embodiment, nucleic acid encoding a hybrid enzyme, said hybrid enzyme 20 comprising a transmembrane region of a first glycosyltransferase and a catalytic region of a second glycosyltransferase. It is preferred that said first and second glycosyltransferases are from different species (and can be from a different genus or even from a different phylum). In one embodiment, said first glycosyltransferase comprises a plant glycosyltransferase. In another embodiment, said plant glycosyltransferase is a xylosyltransferase. In yet another embodiment, 25 said plant glycosyltransferase is a fucosyltransferase. In a preferred embodiment said second glycosyltransferase comprises a mammalian glycosyltransferase. In a particularly preferred embodiment, said mammalian glycosyltransferase is a human galactosyltransferase. It is not intended that the present invention be limited to circumstances where the first and second glycosyltransferases are plant and non-plant, respectively. In one embodiment, said 30 first glycosyltransferase comprises a first mammalian glycosyltransferase and said second glycosyltransferase comprises a second mammalian glycosyltransferase. In a preferred -14a embodiment, said first mammalian glycosyltransferase is a non-human glycosyltransferase and said second mammalian glycosyltransferase is a human glycosyltransferase. It is not intended that the present invention be limited to the type of vector. In one embodiment, the present invention contemplates an expression vector, comprising the nucleic 5 acid encoding the above-described hybrid enzyme. It is also not intended that the present invention be limited to the type of host cells. A variety of prokaryotic and eukaryotic host cells are commercially available for expressing proteins. In one embodiment, the present invention contemplates a host cell containing the vector comprising the nucleic acid encoding the above-described hybrid enzyme (with or without other 10 vectors or other nucleic acid encoding other hybrid enzymes or glycosyltransferases). In a preferred embodiment, the host cell is a plant cell. In a particularly preferred embodiment, the present invention contemplates a plant comprising such a host cell. In one embodiment, the invention contemplates a method, comprising introducing into a plant cell an expression vector of the invention, and culturing the plant cell under conditions 15 allowing expression of the hybrid glycosyltransferase encoded by the expression vector. In another embodiment, the present invention contemplates a method, comprising introducing into a plant host system (i) one or more nucleic acids each coding for a hybrid glycosyltransferase of the invention, and (ii) a nucleic acid coding for a heterologous glycoprotein, wherein each of (i) and (ii) is in operative linkage with a promoter, and collecting a 20 plant or a portion thereof that expresses the hybrid glycosyltransferase(s) and the heterologous protein. It is not intended that the present invention be limited by the method by which host cells are made to express the hybrid enzymes of the present invention. In one embodiment, the present invention contemplates a method, comprising: a) providing: i) a host cell (such as a plant cell, 25 whether in culture or as part of plant tissue or even as part of an intact growing plant), and ii) an expression vector comprising nucleic acid encoding a hybrid enzyme, said hybrid enzyme comprising at least a portion of a CTS region of a plant glycosyltransferase (e.g. the transmembrane domain) and at least a portion of a catalytic region of a mammalian glycosyltransferase; and b) -15 introducing said expression vector into said plant cell under conditions such that said hybrid enzyme is expressed. Again, it is not intended that the present invention be limited to particular transferases. In one embodiment, the plant glycosyitransferase used in the above-described method is a xylosyltransferase. In another embodiment, the plant glycosyltransferase is a N 5 acstylglucosaminyltransferase. In another embodiment, the plant glycosyltransferase is a fucosyltransferse. In a preftrred embodiment, the mammalian glycosyltransferase used in the above-described method is a human galactosyltransferase (such as the human beta 1,4 galactosyltransferase encoded by SEQ IDNO:1 wherein the nucleotides encoding the transmembrane domain are deleted and replaced) (or simply where the nucleotides of SEQ ID NO: 1 10 encoding the catalytic domain, or portion thereof, are taken and linked to nucleotides encoding the CTS region, or portion thereof, of a plant glycosyltransferase.). It is not intended that the present invention be limited to a particular scheme for controlling glycosylation of a heterologous protein using the hybrid enzymes described above. In one embodiment, the present invention contemplates a method, comprising: a) providing: 1) a host cell 15 (such as a plant cell), ii) a first expression vector comprising nucleic acid encoding a hybrid enzyme, said hybrid enzyme comprising at least a portion of a CTS region (e.g. at least a transmembrane domain) of a first (such as a plant) glycosyltransferase and at least a portion of a catalytic region of a second (such as a mammalian) glycosyltransferase, and ii) a second expression vector comprising nucleic acid encoding a heterologous glycoprotein; (or portion thereof, and b) introducing said first 20 and second expression vectors into said plant cell under conditions such that said hybrid enzyme and said heterologous protein are expressed. Alternatively, a single vector with nucleic acid encoding both the hybrid enzyme (or hybrid enzymes) and the heterologous glycoprotein might be used. Regardless of which method is used, the invention contemplates, in one embodiment, the additional step (c) of isolating the heterologous protein - as well as the isolated protein itself as a composition. 25 On the other hand, the present invention also contemplates introducing different vectors into different plant cells (whether they are cells In culture, part of plant tissue, or even part of an intact growing plant). In one embodiment, the present invention contemplates a method, comprising: a) providing: i) a first plant comprising a first expression vector, said first vector comprising nucleic acid encoding a hybrid enzyme (or encoding two or more hybrid enzymes), said hybrid enzyme 30 comprising at least a portion of a CTS region (e.g. the first approximately 40-60 amino acids of the N-terminus) of a plant glycosyltransferase and at least a portion of a catalytic region of a mammalian glycosyltransferase, and ii) a second plant comprising a second expression vector, said second vector comprising nucleio acid encoding a heterologous protein (or portion thereof); and crossing said first plant and said second plant to produce progeny expressing said hybrid enzyme and said heterologous 35 protein. Of course, such progeny can be isolated, grown up, and analyzed for the presence of each (or both) of the proteins. Indeed, the heterologous protein can be used (typically first purified substantially free of plant cellular material) therapeutically (e.g., administered to a human or animal, -16 whether orally, by intravenous, tranadermally or by some other route of administration) to treat or prevent disease. It is not intended that the present invention be limited to a particular heterologous protein. In one embodiment, any peptide or protein that is not endogenous to the host cell (or organism) is 5 contemplated. In one embodiment, the heterologous protein is an antibody or antibody fragment. In a particularly preferred embodiment, the antibody is a human antibody or "humanized" antibody expressed in a plant in high yield. "Humanized" antibodies are typically prepared from non-human antibodies (e.g. rodent antibodies) by taking the hypervariable regions (the so-called CDRs) of the non-human antibodies and "grafting" them on to human frameworks. The entire process can be 10 synthetic (provided that the sequences are known) and frameworks can be selected from a database of common human frameworks. Many times, there is a loss of affinity in the process unless either the framework sequences are modified or the CDRs are modified. Indeed, Increases in affinity can be revealed when the CDRs are systematically mutated (for example, by randomization procedures) and tested. 15 While the present invention Is particularly useful in the context ofheterologous proteins, in one embodiment, the hybrid enzymes of the present invention are used to change the glycosylation of endogenous proteins, La. proteins normally expressed by the host cell or organism. The present invention specifically contemplates the plants themselves. In one embodiment, the present invention contemplates a plant, comprising first and second expression vectors, said first 20 vector comprising nucleic acid encoding a hybrid enzyme, said hybrid enzyme comprising at least a portion of a CTS region (eg. the cytoplasmic tail together with at least a portion of the transmembrane domain) of a plant glycosyltransferase and at least a portion of a catalytic region of a mammalian glycosyltransferase, said second expression vector, said second vector comprising nucleic acid encoding a heterologous protein (or portion thereof). In a preferred embodiment, by 25 virtue of being expressed along with the hybrid enzyme (or hybrid enzymes) of the present invention, the heterologous protein displays reduced (10% to 99%) alpha 1,3 -fucosylation (or even no fucosylation), as compared to when the heterologous protein is expressed In the plant in the absence of the hybrid enzyme (or enzymes). In a preferred embodiment, by virtue of being expressed along with the hybrid enzyme (or hybrid enzymes) of the present invention, the heterologous protein 30 displays reduced (10% to 99%) xylosylation (or even no xylose), as compared to when the heterologous protein is expressed in the plant in the absence of the hybrid enzyme (or enzymes). In a preferred embodiment, by virtue of being expressed along with the hybrid enzyme (or hybrid enzymes) of the present invention, the heterologous protein displays both reduced fucose and xylose, as compared to when the heterologous protein is expressed in the plant in the absence of the hybrid 35 enzyme (or enzymes). It is not intended that the present invention be limited to a particular theory by which reduced fucose and/or xylose is achieved. Very little is known about the sub-Golgi sorting mechanism In plants. The mammalian specific P(1,4)-galactosyltransferase (GalT) has been used (see the -17 Examples below) as an excellent first marker to study this phenomenon since it generates glycan structures not normally found In plants. The glycan structures of plants that express galactosyltransferase has been compared with glycan structures from plants that express a chimeric galactosyltransfbrase of which the CTS domain is exchanged for that of a plant xylosyltransferas (or 5 portion thereof). The change in observed glycan structures show that the galactosyltransferase is, as in mammals, confined to a speciflo sub-compartnent oftho plant Golgi. Without limiting the invention to any particular mechanism, the sorting mechanism of plants and mammals are apparently conserved even to the extent that glycosyltransferases unknown to plants are routed to specific analogous location in the Golgi. This location is later in the Golgi than where the endogenous 10 xylosyl-, fucosyl- and GlcNAcTII (GnTII) transtbrases are located. The finding that N-glycans in these plants that express rolocalised variants of GaIT containing significantly less xylosc and fIoose is also of biotechnological relevance. For glycoproteins intended for therapeutic use in mammals, such as humans, the approach of certain embodiments of the present invention provides methods and compositions for controlling N-linked 15 glycosylation of glycoproteins in plants so that glycoprotein essentially free of xylose and fuoose and containing at least a bi-antennary N-glycans (but not limited to bi-antennary, also include tri antennary, and the like) and (at least one) galactose residue on at least one of the anus of the N glycan can be obtained. Hence, it is not intended that the present invention is limited to bi-antennary N-glycans but also includes bisected bi-antennary N-glycans, tri-antennary N-glyeans, and the like. 20 Furthermore, the invention is not limited to complex-type N-glycans but also includes hybrid-type N-glycans and other type N-glycans. The present invention contemplates such resulting glyco proteins. In addition, the methods and compositions of the present invention may be applicable for plants and non-plant systems where besides xylose, fuhose, Lewits^M type N-glycan modifications (p1-3-GalT, al-4-FucT, other) or other sugars, "interfere" with desired glycoform accumulation. 25 In one embodiment, the invention is directed to controlling N-linked glycosylation of plants by modulating the localization of enzymes involved in glycan biosynthesis in the Golgi apparatus. Specifically, embodiments of the invention are directed to a method of producing in a plant host system a glycoprotein having bi-antennary glycans and containing at least one galactose residues on at least one of the arms and which are devoid (or reduced in) of xylose and fucose, comprising:(a) 30 preventing (or inhibiting) addition of xylose and fucose on the core of the glycan of said glycoprotein and (b) adding one or preferably two galactose residues to said arms. Addition of xylose and fucose to said heterologous glycoproteri may be reduced or even prevented by introducing to said plant host system a nuclei acid encoding a hybrid enzyme comprising a CTS region (or portion thereof) of a protein, particularly an enzyme such as plant 35 xylosyItransferase and catalytic region (or portion thereof) of a galactosyltransferase not normally found in a plant, or a modified galactosyltransforase where its transmembrane portion has been removed and endoplasmic reticulum retention signal have been inserted, wherein said protein or enzyme acts earlier in the Golgi apparatus of a plant cell in said plant host system than said galactosyltransferase. It is preferred that the galactosyltransferase is a mammalian galactosyltransferase and in particular, a human galactosyltransferase. In a most specific embodiment, said galactosyltransferase is human P1,4 galactosyltransferase (GalT). In a preferred embodiment, said xylosyltransferase is a $1,2.xylosyltransferase. The exchange of the CTS region or 5 CTS fragment of a mammalian glycosyltransferase (such as a galactosyltransferase) by one from the group of enzymes that act earlier in the Golgi apparatus than galactosyltransferase including but not limited to those from of XyIT, FucT, GnT 1 , OnTIL GnTI, O nTIV, OnTV, OnTVI, Manf, Man! and ManUI results in strongly reduced amounts of glycans that contain the undosired xylose and fbcose residues (see Figure 2). In addition, galactosylation is improved and the diversity in glycans Is 10 reduced. While not limited to any particular mechanism, the increase in galactosylated glycans that carry neither xylose nor fucose is believed to be mainly attributed to the accumulation of GalGNMan5, GNMan5 or GalGNMan4. Also, galactosylation occurs on one glycan arm only. Apparently, the galactosylation earlier in the Golgi inhibits trimming of the said glycoforms by Mannosidase II (Manu) to GaIGNMan3. Also addition of the second GlcNAc by GloNAcTII 15 (OnTII) is inhibited. Therefore, in one embodiment, a further step is contemplated to obtain the desired glycoprotein that has both arms galactosylated and yet is essentially devoid ofxylose and fucose. Thus, In one embodiment, the method of the invention as noted above further comprises adding galactose residues to the arms of said glycoprotein (see Figure 3). In one embodiment of the 20 Invention, galactose residues are added onto both arms by introducing to said plant host system (a) a nucleic acid sequence encoding a first hybrid enzyme comprising the CTS region (or fragment, such as one including the transmembrane domain) of GnTI and the active domain (or portion thereof) of GnTII; (b) a nucleio acid sequence encoding the second hybrid enzyme comprising the CTS region (or fragment, such as one including the transmembrane of G TI and the active domain of Manfl and 25 (c) a nucleio acid sequence encoding a third hybrid enzyme comprising the CTS region (or fragment, such as one Including the transmembrane domain) of XylT and the active domain (or portion thereof) of human galactosyltransferse (TmXyl-GaIT). In another embodiment of the invention, galactose residues are added onto both arms by introducing to said plant host system (a) a nucleic acid sequence encoding a first hybrid enzyme comprising the CTS region (or fragment, such as one 30 including the transmembrane domain) of ManI and the active domain (or portion thereof) of GnTl; (b) a nucleic acid sequence encoding the second hybrid enzyme comprising the CTS region (or fragment, such as one including the transmembrane domain) of ManI and the active domain (or portion thereof) of GnTII; (c) a nucloic acid sequence encoding the third hybrid enzyme comprising the CTS region (or fragment, such as one including the transmembrane domain) of ManI and the 35 active domain (or portion thereof) of Manl!, and (d) a nucleic acid sequence encoding a fourth hybrid enzyme comprising the CTS region (or fragment, such as one including the transmermbrane domain) of XylT and the active domain (or portion thereof) of human galactosyltransferse (TmXyl Gall).

