In particular, this invention concerns a method of inhibiting or modulating func- tions mediated by lectin domains of polypeptide GaINAc-transferases. More particu- larly, this invention is related to a lectin domain found in the C-terminal region of most GaINAc-transferases, which is structurally and functionally distinct from the catalytic domain of the enzymes and can be expressed in a functional form as a truncated or mu- tated protein that is enzymatically inactive. Still more particularly, this invention is re- lated to the lectin domains of more than 16 GaINAc-transferases, designated GaINAc- T1 to GaINAc-T16.

Further, this invention concerns a method of screening one or more test sub- stances for the ability to inhibit or modulate polypeptide GaINAc-transferase lectins in a cell-free or cell-based assay, in order to find agents which are effective in binding to one or more lectin domains of polypeptide GaINAc-transferases, for example, selective inhibitors of the binding properties of the above mentioned lectin domains and selective inhibitors of the effects that these lectin domains exert on intracellular transport, trans- port to cell surface, and secretion of mucins, glycoproteins, and proteins.

Even further, this invention provides a preferable inhibitor, GalNAcß-benzyl, which is a novel inhibitor and representative of a novel group of inhibitors which dis- play the common characteristic of selectively inhibiting lectins of polypeptide GaINAc- transferases in direct binding assays and not serve as substrates for other glycosyltrans-

ferases involved in O-glycan biosyntheses, while exhibiting inhibitory activity of secre- tion and intracellular transport of mucins and glycoproteins in cells. GaINAcp-benzyl and related inhibitors with the same biological functions represent preferable selective inhibitor compared to GalNAca-benzyl because these do not serve as substrates for glycosyltransferases extending O-glycans and do not provide a general inhibition of the O-glycosylation process in cells.

BACKGROUND OF THE INVENTION Mucin-type O-glycosylation, one of the most abundant forms of protein glyco- sylation, is found on secreted and cell surface associated glycoproteins of all eukaryotic cells except yeast. Mucin-type O-glycans contribute to a number of important molecular functions, including: direct effects on protein conformation, solubility, and stability; specific receptor functions that regulate cell trafficking and cell-cell interactions; and microbial clearance. Mucin-type 0-glycans are synthesised in the Golgi through the se- quential addition of saccharide residues, first to hydroxyl groups on serines and threoni- nes of a protein core and subsequently to hydroxyl groups on the growing saccharide chains that extend from the protein core. There is great diversity in the structures cre- ated by O-glycosylation (hundreds of potential structures), which are produced by the catalytic activity of hundreds of glycosyltransferase enzymes that are resident in the Golgi complex. Diversity exists at the level of the glycan structure and in positions of attachment of 0-glycans to protein backbones. Despite the high degree of potential di- versity, it is clear that O-glycosylation is a highly regulated process that shows a high degree of conservation among multicellular organisms.

The factors that regulate the attachment of O-glycans to particular protein sites and their extension into specific structures are poorly understood. Longstanding hy- potheses in this area propose that mucin-type O-glycosylation occurs in a stochastic manner where structure of acceptor proteins combined with topology and kinetic prop- erties of resident Golgi glycosyltransferases determine the order and degree of glycosy- lation 1. This concept does not fully explain the high degree of regulation and speciali- sation that governs the O-glycosylation process. In particular it is difficult to envision how large mucin molecules with high densities of O-glycans are glycosylated in the Golgi by stochastic mechanisms that also create other sparsely glycosylated proteins.

The first step in mucin-type O-glycosylation is catalysed by one or more mem- bers of a large family of UDP-GaINAc : polypeptide N-acetylgalactosaminyltransferases (GalNAc-transferases) (EC 2.4. 1.41), which transfer GaINAc to serine and threonine acceptor sites 2. To date twelve members of the mammalian GaINAc-transferase family have been identified and characterized 3, and several additional putative members of this gene family have been predicted from analysis of genome databases. The GaINAc- transferase isoforms have different kinetic properties and show differential expression patterns temporally and spatially, suggesting that they have distinct biological functions 2. Sequence analysis of GalNAc-transferases have led to the hypothesis that these en- zymes contain two distinct subunits: a central catalytic unit, and a C-terminal unit with sequence similarity to the plant lectin ricin, designated the"lectin domain"4-7. Previous experiments involving site-specific mutagenesis of selected conserved residues con- firmed that mutations in the catalytic domain eliminated catalytic activity. In contrast, mutations in the"lectin domain"had no significant effects on catalytic activity of the GaINAc-transferase isoform, GalNAc-TI 4. Thus, the C-terminal"lectin domain"was believed not to be functional and not to play roles for the enzymatic functions of Gal- NAc-transferases 4.

However, recent evidence demonstrates that some GalNAc-transferases exhibit unique activities with partially GaINAc-glycosylated glycopeptides. The catalytic ac- tions of at least three GaINAc-transferase isoforms, GaINAc-T4,-T7, and-T10, selec- tively act on glycopeptides corresponding to mucin tandem repeat domains where only some of the clustered potential glycosylation sites have been GaINAc glycosylated by other GalNAc-transferases 8-11. GaINAc-T4 and-T7 recognize different GaINAc- glycosylated peptides and catalyse transfer of GaINAc to acceptor substrate sites in ad- dition to those that were previously utilized. One of the functions of such GaINAc- transferase activities is predicted to represent a control step of the density of O-glycan occupancy in mucins and mucin-like glycoproteins with high density of O- glycosylation. It was hypothesized that such sequential actions of multiple GaINAc- transferase isoforms may be required to complete O-glycan attachments to some mucin peptide sequences allowing for detailed control of density.

One example of this is the glycosylation of the cancer-associated mucin MUC1.

MUC1 contains a tandem repeat O-glycosylated region of 20 residues

(HGVTSAPDTRPAPGSTAPPA) with five potential O-glycosylation sites. Gal- NAc-T1,-T2, and-T3 can initiate glycosylation of the MUC1 tandem repeat and incorporate at only three sites (HGVTSAPDTRPAPGSTAPPA, GalNAc attachment sites underlined). GaINAc-T4 is unique in that it is the only GalNAc-transferase iso- form identified so far that can complete the O-glycan attachment to all five acceptor sites in the 20 amino acid tandem repeat sequence of the breast cancer associated mucin, MUC1. GaINAc-T4 transfers GalNAc to at least two sites not used by other GalNAc-transferase isoforms on the GalNAc4TAP24 glycopeptide (TAPPAHGVTSAPDTRPAPGSTAPP, GalNAc attachment sites underlined)"\ An activity such as that exhibited by GaINAc-T4 appears to be required for production of the glycoform of MUC1 expressed by cancer cells where all potential sites are glycosy- lated 12. Normal MUC1 from lactating mammary glands has approximately 2.6 O- glycans per repeat 13 and MUCI derived from the cancer cell line T47D has 4.8 O- glycans per repeat 12. The cancer-associated form of MUC1 is therefore associated with higher density of O-glycan occupancy and this is accomplished by a GalNAc- transferase activity identical to or similar to that of GaINAc-T4.

The specific mechanism by which GaINAc-T4,-T7, and-T10 recognize and function with GaINAc-glycosylated glycopeptides is not known. However, it was origi- nally demonstrated that the GaINAc-glycopeptide specificity exerted by GalNAc-T4 is directed or at least dependent on its lectin domain. A single amino acid substitution in the T4 lectin domain predicted to inactivate its function abolished the GalNAc- glycopeptide specificity of T4 without adversely affecting the basic catalytic mecha- nism of the transferase 2. This suggests that the lectin domain interacts with GalNAc- glycopeptides and confers a novel catalytic function to the enzyme protein. Despite ex- tensive attempts it has in the past not been possible to demonstrate actual binding of the transferase and lectin to sugars and glycopeptides, but it was possible to demonstrate selective inhibition of the GaINAc-glycopeptide activity of GaINAc-T4 using 230 mM concentration of GalNAc 2. The present inventors and coworkers demonstrated that mM concentrations of GalNAca-benzyl can inhibit the lectin mediated GalNAc- glycopeptide substrate specificity of GaINAc-T4 as well as-T7 (PCT WO 01/85215 A2). Polypeptide GalNAc-transferases, which have not displayed apparent GalNAc-

glycopeptide specificities, also appear to be modulated by their putative lectin domains (PCT WO 0/85215 A2). Recently, it was found that mutations in the GaINAc-T1 puta- tive lectin domain, similarly to those previously analysed in GaINAc-T4, modified the activity of the enzyme in a similar fashion as GaINAc-T4. Thus, while wild type Gal- NAc-Tl added multiple consequtive GaINAc residues to a peptide substrate with multi- ple acceptor sites, mutated GalNAc-TI failed to add more than one GaINAc residue to the same substrate 14. The mechanism is however not understood.

Glycosylation confers physico-chemical properties including protease resis- tance, solubility, and stability to proteins. Glycosylation furthermore confers changes in immunological responses to proteins and glycoproteins. O-glycosylation on mucins and mucin-like glycoproteins protect these molecules found in the extracellular space and body fluids from degradation. Control of O-glycosylation with respect to sites and number (density) of O-glycan attachments to proteins as well as control of the O- glycan structures made at specific sites or in general on glycoproteins, is of interest for several purposes. Diseased cells e. g. cancer cells often dramatically change their O- glycosylation and the altered glycans and glycoproteins may constitute targets for therapeutic and diagnostic measures. Mucins functioning in body fluids may have different properties depending on density and structure of O-glycans attached in protec- tion against disease, including infections by microorganisms. Furthermore, mucins with different glycosylation may change physico-chemical properties including stability and solubility properties that may influence turnover and removal of mucous. A number of lung diseases, e. g. cystic fibrosis, asthma, chronic bronchitis, smokers lungs, are associ- ated with symptomatic mucous accumulation, and it is likely that the nature and structure of mucins play a role in the pathogenesis of such diseases.

Partial inhibitors of O-glycosylation in cells have been reported. Aryl-N-acetyl- a-galactosaminides such as benzyl-, phenyl-, and p-nitrophenyl-GalNAc were origi- nally found to inhibit the second step in 0-glycosylation, the O-glycan processing step, by inhibiting synthesis of core 1 (Galpl-3GalNAcal-R) and more complex structures 23. Benzyl-aGalNAc was also found to inhibit sialylation. It is generally believed that the downstream effects of benzyl-aGalNAc treatment are mediated by substrate compe- tition of biosynthetic glycosylation products of benzyl-aGalNAc. Thus, e. g. the imme-

diate glycosylation product of benzyl-aGaINAc is Gal (31-3GalNAca-benzyl and this serves as an efficient substrate for the core 1 a2-3sialyltransferase ST3Gal-I 2425. Gal- NAca-benzyl has been the most widely used inhibitor of O-glycosylation, but it has only been used in cell culture as effective treatment concentrations lead to intracellular build-up of vesicles with GalNAca-benzyl products and treated cells change morphol- ogy and growth characteristics 26.

Treatment of cells with benzyl-aGalNAc inhibit 0-glycan processing and affect apical sorting of some O-glycosylated proteins 27-29. The mechanism for this is generally believed to be through inhibition of sialylation 26. Inhibition of mucin secretion has also been observed in culture cells, more specifically HT29 MTX cells, but this effect is not generally found in mucin secreting cells True inhibitors of O-glycosylation, i. e. inhibitors of the initial 0-glycan attach- ment process governed by polypeptide GalNAc-transferases have not been identified.

Benzyl-aGalNAc has been shown to exhibit inhibitory effect on the GalNAc- glycopeptide substrate specificity of some polypeptide GalNAc-transferases in test tube enzyme assays (PCT WO 01/85215 A2), however the effect of this compound in vivo on density of O-glycosylation is unknown.

Inhibitors of the initiating step in O-glycosylation could completely or selec- tively block attachment of O-glycans to O-glycosylation sites in proteins. Compounds inhibiting the catalytic function of a selected subset of the polypeptide GalNAc- transferase family may be predicted to only lead to partial inhibition of O-glycosylation capacity of cells. Proteins with no or little O-glycosylation may have entirely different biological properties than their normal glycosylated counterparts. Complete inhibition of O-glycosylation is not desirable because of the many diverse functions of O-glycans, and it is expected to result in cell death. Selective inhibition of O-glycosylation on the other hand is desirable in many cases such as cancer cells producing glycoproteins and mucins with denser O-glycosylation than normal cells. For example breast cancer cells appear to hyperglycosylate the cancer-associated cell surface mucin MUC1 compared to glycosylation in normal cells 12, 30. The overexpression of MUC1 and hyperglycosy- lation found in cancer cells are likely to be important for the pathobiology of cancers.

Methods of inhibiting the hyperglycosylation of mucins in cancer cells are desirable.

It is apparent from the above that inhibitors in the prior art interfere with O- glycan processing, i. e. the glycosylation process that extend GaINAc residues directly attached to proteins at serine and threonine residues. Existing inhibitors of O- glycosylation are not suitable for therapeutic treatment in mammals including man as they profoundly affect O-glycosylation processing as well as lead to undesired morpho- logical and growth effects on culture cells.

Consequently, there exists a need in the art for therapeutic compounds that dis- play selectively and limited inhibition of O-glycosylation without generally affecting the process of O-glycosylation. The present invention meets these needs, and further presents other related advantages.

SUMMARY OF THE INVENTION The present invention provides for isolated polypeptide lectin domains of Gal- NAc-transferases. The present invention also provides for mammalian GaINAc- transferase lectin domains, lectin-functional variants and fragments thereof. The present invention also provides for human GaINAc-T1-T16 lectin domains having the se- quences disclosed in Table III herein.

The present invention also provides for isolated nucleic acids encoding lectin domains of GalNAc-transferases. These nucleic acid sequences can encode for mam- malian GaINAc-transferase lectin domains, lectin-functional variants and fragments thereof. The present invention also provides for isolated nucleic acids encoding human GaINAc-Tl-T16 (these sequences are disclosed in Table III, herein).

The present invention further provides a method of inhibiting or modulating specific functions mediated by lectin domains of polypeptide GalNAc-transferases comprising administering an effective amount of an appropriate agent which is effective in binding to one or more lectin domains of polypeptide GalNAc-transferases and inhib- iting functions mediated by said lectin domains.

The present invention specifically discloses a novel selective inhibitor of poly- peptide GaINAc-transferase lectin domains, which provides a major advancement in that this inhibitor and related inhibitors sharing common characteristics of activity binds lectin domains without serving as acceptor substrate for glycosyltransferases involved in synthesis of O-glycans. For this reason treatment of cells with this and related com-

pounds is not expected to affect O-glycosylation extension including O-glycan branch- ing and sialylation. A preferred embodiment of this invention is a selective inhibitor of GalNAc-transferase lectins that does not serve as substrates for other glycosyltrans- ferases.

In a preferred embodiment, the novel selective GalNAc-transferase lectin in- hibitor is represented by the (3-anomeric configuration of GaINAc-benzyl, GaINAcp- benzyl. Additional preferred selective inhibitors including but not limited to the follow- ing: GalNAcpl-R (R represents any aglycone such as benzyl, phenyl, p-nitrophenyl, or umbrelliferyl, without limitation), selective GaINAcp-peptides, mimetics or compounds derived herefrom. The identifying characteristics of the preferred selective inhibitor is that it retains the binding activity for lectin domains similar to GalNAcp-benzyl and GalNAca-benzyl as determined in binding assays as disclosed in the present invention, and lacks the ability to serve as acceptor substrate for glycosyltransferases involved in O-glycosylation as determined by standard enzyme assays widely known by the skilled in the art.

The lectin domains confer unique properties to the GalNAc-transferases includ- ing but not limited to selective GaINAc-glycopeptide substrate specificity, as well as binding properties for peptides and carbohydrates to enhance catalytic properties and other functions related to the O-glycosylation process. In a preferred embodiment, methods of selectively blocking the lectin mediated acceptor substrate specificities of such GalNAc-transferase isoforms by ßGalNAc and ßGalNAc containing structures are disclosed.

The present invention provides a novel method for large scale screening of test substances for the ability to inhibit lectin-mediated activity of polypeptide GaINAc- transferases in a cell-free assay, which comprises: (i) contacting an isolated polypeptide GaINAc-transferase, an isolated lectin domain from a polypeptide GalNAc-transferase, or a fragment of a polypeptide GalNAc-transferase capable of diplaying lectin-mediated binding on its substrate, with one or more test substances under assay conditions suitable for the detection of said binding ability; and

(ii) measuring whether said lectin-mediated activity is thereby inhibited or modu- lated by one or more of the substances.

The present invention also provides a method of screening one or more test sub- stances for the ability to inhibit or modulate intracellular transport and/or cell surface expression of mucins, O-glycosylated glycoproteins, glycoproteins and proteins in a cell-based assay, which comprises: (i) contacting a cell that expresses mucins, O-glycosylated glycoproteins, glyco- proteins and proteins, with one or more test substances under assay conditions suitable for the detection of inhibition or modulation of said expression; and (ii) measuring whether intracellular transport and cell surface expression of said mucins, O-glycosylated glycoproteins, glycoproteins and proteins are thereby inhibited or modulated by one or more of the substances.

The present invention also provides a method of screening one or more test sub- stances for the ability to inhibit or modulate secretions of mucins, O-glycosylated gly- coproteins, glycoproteins and proteins in a cell-based assay, which comprises: (i) contacting a cell that secretes mucins, O-glycosylated glycoproteins, glyco- proteins with one or more test substances under assay conditions suitable for the detection of inhibition or modulation of said secretion; and (ii) measuring whether secretion of said mucins, O-glycosylated glycoproteins, glycoproteins and proteins are thereby inhibited or modulated by one or more of the substances.

Substances identified as agents which are effective in binding to one or more lectin domains of polypeptide GaINAc-transferases using the above method of the in- vention may e. g. be selected from the group consisting of naturally or non-naturally oc- curring carbohydrates, peptides, glycopeptides, glycoconjugates and portions and frag- ments thereof. They may also be found among nucleic acids as well as small organic or inorganic molecules. They include but are not limited to peptides such as soluble pep- tides including Ig-tailed fusion peptides, members of random peptide libraries and com- binatorial chemistry-derived molecular libraries made of D-and/or L-configuration

amino acids, phosphopeptides (including members of random or partially degenerate, directed phosphopeptide libraries), antibodies [e. g. polyclonal, monoclonal, humanized, anti-idiotypic, chimeric, single chain antibodies, fragments, (e. g. Fab, F (ab) 2, and Fab expression library fragments, and epitope-binding fragments thereof)], and polypep- tides. A substance to be used as an agent according to the invention may be an endoge- nous physiological compound or it may be a natural or synthetic compound.

Agents in accordance with the present invention are useful for changing the density and sites of O-glycan occupancy in mucins and 0-linked glycoproteins. Further uses are in changing Golgi-transport and intracellular sorting events conferred by the lectin domains of GaINAc-transferases. For example, inhibitors of lectin domains of GaINAc-transferases may be useful in manipulating disease associated O-glycosylation to augment immunity and to prepare vaccines. Further use may be found in manipulat- ing mucin secretion and O-glycan density in diseases associated with mucous accumu- lation to decrease secretion and enhance clearance of mucins. Further use may entail modulating O-glycosylation of recombinant glycoproteins by inhibition of polypeptide GalNAc-transferases in host expression cells. These and other aspects of the present in- vention will become evident upon reference to the following detailed description and drawings.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a schematic of the predicted domain structure of polypeptide Gal- NAc-transferases 31. Polypeptide GalNAc-transferases are predicted to be type II trans- membrane proteins with a cytoplasmic N-terminal tail, a non-cleaved hydrophobic transmembrane retention signal (TM, grey box), a stem region of variable size, a well conserved catalytic unit of approximately 340 amino acid residues, and a poorly con- served C-terminal domain of approximately 130 amino acids which display structural similarity to the lectin ricin and have been designated the"lectin domain". The putative lectin domains consist of a-, (3-, and y-ricin-like repeats. C indicates conserved cystine residues among sequences.

Figure 2 is a multiple sequence alignment (ClustalW) of putative lectin do- mains derived from 16 human polypeptide GalNAc-transferases. Potitions of conserved motifs CLD and QxW in the a, (3, and y repeats are indicated. The numbering indicated in the margins reflects numbering of the analysed sequence region of each GaINAc- transferase. Conserved residues are indicated by black box'ing.

Figures 3A and B are schematic representations of human soluble secreted MUC1 expression constructs used for stable transfectants of CHO cells. Panel A: IgG2A His-tag was inserted into Bsu36I/Xbal site of MUC1FL, generating a His- tagged MUC I construct containing the endogenous MUC1 secretion signal peptide.

Panel B: MuclFL Sau3AI insert was inserted into the BamHI site of pcDNA-inf., gen- erating a non-tagged MUC1 construct containing the y-interferon secretion signal pep- tide.

Figure 4 is a plot of absorbance v. ligand dilution showing GalNAc-transferase binding to GaINAc-MUC1 glycopeptide. A direct binding assay (ELISA) mediated by the lectin domain was developed and validated with soluble secreted GaINAc-T4 and- T2 enzyme proteins. ELISA plates were coated with peptides or glycopeptides at 1 pg/ml, blocked with BSA, and incubated with biotinylated enzymes. After washing, bound enzyme proteins were detected with HRP-Streptavidin as described in detail in Example 2. Secreted soluble constructs of GaINAc-transferases which are enzymati- cally active may bind to (glyco) peptide substrates through their catalytic units as origi- nally described for GalNAc-T2 32. However, GaINAc-transferase binding to substrates by the catalytic domain requires UDP and divalent cat-ions (binding destroyed by EDTA treatment), in accordance with previous experience 32. Panel A: GaINAc-T4 wild type enzyme proteins (o) and GaINAc-T2 (a) selectively bind GaINAc-MUCI glycopeptides, with no significant binding observed to the unglycosylated peptide (GaINAc-T4 wt (A) and GalNAcT2 (A)). Panel B: Furthermore, the GaINAc-T4 lectin mutant did not bind to either glycosylated Mucl (GaINAc-Mucl) (A) or non gly- cosylated Mucl (Mucl) (o), whereas GaINAc-T4 wild type binds GaINAc-Mucl (o) but not nonglcosylated Mud (t). Binding was not affected by 10 mM EDTA. Soluble

secreted GaINAc-T4 mutant, GalNAc-T4459H 2, in which the lectin domain has been se- lectively inactivated by a single amino acid substitution, showed no binding demon- strating that the binding observed with the wild type enzyme is mediated through the lectin domain.

Figure 5 is a plot (absorbance at 495 v. concentration of inhibitor) showing in- hibition of GaINAc-T4 lectin binding. Direct binding assays were performed with pre- incubation of GaINAc-T4 with inhibitors followed by incubation of GaINAc-T4 in ELISA plates activated with GaINAc-Mucl as described in detail in Example 2. Gal- NAco-benzyl (t) as well as GaINAcp-benzyl (o) inhibit at 3-6 mM, whereas the con- trol GIcNAca-benzyl (A) showed no inhibition. This demonstrates that the GaINAc- transferase lectin domains show no specificity for the anomeric configuration of Gal- NAc, and identifies a novel inhibitor, GaINAcp-benzyl, of GaINAc-transferase lectins.

Figures 6 A-J are a series of photomicrographs showing immunostaining of wild type CHO and transfected wild type CHO cells with a secreted MUC1 construct.

CHOldID/MUCsoI-cloneD5 was established from this population. MUC1 expression in the cytoplasm of 10-20 % cells is visualized by HMFG2, SM3, and vu-4H5 antibodies.

Anti-T antibody HH8 reacted only after neuraminidase pretreatment and the anti-Tn antibody reacted similarly before and after neuraminidase treatment. This suggests that cells grown in GaINAc alone produce mainly the Tn glycoform of MUC1, while cells grown in Gal and GaINAc produce mainly the sialylated T (core 1) glycoforms.

Figure 7 is a series of SDS-PAGE Western analysis of MUC1 secreted from wild type CHO cells stably transfected cells with a secreted MUC1 construct (CHOldlD/MUCsol-cloneD5). Neu + indicates pretreatment of samples with neura- minidase as described in Examples. Cells were grown in culture medium after 24 or 48 hours analysed directly or after neuraminidase treatment.

Figure 8 A-X (left to right from top to bottom) is a series of photomicrographs showing immunostaining of CHO Idled cells stably transfected cells with a full coding

cell surface secreted MUCI construct. CHOIdID/MUClF-clone2 cells were grown in Optimem medium without and with 1 mM GaINAc, and 1 mM GaINAc plus 0.1 mM Gal for 24-48 hours. Cells were trypsinized, washed, air-dried on cover slides, and im- munostained as described in Examples with antibodies to MUC1 and T and Tn carbo- hydrates. Reactivity was evaluated before and after neuraminidase treatment of dried acetone fixed cells. +/-neu indicates that the staining was identical with or without neuraminidase pretreatment.

Figures 9A and B are SDS-PAGE Western analysis of MUCI secreted from CHO Idled cells stably transfected cells with a secreted MUC I construct.

CHOldlD/MUCsol-cloneD5 cells were grown in the presence or absence of sugars in- dicated, and samples of the culture medium analysed directly after 24-48 hours. Positive control (GaINAc-peptide) is a 60-mer MUC1 tandem repeat GaINAc-glycopeptide gly- cosylated with human polypeptide GalNAc-transferase GaINAc-T2. Lane labeled con- trol includes medium from CHO Idled cells. Anti-MUC1 monoclonal antibodies 5E5 and HMFG2 were used.

Figures 10A and B are SDS-PAGE Western analysis of MUC1 secreted from CHO Idled cells stably transfected cells with a secreted MUC1 construct. Same experi- ment as Figure 9 using anti-MUC1 monoclonal antibodies VU-4H5 and VU-2G7.