-19 It is not intended that the present invention be limited to particular combinations of hybrid enzymes or the number of such hybrid enzymes employed in a single cell, plant tissue or plant. In a preferred embodiment the present invention contemplates host cells expressing TmXyl-GalT plus TmGnTI-GnTII plus TmGnTI-ManU. In one embodiment of the invention, galactose residues are 5 added to said arms by introducing to said plant host system (a) a nucleic acid sequence encoding a first hybrid enzyme comprising a CTS region (or fragment thereof) of a protein, particularly an enzyme, including but not limited to N-acetylglucosaminyltransferase I (GuTI) and a catalytic region (or portion thereof) of a mannosidase U (ManI), wherein said enzyme acts earlier in the Golgi apparatus of a plant cell in said plant host system than said mannosidase II or modified mannosidase 10 II where its transmembrane portion has been deleted and endoplasmic reticulun retention signal have been inserted and (b) a nucloic acid sequence encoding a second hybrid enzyme comprising a CTS region (or fragment, such as one including the transmrembrane domain) of an enzyme including but not limited to N-acetyl-glucosaminyltransferase I (GnTI) and a catalytic region (or portion thereof) of a N-acetylglucosaminyl-transferasc U (GnTl), wherein said enzyme acts earlier in the 15 Golgi apparatus of a plant cell in said plant host system than saidN acetylglucosaminyl-transferaseHl (GnTH) or modified N-acetylglucosaminyltransferase II (GnTU1) where its transmembrane portion has been deleted and an cndoplasmic reticulum retention signal have been inserted. The sequences encoding N-acetylglucosaminyltransferases or mannosidase U or the said transmembrane fragments can originate form plants or from eukaryotic non-plant organisms (e.g., mammals). 20 In yet another preferred embodiment, the present invention contemplates a host cell expressing TmXyl-GaIT plus TmManI-GnTI plus TmManI-ManHI plus TmMan-OnTl, In another embodiment of the invention, galactose residues are added to said arms by introducing to said plant, host system (a) a nucleic acid sequence encoding a first hybrid enzyme comprising a CTS region (or fragment, such as one including the transmembrane domain) of a protein, particularly an enzyme, 25 including but not limited to Mannosldase I (Mani) and a catalytic region (or portion thereof) of a N acetylglucosaruinyltransferase I (GnTI), wherein said enzyme acts earlier in the Golgi apparatus of a plant cell in said plant host system than said N-acotylglucosaminyl-transtfrase I (OnTl) or modified N acctylglucosaminyltransferase I (GnTI) where its transmembrane portion has been deleted and endoplasmic reticulum retention signal have been inserted and (b) a nucleic acid sequence encoding 30 a second hybrid enzyme comprising a CTS region (or fragment, such as one including the transmembrane domain) of an enzyme including but not limited to Mannosidase I (ManI) and a catalytic region (or portion thereof) of a Mannosidase H (ManlI), wherein said enzyme acts earlier in the Golgi apparatus of a plant call in said plant host system than said Mannosidase II (ManH) or modified Mannosidase f (ManII) where its transmembrane portion has been deleted and an 35 endoplasmic reticulum retention signal have been inserted and (c) a nucleic acid sequence encoding a third hybrid enzyme comprising a CTS region (or fragment, such as one including the transmembrane domain) of an enzyme including but not limited to Mannosidase I (Man!) and a catalytic region (or portion thereof) of a N-acetylglucos-amlnyltransferase H (OnTE), wherein said -20 enzyme acts earlier in the Golgi apparatus of a plant cell in said plant host system than said N acetylglucosaminyltransferase 1 (OnTIl) or modified N-acetylglucosaminyltransferase U (UnTII) where its transmembrane portion has been deleted and an endoplasmic reticulum retention signal have been inserted. The sequences encoding N-acetylgluoosaminyltransferases or mannosidases or 5 the said transmembrano fragments can originate from plants or from eukaryotic non-plant organisms (e.g., mammals). In still another preferred embodiment, the present invention contemplates host cells expressing TmXyl-GalT plus ManlI, In another embodiment ofthe invention, galactose residues are added to said arms by introducing to said plant host system (a) a nucleic acid sequence encoding 10 a Mannosidase III (ManlH, wildtype gone sequence but not limited to: also Manfl with endoplasmic reticulum retention signal; Man1I with transmembrane fragment of early (cis-) Golgi apparatus glycosyltransferase (OnTI, Mani, OnTJII). The sequences encoding Mannosidase III oan originate form insects, preferably from Spodopterafrugiperda or Drosophila melanogaster (but not limited to), human or from other organisms, 15 In still another preferred embodiment the present invention contemplates a host cell expressing TXyl-GalT plus ManIH plus TmGnTI-GnTI, In yet another preferred embodiment, the present invention contemplates a host cell expressing TmXyl-GalT plus Manll plus TmManI GnTI plus TmManI-UnTH. The method of the invention may optionally comprise, In one embodiment, introducing into 20 said plant host system a mammalJan N-acetylglucosaminyltransferase GnTMl, particularly a human GnTHI or hybrid protein comprising a catalytic portion of mammalian OnTIUI and a transmembranc portion of a protein, said protein residing in the ER or earlier compartment of the Golgi apparatus of a eukaryotic cell. For example, in one embodiment, the hybrid enzyme TmXyl-GnTfIl is contemplated (along with nucleic acid coding for such a hybrid enzyme, vectors containing such 25 nucleic acid, host cells containing such vectors, and plants - or plant parts - containing such host cells). In another embodiment, the hybrid enzyme TmFuc-GnTI is contemplated (along with nucleic acid coding for such a hybrid enzyme, vectors containing such nucleic acid, host cells containing such vectors, and plants - or plant parts - containing such host cells). The present invention specifically contemplates host cells expressing such hybrid enzymes (with or without 30 additional hybrid enzymes or other glycosyltmnsforases). The invention is farher directed to said hybrid and modified enzymes, nucleic acid sequences encoding said hybrid enzymes, vectors comprising said nucleic acid sequences and methods for obtaining said hybrid enzymes. Furthermore, the invention is directed to a plant host system comprising a hetorologous glycoprotein having preferably complex type bi-antennary glycans 35 and containing at least one galactose residue on at least one of the arms and are devoid of xylose and fuoose. A "heterologous glycoprotein" is a glycoprotein originating from a species other than the -21 plant host system. The glycoprotein may include but is not limited to antibodies, hormones, growth factors and growth factor receptors and antigens. Indeed, the present invention is particularly useful for controlling the glycosylation of heterologous glycoproteins, such as antibodies or antibody fragments (single chain antibodies, Fab 5 fragments, Fab 2 fragments, Fv fragments, and the like). To control the glycosylation of an antibody, the gene construct encoding a hybrid enzyme of the present invention (e.g., the TmXyl-GalT gene construct) can be introduced in transgenic plants expressing an antibody (eg., monoclonal antibody) or antibody fragment. On the other hand, the gene(s) encoding the antibody (or antibody fragment) can be introduced by retransformation of plant expressing TmXyl-GalT gene construct. In still 10 another embodiment, the binary vector harbouring the TriXyl-GaIT expression cassette can be co transformed to plants together with a plant binary vector harbouring the expression cassettes comprising both light and heavy chain sequences of a monoclonal antibody on a single T-DNA or with binary vectors harbouring the expression cassettes for light and heavy chain sequences both separately on independent T-DNA's but both encoding a monoclonal antibody. The present 15 invention specifically contemplates, in one embodiment, crossing plants expressing antibodies with plant expressing the hybrid glycosyltransferaso(s) of the present invention. A "host system" may include but is not limited to any organism containing glycoproteins with N-glycans. A "plant host system" may include but is not limited to a plant or portion thereof, which 20 includes but is not limited to a plant cell, plant organ and/or plant tissue, The plant may be a monocotyledon (monocot) which is a flowering plant whose embryos have one cotyledon or seed leaf and includes but is not limited to lilies, grasses, corn (Zea mays), rice, grains Including oats, wheat and barley, orchids, irises, onions and palms. Alternatively, the plant may be a dicotyledenon (dicot) which includes but is not limited to tobacco (Nicorfana), tomatoes, potatoes, legumes (e.g, 25 alfalfa and soybeans), roses, daises, cacti, violets and duckweed. The plant may also be a moss which includes but is not limited to Physcomitrella patens. The invention is further directed to a method for obtaining said plant host system. The method comprises crossing a plant expressing a heterologous glycoprotein with a plant comprising (a) a hybrid enzyme comprising a catalytic region (or portion thereof) of a 30 galactosyltransferase not normally found in a plant and a CTS region (or fragment, such as one including the transmembrane domain) of a protein, wherein said protein acts earlier in the Golgi apparatus of a plant cell in said plant host system than said galactosyltransferase or a modified galactosyltransferase where its transmembranc portion has been deleted and endoplasmic reticulum retention signal has been inserted; (b) a hybrid enzyme comprising a CTS region (or portion thereof, 35 such as one including the transmembrane domain) of a protein, particularly an enzyme, including but not limited to N-aoetylglucosaminyltransferase I (GnTI) and a catalytic region (or portion thereof) of a mannosidase 1D (ManIl), wherein said enzyme acts earlier in the Golgi apparatus of a plant cell In said plant host system than said mannosidase II or modified mannosidase II where Its transmenmbrane -22 portion has been deleted and endoplasmic reticulum retention signal have been inserted and (a) a hybrid enzyme comprising at least a transmembrane region of an enzyme (such as the first 40-60 amino acids of the N-terminus) of a glyeosyltransferase including but not limited to N acetylglucosaminyltransferase I (OnTi) and a catalytic region of a N-acetylglucos-aminyltransferase 5 11 (GnTI), wherein said enzyme acts earlier in the Golgi apparatus of a plant cell in said plant host system than said N acetylglucosaminyltransferase 11 (GnTil) or modified N-acetylglucosaminyl transferase 11 (GnTIl) where its transmembrane portion has been deleted and an endoplasmic reticulum retention signal have been inserted., harvesting progeny ftom said crossing and selecting a desired progeny plant expressing said heterologous glycoprotein. 10 The Invention is further directed to said plant or portion thereof which would constitute a plant host system. Said plant host system may further comprise a mammalian GnTI enzyme or hybrid protein comprising a catalytic portion of mammalian OnTIIH and a tranamembrane portion of a protein, said protein residing in the ER or earlier compartment of the Golgi apparatus of a eukaryotic cell. 15 Additionally, the invention also provides the use of a plant host system to produce a desired glycoprotein or functional fragment thereof. The invention additionally provides a method for obtaining a desired glycoprotein or functional fragment thereof comprising cultivating a plant according to the invention until said plant has reached a harvestable stage, for example when sufficient biomass has grown to allow profitable harvesting, followed by harvesting said plant with 20 established techniques known in the art and fractionating said plant with established techniques known in the art to obtain fractionated plant material and at least partly isolating said glycoprotein from said fractionated plant material. Alternatively, said plant host cell system comprising said heterologous glycoprotein may also be obtained by Introducing into a plant host cell system or portion thereof (a) a nucleic acid 25 sequence encoding a hybrid enzyme comprising a catalytic region of a galactosyltransferase not normally found In a plant and at least the transmembrane region (or more of the CTS) of a protein, wherein said protein acts earlier in the Golgi apparatus of a plant cell in said plant host system than said galactosyltransferase or a modified galactosyltransferase where its transmembrane portion has been deleted and endoplasmic reticulum retention signal have been inserted; (b) a nucleic acid 30 sequence encoding a first hybrid enzyme comprising at least the transmernbrane region (or more of the CTS if desired) of a protein, particularly an enzyme, including but not limited to N acetylglucosaminyltransferase I (GnTI) and a catalytic region of a mannosidase 11 (Manil), wherein said enzyme acts earlier in the Golgi apparatus of a plant cell in said plant host system than said rnannosidase II, or modified mannusidase 11 where its transmembrane portion has been deleted and 35 endoplasmic reticulum retention signal have been inserted and (o) a nucleic acid sequence encoding a second hybrid enzyme comprising at least a transmembrane region (more of the CTS if desired) of an enzyme including but not limited to N-acetylglucosaminyl-transferase I (GnTI) and a catalytic region of a N-acetylglucosaminyltransferase H (GnTIl), wherein said enzyme acts earlier in the -23 Golgi apparatus of a plant cell In said plant host system than said N- acetylglucos-aminyltransferase 11 (OnTI) or modified N-acetylglucosaminyltransferase 1 (OnTil) where its transmombrane portion has been deleted and an endoplasmic reticulum retention signal have been inserted. and isolating a plant or portion thereof expressing said heterologous glycoprotein (or portion thereof). In one 5 embodiment, one vector comprising all of the nucleio acid sequences is introduced into said plant host system. In another embodiment, each nucleic acid sequence is inserted Into separate vectors and these vectors are introduced into said plant host system. In another embodiment combinations of two or more nucleic acid sequences are inserted into separate vectors which are than combined into said plant host system by retransformation or co-transformatlon or by crossing. 10 The invention also provides use of such a plant-derived glycoprotein or functional fragment thereof according to the invention for the production of a composition, particularly, pharmaceutical composition, fbr example for the treatment of a patient with an antibody, a hormone, a vaccine antigen, an enzyme, or the like. Such a pharmaceutical composition comprising a glycoprotein or functional f-agment thereof is now also provided. 15 Finally, it is contemplated that the above-described approach may be useful in reducing the overall diversity In glycans in plants expressing one or more of the hybrid enzymes of the present Invention (as compared to wild-type plants or plants simply transformed with only mammalian GalT). 20 BRIEF DESCRITON OF THE FIGURES Figure 1 compares the glycosylation pathway of glycoproteins in plants and in mammals. Figure 2 shows the effect of exchanging the CTS fragment of galactosyltransferase with xylosyltransferase Figure 3 shows the further effect of relocalizing mannosidase II and GlcNAcTh. 25 Figure 4 top panel shows a T-DNA construct carrying the genes encoding glycan modifying enzymes to produce efficiently galactosylated glycans that are devoid of immunogenic xylose and fucose and the bottom panel shows a T-DNA construct carrying antibody light chain and heavy chain genes. Figure 5 shows the nucleic acid sequence (SEQ ID NO:1) for a human galactosyltransferase 30 (human B I,4-galactosyltransfrase - GaIT). Figure 6 shows the nucleic acid sequence of Figure 5 along with the corresponding amino acid sequence (SEQ ID NO:2). Figure 7 shows an illustrative mutated sequence (SEQ ID NO:59) derived the wild type amino acid sequence (SEQ ID NO:2) for a human galactosyltransferase, wherein a seine has been 35 deleted from the cytoplasmic tail and a G-I-Y motif has been repeated. Of course, such changes are merely illustrative of the many possible changes within the scope of the present invention. For example, in one embodiment, the present invention contemplates mutated sequences wherein only deletions (one or more) are employed (eg. deletions in the cytoplasmic tail domain or the stem -24 domain) - with no insertions or repeats. Similarly, in one embodiment, the present invention contemplates mutated sequences wherein only (one or more) insertions or replacements (e.g. In the transmembrane domain) are employed - with no deletions. Figure 8 shows the nucleic acid sequence (SEQ ID NO:3) encoding a hybrid enzyme 5 comprising human galactosyltransferase (human B 1,4-galactosyltransferase - GaiT), The upper case letters are nucleotides of Arabidopsis thaltana nRNA for beta ,2-xylosyltransferase (database entry: EMBL:A TH277603, the TmXyl-fragment used Involves nucleotides 135-297 of this database sequence). Figure 9 shows the nucleic acid sequence of Figure 8 along with the corresponding amino 10 acid sequence (SEQ ID NO:4). Figure 10 shows the amino acid sequence (SEQ ID NO:4) for the hybrid enzyme encoded by the nucleic acid shown in Figure 8. Figure 11 shows the nucleic acid sequence (SEQ ID NO:5) for the human glycosyltransferaso GnTII (along with additional sequence encoding a myo-tag) (primary accession number Q09327 15 GNT3 HUMAN). Figure 12 shows the nucleio acid sequence ofFigure 11 along with the corresponding amino acid sequence (SEQ ID NO:6). Figure 13 shows the amino acid sequence (SEQ II) NO:6) for a human GnTI! (along with additional amino acid sequence of the myc epitope tag SEQ ID NO:7). 20 Figure 14 shows the nucleic acid sequence (SEQ ID NO:9) encoding one embodiment of a hybrid enzyme of the present invention, said hybrid enzyme comprising the tranamembrane domain of a plant xylosyltransferase (TmXyl-) and the catalytic domain (along with other regions) fbr human GnTIl (TmnXyl-UnTIII) (along with additional sequence encoding a myc-tag). Figure 15 shows the nucleic acid sequence of Figure 14 along with the corresponding amino 25 acid sequence (SEQ ID NO:10). Figure 16 shows the amino acid sequence (SEQ ID NO: 10) for hybrid enzyme encoded by the nucleio acid of Figure 14 (along with additional sequence for the myo epitope tag SEQ ID NO:7). Figure 17 shows the complete nuclei acid sequence (SEQ D) NO;27) for a cassette encoding the hybrid enzymes TmXyl-GalT plus TmOnTI-GnTU plus TmGnTI-ManlI). 30 Figure 18 shows the complete nucleic acid sequence (SEQ ID NO:28) for a cassette encoding the hybrid enzyme TmrGnT-ManUI (with the RbcS1 promoter sequence SEQ ID NO:39 shown). Figure 19 shows the nucleic acid sequence (SEQ ID NO:29) encoding the hybrid enzyme TmnT-Manfl. Figure 20 shows the nucleic acid sequence (SEQ ID NO:30) encoding the hybrid enzyme 35 TmGnTn-GnTII. Figure 21 shows the nucleic acid sequence (SEQ ID NO:31) encoding the hybrid enzyme TmGnThuGnTfI, wherein the transmembrane fragment used (designated TmGntl) has the nucleic acid sequence set fbrth in SEQ ID NO:32.

-25 Figure 22A shows the nucleic acid sequence (SEQ ID NO:32) encoding one embodiment of a tranumembrane domain fragment (TmGnTl). Figure 22B shows the nuclei acid sequence (SEQ ID NO:33) encoding another embodiment of a transmembrane domain fragment (TmManl). Figure 23 shows the complete nucleic acid sequence (SEQ ID NO:34) for a triple cassette 5 embodiment of the present Invention. Figure 24 shows the nucleic acid sequence (SEQ ID NO:35) for a hybrid gene expression cassette (TmManI-GnTI). Figure 25 shows the nucleic acid sequence (SEQ ID NO:36) for the histone 3.1 promoter. Figure 26 shows the nuclei acid sequence (SEQ ID NO:37) for the hybrid gene fusion 10 (TmManl-TmGnTI), Figure 27 shows the nucleic acid sequence (SEQ ID NO:38) for the hybrid gene fusion TmManI-ManlI (with the RbcSl promoter sequence SEQ ID NO:39 shown). Figure 28 shows the nucleic acid sequence (SEQ ID)NO:39) for the RbcS 1 promoter. Figure 29 shows the nucleic acid sequence (SEQ ID NO:40) for the hybrid gene TmMatl 15 ManUi wherein the nucleic acid sequence (SEQ ID NO;33) encoding the tranumembrane fragment is shown, Figure 30 shows the nucleic acid sequence (SEQ ID NO:41) for the hybrid gene TmMani GnTIL. Figure 31 shows the nucloe acid sequence (SEQ ID NO:42) fbr the Lhca promoter. 20 Figure 32 shows the nucleic acid sequence (SEQ ID NO:43) for the hybrid gene TmManI GnTII wherein the nucleic acid sequence (SEQ ID NO:33) encoding the transmrmbrane fragment is shown Figure 33 shows the nucleic acid sequence (SEQ ID NO:44) for the terminator sequence used (see below). 25 Figure 34 is a Western Blot which examines total protein glycosylation of plants of the present invention compared to control plants. Figure 35 is a lectin blot with RCA on F1 progeny of crossed plants, said progeny made according to one embodiment of the present Invention Figure 36 is a Western Blot. Panel A was assayed with anti-IgO antibody. Panel B was 30 assayed with an anti-HRP antibody. Panel C was assayed with a specific anti-Xyl antibody fraction. Panel D was assayed with a specific anti-Fucose antibody fraction. Panel E was assayed with the lectin RCA, Figure 37 shows the nucleic acid sequence (SEQ ID NO:49) of a hybrid gene wherein the aminoterminal CTS region of an Insect Mannosidase III gene is replaced by a mouse signal peptide 35 and a carboxyterminal endoplasmic reticulum retention signal (KDEL) was added. Figure 38 shows the corresponding amino acid sequence (SEQ ID NO:50) for the nucleic acid sequence of Figure 37.