Figures 11A-D are a series of photomicrographs (left to right from top to bot- tom of) anti-MUC1 antibody immunofluorescense staining of CHO IdID cells stably transfected with a full coding cell surface MUC1 (CHOIdID/MUClF-clone2). Cell grown in the presence of GaINAc were treated with the 0-glycosylation inhibitor Gal- NAca-benzyl or control GlcNAca-benzyl. Cells were grown in plates and stained without permeabilization as described in Example 4.

Figure 12 is an SDS-PAGE Western analysis of GalNAca-benzyl inhibition of MUC1 expression in CHO IdID cells stably transfected with a full coding MUC1 con- struct. Cells were grown for 24 hours (lanes 1-6) or 48 hours (lanes 7-12) in the pres-

ence of 1 mM GaINAc (lanes 1-3 and 7-9) or 1 mM GaINAc and 0.1 mM Gal (lanes 4- 6 and 10-12) to limit core O-glycosylation to GalNAcal-O-Ser/Thr and Gal (31- 3GalNAcal-O-Ser/Thr, respectively. Cells were further treated with 2 mM GalNAca- benzyl (lanes 1, 4,7, 10), 2 mM GIcNAca-benzyl (lanes 2,5, 8,11) or no inhibitor (lanes 3,6, 9,12). Cells were washed and lysed at 24 or 48 hours and the lysates sub- jected to immunoprecipitation with monoclonal antibody HMFG2, which broadly rec- ognize MUC 1 glycoforms. Immunoprecipitates were analysed by SDS-PAGE and western blot using HMFG2 antibody to detect MUC1 expression. Lane M indicates mo- lecular markers with assigned mw. Lane C includes a control MUC1 180-mer tandem repeat peptide which has been GaINAc-glycosylated with 3 moles GaINAc per repeat using GalNAc-T2. The sharp bands migrating at 100-200 kd are immunoglobulins indi- cated by IgG. At 24 hours the MUCI glycoforms expressed by cells grown in GaINAc or Gal and GalNAc migrated similarly, indicating that synthesis of sialylated core 1 O- glycans were time limited (lanes 1-6). At 48 hours, MUC1 glycoforms migrating as higher molecular weight species were expressed by cells grown in Gal and GalNAc (lanes 11-12). Treatment with GalNAca-benzyl had no significant effect at 24 hours (lanes 1 and 4), but after 48 hours a significant reduction in MUC1 expression was found in cells grown in GalNAc as well as in Gal and GalNAc (lanes 7 and 10). In the latter case a significant shift in migration further confirmed that GalNAca-benzyl also serves as an inhibitor of O-glycan extension and reduces O-glycosylation to Gal- NAcal-O-Ser/Thr. GIcNAca-benzyl served as a control and had no effect on MUC1 expression and O-glycosylation compared to untreated cells (lanes 8 and 11).

Figure 13 is a SDS-PAGE Western analysis resulting from the same experi- ment as in Figure 12, but using a novel monoclonal antibody, 5E5, to visualize MUC1 expression. Cells were grown for 48 hours in the presence of 1 mM GalNAc (lanes 1-3) or 1 mM GalNAc and 0.1 mM Gal (lanes 4-6) to limit core O-glycosylation to Gal- NAcal-O-Ser/Thr and Galßl-3GalNAcal-O-Ser/Thr, respectively. Cells were further treated with 2 mM GalNAca-benzyl (lanes 1,4), 2 mM GIcNAca-benzyl (lanes 2,5) or no inhibitor (lanes 3,6). Cells were washed and lysed at 48 hours and the lysates sub- jected to immunoprecipitation with monoclonal antibody HMFG2, which broadly rec-

ognize MUC1 glycoforms. Immunoprecipitates were analysed by SDS-PAGE and western blot using 5E5 antibody, which selectively recognize GaINAc-glycosylated MUCI expression and show no reactivity with unglycosylated MUC1 peptides. Lanes M and C as described in legend to Figure 12. Treatment with GalNAca-benzyl pro- duced a significant reduction in MUC1 expression in cells grown in GalNAc as well as in Gal and GalNAc (lanes 1 and 4). In cells grown in Gal and GalNAc (lanes 4-6) only weak expression of MUC1 was detected, but treatment of cells with GalNAca-benzyl still produced a marked shift in migration to lower molecular weight migrating species Figure 14 is an SDS-PAGE Western analysis showing the identification of a novel inhibitor, GaINAcp-benzyl, which exhibits the same effect on mucin transport as GalNAca-benzyl, but does not affect 0-glycan extension and O-glycosylation in gen- eral. CHO Idled cells stably transfected with a full coding MUC1 construct were grown for 36 hours in the presence of 1 mM GalNAc (lanes 1-3 and 7-9) or 1 mM GalNAc and 0.1 mM Gal (lanes 4-6 and 10-12) to limit core O-glycosylation to GalNAcal-O- Ser/Thr and Galß1-3GalNAcα1-O-Ser/Thr, respectively. Cells were further treated with 2 mM GalNAca-benzyl (lanes 1, 4,7, 10), 2 mM GaINAcp-benzyl (lanes 2,5, 8,11) or 2 mM GIcNAca-benzyl (lanes 3,6, 9,12). Cells were washed and lysed at 36 hours and the lysates subjected to immunoprecipitation with monoclonal antibodies HMFG2 (lanes 1-6) or 5E5 (lanes 7-12). Immunoprecipitates were analysed by SDS-PAGE and western blot using HMFG2 antibody. Lanes M and C as described in legend to Figure 7. Treatment with GaINAcp-benzyl produced the same or better reduction in MUC1 expression as treatment with GalNAca-benzyl in cells grown in GalNAc as well as in Gal and GalNAc (lanes 2,5, 8). In cells grown in Gal and GalNAc (lanes 4-6) MUC1 expression was reduced with GalNAcß-benzyl treatrnent (lane 5), but in contrast to cells treated with GalNAca-benzyl (lane 4), GalNAcp-benzyl produced no change in the mi- gration of MUCI demonstrating that this inhibitor does not affect O-glycosylation. The lack of immunoprecipitation of MUCI by 5E5 in cells grown in Gal and GalNAc (lanes 10-12) indicates that MUC1 is glycosylated with more complex structures than Gal- NAcal-O-Ser/Thr as recognized by this antibody.

Figure 15 is an SDS-PAGE Western analysis resulting from the same experi- ment as Figure 14, except that expression is visualized by the monoclonal antibody 5E5. Cells were grown for 36 hours in the presence of 1 mM GalNAc (lanes 1-3 and 7- 9) or 1 mM GalNAc and 0.1 mM Gal (lanes 4-6 and 10-12) to limit core O- glycosylation to GalNAcal-O-Ser/Thr and Galpl-3GalNAcal-O-Ser/Thr, respec- tively. Cells were further treated with 2 mM GalNAca-benzyl (lanes 1,4, 7,10), 2 mM GalNAcp-benzyl (lanes 2,5, 8,11) or 2 mM GlcNAca-benzyl (lanes 3,6, 9,12). Cells were washed and lysed at 36 hours and the lysates subjected to immunoprecipitation with monoclonal antibodies HMFG2 (lanes 1-6) or 5E5 (lanes 7-12). lmmunoprecipi- tates were analysed by SDS-PAGE and western blot using 5E5 antibody. Lanes M and C as described in legend to Figure 12. Treatment with GaINAcp-benzyl produced the same or better reduction in MUC1 expression as treatment with GalNAca-benzyl in cells grown in GalNAc (lanes 2 and 8). The lack of immunostaining of MUC1 by 5E5 in cells grown in Gal and GalNAc indicates that MUC1 is glycosylated with more com- plex structures than GalNAcal-O-Ser/Thr as recognized by this antibody.

Figures 16A-O (left to right from top to bottom) is a series of photomicro- graphs showing that the main O-glycan phenotype of CHO ldlD cells grown in Gal and GalNAc is sialylated T and that the 5E5 antibody does not react with MUC1, with T or silaylated T glycoforms of MUC1.

Figure 17 illustrates SDS-PAGE Western analysis of MUC5AC secretion from HT29MTX cells using polyclonal anti-MUC5AC antibody LUM5-1. In order to evalu- ate GalNAcß-benzyl as an inhibitor of mucin glycosylation and secretion we used the HT-29 metotrexate (MTX) cultured colon carcinoma cells of mucin secreting pheno- type. HT-29 MTX cells have a goblet cell like phenotype with constitutive production of MUC5AC. Cells were continuously grown up to 21 days in the presence of 5 mM GalNAca-benzyl, 5 mM GaINAcp-benzyl, 5 mM GIcNAca-benzyl and without inhibi- tor. SDS-PAGE Western blotting of media (10 pL) using the polyclonal anti-MUC5AC antibody LUM5-1 detected MUC5AC in media from untreated cells (Lane 1) and cells treated with GlcNAca-benzyl (Lane 2). In striking contrast, MUC5AC was not detected

in the media from cells treated with GalNAca-benzyl (Lane 3) or GaINAc (3-benzyl (Lane 4). Results shown are from media collected at 14 days.

Figure 18 illustrates immuno-fluorescence staining of HT29MTX cells treated with inhibitors as indicated and stained with the anti-MUC5AC monoclonal antibody, CLH2. Treatment of cells with GalNAca-benzyl (Panel A) induced a storage phenotype with increased intracellular staining of MUC5AC. Cells treated with GaINAcp-benzyl (Panel B) did not lead to a storage disease phenotype, and furthermore a clearly dimin- ished intracellular staining of MUC5AC was found compared to cells treated with GlcNAca-benzyl (Panel C) and untreated cells (Panel D).

DETAILED DESCRIPTION OF THE INVENTION All patent applications, patents, and literature references cited in this specifica- tion are hereby incorporated by reference in their entirety. In the case of conflict, the present description, including definitions, is intended to control.

1. Definitions The terms used in this specification generally have their ordinary meanings in the art, within the context of this invention and in the specific context where each term is used. Certain terms are discussed below, or elsewhere in the specification, to provide additional guidance to the practitioner in describing the compositions and methods of the invention and how to make and use them.

As used herein, the term"about"or"approximately"means within 50% of a given value, preferably within 20%, more preferably within 10%, more preferably still within 5%, and most preferably within 1 % of a given value. Alternatively, the term "about"or"approximately"means that a value can fall within a scientifically acceptable error range for that type of value, which will depend on how qualitative a measurement can be given the available tools. "About"or"approximately"may define a distribution around a mean value, rather than a single value.

Molecular Biology Definitions. In accordance with the present invention, there may be employed conventional molecular biology, microbiology and recombinant DNA techniques within the skill of the art. Such techniques are explained fully in the

literature. See, for example, Sambrook, Fitsch & Maniatis, Molecular Cloning: A Laboratory Manual, Second Edition (1989) Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York (referred to herein as"Sambrook et al. , 1989"); DNA Clon- ing: A Practical Approach, Volumes I and II (D. N. Glover ed. 1985); Oligonucleotide Synthesis (M. J. Gait ed. 1984); Nucleic Acid Hybridization (B. D. Hames & S. J. Hig- gins, eds. 1984); Animal Cell Culture (R. I. Freshney, ed. 1986); Immobilized Cells and Enzymes (IRL Press, 1986); B. E. Perbal, A Practical Guide to Molecular Cloning (1984); F. M. Ausubel et al. (eds.), Current Protocols in Molecular Biology, John Wiley & Sons, Inc. (1994).

"Nucleic acid"or"polynucleotide"as used herein refers to purine-and pyrimidine-containing polymers of any length, either polyribonucleotides or polydeoxy- ribonucleotides or mixed polyribo-polydeoxyribo nucleotides. This includes single- and double-stranded molecules, i. e. , DNA-DNA, DNA-RNA and RNA-RNA hybrids, as well as"protein nucleic acids" (PNA) formed by conjugating bases to an amino acid backbone. This also includes nucleic acids containing modified bases (see below).

"Complementary DNA or cDNA"as used herein refers to a DNA molecule or sequence that has been enzymatically synthesised from the sequences present in an mRNA template, or a clone of such a DNA molecule. A"DNA Construct"is a DNA molecule or a clone of such a molecule, either single-or double-stranded, which has been modified to contain segments of DNA that are combined and juxtaposed in a manner that would not otherwise exist in nature. By way of non-limiting example, a cDNA or DNA which has no introns are inserted adjacent to, or within, exogenous DNA sequences.

A plasmid or, more generally, a vector, is a DNA construct containing genetic information that may provide for its replication when inserted into a host cell. A plas- mid generally contains at least one gene sequence to be expressed in the host cell, as well as sequences that facilitate such gene expression, including promoters and tran- scription initiation sites. It may be a linear or closed circular molecule.

Nucleic acids are"hybridizable"to each other when at least one strand of one nucleic acid can anneal to another nucleic acid under defined stringency conditions.