-26 Figure 39 shows the nucleic acid sequence (SEQ ID NO:51) of a hybrid gene wherein the aminoterminal CTS region of a human beta-1,4-galactosyltransferase (GalT) gene Is replaced by a mouse signal peptide and a carboxyterminal endoplasmic reticulum retention signal (KDEL) was added. 5 Figure 40 shows the corresponding amino acid sequence (SEQ ID NO:52) for the nucleic acid sequence of Figure 39. Figure 41 shows the nucleic acid sequence (SEQ ID NO:53) of a hybrid gene wherein the aminotenninal CTS region of an Arabidopsts thallana OnT! gene is replaced by a mouse signal peptide and a carboxyterminal endoplasmic reticulum retention signal (KDEL) was added. 10 Figure 42 shows the corresponding amino acid sequence (SEQ ID NO:54) for the nucleic acid sequence of Figure 41. Figure 43 shows the nucleic acid sequence (SEQ ID NO:55) of a hybrid gene wherein the aminoterminal CTS region of an Arabidopsis thaliana GnTl gene is replaced by a mouse signal peptide and a carboxyterminal endoplasmic reticulum retention signal (KDBL) was added. 15 Figure 44 shows the corresponding amino acid sequence (SEQ ID NO:56) for the nuclei; acid sequence of Figure 43. Figure 45 shows the nucleic acid sequence (SEQ ID NO:57) of a hybrid gene wherein the aminoterminal CTS region of a human beta-1,4-galactosyltransferase (GalT) gene is replaced by the CTS region of the human gene for GnTL. 20 Figure 46 shows the corresponding amino acid sequence (SEQ ID NO:58) for the nuclei; acid sequence of Figure 45. Figure 47 is a schematic of how enzymes might be localized to the Golgi. Figure 48 is a non-limiting speculative schematic of how the "swapping" of regions of transferases might cause relocalization. 25 DETAILED DESCRIPTION OF THE INVNTION Hybrid Enzymes The nucleio acid sequences encoding the various glycosylation enzymes such as 30 mannosidases, OloNAcTs, galactosyltransforases may be obtained using various recombinant DNA procedures known in the art, such as polymerase chain reaction (PCR) or screening of expression libraries to detect cloned DNA fragments with shared structural features. See, e.g., Innis et al., 1990, PCR: A Guide to Methods and Application, Academic Press, New York. Other nucleic acid amplification procedures such as ligase chain reaction (LCR), ligated activated transcription (LAT) 35 and nucleic acid sequence-based amplification (NASBA) or long range PCR may be used. Once the DNA fragments am generated, identification of the specific DNA fragment containing the desired gene may be accomplished in a number of ways. For example, if an amount of a portion of a gene or its specific RNA, or a fragment thereof, is available and can be purified and -27 labeled, the generated DNA fragments may be screened by nucleic acid hybridization to the labeled probe [Benton and Davis, Science 196:180 (1977); Grunstein and Hogness, Proc. Natl. Acad Sot. U.S.A. 72:3961 (1975)3. Alternatively, the presence ofthe gene may be detected by assays based on the physical, chemical, or immunological properties of its expressed product. For example, cDNA 5 clones, or DNA clones which hybrid-select the proper mRNAs, can be selected which produce a protein that e.g., has similar or identical electrophoretic migration, isoelectric focusing behavior, proteolytic digestion maps, or antigenic properties as known for the protein of interest A nucleic acid sequence encoding a hybrid enzyme comprising a transmembrane portion of a first enzyme and a catalytic portion of a second enzyme may be obtained as follows. The sequence 10 encoding the tranamembrane portion is removed from the second enzyme, leaving a nucleic acid sequence comprising a nucleic acid sequence encoding the C-terminal portion of the second enzyme, which encompasses the catalytic site. The sequence encoding the tranamembrane portion of the first enzyme is isolated or obtained via PCR and ligated to the sequence encoding a sequence comprising the C-terminal portion of the second enzyme. 15 Modified Enzymes A nucleic acid sequence encoding a protein, particularly enzymes such as galactosyltransferses, mannosidases and N-acetylglucosamine transferases that are retained in the FR may be obtained by removing the sequence encoding the transmembrane fagment and 20 substituting it for a methionine (initiation of translation) codon and by inserting between the last codon and the stop codon of galactosyltransferase the nucleic acid sequence encoding an ER retention signal such as the sequence encoding KDEL (amino acid residue sequence: lysine-aspartic acid-glutamic acid-leucine) [Rothman Cell 50:521 (1987)]. 25 Using Domains and Portions Thereof As noted above, the phrases "at least a portion of " or a "fragment of' refers to the minimal amino acid sequence necessary fbr a protein or a peptide to retain its natural or native function. For example, the function of an enzyme could refer to its enzymatic or catalytic role, its ability to anchor a protein in the Golgi apparatus, or as a signal peptide. Thus, the phrases "at least a portion of a 30 tranamembrane domain" or "a fragment of a transmembrane domain" each refer to the smallest amino acid segment of a larger transmrnmbrane domain that still retains at least part of the native transmembrane functionality (for example, the function may be evident, albeit decreased). As another example, the phrases "at least a portion of a catalytic region" or "a fragment of a catalytic region" each refer to the smallest amino acid segment of a larger catalytic region that still retains at 35 least part of the native catalytic functionality (again, even if somewhat decreased). As discussed herein, one skilled in the art will know the minimal amino acid segment that is necessary for a protein or a peptide to retain at least some of the functionality of the native protein or peptide.

-28 The glycosyltransferase enzymes are typically grouped into families based on the type of sugar they transfer (galactosyltransferases, sialyltransferases, etc.). Based on amino-acid sequence similarity and the stereoehemical course of the reaction, glycosyltransferases can be classified Into at least 27 and perhaps as many as 47 different families [Campbell et al., Biochem. J. 326:929-939 5 (1997), Biochem. J. 329:719 (1998)]. The majority of glycosyltransferases cloned to date are type fl transmembrane proteins (i.e., single transmembrane domain with the NH 2 terminus in the cytosol and the COOH terminus in the lumen of the Golgi apparatus). Regardless of how they are classified, all glycosyltransferases share some common structural features: a short NHrterminal cytoplasmic tail, a 16-20 amino acid signal-anchor or transaembrane domain, and an extended stem region 10 which is followed by the large COOH-terminal catalytic domain. The cytoplasmic tail appears to be involved in the specific localization of some types of glycosyltransferases to the Golgi [Milland et al., J,. iol. Chem. 277:10374-10378]. The signal anchor domains can act as both uncleavable signal peptides and as membrane-spanning regions that orient the catalytic domains of the glycosyltransferases within the lumen of the Golgi apparatus. 15 In one embodiment of the present invention, a portion defined by the N-terminal 77 amino acids of Nicotiana benthamiana (tobacco) acetylglucosaminyltransferase I are contemplated for use in the hybrid enzymess, since this portion has been found to be sufficient to target to and to retain a reporter protein in the plant Golgi apparatus [Essl et al., FEBSLett 453:169-173 (1999)]. Subcellular localization in tobacco of various fusion proteins between the putative cytoplasmic, 20 transmembrane and stem domains revealed that the cytoplasmic-transmembrane domains alone were sufficient to sustain Golgi retention of P 1,2-xylosyltransferase without the contribution of any luminal sequences [Dirnborger et al., Plar Mo. Biol. 50:273-281 (2002)]. Thus, as noted above, certain embodiments of the present invention utilize portions of the CTS region which involve only the cytoplasmic-transmembrane domains (or portions thereof) without utilizing the stem region of 25 the CTS region. However, while some types of glycosyltransferases rely primarily on their transmembrane domain for Golgi retention, other types require their transmembrmne region and sequences flanking one or both sides of this region [Colley, Glycobiology 7:1-13 (1997)]. For example, the N-terminal peptide encompassing amino acids 1 to 32 appears to be the minimal targeting signal sufficient to localize P 1,6 N-acetylglucosaminyltransferase to the Golgi. This 30 peptide makes up the cytoplasmic and transmembrane domains of this enzyme [Zerfaoui et al., Glycobiolog 12:15-24]. A great deal of information is available on the amino acid sequences of the domains for specific glycosyltransferases. For example, the amino acid sequence of the mammalian galactosyltransferase provided in GenBank Accession No. AAMI 7731 has the "stem" and 35 "catalytic" domains spanning residues 19 to 147 and residues 148 to 397, respectively [U.S. Patent No, 6,416,988, hereby incorporated by reference] - and the present invention, in certain embodiments, specifically contemplates such portions for use in the hybrid enzyme(s). The amino acid sequence of the rat liver sialyltransferase provided in GenBank Accession No. AAC91156 has a -29 9-amino acid NHrterminal cytoplasmic tall, a 17-amino acid signal-anchor domain, and a luminal domain that includes an exposed stem region followed by a 41 kDa catalytic domain (Hudgin et al., Can, J. Biochem. 49:829-837 (1971); US, Patent Nos. 5,032,519 and 5,776,772, hereby incorporated by reference]. Known human and mouse 0 1,3-galactosyltransferases have a catalytic domain with 5 eight conserved regions [Kolbinger et al., J Biol. Chem, 273:433-440 (1998); Hennet et al., J Blot. Chem. 273;58-65 (1998); U.S. Patent No. 5,955,282, hereby incorporated by reference). For example, the amino acid sequence of mouse UDP-galactose: P-N-acetylglucosamine P 1,3 galactosyltransferase-I provided In GenBank Accession No. NM020026 has the following catalytic regions: region I fom residues 78-83; region 2 from residues 93-102; region 3 from residues 116 10 119; region 4 from residues 147-158; region 5 from residues 172-183; region 6 fom residues 203 206; region 7 from amino acid resIdues 236-246; and region 8 from residues 264-275. [Hennet e al., supra.) - all of which are contemplated in certain embodiments of the present invention as useful portions in the context of the hybrid enzyme(s) discussed above. While earlier comparisons amongst known cDNA clones of glycosyltransferases had 15 revealed very little sequence homology between the enzymes [Paulson et al., J. Biol. Chem. 264:17615-618 (1989)], more recent advances have made it possible to deduce conserved domain structures in glycosyltransferases of diverse specificity [Kapitonoy et al., Glycobiology 9:961-978 (1999)]. For example, the nucleic acid and amino acid sequences of a number of glycosyltransferases have been identified using sequence data provided by the complete genomic 20 sequences obtained for such diverse organisms as Homo sapiens (humans), Caenorhabdtis elegans (soil nematode), Arabidopsis thaliana (thale cress, a mustard) and Oryza sativa (rce). As a result of extensive studies, common amino acid sequences have been deduced for honiologous binding sites of various families of glycosyltransferases. For example, sialyltransferases have sialyl motifs that appear to participate in the recognition of the donor substrate, CMP-sialic acid 25 [Paulson er aL, J. Blol Chem., 264:17615-17618 (1989); Datta et al., J. Biot Chem., 270:1497-1500 (1995); Katsutoshi, Trends Glycosci. Glycotech. 8:195-215 (1996)]. The hexapeptide RDKXNJ) in Gal al-3 galactosyltransferase and RDKKNE in GleNAc 01-4 galactosyltransferase have been suggested as the binding site for UDP-Gal [(Joziasse et al., J. Biol. Chem., 260:4941-4951 (1985), J Biol, Chem., 264:14290-14297 (1989); Joziasse, Glycobiology, 2:271-277 (1992)]. 30 A small, highly-conserved motif formed by two aspartie acid residues (DXD), which is frequently surrounded by a hydrophobic region, has been identified in a large number of different eukaryotic transferases, including c-1, 3-mannosyltransferase, P-1, 4-galactosyltransfereases, m-1, 3 galactosyltransferases, glucuronyltransferases, fucosyltransferases, glycogenins and others [Wiggins et al., Proc. Nat. Acad. Sci. USA. 95:7945-7950 (1998)]. Mutation studies indicate that this motif 35 is necessary for enzymatic activity [Busch et al., J. Bido. Chem. 273:19566-19572 (1998); Wang et al., J. Blol. Chem. 277:18568-18573 (2002)). Multiple peptide alignment showed several motifs corresponding to putative catalytic domains that are conserved throughout all members of the P 3galactosyltransferase family, namely, a type fl transmembrane domain, a conserved DxD motif, an N-glycosylation site and five conserved cysteines [Gromova et 4, Mol. Carcinog. 32:61-72 (2001)]. Through the use of BLAST searches and multiple alignments, the E-Xrfl motif was found to be a highly conserved among the members of four ihmilies of retaining glycosyltransferases [Cid et 5 al, J. Biot. Chem, 275:33614-33621 (2000)]. The 0-linked acetylglucosaminyltransferases (GlaNAc) add a single p.N-acetylglucosamine moiety to specific serine or threonine hydroxyls. BLAST analyses, consensus secondary structure predictions and fold recognition studies indicate that a conserved motif in the second Rossmann domain points to the UDP-GlcNAc donor-binding site [Wrabl t al., J. Mo. Bol. 314:365-374 (2001)). The J1, 3-glycosyltransferase enzymes 10 identified to date share several conserved regions and conserved cysteine residues, all being located in the putative catalytic domain. Site-directed mutagenesis of the marine #3GatT-I gene (Accession No. AF029790) indicate that the conserved residues W1 01 and W1 62 are involved in the binding of the UDP-galactose donor, the residue W315 in the binding of the N-acetylglucosamine- p-p nitrophenol acceptor, and the domain including E264 appears to participate in the binding of both 15 substrates [Malissard et al., E, J. Blochem. 269:233-239 (2002)]. Expression of Proteins of Interest in Plant Host System The nucleic acid encoding the hybrid or modified enzymes or other heterologous proteins, such as a heterologous glycoprotein may be inserted according to certain embodiments of the present 20 invention into an appropriate expression vector, is., a vector which contains the necessary elements for the transcription and translation of the inserted coding sequence, or in the case of an RNA viral vector, the necessary elements fbr replication and translation, as well as selectable markers. These include but are not limited to a promoter region, a signal sequence, 5' untranslated sequences, initiation codon (depending upon whether or not the structural gene comes equipped with one), and 25 transcription and translation termination sequences. Methods for obtaining such vectors are known in the art (see WO 01/29242 for review). Promoter sequences suitable for expression in plants are described in the art, e.g., WO 91/198696. These Include non-constitutive promoters or constitutive promoters , such as, the nopaline synthetase and octopine synthetase promoters, cauliflower mosaic virus (CaMV) 19S and 30 35S promoters and the figwort mosaic virus (FMV) 35 promoter (see U.S. Pat. Nos. 5, 352,605 and 6,051,753, both of which are hereby incorporated by reference). Promoters used may also be tissue specific promoters targeted for example to the endospenn, aleurone layer, embryo, pericarp, stem, leaves, tubers, roots, and the like. A signal sequence allows processing and translocation of a protein where appropriate. The 35 signal can be derived ftom plants or could be non-plant signal sequences. The signal peptides direct the nascent polypeptide to the endoplasmic reticulum, where the polypeptide subsequently undergoes post-translational modification. Signal peptides can routinely be identified by those of skill in the art. They typically have a tripartite structure, with positively charged amino acids at the N-terminal end, -31 followed by a hydrophobic region and then the cleavage site within a region of reduced hydrophobicity. The transcription termination is routinely at the opposite end from the transcription initiation regulatory region. It may be associated with the transcriptional initiation region or from a different 5 gene and may be selected to enhance expression. An example Is the NOS terminator from Agrobacterium Ti plasmid and the rice alpha-amylase terminator. Polyadenylation tails may also be added. Examples include but are not limited to Agrobacterum octopine synthetase signal, [Gielen et at., RMBO J. 3:835-846 (1984)] or nopaline synthase of the same species [Depicker et al., MoL. AppL. Genet. 1:561-573 (1982)]. 10 Enhancers may be included to increase and/or maximize transcription of the heterologous protein. These include, but are not limited to peptide export signal sequence, codon usage, introns, polyadenylation, and transcription termination sites ( see WO 01/29242). Markers include preferably prokaryote selectable market. Such markers include resistance toward antibiotics such as ampicillin, tetracycline, kanamycin, and spectinomycin. Specific 15 examples include but are not limited to streptomycin phosphotransferase (spt) gene coding for streptomycin resistance, neomycin phosphotransferase (nptff) gene encoding kanamycin or geneticin resistance, hygromycin phosphotransferase (hpt) gene encoding resistance to hygromycin. The vectors constructed may be introduced into the plant host system using procedures known In the art (reviewed in WO 01/29242 and WO 01/31045). The vectors may be modified to 20 intermediate plant transformation plasmids that contain a region of homology to an Agrobacterium tumefaciens vector, a T-DNA border region from A. tumefaciens. Alternativly, the vectors used in the methods ofthe present invention may be Agrobacteriun vectors. Methods for introducing the vectors Include but are not limited to microinjection, velocity ballistic penetration by small particles with the nucleic acid either within the matrix of small beads or particles, or on the surface and 25 electroporation. The vector may be introduced into a plant cell, tissue or organ. In a specific embodiment, once the presence of a heterologous gene is ascertained, a plant may be regenerated using procedures lown in the art. The presence of desired proteins may be screened using methods known in the art, preferably using screening assays where the biologically active site is detected in such a way as to produce a detectable signal. This signal may be produced directly or indirectly. 