Stringency of hybridization is determined, e. g. , by a) the temperature at which hybridi- zation and/or washing is performed, and b) the ionic strength and polarity (e. g., forma-

mide) of the hybridization and washing solutions, as well as other parameters. Hybridi- zation requires that the two nucleic acids contain substantially complementary se- quences; depending on the stringency of hybridization, however, mismatches may be tolerated. Typically, hybridization of two sequences at high stringency (such as, for example, in an aqueous solution of 0. 5X SSC, at 65°C) requires that the sequences ex- hibit some high degree of complementarity over their entire sequence. Conditions of intennediate stringency (such as, for example, an aqueous solution of 2X SSC at 65°C) and low stringency (such as, for example, an aqueous solution of 2X SSC at 55°C), re- quire correspondingly less overall complementarily between the hybridising sequences.

Hybridization stringency has been defined in numerous publication known to the skilled in the art (Meinkoth and Wahl, Anal. Biochem. 138, 267-284, 1984; Maniatis et al., Mo- lecular cloning: a laboratory manual. Cold Spring Harbor Laboratory Press, 1989; J. Q. Zhang, Eur. J. Biochem. 239,835-841 : 1996; M. Friedman-Einat, General and Com- parative Endocrinology 115, 354-363: 1999: M. Szabo, J. Bacteriology, 1544-1553: 1995; S. Sau, J. Bacteriology, 21182126,1996). Nucleic acids are"hybridizable"to each other when at least one strand can anneal to another nucleic acid under defined stringency conditions. High stringency hybridization is defined as 42°C over night hybridization under standard conditions (Maniatis et al. , Molecular cloning: a laboratory manual.

Cold Spring Harbor Laboratory Press, 1989), followed by 5 washes with 2XSSC, 0. 1% SDS at 42°C, once with 0. 5XSSC, 0. 1% SDS at 55°C and once with 0. 1XSSC, 0. 1% SDS at 55°C. (1XSSC is 0. 15M NaCI, 0. 015M Na citrate). Northern and Southern nucleic acid blotting hybridization techniques, especially for the purpose of investigat- ing hybridization specificity, is well known to those skilled in the field of the invention.

An"isolated"nucleic acid or polypeptide as used herein refers to a component that is removed from its original environment (for example, its natural environment if it is naturally occurring). An isolated nucleic acid or polypeptide contains less than about 50%, preferably less than about 75%, and most preferably less than about 90%, of the cellular components with which it was originally associated.

A"probe"refers to a nucleic acid that forms a hybrid structure with a sequence in a target region due to complementarily of at least one sequence in the probe with a sequence in the target region.

A nucleic acid that is"derived from"a designated sequence refers to a nucleic acid sequence that corresponds to a region of the designated sequence. This encom- passes sequences that are homologous or complementary to the sequence, as well as "sequence-conservative variants"and"function-conservative variants". Sequence- conservative variants are those in which a change of one or more nucleotides in a given codon position results in no alteration in the amino acid encoded at that position.

Function-conservative variants of polypeptide GalNAc-transferases are those in which a given amino acid residue in the polypeptide has been changed without altering the overall conformation and enzymatic activity (including substrate specificity) of the native polypeptide; these changes include, but are not limited to, replacement of an amino acid with one having similar physico-chemical properties. This includes but is not limited to, replacement of an amino acid with one having similar structural or physical properties, including polar or non-polar character, size, shape and charge (see, e. g., Ta- ble A).

A"polypeptide"is a chain of chemical building blocks called amino acids that are linked together by chemical bonds called"peptide bonds". The term"protein"refers to polypeptides that contain the amino acid residues encoded by a gene or by a nucleic acid molecule (e. g. , an mRNA or a cDNA) transcribed from that gene either directly or indirectly. Optionally, a protein may lack certain amino acid residues that are encoded by a gene or by an mRNA. For example, a gene or mRNA molecule may encode a se- quence of amino acid residues on the N-terminus of a protein (i. e. , a signal sequence) that is cleaved from, and therefore may not be part of, the final protein. A protein or polypeptide, including an enzyme, may be a"native"or"wild-type", meaning that it occurs in nature; or it may be a"mutant", "variant"or"modified", meaning that it has been made, altered, derived, or is in some way different or changed from a native pro- tein or from another mutant.

A"mutation"means any process or mechanism resulting in a mutant protein, enzyme, polypeptide, polynucleotide, gene, or cell. This includes any mutation in which a protein, enzyme, polynucleotide, or gene sequence is altered, and any detect- able change in a cell arising from such a mutation. The altered protein, enzyme, poly- peptide or polynucleotide is a Amutant@, also called a Avariant. @ Typically, a mutation occurs in a polynucleotide or gene sequence, by point mutations (substitutions), dele-

tions, or insertions of single or multiple nucleotide residues. A mutation includes polynucleotide alterations arising within a protein-encoding region of a gene as well as alterations in regions outside of a protein-encoding sequence, such as, but not limited to, regulatory or promoter sequences. A mutation in a gene can be"silent", i. e. , not re- flected in an amino acid alteration upon expression, leading to a"se- quence-conservative"variant of the gene. This generally arises when one amino acid corresponds to more than one codon. Table A outlines which amino acids correspond to which codon (s).

Thus, due to the degeneracy of the genetic code, any three-nucleotide codon that encodes the GalNAc-transferase lectin domain polypeptides described herein is within the scope of the invention.

The terms Amutant@ and Avariant@ may also be used to indicate a modified or altered gene, DNA or RNA sequence, enzyme, cell, etc., i. e., any kind of mutant. Such changes also include changes in the promoter, ribosome binding site, etc.

As outlined above, amino acid substitutions are generally based on the relative similarity of the amino acid side-chain substituents, for example, their hydrophobicity, hydrophilicity, charge, size, and the like. Exemplary substitutions which take various of the foregoing characteristics into consideration are well known to those of skill in the art and include: arginine and lysine ; glutamate and aspartate; serine and threonine; glutamine and asparagine ; and valine, leucine and isoleucine.

In addition, modifications, which do not normally alter the primary sequence of the GalNAc-transferase lectin domain polypeptides, include in vivo or in vitro chemical derivatization of polypeptides, e. g. , acetylation, methylation, or carboxylation. Also included as variant polypeptides of this invention are these polypeptides modified by glycosylation, e. g. , those made by modifying the glycosylation patterns of a polypeptide during its synthesis and processing or in further processing steps; or by exposing the polypeptide to enzymes which affect glycosylation, such as mammalian glycosylating or deglycosylating enzymes. Also embraced as variant polypeptides are the above- identified mutagenized sequences, which have phosphorylated amino acid residues, e. g. , phosphotyrosine, phosphoserine, or phosphothreonine.

TABLE A Amino Acids, Corresponding Codons, and Functionality/Property Amino Acid SLC DNA codons Side Chain Property Isoleucine 1 ATT, ATC, ATA Hydrophobic Leucine L CTT, CTC, CTA, CTG, TTA, TTG Hydrophobic Valine V GTT, GTC, GTA, GTG Hydrophobic Phenylalanine F TTT, TTC Aromatic side chain Methionine M ATG Sulphur group Cysteine C TGT, TGC Sulphur group Alanine A GCT, GCC, GCA, GCG Hydrophobic Glycine G GGT, GGC, GGA, GGG Hydrophobic Proline P CCT, CCC, CCA, CCG Secondary amine Threonine T ACT, ACC, ACA, ACG Aliphatic hydroxyl Serine S TCT, TCC, TCA, TCG, AGT, AGC Aliphatic hydroxyl Tyrosine T TAT, TAC Aromatic side chain Tryptophan W TGG Aromatic side chain Glutamine Q CAA, CAG Amide group Asparagine N AAT, AAC Amide group Histidine H CAT, CAC Basic side chain Glutamic acid E GAA, GAG Acidic side chain Aspartic Acid D GAT, GAC Acidic side chain Lysine K AAA, AAG Basic side chain Arginine R CGT, CGC, CGA, CGG, AGA, AGG Basic side chain Stop codons Stop TAA, TAG, TGA

As referred to herein, "sequence similarity"means the extent to which nucleo- tide or protein sequences are related. The extent of similarity between two sequences can be based on percent sequence identity and/or conservation. Amino acids other than those indicated as conserved may differ in a protein or enzyme so that the percent pro-

tein or amino acid sequence similarity between any two proteins of similar function may vary and can be, for example, at least 70%, preferably 75%, more preferably 80%, even more preferably 85%, and most preferably at least 90%, as determined according to an alignment scheme.

"Sequence identity"herein means the extent to which two nucleotide or amino acid sequences are invariant.

"Sequence alignment"means the process of lining up two or more sequences to achieve maximal levels of sequence identity (and, in the case of amino acid sequences, conservation), e. g., for the purpose of assessing the degree of sequence similarity. Nu- merous methods for aligning sequences and assessing similarity and/or identity are known in the art such as, for example, the ClustalW method, the Cluster Method, wherein similarity is based on the MEGALIGN algorithm, as well as BLASTN, BLASTP, and FASTA (Lipman and Pearson, 1985; Pearson and Lipman, 1988). When using all of these programs, the preferred settings are those that result in the highest se- quence similarity.

The term Ahost cell@ means any cell of any organism that is selected, modified, transformed, grown, or used or manipulated in any way, for the production of a sub- stance by the cell, for example the expression by the cell of a gene, a DNA or RNA se- quence, a protein or an enzyme.

A"donor substrate"is a molecule recognised by, e. g. , a polypeptide GalNAc- transferees and that contributes a N-acetylgalactosamine moiety for the transferase reac- tion. For polypeptide GaINAc-transferases, a donor substrate is UDP-N- acetylgalactosamine or with some GalNAc-transferase isoforms UDP-galactose. An "acceptor substrate"is a molecule, preferably a peptide, protein, glycopeptide, and gly- coprotein, that is recognised by, e. g. , a polypeptide GalNAc-transferase and that is the target for the modification catalysed by the transferase, i. e. , receives the carbohydrate moiety. For polypeptide GaINAc-transferases, acceptor substrates include without limi- tation peptides, proteins, glycopeptides, and glycoproteins.

The term"agonist"refers to a molecule that increases the amount of, or pro- longs the duration of, the activity of the polypeptide. The term"enhancer"refers to a molecule that similarly increases the amount of, or prolongs the duration of, the activity of the polypeptide. The term"antagonist"refers to a molecule, which decreases the bio-

logical or immunological activity of the polypeptide. The term"inhibitor"similarly re- fers to a molecule, which decreases the biological or immunological activity of the polypeptide. Agonists, antagonists, and inhibitors may include proteins, nucleic acids, carbohydrates, or any other molecules that associate with a polypeptide GaINAc- transferase.

The term"agent"includes small molecules, peptide mimetics and polypeptides.

"Mimetics"of GaINAc-transferase lectin-domain inhibitors are molecules that functionally mimic the structure or function of a GalNAc-transferase lectin-domain in- hibitor. Molecular mimetics include, but are not limited to: small organic compounds; nucleic acids and nucleic acid derivatives; saccharides or oligosaccharides; peptide mi- metics including peptides, proteins, and derivatives thereof, such as peptides containing non-peptide organic moieties, synthetic peptides which may or may not contain amino acids and/or peptide bonds, but retain the structural and functional features of a peptide ligand; pyrrolidines; peptoids and oligopeptoids which are molecules comprising N- substituted glycine, such as those described by Simon et al. (1992) Proc. Natl. Acad.

Sci. USA 89: 9367.

The human N-acetylgalactosaminyltransferase T1 gene (GALNTI) has been de- scribed previously 32. The sequence of the GALNT1 mRNA and the sequence of the GaINAc-T1 polypeptide have been submitted to GenBank/EBI Data Bank and assigned accession numbers X85018 and CAA59380, respectively.

The human N-acetylgalactosaminyltransferase T2 gene (GALNT2) has been de- scribed previously 32. The sequence of the GALNT2 mRNA and the sequence of the GalNAc-T2 polypeptide have been submitted to GenBank/EBI Data Bank and assigned accession numbers X85019 and CAA59381, respectively.

The human N-acetylgalactosaminyltransferase T3 gene (GALNT3) has been de- scribed previously 33. The sequence of the GALNT3 mRNA and the sequence of the GaINAc-T3 polypeptide have been submitted to GenBank/EBI Data Bank and assigned accession numbers X92689 and CAA63371, respectively.

The human N-acetylgalactosaminyltransferase T4 gene (GALNT4) has been de- scribed previously 10. The sequence of the GALNT4 mRNA and the sequence of the GalNAc-T4 polypeptide have been submitted to GenBank/EBI Data Bank and assigned accession numbers Y08564 and CAA69875, respectively.

The human N-acetylgalactosaminyltransferase T5 gene (GALNTS) has been de- scribed previously. The sequence of the GALNT5 mRNA and the sequence of the GaINAc-T5 polypeptide have been submitted to GenBank/EBI Data Bank and assigned accession numbers AJ245539 and CAB65104, respectively.

The human N-acetylgalactosaminyltransferase T6 gene (GALNT6) has been de- scribed previously 34. The sequence of the GALNT6 mRNA and the sequence of the GaINAc-T6 polypeptide have been submitted to GenBank/EBI Data Bank and assigned accession numbers AJ133523 and CAB55325, respectively. The human N- acetylgalactosaminyltransferase T7 gene (GALNT7) has been described previously 8 The sequence of the GALNT7 mRNA and the sequence of the GaINAc-T7 polypeptide have been submitted to GenBank/EBI Data Bank and assigned accession numbers AJ002744 and CAB60270, respectively.

The human N-acetylgalactosaminyltransferase T8 gene (GALNT8) has been de- scribed previously 35. The sequence of the GALNT8 mRNA and the sequence of the GaINAc-T8 polypeptide have been submitted to GenBank/EBI Data Bank and assigned accession numbers AJ271385 and CAB89199, respectively.

The human N-acetylgalactosaminyltransferase T9 gene (GALNT9) has been de- scribed previously 36. The sequence of the GALNT9 mRNA and the sequence of the GaINAc-T8 polypeptide have been submitted to GenBank/EBI Data Bank and assigned accession numbers AB040672 and BAB 13699, respectively.

The human N-acetylgalactosaminyltransferase T10 nucleic acid sequence (GALNT10) and the sequence of the encoded GalNAc-T10 polypeptide have been sub- mitted to GenBank/EBI Data Bank. The nucleic acid accession number is AJ505950.

The human N-acetylgalactosaminyltransferase Tl 1 gene (GALNTl l) has been described previously 3. The sequence of the GALNTIl mRNA and the sequence of the GaINAc-Tl 1 polypeptide have been submitted to GenBank/EBI Data Bank and as- signed accession numbers Y12434 and CAC79625, respectively.

The human N-acetylgalactosaminyltransferase T12 nucleic acid sequence (GALNT12) and the sequence of the GaINAc-T12 polypeptide have been submitted to GenBank/EBI Data Bank. The nucleic acid accession number is AJ132365. This se- quence is disclosed herewith.

The human N-acetylgalactosaminyltransferase T13 nucleic acid sequence (GALNT13) and the sequence of the GalNAc-T13 polypeptide have been submitted to GenBank/EBI Data Bank. The nucleic acid accession number is AR153422.

The references cited above for describing human GalNAc-TI-13 are incorpo- rated herein by reference in their entirety and to the same extent as if each reference was individually incorporated by reference.

Expression to produce enzymatically-active polypeptide GalNAc-transferases can be carried out in any number of conventional expression systems familiar to those skilled in the art. In one embodiment, GalNAc-transferases are expressed in a secreted soluble form, which can be recovered from the culture medium. Such secreted enzymes lack the N-terminal cytoplasmic tail and transmembrane retention sequence, and have N-terminal sequence starting in the predicted stem region (depicted domain structures of polypeptide GalNAc-transferases shown in Figure 1). The boundaries of the stem is N-terminally defined by the hydrophobic signal sequence, while the C-terminal bound- ary is less clearly defined but limited to the conserved catalytic unit of the enzymes as defined by multiple sequence alignments. For some isoforms including GaINAc-T2 the N-terminal sequence have been determined in naturally occurring soluble proteins de- rived from proteolytic cleavage. In another embodiment, host cells (e. g. CHO cells) are engineered to express full coding GalNAc-transferases and glycosylate substrates in vivo in host cells.

Expression to produce functional lectin domains of polypeptide GaINAc- transferases without the catalytic unit (or activity) can be carried out in any number of conventional expression systems familiar to those skilled in the art. In one embodiment, GaINAc-transferase lectins are expressed in a secreted soluble form, which can be re- covered from the culture medium. Such secreted soluble forms lack the N-terminal cy- toplasmic tail, transmembrane retention sequence, stem region and the catalytic unit.

The boundaries of the catalytic units and lectin domains are defined by multiple se- quence alignments and experimentation of lectin binding activity (multiple sequence alignment analysis of the C-terminal sequences polypeptide GalNAc-transferases in- cluding the most C-terminal boundaries of the catalytic domains and the entire lectin domains shown in Figure 2). The boundaries cannot be clearly defined but the most C- terminal well-conserved sequence motif of the catalytic units (WYLENVYP) can be

excluded from the lectin domains. Parts of or the entire catalytic domains may be in- cluded to produce functional lectin domains, and inclusion of inactivating mutations in the catalytic units (e. g. mutations in the DxH motif important for donor substrate bind- ing, or residues important for acceptor substrate binding 4) may be used to avoid addi- tional binding activity mediated through the catalytic units. In another embodiment, host cells (e. g. CHO cells) are engineered to express full coding polypeptide GaINAc- transferases with or without mutations in their catalytic units and binding mediated through lectin domains are detremiend in vivo in host cells.

Cells stably or transiently transfected with full coding or secreted expression constructs of mucins, mucin-like glycoproteins, O-glycosylated proteins, or proteins can be carried out by any number of conventional methods familiar to those skilled in the art. In one embodiment, the mucin MUC1 is expressed in a soluble form, which can be recovered from the culture medium (Figure 3 illustrates MUC1 expression constructs used in this invention; the DNA sequence is available from GenBank accession number M61170). In another embodiment, host cells (e. g. CHO or CHO Idled cells) are engi- neered to express MUC1 on the cell surface. In a preferred embodiment of the invention the cells are mammalian and more preferably, the cells are human.

Human cell lines expressing cell surface mucins or secreting mucins can be se- lected, cultured and treated by any number of conventional methods familiar to those skilled in the art. In one embodiment, mucins are expressed in a secreted soluble form without transmembrane retention sequence, which can be recovered from the culture medium. In another embodiment, host cells (e. g. CHO IdID cells) are engineered to ex- press full coding mucins on the cell surface. 2.

2. General Aspects of the Invention The putative lectin domains of some GalNAc-transferases notably GaINAc-T4 and-T7 have previously been shown to be important for the GaINAc-glycopeptide sub- strate specificities exhibited by these GalNAc-transferase isoforms (PCT WO 01/82215). The mechanism by which the putative lectin domains exert this effect on the enzyme activities is unknown. However, because GaINAc and GalNAca-benzyl were found to selectively inhibit these activities it was hypothesized that the putative lectin domains functioned by recognizing the sugar or glycopeptide in a lectin-like inter-

action2. Considerable efforts have been applied to demonstrate actual binding without success in the past (Bennett et al. unpublished, personal communications).

In the present invention a direct binding assay was developed using secreted soluble GaINAc-T4 and-T2, and chemoenzymatically produced multimeric MUC1 tandem repeat GaINAc-glycopeptides (Figure 4). Short MUCI glycopeptides of tradi- tional length of 15-20 amino acids have failed to provide significant binding in the same assay system, and one improvement leading to the success was the application of ex- tended multimeric MUC1 GaINAc-glycopeptides. Binding was also found to an enzy- matically GalNAc-glycosylated fusion protein expressed in E. coli and containing 30 amino acids of the MUC2 tandem repeat. Another improvement was the use of bioti- nylation of the enzymes, which provided an improved signal compared to previous at- tempts with identifying retained or bound enzyme by measuring activity. The specific activities of GaINAc-transferases as measured in in vitro assays are relatively low, and in past attempts to use binding and elution of enzyme acitivity presumably the detection level was not sufficient to detect binding. The developed assay was validated to demon- strate binding through the lectin domains by several ways: i) binding was selective for GaINAc-glycosylated glycopeptides with no significant binding to unglycosylated pep- tides; ii) a single amino acid substitution in the lectin domains of GaINAc-T4 (and- T2), known to selectively destroy GaINAc-glycopeptide specificity of these enzymes without affecting the catalytic unit 2 abolished binding; iii) binding was not affected by EDTA treatment which is known to destroy catalytic activity of GaINAc-transferases 32, 37-39 ; iv) binding was selectively inhibited by the monosacharide GaINAc and not by other sugars.

In order to minimize the size and functional complexity of GaINAc-transferase lectins to be used as probes for binding studies, we used multiple sequence alignment analysis to predict and design suitable expression constructs for isolated lectin domains (Figure 2). In the present invention a direct binding assay was developed using isolated lectin domains of GaINAc-T4 and-T2 with minimal size (Figure 5). Analysis of the fine specificity of the binding by inhibition studies showed that GaINAc-T2 and-T4 lectins exhibit restricted specificity for GaINAc structures, and surprisingly that the anomeric configuration of the GaINAc residue is not important. Thus, both GalNAca-

and GaINAcp-benzyl inhibited binding to the same degree. The lectin Helix Pomatia (HP) was used as a control plant lectin with known binding specificity for GalNAca- structures. HP lectin showed a very different highly preferred binding specificity for GalNAca-structures.

Studies with GalNAca-benzyl have shown that this compound is effectively taken up by cells and used in the Golgi compartments. It is also well known in the art that sugar-aryl compounds are taken up by the cell and used in the Golgi compartments.

Thus, the surprising finding that GalNAc-transferase lectins can be inhibited by ßGal- NAc (GaINAcp-benzyl), provides a new tool to study the function of polypeptide Gal- NAc-transferases in vivo ; GaINAcp-benzyl, because it, too, will enter the cell and be used in the Golgi compartments.

Availability of a binding assay is a useful tool to identify and characterize in- hibitors of GalNAc-transferase lectins. In accordance with the binding assay method of the invention, GaINAc-transferases are contacted with a glycopeptide, glycoprotein, fu- sionprotein, or other appropriate structure or polymer containing the sugar hapten struc- ture recognized by the GalNAc-transferase lectins, preferably N-acetylgalactosamine, and the GalNAc-transferase protein or lectin bound is quantitatively measured. The GaINAc-transferases may be in the form of a secreted soluble construct as applied in this invention, and any extended or truncated construct of a GalNAc-transferase as well as hybrid fusion protein that maintains the binding properties of the lectin. The ligand may be in the form of a chemoenzymatically produced MUC1 GalNAc-glycopeptide as applied in this invention, and any glycopeptide, glycoprotein, fusionprotein of any size or sequence, or other appropriate structure or polymer containing the sugar hapten structure recognized by the GalNAc-transferase lectins. Synthesis and chemoenzymatic synthesis of glycopeptides are familiar to those in the art, and are described in the lit- erature cited above and in the Examples provided below. The ligand sugar may be GaINAc, N-acetylgalactosamine, or any other sugar in any linkage and sequence recog- nized by a GalNAc-transferase lectin. The binding assay may be an enzyme-linked solid phase immunoadsorption assay (ELISA) as applied in the invention, and any vari- ant assay hereof where binding to ligand can be detected including without limitation radioimmunoassay (RIA), surface plasmon resonance (SPR), chemoluminescense, nu-

clear magnetic resonance spectroscopy ('H-NMR), and other methods know in the art.

Binding may be detected by horse-radish-peroxidase HRP-Avidin biotin as applied in this invention, and any other detection system including without limitation enzyme re- actions, fluorescence, radioactivity, spectroscopy, spectrometry and other methods. The GaINAc-transferases may be labelled by biotinylation as applied in this invention, and any other labelling including without limitation antibody tags, enzymes, fluorochromes, radioisotopes and other methods know in the art, as well as detected by antibodies, phage antibody fragments or other binding proteins. The assay may used to characterize binding specificities of GaINAc-transferase lectins, screen and identify inhibitors of GalNAc-transferase lectins, and screen and identify competitive binders such as differ- ent GalNAc-transferase lectins and other lectins and proteins with binding properties for carbohydrates.

An in vivo model system for secretion of mucins was developed. A truncated secreted expression construct of the human cell surface mucin MUC 1 containing 32 tandem repeats (Figure 3), was stably transfected into CHO wild type and CHO Idled cells. Figure 6 illustrates intracellur expression of soluble MUC1 in wild type CHO transfectant clone, wtCHO/MUCl sol-clone-C4, visualized by multiple monoclonal anti-MUCI antibodies. Analysis of glycosylation was performed with a panel of anti- bodies with well-defined specificities for carbohydrate structures, and reactivity was mainly found with anti-T after pre-treatment with neuraminidase to remove sialic acids.

Weak staining with anti-Tn was also found in some cells. Figure 7 illustrates western blot analysis of secreted MUC1 from the same cells. High molecular weight MUC1 mi- grating with apparent mw higher than 300 Kd is labelled by HMFG2, SM3, and VU- 2G7, while all antibodies including VU-4H5 label a low molecular weight MUC1 mi- grating with apparent mw of 130 Kd and presumed to represent virtually unglycosylated MUC1. Pretreatment with neuraminidase decreased migration of the high molecular weight MUC1 bands, and anti-T antibody reactivity emerged. Stable MUC1 transfec- tants in CHO Idled showed similar patterns of reactivity when grown in Gal and Gal- NAc.

An in vivo model system for cell surface expression of mucins was developed.