30 Examples of such assays include ELISA or a radioimnunoassay. Transient Expression The present invention specifically contemplates both stable and transient expression of the above-described hybrid enzymes. Techniques for transforming a wide variety of higher plant species 35 for transient expression of an expression cassette are well known [see, for example, Weising et al., Ann. Rev. Genet. 22:421-477(1988)]. Variables of different systems include type nucleic acid transferred (DNA, RNA, plasmid, viral), type of tissue transformed, means of introducing transgene(s), and conditions of transformation. For example, a nucleic acid construct may be -32 introduced directly into a plant cell using techniques ranging from electroporation, PEG poration, particle bombardment, silicon fiber delivery, microinjeetion of plant cell protoplasts or embryogenic callus or other plant tissue, or Agrobacterium-mediated transformation [H1el t al., Plant J. 6:271 282 (1994)). Because transformation efficiencies are variable, internal standards (eg, 35S-LuC) aro 5 often used to standardize transformation efficiencies. Expression constructs for transient assays include plasmids and viral vectors, A variety of plant viruses that can be employed as vectors are known in the art and include cauliflower mosaic virus (CaMV), geminivirus, brome mosaic virus, and tobacco mosaic virus. Plant tissues suitable for transient expression include cultured cells, either intact or as 10 protoplasts (in which the cell wall is removed), cultured tissue, cultured plants, and plant tissue such as leaves. Some transient expression methods utilize gene transfer into plant cell protoplasts mediated by electroporation or polyethylene glycol (PEG). These methods require the preparation and culture of plant protoplasts, and involve creating pores in the protoplast through which nucleio acid is 15 transferred into the interior of the protoplast. Exemplary electroporation techniques are described in Fromm er al., Proc. Natl. Acad. Scl. 82: 5824 (1985). The introduction of DNA constructs using polyethylene glycol precipitation is described in Paszkowski et al., EMBO .. 3: 2717-2722 (1984). PEG-mediated transfbrmation of tobacco protoplasts, which includes the steps of isolation, purification, and transformation of the 20 protoplasts, are described in Lyck at at, (1997) Planta 202: 117-125 and Scharf et al., (1998) Mol Cell Biol 18: 2240-2251, and Kirschner et al., (2000) The Plant J24(3): 397-411. These methods have been used, for example, to identify cis-acting elements in promoters activated by external stimuli, Abel and Theologis (1994) Plant J5: 421-427; Hattori at al., (1992) Genes Dev 6: 609 618; Sablowski et at, (1994)EMBOJ13: 128-137; and Solano stat., (1995) EMBOJ14: 1773 25 1784), as well as for other gone expression studies (U. S. Patent 6,376,747, hereby incorporated by reference). Ballistic transformation techniques are described in Klein et at, (1987) Natune 327:70-73. Biolistic transient transfbrnation is used with suspension colls or plant organs. For example, it has been developed for use In Nicotiana tabacum leaves, Godon et al (1993) Biochimie 75(7): 591-595. 30 It has also been used in investigating plant promoters, (Baum et at, (1997) Plant J 12: 463-469; Stromvik et at, (1999) Plant Mol Biol 41(2): 217-31, Tuerck and Fromm (1994) Plant Cell 6; 1655-1663; and U. S. Patent 5,847,102, hereby incorporated by reference), and to characterize transcription factors (Goff et at, (1990) RMBO J9: 2517-2522; Gubler et at., (1999) Plant J17: 1 9; and Sainz at al., (1997) Plant Cell 9: 611.-625). 35 Other methods allow visualization of transient expression of genes in situ, such as with onion epidermal peels, in which OF? expression in various cellular compartments was observed (Scott at al., (1999) Biotechniques 26(6): 1128-1132 -33. Nucleic acids can also be introduced into plants by direct injection. Transient gene expression can be obtained by injection of the DNA Into reproductive organs of a plant (see, for example, Pena et al., (1987) Nature, 325.:274), such as by direct DNA transfer into pollen (see, for example, Zhou et al,, (1983) Methods in Enzymology, 101:433; D. Hess (1987) Intern Rev. Cytol., 5 107:367; Luo et al., (1988) Plant Mol. Biol. Reporter, 6:165. DNA can also be injected directly into the cells of immature embryos (see, for example, Neuhaus et al., (1987) Theor. Apple, Genet: 75:30; and Benbrook et al., (1986) in Proceedings Bio Expo 1986, Butterworth, Stoneham, Mass., pp. 27-54). Agrobacterium-mediated transformation is applicable to both dicots and monocots. 10 Optimized methods and vectors for Agrobacterium-mediated transformation of plants in the family Graminae, such as rice and maize have been described (see, for example, Heath et a., (1997) MCI. Plant-Microbe Interact. 10:221-227; Hici et al., (1994) Plant J. 6:271-282 and Ishida et al., (1996) Nat. Biotech. 14:745-750). The efficiency ofmaize transformation is affected by a variety oftotors including the types and stages of tissue infected, the concentration of Agrobacterium, the tissue 15 culture media, the Ti vectors and the maize genotype. Another useful basic transformation protocol involves a combination of wounding by particle bombardment, followed by use of Agrobacterium for DNA delivery (see, for example, Bidney et al., (1992) Plant Mo. Biol. 18:301-313). Both intact meristem transformation and a split meristem transformation methods are also known (U. S. Patent 6,300,545, hereby incorporated by reference). 20 Additional methods utilizing Agrobacteria include agroinfection and agroinfiltration. By inserting a viral genome into the T-DNA, Agrobacterium can be used to mediate the viral infection of plants (see, for example, U. S. Patent 6,300,545, hereby incorporated by reference). Following transfer of the T-DNA to the plant cell, excision of the viral genome from the T-DNA (mobilization) is required for successful viral infection. This Agrobacterium-mediated method for introducing a 25 virus into a plant host is known as agroinfection (see, for example, Grimsley, "Agroinfection" pp. 325-342, in Methods in Molecular Biology, vol 44: Agrobacterlum Protocols, ed. Gartland and Davey, Humana Press, Inc., Totowa, N.J.; and Grimsley (1990) Physlol. Plant. 79:147-153). The development of plant virus gene vectors for expression of foreign genes in plants provides a means to provide high levels of gene expression within a short time. 30 Suitable viral replicons include double-stranded DNA from a virus having a double stranded DNA genome or replication intermediate. The excised viral DNA is capable of acting as a replicon or replication intermediate, either independently, or with factors supplied in trans. The viral DNA may or may not enoode infectious viral particles and furthermore may contain insertions, deletions, substitutions, rearrangements or other modifications. The viral DNA may contain heterologous 35 DNA, which is any non-viral DNA or DNA fion a different virus. For example, the heterologous DNA may comprise an expression cassette for a protein or RNA of interest. Super binary vectors carrying the vir gonns of Agrobacterium strains A281 and A348 are useful for high efficiency transformation of monocots. However, even without the use of high -34 efficiency vcotors, it has been demonstrated that T-DNA is transferred to maize at an efficiency that results in systemic infection by viruses introduced by agroinfection, although tumors are not formed (Grimsley et al., (1989) Mol. Gen. Genet. 217:309.316). This is because integration of the T-DNA containing the viral genome is not required for viral multiplication, since the excised viral genome 5 acts as an independent replicon. Another Agrobacteria-mediated transient expression assay is based on Agrobacterum mediated transformation of tobacco leaves in plant (Yang et al,, (2000) The PlantJ22(6): 543 551). The method utilizes infiltration of agrobacteria carrying plasmid constructs into tobacco leaves, and is referred to as agroinfiltration; it has been utilized used to analyze in vivo expression of 10 promoters and transcription factors in as little as 2-3 days. It also allows examination of effects of external stimuli such as pathogen infections and environmental stresses on promoter activity In situ, Example 1 An Arabidopsis thallana cDNA encoding $1,2-xylosyltransferase was isolated from a cDNA 15 library by a previously described PCR based sibling selection procedure [Bakker et al, BBRC 261:829 (1999)]. Xylosyltransferase activity was confirmed by immunostaining of transfected CHO cells with a xylose specific antibody purified from rabbit-anti horseradish-peroxidass antiserum. A DNA fragment covering the N-terminal part of the xylosyltransferase was amplified using primers: 20 XylTpvuF:ATACTCGAGTTAACAATGAGTAAACGGAATC (SEQ ID NO:45) and XylTpvuR:TTCTCGATCGCCQATTGGTTATTC (SEQ ID NO:46) Xhol and Hps 1 restriction sites were introduced In front ofthe start codon and a PvuI was introduced at the reverse end. A C-terminal fragment from Human p1,4galaotosyltransferase (acc.no. x55415, Aoki 1992) was amplified using primers GalTpvuF:GCCGCCGCGATCGGGCAGTCCTCC (SEQ 25 ID NO:47) and GalTrev;AACGGATCCACOCTAGCTCGGTOTCCCGAT (SEQ ID NO:48) thus introducing Pvul and BamHI sites. The XhoI/PvuI and PvuI/BamHI digested PCR fragments were ligated in Xhol/BamHI digested pBluesorlptSK+ and sequenced. The resulting open reading frame encodes a fusion protein containing the first 54 amino acids ofA. thaliana 1,2-xylosyltransferase fused with amino acid 69 to 398 of human j1,4galactosyltransferase and is designated as TmXyl 30 GaIT. The fragment was cloned into a plant expression vector between the CaMV358 promoter and Nos terminator, using Hpal/BamHI. The clone was introduced into Nicotiana tabacum (samsun NN) as described for native human p1,4galactosyltransferase [Bakker et aL, Proc. Nat. Acad Sct. USA 98:2899 (2001)). Protein extract of transgenic plants and Wester Blots were made as described [Bakker et al., 35 Proc. Nat. Acad. Sct. USA 98:2899 (2001)). Based on reaction with the lectin RCA, a transgenic plant expressing TmXylGalT was selected for further glycan analysis by MALDI-TOP [Elbers et al., Plant Physiology 126:1314 (2001] and compared with glycans isolated from plants expressing native P1,4galavtosyltransferase and with glycans from wild-type plants. Relative peak areas of the -35 MALDI-TOF spectrum are given in Table 1. That is to say, Table 1 is a comparison of the results of mass spec (MALDI-TOF) analysis of N-glycans of endogenous glycoproteins of control tobacco ('Tobacco'", transgenic tobacco expressing human beta.1,4-galactosyltransferase ("GalT") and transgenic tobacco plants expressing the beta-1,4-galactosyltransferase gene of which the CTS 5 region has been replaced with that of beta-1,2-xylosyhtransfbrase ("TmXyl-GalT'". TABLE I miz Typ Tobacco GaiT TmXyl-GaIT 933 M3 3 7 1065 XM3 10 16 3 1079 FM3 4 1095 M4 9 1211 FXM3 41 27 1257 M5 4 5 23 1268 GNXM3 4 1298 GaIGNM3 6 1298 GNM4 1414 GNFXM3 27 13 5 1419 me 7 a 10 1460 GalGNM4 11 1460 GNM5 1485 GN2FM3 4 1576 GaIGNFXM3 5 15761 GNFXM4 1581 M7 3 4 1600 GNFM5 3 1606 GalGNPM4 1617 GN2FXM3 8 9 1622 GalGNM5 8 1622 IGNM6 1743 MS 2 3 1768 IGaIGNFM5 3 1768 GNFM6 1779 GaIGN2FXM3 2 1905 M9 1 1941 Ga2GN2FXM3 2 TOTAL 100 100 101 These data show that; 1. In TmXylGalT plants, xylosylation and fucosylation of the glycans is dramatically reduced: 82% of the glycans do not carry xylose nor fTcose as compared to 14% in wild 10 type plants. 2. Galactosylation has increased from 9% in GaT plants to 32% in TmXylGalT plants.