A full coding expression construct of the human cell surface mucin MUC1 containing 32 tandem repeats (Figure 3), was stably transfected into CHO wild type and CHO Idled

cells. CHO IdID cells were originally established by Krieger et al. 40 and found to have a defect in UDP-Gal/GaINAc epimerase that renders the cells incapable of producing UDP-Gal and UDP-GaINAc. Lack of UDP-Gal limits the synthesis of all types of gly- coconjugates including glycosphingolipids, N-linked and 0-linked glycoproteins. The synthesis of 0-linked glycoproteins will be arrested at GalNAcal-O-Ser/Thr with or without addition of α2, 6 linked sialic acid. In the absence UDP-GaINAc mainly O- linked mucin-type glycoconjugates are affected, and essentially no glycosylation occur, as the first sugar attached is GaINAc. The defect in CHO Idled cells can be selectively restored by addition of 1 mM GalNAc and or 0.1 mM Gal to the growth medium Addition of both sugars essentially restores normal glycosylation, whereas addition of GalNAc alone limits O-glycosylation to GalNAcal-O-Ser/Thr with or without addition of α2, 6 linked sialic acid, and also affects galactosylation of N-linked glycosylation and glycolipid biosynthesis. Altschuler et al. 41 have previously shown that cell surface ex- pression of MUC I in CHO Idled cells requires addition of GalNAc.

Cell surface expression of MUC1 was established in stably transfected CHO wildtype and CHO Idled cells. MUC1 was detected at the surface of non-permeabilized cells using monoclonal anti-MUCI antibodies (Figure 8). In accordance with Alschuler et al. 41 MUC1 surface expression in CHO Idled cells was only found in cells grown in GalNAc or Gal and GaINAc, whereas cells grown without sugars or only in Gal failed to express MUC1 at the surface. In agreement with the conclusion drawn by Altchuler et al. 41 surface expression of MUC1 was dependent only on the first step in O- glycosylation, the addition of GalNAc.

MUC1 produced in CHO IdID cells grown without GalNAc is not accumulated in Golgi, but degraded in lysosomes 41. This indicates that measuring total MUC1 in cell lysates rather than exclusively at the cell surface may be used as a measure of MUC1 expression. The experiments shown in Figures 9-10 use immunoprecipitation of total cell lysates with anti-MUCl antibody followed by western blot analysis with the same or different anti-MUC1 antibody to quantify and characterize MUC1 expression in cells. MUC1 produced in cell grown without GalNAc or only in the presence of Gal migrate close to the predicted mass of the protein core. With the addition of GalNAc to the medium high molecular weight forms of MUC1 are found, and these react with all

antibodies except VU-2G7. The antibody 5E5 only reacts with Tn glycoforms and lack of reactivity with MUC1 from cells grown in Gal and GalNAc indicate that the majority of MUC1 produced is glycosylated with sialyl-T structures.

GalNAca-benzyl is a well-known inhibitor of 0-glycosylation extension Treatment of cells with 1-2 mM GalNAca-benzyl partially blocks core 1 O- glycosylation including a2, 3 sialylation. Treatment of cells with GalNAca-benzyl is also known to affect surface expression of mucins and O-glycosylated glycoproteins, as well as in some cases secretion of mucins. A number of mammalian cell lines have been treated with 1-2 mM GalNAca-benzyl in the past and the resulting effects on O- glycosylation as well as mucin secretion have varied with cell type (for a detailed re- view see 42).

The effect of GalNAca-benzyl on mucin transport and secretion has been con- cluded to be due to blockage of O-glycosylation extension 42. Figures 11-13 illustrate that CHO Idled cells grown in Gal and GalNAc (or wild type CHO cells), and treated with 2 mM GalNAca-benzyl in agreement with this, exhibits reduced expression of MUC1 as well as altered O-glycosylation as judged by an altered SDS-PAGE migration pattern. Wildtype CHO cells as well as CHO Idled cells grown in Gal and GalNAc pro- duce 0-glycans of the mono-and disialylated core 1 structures (NeuAca2-3Galß 1- 3 [NeuAca2-6] +/GalNAca-O-Ser/Thr). Treatment with GalNAca-benzyl results in some exposure of unsialylated core 1 as evaluated by staining with anti-T monoclonal antibody HH8, whereas only very little Tn is exposed as evaluated with anti-Tn mono- clonal antibody 5F4. The altered SDS-PAGE migration of MUC1 produced in CHO ldlD cells grown in Gal and GalNAc (or wild type CHO cells, not shown) shown in Figures 12-13 (lanes 4 and 10 when indicated) reflects mainly loss of sialic acids.

If the effect of GalNAca-benzyl treatment on mucin transport and secretion is due to inhibition of sialylation, then treatment of CHO Idled cells grown only in Gal- NAc and hence producing only GalNAcal-O-Ser/Thr O-glycosylation (neglible STn is produced as evidenced by lack of staining with anti-STn monoclonal antibodies 3F 1 and TKH2, while cells stain very strongly with anti-Tn monoclonal antibodies SF4 and 1E3), should have no effect on expression of MUC1 in these cells. Surprisingly as shown in Figures 12-13 (lanes 1 and 7 when indicated), GalNAca-benzyl treatment

does inhibit MUC1 expression in CHO Idled cells with O-glycosylation controlled and limited to Tn glycoforms. This result shows for the first time that mucin transport and secretion may be directly affected by treatment with GalNAca-benzyl and not through a mechanism involving inhibtion of sialylation or the O-glycosylation extension path- ways. Combined with the findings of Altchuler et al. 41, these results indicate that mucin transport and secretion requires some degree of GaINAc O-glycosylation, whereas O- glycan extension including sialylation seems to be of less importance for this process.

An appropriate control for GalNAca-benzyl treatment has not previously been studied. Selection of a benzyl monosaccharide that is not involved in and does not affect glycosylation pathways in cells is problematic. We chose to use GlcNAca-benzyl as such a control as this structure is not used in glycosylation pathways of mammalian glycoproteins and glycosphingolipids. As shown in Figures 12-13 (lanes 2,5, 8, and 11, when indicated) treatment of CHO Idled cells with 2 mM GlcNAca-benzyl had no ef- fect on MUC1 expression and glycosylation. GlcNAca-benzyl thus serves as a control for treatment of cells with benzyl sugars, and this is important because benzyl sugars and their biosynthetic products appear to aggregate in cells and cause morphological changes with prolonged treatment 26.

Since transport of mucin in cells was selectively inhibited by GalNAca-benzyl (and not GlcNAca-benzyl), even in cells limited to GalNAcal-O-Ser/Thr O- glycosylation, we hypothesized that polypeptide GalNAc-transferases and in particular their lectin domains could be involved in ensuring mucin transport and preventing di- rection to lysosomes. On the one hand, GIcNAca-benzyl does not in general inhibit GaINAc O-glycosylation and polypeptide GalNAc-transferase enzyme activity. On the other, we had previously discovered that the GaINAc-glycopeptide acceptor substrate specificities of some GaINAc-transferase isoforms including GaINAc-T4 and-T7 2 (PCT WO 01/85215), which suggested that these enzymes could be involved. These precepts thus give rise to plural hypothesis. One hypothesis would suggest that inhibi- tion of the GaINAc-glycopeptide acceptor substrate specificity of GalNAc-transferases leads to mucin glycoforms with lower density of O-glycan occupancy (shown in vitro for e. g. MUC1 tandem repeats 2), and that this decrease in O-glycan density results in increased targeting to lysosomal degradation and hence decrease in expression. Another

hypothesis would suggest that the lectin domains of GaINAc-transferases in general have the capacity to bind GaINAc and hence provide a lectin mediated chaperone-like function, which is required for Golgi transport of O-glycosylated proteins. Lectin chaperones are well known to function ER transport as well as in lysosomal targeting43, but the existence of such lectin chaperones for cell surface expression and secretion have not been demonstrated in the Golgi or trans-Golgi network.

As described above we found in the present invention that GalNAc-transferase lectins in a binding assay to GaINAc-glycopeptides surprisingly showed similar inhibi- tion with GalNAca-benzyl and GaINAcp-benzyl. This indicates that these lectins in contrast to many lectins including Helix Pomatia fail to distinguish the anomeric con- figuration of the monosaccharide hapten recognized. ßGalNAc is a rare linkage in mammalian glycoproteins and is found only in N-linked glycoproteins in man associ- ated with the hormone specific glycosylation pattern where it generally is sulphated.

Although, ßGalNAc is found in both ganglioseries (GalNAcpl-4GaIpl-4Glcpl-Cer) and globoseries (GalNAcßl-3Galal-4Galßl-4Glcßl-Cer) it is expected that treatment of cells with 2 mM GaINAcp-benzyl will not interfere significantly with glycosylation.

This is based on the findings that the ß1, 3galactosyltransferases (p3Gal-T4 and 03Gal- T5, respectively) involved in extending ßGalNAc in these two glycolipid structures show no or very poor activity with GalNAcß-benzyl 44'45. The only ßGalNAc contain- ing structure in 0-linked glycosylation is found in the blood group related Sda structure (GalNAcß1-4 (Neua2-3) Galpl-3GalNAcal-O-Ser/Thr) which has very restricted ex- pression 46.

We therefore tested if GalNAcß-benzyl treatment of cells showed the same ef- fects as GalNAca-benzyl treatment. As shown in Figures 14-15 (lane 5) GalNAcß- benzyl treatment does not interfere with O-glycosylation in contrast to GalNAca- benzyl (lane 4), as no difference in SDS-PAGE migration is observed. However, Gal- NAca-benzyl, as well as GalNAcß-benzyl treatment of cells, produces similar signifi- cant reduction in expression of MUC1 Figures 14-15. This shows surprisingly and for the first time that GalNAcß-benzyl represents a selective inhibitor of mucin transport and secretion. GalNAcß-benzyl is a novel preferred inhibitor of transport, surface ex- pression and secretion of O-glycosylated proteins and mucins, because it does not inter-

fere with the O-glycosylation extension process. GalNAcp-benzyl is not expected to accumulate biosynthetic oligosaccharide products similar to those found with Gal- NAca-benzyl treatment 26. The finding that GaINAcp-benzyl exerts these effects on mucin expression combined with the finding that it inhibits polypeptide GaINAc- transferases strongly indicate that the mechanism by which GalNAca-and GalNAcp- benzyl inhibits mucin expression is through inhibition of GalNAc-transferase lectins.

This supports the second hypothesis articulated above. Polypeptide GalNAc-transferase lectins thus represent prime targets for intervention with mucin secretion and cell sur- face expression, and GaINAcp-benzyl represents a novel selective prototype inhibitor for such intervention.

Preferred compounds for inhibition of GalNAc-transferase lectins are inactive as acceptor substrates for glycosyltransferases. In particular, the following glycosyl- transferase activities: core 1 UDP-Gal : GalNAc-peptide 31, 3galactosyltransferases, CMP-NeuAc: GaINAc-peptide a2,6sialyltransferases, and UDP-GIcNAc : pl, 3N- acetylglucosaminyltransferases involved in O-glycosylation, are inactive with the pre- ferred inhibitory compounds. Examples of such inhibitory compounds are GalNAcal- 0-benzyl with substitution of hydroxyl groups at C3 and/or C6 with methyl or acetyl groups to block acceptor sites.

The methods described herein are designed to identify substances and com- pounds that bind to and or modulate the biological activity of a polypeptide GaINAc- transferase lectin, including substances that interfere with or enhance the activity of a polypeptide GalNAc-transferase lectin.

GalNAc-transferase lectins may be used in the form of a truncated lectin do- main as shown in Example 3, as a secreted GalNAc-transferase enzyme as shown in Example 2, or as a truncated protein or fusion protein with or without catalytic activity but with retained lectin domain and carbohydrate binding activity.

Agents that modulate a polypeptide GalNAc-transferase lectin can be identified based on their ability to associate with such a lectin. Therefore, the invention also pro- vides a method of identifying agents that associate with a polypeptide GalNAc- transferase lectin. Agents identified using the method of the invention may be isolated, cloned and sequenced using conventional techniques. An agent that associates with a

polypeptide GalNAc-transferase lectin may be an agonist or antagonist of the biological or immunological activity of the lectin.

Agents that can associate with a polypeptide GalNAc-transferase lectin may be identified by reacting such GalNAc-transferase lectin with a test substance, which po- tentially associates with a polypeptide GalNAc-transferase lectin under conditions which permit the association, and removing and/or detecting the associated GaINAc- transferase lectin and substance. The substance-lectin complex, free substance, or non- complexed lectin may be assayed. Conditions, which permit the formation of substance- lectin complexes, may be selected having regard to factors such as the nature and amounts of the substance and the lectin.

The substance-lectin complex, free substance or non-complexed lectin may be isolated by conventional isolation techniques, for example, salting out, chromatography, electrophoresis, gel filtration, fractionation, absorption, polyacrylamide gel electropho- resis, agglutination, or combinations thereof. To facilitate the assay of the components, a labelled antibody against the transferase or the substance or a labelled lectin or a la- belled substance may be utilized. The antibodies, lectins, or test substances may be la- belled with a detectable substance as described above.

A polypeptide GalNAc-transferase lectin, or a test substance used in the method of the invention may be insolubilized. For example, a lectin, or a test substance may be bound to a suitable carrier such as agarose, cellulose, dextran, Sephadex, Sepharose, carboxymethyl cellulose polystyrene, filter paper, ion-exchange resin, plastic film, plas- tic tube, glass beads, polyamine-methyl vinyl-ether-maleic acid copolymer, amino acid copolymer, ethylene-maleic acid copolymer, nylon, silk, etc. The carrier may be in the shape of, for example, a tube, test plate, beads, disc, sphere etc. The insolubilized lectin or substance may be prepared by reacting the material with a suitable insoluble carrier using known chemical or physical methods, for example, cyanogen bromide coupling.

The invention also contemplates a method for evaluating an agent for its ability to modulate the biological activity of a polypeptide GalNAc-transferase lectin by assay- ing for an agonist or antagonist (i. e. enhancer or inhibitor) of the association of the lectin with a substance that interacts with the polypeptide (e. g. carbohydrate binding site or parts thereof). The basic method for evaluating whether an agent is an agonist or an- tagonist of the association of a polypeptide GalNAc-transferase lectin and a substance

that associates with the lectin is to prepare a reaction mixture containing the lectin and the substance under conditions which permit the formation of substance-lectin com- plexes, in the presence of a test agent. The test agent may be initially added to the mix- ture, or may be added subsequent to the addition of the lectin and substance. Control reaction mixtures without the test agent or with a placebo are also prepared. The forma- tion of complexes is detected and the formation of complexes in the control reaction, but not in the reaction mixture, indicates that the test agent interferes with the interac- tion of the lectin and substance. The reactions may be carried out in the liquid phase or the lectin, substance, or test agent may be immobilized as described herein.

It will be understood that the agonists and antagonists, i. e. enhancers and inhibi- tors, that can be assayed using the methods of the invention may act on one or more of the interaction sites on the lectin or substance including agonist binding sites, competi- tive antagonist binding sites, non-competitive antagonist binding sites or allosteric sites.

It will also be understood that competitive assays, in addition to direct assays, can be used to screen for and identify the agents of the present invention.

The invention also makes it possible to screen for antagonists that inhibit the ef- fects of an agonist of the interaction of a polypeptide GaINAc-transferase lectin with a substance capable of associating with the lectin. Thus, the invention may be used to as- say for an agent that competes for the same interacting site of a polypeptide GalNAc- transferase lectin.

Test compounds are screened from, for example, large libraries of synthetic or natural compounds. Numerous means are currently used for random and directed syn- thesis of saccharide, peptide, and nucleic acid based compounds. Examples of available libraries are synthetic compound libraries are commercially available from Maybridge Chemical Co. (Trevillet, Cornwall, UK), Comgenex (Princeton, NJ), Brandon Associ- ates (Merrimack, NH), and Microsource (New Milord, CT).

Agents which are effective in modulating a polypeptide GalNAc-transferase lectin can be identified, based on their ability to interfere with or enhance the lectin me- diated binding capacity of the GalNAc-transferase protein or fragment hereof contain- ing the lectin region. Therefore, the invention provides a method for evaluating a test substance for its ability to modulate the binding capacity of a polypeptide GalNAc- transferase lectin comprising

(a) reacting a binding substrate with a GalNAc-transferase or lectin polypeptide or fragment hereof in the presence of a test substance; (b) measuring the amount of binding substrate bound to the GalNAc-transferase polypeptide or fragment hereof, and (c) carrying out steps (a) and (b) in the absence of the test substance to determine if the substance interferes with or enhances the binding by the polypeptide GalNAc-transferase.

Suitable binding substrates for use in the methods of the invention are polypep- tides, glycopolypeptides, or glycoproteins, which are either synthetic or naturally occur- ring structures. The GalNAc-transferase lectin polypeptide may be obtained from natu- ral sources or produced using recombinant methods as described and referenced herein.

The binding or modifying substrates may be labelled with a detectable sub- stance as described herein, and the interaction of the polypeptide of the invention with the binding or modifying substrates will give rise to a detectable change. The detectable change may be colorimetric, photometric, radiometric, potentiometric, etc. The Gal- NAc-transferase lectin polypeptide is reacted with the binding or modifying substrates at a pH and temperature effective for the polypeptide to bind the substrates, and where preferably one of the components is labeled, to produce a detectable change. It is pre- ferred to use a buffer with the substrates to maintain the pH within the pH range effec- tive for the polypeptides. The buffer and substrates may be used as an assay composi- tion. Other compounds such as EDTA and detergents may be added to the assay com- position.

The reagents suitable for applying the methods of the invention to evaluate agents that modulate a polypeptide GaINAc-transferase lectin may be packaged into convenient kits providing the necessary materials packaged into suitable containers. The kits may also include suitable supports useful in performing the methods of the inven- tion.

Agents that modulate polypeptide GalNAc-transferase lectin (s) can also be identified by treating immortalized cells which express the transferase (s) with a test substance, and comparing the intracellular transport, degradation, surface expression, or

secretion of O-glycosylated proteins, mucins, and glycoproteins performed of the cells with those of the cells in the absence of the test substance and/or with immortalized cells which do not express the transferase (s). Examples of immortalized cells that can be used include human cell lines, Chinese hamster ovary (CHO) cells and mutant cells CHO Idled 40, which express polypeptide GaINAc-transferase (s) or lectin (s) and produce cell membrane bound or secreted forms of the human mucin MUC1. In the absence of an inhibitor the cells will produce and transport MUC1 to the cell surface or secrete MUC1 into the growth medium. Substances that reduce the cell surface expression or the quantity of MUC 1 in the medium may be considered an inhibitor.

The agents identified by the methods described herein, may be used for modu- lating the binding activity of a polypeptide GaINAc-transferase lectin, and they may be used as prototype drugs in the treatment of conditions mediated by a polypeptide Gal- NAc-transferase lectin and in designing further substances effective to treat such condi- tions. In particular, they may be used to alter density of O-glycosylation on glycopro- teins and mucins produced by cells, the intracellular transport and surface expression of glycoproteins and mucins, the secretion of glycoproteins and mucins, and other func- tions governed by the polypeptide GalNAc-transferases and their lectins in transport and secretion of glycoproteins and mucins.

Therefore, the present invention has potential application in the treatment of various disorders associated with aberrant O-glycosylation and/or mucin production in mammals, preferably humans. Such disorders include the following: tumors and can- cers, lungs diseases associated with mucous accumulation such as asthma, chronic bronchitis, smoker's lung, cystic fibrosis, diseases of exocrine glands associated with increased or decreased mucin secretion such as Sjgrens syndrome, dry mouth etc.

Other disorders include dysregulation of selectin-mediated leukocyte trafficking and would include but not be limited to disorders involving autoimmunity, arthritis, leu- kaemia's, lymphomas, immunosuppression, sepsis, wound healing, acute and chronic in action, cell mediated immunity, and the like.

The agents identified by the methods described herein, have potential applica- tion in treatment of tumors including inhibition of tumor metastasis and growth and/or regression of same. Tumor metastasis may be inhibited by inhibiting the adhesion of circulating cancer cells. The agents of the invention have particular potential application

may be especially useful in the treatment of various forms of neoplasia such as leukae- mias, lymphomas, melanomas, adenomas, sarcomas, and carcinomas of solid tissues in patients. In particular the composition may be used for treating malignant melanoma, pancreatic cancer, cervico-uterine cancer, cancer of the liver, kidney, stomach, lung, rectum, breast, bowel, gastric, thyroid, neck, cervix, salivary gland, bile duct, pelvis, mediastinum, urethra, bronchogenic, bladder, esophagus and colon, and Kaposi's Sar- coma which is a form of cancer associated with HIV-infected patients with Acquired Immune Deficiency Syndrome (AIDS). The substances etc. are particularly useful in the prevention and treatment of tumors of lining mucosa and glands and the metastases derived from these tumors.

Accordingly, the various agents may be formulated into pharmaceutical compo- sitions for administration to subjects in a biologically compatible form suitable for ad- ministration in vivo. By biologically compatible form suitable for administration in vivo is meant a form of the agent to be administered in which any toxic effects are out- weighed by the therapeutic effects. The agents may be administered to living organisms including humans, and animals. Administration of a therapeutically active amount of the pharmaceutical compositions of the present invention is defined as an amount effec- tive, at dosages and for periods of time necessary to achieve the desired result. For ex- ample, a therapeutically active amount of an agent may vary according to factors such as the disease state, age, sex, and weight of the individual, and the ability of pharmaceu- tical composition or polypeptide to elicit a desired response in the individual. Dosage regima may be adjusted to provide the optimum therapeutic response. For example, several divided doses may be administered daily or the dose may be proportionally re- duced as indicated by the exigencies of the therapeutic situation.

The active agent may be administered in a convenient manner such as by injec- tion (subcutaneous, intravenous, etc. ), oral administration, inhalation, transdermal ap- plication, or rectal administration. Depending on the route of administration, the active agent may be coated in a material to protect the agent from the action of enzymes, acids and other natural conditions that may inactivate it.

The compositions described herein can be prepared by methods known per se for the preparation of pharmaceutically acceptable compositions which can be adminis- tered to subjects, such that an effective quantity of the active agent is combined in a

mixture with a pharmaceutically acceptable vehicle. Suitable vehicles are described, for example, in Remington's Pharmaceutical Sciences (Remington's Pharmaceutical Sci- ences, Mack Publishing Company, Easton, Pa. , USA 1985). On this basis, the composi- tions include, albeit not exclusively, solutions of the agents in association with one or more pharmaceutically acceptable vehicles or diluents, and contained in buffered solu- tions with a suitable pH and iso-osmotic with the physiological fluids.

The phrase"pharmaceutically acceptable"refers to molecular entities and com- positions that are physiologically tolerable and do not typically produce an allergic or similar untoward reaction (for example, gastric upset, dizziness and the like) when ad- ministered to an individual. Preferably, and particularly where a immunogenic compo- sition is used in humans, the term"pharmaceutically acceptable"denotes approved by a regulatory agency (for example, the U. S. Food and Drug Agency) or listed in a gener- ally recognized pharmacopeia for use in animals (for example, the U. S. Pharmacopeia).

Toxicity and therapeutic efficacy of compounds can be determined by standard pharmaceutical procedures, for example in cell culture assays or using experimental animals to determine the LD50 and the ED50. The parameters LD50 and ED50 are well known in the art, and refer to the doses of a compound that are lethal to 50% of a population and therapeutically effective in 50% of a population, respectively. The dose ratio between toxic and therapeutic effects is referred to as the therapeutic index and may be expressed as the ratio: LD50/ED50. Compounds that exhibit large therapeutic indices are preferred. While compounds that exhibit toxic side effects may be used.

However, in such instances it is particularly preferable to use delivery systems that spe- cifically target such compounds to the site of affected tissue so as to minimize potential damage to other cells, tissues or organs and to reduce side effects.

Data obtained from cell culture assay or animal studies may be used to formu- late a range of dosages for use in humans. The dosage of compounds used in therapeu- tic methods of the present invention preferably lie within a range of circulating concen- trations that includes the ED50 concentration but with little or no toxicity (e. g. , below the LD50 concentration). The particular dosage used in any application may vary within this range, depending upon factors such as the particular dosage form employed, the route of administration utilized, the conditions of the individual (e. g. , patient), and so forth.

Non-human animals include, without limitation, laboratory animals such as mice, rats, rabbits, hamsters, guinea pigs, etc.; domestic animals such as dogs and cats; farm animals such as sheep, goats, pigs, horses, and cows; and non-human primates.

A therapeutically effective dose may be initially estimated from cell culture as- says and formulated in animal models to achieve a circulating concentration range that includes the IC50. The IC50 concentration of a compound is the concentration that achieves a half-maximal inhibition of symptoms (e. g. , as determined from the cell cul- ture assays). Appropriate dosages for use in a particular individual, for example in hu- man patients, may then be more accurately determined using such information.

Measures of compounds in plasma may be routinely measured in an individual such as a patient by techniques such as high performance liquid chromatography (HPLC) or gas chromatography.

Pharmaceutical compositions for use in accordance with the present invention may be formulated in conventional manner using one or more physiologically accept- able carriers or excipients.

Thus, the compounds and their physiologically acceptable salts and solvates may be formulated for administration by the routes described above.

For oral administration, the pharmaceutical compositions may take the form of, for example, tablets or capsules prepared by conventional means with pharmaceutically acceptable excipients such as binding agents (e. g. , pregelatinised maize starch, polyvi- nylpyrrolidone or hydroxypropyl methylcellulose); fillers (e. g. , lactose, microcrystalline cellulose or calcium hydrogen phosphate); lubricants (e. g. , magnesium stearate, talc or silica); disintegrants (e. g. , potato starch or sodium starch glycolate); or wetting agents (e. g. , sodium lauryl sulphate). The tablets may be coated by methods well known in the art. Liquid preparations for oral administration may take the form of, for example, solu- tions, syrups or suspensions, or they may be presented as a dry product for constitution with water or other suitable vehicle before use. Such liquid preparations may be pre- pared by conventional means with pharmaceutically acceptable additives such as sus- pending agents (e. g. , sorbitol syrup, cellulose derivatives or hydrogenated edible fats); emulsifying agents (e. g. , lecithin or acacia); non-aqueous vehicles (e. g. , almond oil, oily esters, ethyl alcohol or fractionated vegetable oils); and preservatives (e. g. , methyl

or propyl-p-hydroxybenzoates or sorbic acid). The preparations may also contain buffer salts, flavoring, coloring and sweetening agents as appropriate.