-36 Example 2 A transgenic plant expressing said TmXyl-GalT gene (TmXyl-GalT-12 plant) was selected (above) based on lectin blotting using biotin-labolled RCA (Vector Laboratories, Burlingame, California). Comparison of protein extracts of MGR48 transgenic (control) plant, a selected 5 transgonio plant expressing the unmodified human pl,4-galactosyltransferase gene and TmXyl GaT-12 plant for the presence of xylose and fucose using anti-ImP (horseradish peroxidase) polyolonal antibody (known for high anti-xylose and antl-fucose reactivity) clearly showed reduced xylose and fucose (Figure 34: "Anti-HRI'"). Western blotting using an antl-xylose fraction of the anti-HRP and an anti-fucose fraction (each of which can be prepared by affinity chromatography 10 over the appropriate ligand) showed that especially xylose was reduced compared to control plants (Figure 34: anti-Fuc" and "anti-Xyl'". Example 3 The TmXyl-GalT-12 plant was crossed with a transgenic plant expressing the monoclonal 15 antibody MQR48 from a single T-DNA integration event (MGR48-3 1) and which was first made homozygous by selecting offspring plants not segregating for the kanamycin resistance marker and antibody production (MOR48-31-4). Pollen of MGR48-31-4 was used for pollination of emasculated TmXyl-GalT-12 plants. Vice versa, pollen of TmXyl-GalT-12 plant was used for fertilization on emasculated MGR48-31-4 plants. A number of FI plants were analyzed for the presence of MGR48 20 by western blotting and for galactosylation of endogenous glycoproteins by lectin blotting using RCA (Figure 35), One plant expressing MGR48 and showing galaotosylation of endogenous glycoproteins was selected for further analysis. This plant was identified as XGM8. Seeds from TmXyl-GalT-12 (9) x MGR48-31-4 (d) were sown and FI offspring plants (XGM) were analysed for antibody production by Western blotting and for galactosylation by lectin 25 blotting using biotinylated RCA120 (Vector Labs., Burlingane, California) using standard techniques as described before. All plants as expected expressed the monoclonal antibody MOR48 and the majority also had galactosylated glycans as depicted from lectin blotting using RCAI20. A single plant expressing both antibody MGR48 and having galactosylated N-glycans was chosen for further analysis (XGMS) (TmXyl-GalT-12 X MGR48-31-4 offering plant 8). The monoclonal 30 recombinant MGR48 antibody was purified from this plant as described before and submitted to N glycan analysis by MALDI-TOF. Briefly, XGM8 plant was grown in greenhouse for .itibody production under optimal conditions [Elbers or al., Plant Physiology 126:1314 (2001)]. Protein extract of leaves of transgenic XGM8 plant was made and monoclonal antibody was purified using protein G chromatography as 35 described [Bakker St al, Proc. Nat. Acad. Sat. USA 98:2899 (2001)]. MALDI-TOF of N-glycans of purified monoclonal antibody was as described (Elbers et aL, 2001, supra). The presence of galactose on glycans was established by enzyme sequencing using bovine testis p-galactosidase as described (Bakker et al., 2001, supra; Table 2). Table 2 (below) is a comparison of the results of -37 mass spec (MALDI-TOF) analysis of N-glyoans- of endogenous glycoproteins ("Xyl-GalT Endo") of a Fl hybrid of TnXyl-GaIT-12plant and plant producing rec-mAb (MGR48) and ofN-glycans of rec-mAB purified by protein 0 chromatography from said F1 hybrid. TABLE 2 Xyl-GaIT Xyl-GaIT mlz Type Endo IgG 933 M3 8 4 1065 XM3 2 2 1079 FM3 2 3 1095 ____M4 5 5 1136 GNM3 1 2 1211 FXM3 6 3 1241 FM4 3 2 1257 M5 17 12 1268 GNXM3 1 2 1282 GNFM3 2 3 1298 GaIGNMS 3 4 1403 FM5 4 3 1414 GNFXM3 2 4 1419 Me 5 4 1430 GNXM4 2 2 1430 GaIGNXM3 1444 GNFM4 1 3 1444 GaIGNFM3 1460 GaIGNM4 8 10 1460 GNM5 1471 GN2XM3 1 1485 GN2FM3 1 1 1501 GaIGN2M3 1 1 1576 GaIGNFXM3 2 3 1576 GNFXM4 1581 M7 2 2 1593 GaIGNXM4 1 2 1593 GNXMS 1806 GNFM5 3 4 1606 GaIGNFM4 1617 GN2FXM3 2 1 1622 GaIGNM5 6 8 1622 GNM6 1647 GaeGN2FM3 i 1 1683 GsI2GN2M3 I I 1738 GNFXM5 1 2 1738 GaIGNFXM4 1743 Ma 1 2 1754 Ga!GNXM5 1 2 1768 GaIGNFM5 2 3 1768 GNFMB -38 1784 GNM7 1 1 1784 GaiGNMB 1809 Ga2GN2FM3 2 1 1900 GNFXM8 I 1900 GaIGNFXM5 1906 M9 1 1 TOTAL 101 102 These data show that: 1. In the Fl hybrid, xylosylation and fucosylation of the glycans is dramatically reduced: 43% of the glycans of endogenous glycoproteins lack xylose and lUcose as compared to only 14% In 5 wild-type tobacco plants. 2. The glycans of purified mAb of this F 1 hybrid have reduced xylosc and fucose, 47% compared to 14% for wildtype tobacco, See also Figure 36, panels B-D. 3. Galactosylation of endogenous glycoproteina of Fl hybrid has increased from 9% in GaIT plants to 37% in Fl TmXyl-GaiT X MGR48 plant. See also Figure 35. 10 4. Purified ree-mAb ftom said F (see Figure 36, panel A) shows increased galactosylation; that is to say, 46% has galactose. See also Figure 36, panel E. It should however be noted that the observed quantities (MALDI-TOF) do not necessarily reflect the molar rations of said glycoforms in vivo. Quantification based on MALDI-TOF can be under- or overestimated depending on the specific glycofbrmi under study. Also, since there is no molecular 15 weight difference between Gal and Man, some peaks can not be annotated unambiguously unless there are clear differences in relative height of specific molecules before and after galactosidase treatment. Example 4 20 A more direct comparison of xylose, ifcose and galactose content was done by examining the MGR48 IgG antibodies from hybridoma, transgenic tobacco and TmXy-GaIT transgenic tobacco. As mentioned above, the TmXyl-GaIT-12 plant was crossed with tobacco plant expressing MGR48 IgG (M0R48 tobacco) resulting in an F1 hybrid harbouring MGR48 TmXyl-GalT. An Fl plant was chosen for extraction and purification of MGR48 IgO. Antibodies ftom said plants 25 (tobacco and TmXyl-GalT) were isolated and purified using protein G chromatography (Elbers et a/., 2001. Plan; Physloogy 126: 1314-1322). 300 nanograms amounts of each, hybridoma MGR48 and plant-dorived recMGR48, were loaded on precoast 12% SDS-PAGE gels (BioRad) and run. The contents of each lane were as follows: Lane 1, MGR48 from hybridoma; Lane 2, purified recMGR48 from normal transgenic tobacco plant; and Lane 3, purified recMGR48 from TmXyl-GalT transgenio 30 plant. Following SDS-PAGE proteins were transferred to nitrocellulose using CAPS buffer. Blots we're incubated with A, anti-mouse IgO; B, polyclonal rabbit antI-HRP (anti-xylose/(alpha 1,3 fucose); C, anti-xylose; D, anti-(alpha 1,3-) ftcose antibodies; and B, biotinylated RCA. Detection -39 was with LumiLight on Lumi Imager following incubation with FRP-labelled sheep anti-mouse (panel A) or goat-anti-rabbit (panels B-)) antibodies and HRP-labeled streptavidin (B). Panel A shows that approximately similar amounts of the MOR48 IgO was loaded for all lanes (1-3). L refers to Light chain and H, heavy chain of MGR48 IgG. 5 Panel B shows that the heavy chain of MGR4 antibody in lane 2 (tobacco) strongly reacts with anti HRP as expected, whereas the heavy chain ofhybridoma derived MGR48 (lane 1) does not (as expected). Hybridoma derived antibodies do not carry xylose and alpha 1, 3-flictose residues. Remarkably, MGR48 antibodies from TmXyl-GalT tobacco plant also do not react, suggesting that the heavy chain of antibody from this plant have significantly reduced (perhaps by 90% or more) the 10 amounts of xylose and flcose residues on the N-glycans. This is confirmed by experiments depicted in panels C (anti-xylose) and D (anti-fuoose). Panel B shows that the heavy chain ofMGR4S antibody of hybridoma (lane 1) has a galactosylated N-glycan, whereas tobacco-derived MGR48 (lane 2) has not, both as expected. Heavy chain of MGR48 from the TmXyl-Ga1T plant (lane 3) also has galactosylated N-glycan due to the presence of the construct expressing the hybrid enzyme. 15 These data are in agreement with the data obtained from similar experiments using total protein extracts from similar plants (tobacco and TmXyl-GalT-12 plant) as shown previously and confirm that the novel trait introduced in tobacco from expression of TmXyl-GalT gene can be stably transmitted to offspring and a recombinant monoclonal antibody. 20 Example 5 Further characterization of the above-described F1 hybrid was performed by treatment with beta-galactosidase. Table 3 is a comparison of the results of mass speo (MALDI-TOF) analysis of N-glycans of rco-mAbs purifled by protein 0 chromatography from an F1 hybrid of TmXyl-GalT and MOR48 plant before and after treatment of the glycans with beta-galactosidase. 25 TABLE 3 Xyl-GaIT Xyl-GaIT m/z Type IgG- IgG+bet-galactosidase 933 M3 4 4 1065 XM3 2 2 1079 FM3 3 3 1095 M4 5 4 1136 GNM3 2 3 1211 FXM3 3 4 1241 FM4 2 2 1257 M6 12 13 1268 GNXM3 2 3 1282 GNFM3 3 3. 1298 GaIGNM3 4 4 1403 FM5 3 2 1414 GNFXM3 4 5 1419 M6 4 3 40 1430 GNXM4 2 2 1430 GaIGNXM3 1444 GNFM4 3 3 1444 GaIGNFM3 1460 GaIGNM4 10 14 1460 GNM5 1471 GN2XMS 1 14865 GN2FM3 I I 1501 GaIGN2M3 1 1576 GaIGNFXM3 3 3 1578 GNFXM4 1581 M7 2 2 1593 GaIGNXM4 2 2 1593 GNXM5 1606 GNFM5 4 6 1606 GaIGNFM4 1617 _ GN2FXM3 1 1 1622 GaIGNM5 0 1622 GNM6 1647 GaIGN2FM3 1 1663 GaI2GN2M3 1 1738 _ _GNFXM5 2 2 1738 GaIGNFXM4 1743 MS 2 2 1754 GaIGNXMS 2 1 1768 GaIGNFM5 3 1768 GNFM6 1784 GNM7 1 1784 GaIGNM6 1809 Gal2GN2FM3 1 1900 GNFXMO 1900 GalGNFXM5 1905 M9 1 TOTAL 102 100 These data show that: 1. Rec-mAbs from F] hybrid contain galactose which can be deduced from the observed reduction of specific (galactose-containing) glycoforms after beta-galactosidase treatment and increase of 5 glycofons lacking galactose. Note the observed reduction of m/z 1622 from 6 to 1% and simultaneous increase of m/z 1460 from 10 to 14% which is the result of the removal of galactose from GaIGNM5 to give rise to NM5. The same is true for m/z 1768 (3 to 1% decrease) and corresponding n/z 1606 peak (4 to 6% Increase). See also Figure 36, panel E. 2. Similarly a number of peaks that can be attributed to galactose containing glycans vanish upon 10 treatment with galactosidase, especially m/z 1501, 1647 and 1663 confirming the presence of galactose, -41 Example 6 In another embodiment, the aminoterminal CTS region of an insect Mannosidase II gene (accession number: AF005034; mistakenly annotated as a Mannosidase H genel) is replaced by a 5 mouse signal peptidc coding sequence for import into the endoplasmic reticulum (see Figure 37). The signal peptide, sequence encodes a fully active signal peptide normally present at the aminoterminus of IgG sequences and has been used successfully in plants and other organisms before. Furthennore a synthetic sequence coding for a so-called endoplasmic reticulum retention sequence (KDEL) is added to the carboxyterninus of the gene part encoding the catalytic fragment 10 for ER retention. The hybrid Marmosidase II protein encoded by this gene sequence will hence accumulate preferentially in the endoplasmic reticulum. Example 7 In another embodiment, the aminoterminal CTS region of the human beta-1,4 15 galactosyltransferase (GaIT) gene (accession A52551) is replaced by a mouse signal peptide coding sequence for import into the endoplasmie reticulum (see Figure 39). The signal peptide sequence encodes a fully active signal peptide normally present at the aminoterminus of IgG sequences and has been used successiblly in plants and other organisms before. Furthermore a synthetic sequence coding for a so-called endoplasmic reticulum retention sequence (KDEL) is added to the 20 carboxyterminus of the gene part encoding the catalytic fragment for ER retention. The hybrid beta 1,4-galactosyl-transferase protein encoded by this gene sequence will hence accumulate preferentially In the endoplasmic reticulum. Example S 25 In another embodiment, the aminoterminal CTS region of Arabidopsis thaliana GnTI (aco. AJ243198) is replaced by a mouse signal peptide coding sequence for Import Into the endoplasmic reticulum (see Figure 41). The signal peptide sequence encodes a fully active signal peptide normally present at the aminoterminus of IgG sequences and has been used successfully in plants and other organisms before. Furthennore a synthetic sequence coding for a so-called ndoplasmic 30 reticulum retention sequence (KDEL) is added to the carboxyterminus of the gene part encoding the catalytic fragment for ER retention. The hybrid GnTI protein encoded by this gene sequence will hence accumulate preferentially in the endoplasmic reticulum. Example 9 35 In another embodiment, the aminoterninal CTS region of an Arabidopsis thaliana (InTJI (aoc. A.249274) is replaced by a mouse signal peptide coding sequence for import into the endoplasmic reticulum (see Figure 43). The signal peptide sequence encodes a fully active signal peptide normally present at the aminoterminus of IgG sequences and has been used successfUlly in -42 plants and other organisms before, Furthermore a synthetic sequence coding for a so-called endoplasmic reticulum retention sequence (KDEL) is added to the carboxyterminus of the gene part encoding the catalytic fragment for ER retention. The hybrid GnTIl protein encoded by this gene sequence will hence accumulate prefirentially in the endoplasmic reticulum. 5 Example 10 In another embodiment, the aminoterminal CTS region of the human gene for beta-1,4 galactosyltransfbrase (Gal) gene is replaced by the CTS region of the human gene for GnTI (TmhuGnTI-GalT) (see Figure 45). 10 It Is understood that the present invention is not limited to any particular mechanism. Nor is it necessary to understand the mechanism in order to successfully use the various embodiments of the invention. Nonetheless, it is believed that there Is a sequential distribution of Golgi enzymes (Figure 47) and that the swapping in of tranamembrane domains of plant glycosyltransferases causes 15 relocalization (Figure 48). It Is understood that the present invention is not limited to the particular methodology, protocols, col lines, vectors, and reagents described herein, as these may vary, It is also to be understood that the terminology used herein is used for the purpose of describing particular embodiments only, and is not intend to limit the scope of the present invention. It must be noted that 20 as used herein and In the appended claims, the singular forms "a", "an", and "the" include plural reference unless the context clearly dictates otherwise, Unless defined otherwise, all technical and scientific terns used herein have the same meanings as commonly understood by one of ordinary sidll in the art to which this invention belongs. The invention described and claimed herein is not to be limited in scope by the specific 25 embodiments herein disclosed, since these embodiments are intended as illustrations of several aspects of the invention. Any equivalent embodiments are intended to be within the scope of this invention. Indeed, various modifications of the invention in addition to those shown and described herein will become apparent to those sllled in the art from the foregoing description. Such modifications are also intended to fall within the scope of the appended claims. 30 Various references are cited herein, the disclosures of which are incorporated by reference in their entireties.

Claims (20)

1. A hybrid glycosyltransferase comprising a first fragment that comprises the transmembrane domain of a plant glycosyltransferase and a second fragment that comprises the catalytic domain of a mammalian glycosyltransferase, wherein the first fragment is from an N 5 acetylglucosaminyltransferase I (GnTI) and the second fragment is from a mannosidase II (ManlI) or N-acetylglucosaminyltransferase II (GnTII).

2. A hybrid glycosyltransferase according to Claim 1, wherein the first fragment comprises the cytoplasmic tail-transmembrane-stem (CTS) region in the plant glycosyltransferase.

3. A hybrid glycosyltransferase according to Claim 1, wherein the hybrid 10 glycosyltransferase comprises the amino acid sequence encoded by SEQ ID NO: 29, 30 or 31.

4. A nucleic acid comprising a nucleotide sequence coding for a hybrid glycosyltransferase according to any one of Claims 1 to 3.

5. A nucleic acid according to Claim 4, wherein the nucleotide sequence is selected from the group consisting of SEQ ID NO: 29, 30, and 31. 15

6. An expression vector comprising a nucleic acid according to Claim 4 or 5.

7. A host cell comprising a nucleic acid according to Claim 4 or 5, or an expression vector according to Claim 6.

8. A host cell according to Claim 7, wherein the host cell is a bacterial cell.

9. A host cell according to Claim 7, wherein said host cell is a plant cell. 20

10. A host cell according to Claim 9, further comprising a nucleotide sequence encoding a heterologous glycoprotein.

11. A host cell according to Claim 10, wherein the heterologous glycoprotein is an antibody or a fragment thereof.

12. A method comprising: 25 introducing into a plant cell the expression vector of Claim 6, and 44 culturing the plant cell under conditions allowing expression of the hybrid glycosyltransferase encoded by the expression vector.

13. A method according to Claim 12, further comprising introducing another expression vector encoding a heterologous glycoprotein. 5

14. A method according to Claim 13, wherein the heterologous glycoprotein is an antibody or a fragment thereof.

15. A method according to Claim 13 or 14, further comprising isolating the heterologous glycoprotein thus produced.

16. A method comprising: 10 introducing into a plant host system (i) one or more nucleic acids each coding for a hybrid glycosyltransferase as set forth in Claim 1, 2 or 3, and (ii) a nucleic acid coding for a heterologous glycoprotein, wherein each of (i) and (ii) is in operative linkage with a promoter, and collecting a plant or a portion thereof that expresses the hybrid glycosyltransferase(s) and 15 the heterologous protein.

17. A method according to Claim 16, wherein the one or more nucleic acids set forth in (i) comprise the nucleotide sequence SEQ ID NO: 29, 30 and/or 31.

18. A method according to Claim 16 or 17, further comprising isolating the heterologous glycoprotein. 20

19. A hybrid glycosyltransferase according to Claim 1, substantially as hereinbefore described with reference to any one of the examples or figures.

20. A method according to Claim 12 or 16, substantially as hereinbefore described with reference to any one of the examples or figures.