Preparations for oral administration may be suitably formulated to give con- trolled release of the active compound. For buccal administration the compositions may take the form of tablets or lozenges formulated in conventional manner. For ad- ministration by inhalation, the compounds for use according to the present invention are conveniently delivered in the form of an aerosol spray presentation from pressurized packs or a nebuliser, with the use of a suitable propellant, e. g. , dichlorodifluoromethane, trichlorofluoromethane, dichlorotetrafluoroethane, carbon dioxide or other suitable gas.

In the case of a pressurized aerosol the dosage unit may be determined by providing a valve to deliver a metered amount. Capsules and cartridges of e. g. , gelatin for use in an inhaler or insufflator may be formulated containing a powder mix of the compound and a suitable powder base such as lactose or starch.

The compounds may be formulated for parenteral administration by injection, e. g. , by bolus injection or continuous infusion. Formulations for injection may be pre- sented in unit dosage form, e. g. , in ampoules or in multi-dose containers, with an added preservative. The compositions may take such forms as suspensions, solutions or emul- sions in oily or aqueous vehicles, and may contain formulatory agents such as suspend- ing, stabilizing and/or dispersing agents. Alternatively, the active ingredient may be in powder form for constitution with a suitable vehicle, e. g. , sterile pyrogen-free water, before use.

The compounds may also be formulated in rectal compositions such as supposi- tories or retention enemas, e. g. , containing conventional suppository bases such as co- coa butter or other glycerides.

In addition to the formulations described previously, the compounds may also be formulated as a depot preparation. Such long acting formulations may be adminis- tered by implantation (for example subcutaneously or intramuscularly) or by intramus- cular injection. Thus, for example, the compounds may be formulated with suitable polymeric or hydrophobic materials (for example as an emulsion in an acceptable oil) or ion exchange resins, or as sparingly soluble derivatives, for example, as a sparingly soluble salt.

The compositions may, if desired, be presented in a pack or dispenser device that may contain one or more unit dosage forms containing the active ingredient. The pack may for example comprise metal or plastic foil, such as a blister pack. The pack or dispenser device may be accompanied by instructions for administration.

After pharmaceutical compositions have been prepared, they can be placed in an appropriate container and labelled for treatment of an indicated condition. For admini- stration of an inhibitor of a polypeptide GaINAc-transferase, such labelling would in- clude amount, frequency, and method of administration.

The use of inhibitors of the lectin domain mediated activities of the above men- tioned polypeptide GalNAc-transferase isoforms and other isoforms allows for unique selective inhibition of these functions in vitro and in vivo in cells and organisms. This is desirable in manipulating the density of O-glycans, e. g. changing high density O- glycosylated tumour-associated MUC1 to low density normal MUC1 in cells. Further this is desirable in inhibiting any adhesive role the lectin domains may play in Golgi transport and intracellular sorting.

EXAMPLES 1. Cloning, Expression, and Purification of Soluble GaINAc-Transferase Proteins and Soluble GaINAc-Transferase Lectins.

Polypeptide GalNAc-transferases are highly conserved throughout evolution.

Orthologous relationships can be defined from man to Drosophila, 48 and ortholgous members of all human polypetide GalNAc-transferase isoforms are clearly identifiable in mouse and rats, and likely all mammals.

Polypeptide GalNAc-transferases are predicted to be type II transmembrane Golgi-resident proteins with a domain structure depicted in Figure 1 2. The N-terminal cytoplasmic tail, the hydrophic transmembrane signal sequence, and the stem region may be involved in directing Golgi-localization 47. The catalytic unit of the enzymes is approximately 300-350 amino acid residues and highly conserved in primary sequence among isoforms and also throughout evolution of the gene family 3,48. The C-terminal region of approximately 130 amino acids exhibits similarity with the galactose binding

lectin, ricin. This region show little sequence similarity among isoforms and is poorly conserved in evolution 3.

Soluble, secreted expression constructs of human GaINAc-transferases Gal- NAc-T1,-T2,-T3,-T4,-T6,-T7, and-Tl 1 for baculo-virus mediated expression in in- sect cells have been described in detail previously 3, 8, i0, 32-34 His-tagged soluble expres- sion constructs for all human ppGalNAc-transferases, including novel genes designated GalNAc-T12,-T13,-T14,-T15, and-T16, were prepared using PCR primers as listed in Table I below.

TABLE 1. Primers used for PCR of soluble secreted GaINAc-transferase expression constructs.

Expression constructs were amplified by PCR using 20 ng plasmid DNA as template. Expand High Fidelity-kit (Roche) was used as recommended by the manufac- turer using an ABI2700 thermocycler (Applied Biosystems). Products were digested <BR> <BR> <BR> <BR> with EcoRI (GalNAc-T2,-T11,-T12 and-T16), BamHI (GalNAc-T1,-T3,-T4,-T5,- T6,-T8,-T9,-T 10,-T 13,-T 14 and-T 15) and HindlII (GaINAc-T7), and sub-cloned into the EcoRI or HindlII site of pBKS-HistagI or the BamHI site of pBKS-HistaglI.

PBKS-Histag-I and-II vectors were generated from pBluescrip (Stratagene), by insert- ing a fragment encoding 6xHis, a thrombin cleavage site, and a T7 antibody site. pBKS- Histag-I was modified with the sequence 5'-GCGGCCGCTCTAGAACTAGTGGATCCAGCAGCCATCATCATCATCATCACAGCAGCG GCC TGGTGCCGCGCGGCAGCCATATGGCTAGCATGACTGGTGGACAGCAAATGGGTCGCGGAA TTC CGATATCAAGCTTATCGATACCGTCGACCTCGAG-3'. pBKS-Histag-11 was modified with the sequence: 5'-GAATTCGCGGCCGCAGCAGCCATCATCATCATCATCACAGCAGCGGCCTGGTGCCGC GCG GCAGCCATATGGCTAGCATGACTGGTGGACAGCAAATGGATCCACTAGTTCTAGAGCGGC CGC -3 All construct were fully sequenced. His-tagged GalNAc-transferase pBKS-HIS- tag-I constructs were excised with Notl and XhoI (blunt-ended) or as for GaINAc-Tl 1, with NotI and HindIII (blunt-ended) and sub-cloned into the NotIBgllI (blunt-ended) site of the pAcGP67A Baculo expression vector (Pharmingen). His-tagged GaINAc- transferase pBKS-HIS-tag-II constructs were excised with Notl and inserted into the Notl site of pAcGP67A Baculo expression vector.

The coding region for human polypeptide GalNAc-T12 has been submitted to GenBank/EBI Data Bank and assigned accession number AJ132365 : Human GalNAc-T12 DNA sequence: ATGTGGGGGCGCACGGCGCGGCGGCGCTGCCCGCGGGAACTGCGGCGCGGCCGGGAGGCG CTGTTGGTGC TCCTGGCGCTACTGGCGTTGGCCGGGCTGGGCTCGGTGCTGCGGGCGCAGCGTGGGGCCG GGGCCGGGGC TGCCGAGCCGGGACCCCCGCGCACCCCGCGCCCCGGGCGGCGCGAGCCGGTCATGCCGCG GCCGCCGGTG CCGGCGAACGCGCTGGGCGCGCGGGGCGAGGCGGTGCGGCTGCAGCTGCAGGGCGAGGAG CTGCGGCTGC <BR> <BR> <BR> <BR> AGGAGGAGAGCGTGCGGCTGCACCAGATTAACATCTACCTCAGCGACCGCATCTCACTGC ACCGCCGCCT GCCCGAGCGCTGGAACCCGCTGTGCAAAGAGAAGAAATATGATTATGATAATTTGCCCAG GACATCTGTT ATCATAGCATTTTATAATGAAGCCTGGTCAACTCTCCTTCGGACAGTTTACAGTGTCCTT GAGACATCCC <BR> <BR> <BR> CGGATATCCTGCTAGAAGAAGTGATCCTTGTAGATGACTACAGTGATAGAGAGCACCTGA AGGAGCGCTT

Human GalNAc-T12 amino acid sequence: The coding region for human polypeptide GalNAc-T13 has been submitted to GenBank/EBI Data Bank and assigned accession number AR153422 : Human GalNAc-T13 DNA sequence:

Human GalNAc-T13 amino acid sequence: The coding region for human polypeptide GalNAc-T14 has been submitted to GenBank/EBI Data Bank and assigned accession number AJ505991: Human GalNAc-T14 DNA sequence:

Human GalNAc-T14 amino acid sequence: The coding region for human polypeptide GalNAc-T15 has been submitted to GenBank/EBI Data Bank and assigned accession number Y09324: Human GalNAc-T15 DNA sequence:

Human GalNAc-T15 amino acid sequence: The coding region for human polypeptide GalNAc-T16 has been submitted to GenBank/EBI Data Bank and assigned accession number AJ505951 : Human GalNAc-T16 DNA sequence:

Human GalNAc-T16 amino acid sequence: Additional homologous polypeptide GalNAc-transferase genes have been iden- tified and cloning and expression are in progress, and it follows from the descriptions that similar methods as outlined above will yield soluble secreted proteins for study.

Expression constructs may have immunoaffinity tags or purification tags at the N

terminal and/or C-terminal region. These may include myc, FLAG, HIS, GST, and other (Stratagene, Qiagen, Amersham Biosciences).

Soluble secreted expression constructs of GaINAc-transferase lectin domains were prepared from the GaINAc-transferase expression constructs described above by PCR using primer pairs as listed Table 11 below.

Table II. Primers used for amplification of GaINAc-transferase lectin domains GaINAc-Tl lectin domain : T1LECFOR : 5'-CAAAGGAAGCTTATGGAGATATATCGTCAAGAG-3' T1LECREV : 5'-GCAAGCTCGAGGCGGCCGCTCAGAATATTTCTGGAAGGGTGAC-3' GaINAc-T2 lectin domain : T2LECFOR : 5'-CAAGGAAGCTTCTTATGGAAATATTCAGAGCAGATTG-3' T2LECREV : 5'-GCAAGCTCGAGGCGGCCGCCTACTGCTGCAGGTTGAGC-3' GaINAc-T3 lectin domain : T3LECFOR : 5'-CAAGGAAGCTTCATTTGGTGATCTTTCAAAAAGATTT-3' T3LECREV : 5'-GCAAGCTCGAGGCGGCCGCAGGAACACTTAATCATTTTGG-3' GaINAc-T4 lectin domain : T4LECFOR : 5'-AGAAAAGAAGCTTATGGTGATATTTCTG-3' EBHC307 : 5'-AGCGGATCCGACGAAGTGCTGTTGTGCT-3' GaINAc-TS lectin domain : T5LECFOR : 5'-CAAGGAAGCTTTAGATGTTGGCAACCTCACCCAGC-3' T5LECREV : 5'-GCAAGCTCGAGGCGGCCGCAAGCATCAGTTACACTTCAGGCTTC-3' GaINAc-T6 lectin domain : T6LECFOR : 5'-CAAGGAAGCTTCCTTCGGTGACATTTCGGAACG-3' T6LECREV : 5'-GCAAGCTCGAGGCGGCCGCTGGGTCCTAGACAAAGAGCC-3' GaINAc-T7 lectin domain : T7LECFOR : 5'-AGAAAAGAAGCTTATGGGGATATATCGGAGCTG-3' T7LECREV : 5'-GCAAGCTCGAGGCGGCCGCTCTCTAAACACTATGGATGTTATTC-3' GaINAc-T8 lectin domain : T8LECFOR : 5'-CAAGGAAGCTTTTGGAGACGTTTCTTCCAGAATG-3' T8LECREV : 5'-GCAAGCTCGAGGCGGCCGCTCACTGGCTGTTGGTCTGACCCC-3' GaINAc-T9 lectin domain : T9LECFOR : 5'-CAAGGAAGCTTTCGGGGACGTGTCTGAGAGGCTG-3' T9LECREV : 5'-GCAAGCTCGAGGCGGCCGCTCAGTGCCGTGCGTGTTTGATCC-3' GaINAc-T10 lectin domain : T10LECFOR : 5'-CAAGGAAGCTTCCGCTGGGGATGTCGCAGTCCAG-3' T1OLECREV : 5'-GCAAGCTCGAGGCGGCCGCTCAGTTCCTATTGAATTTTTCC-3' GaINAc-T11 lectin domain : T11LECFOR : 5'-CAAGGAAGCTTGCAATATCAGTGAGCGTGTGG-3' T11LECREV : 5'-GCAAGCTCGAGGCGGCCGCCCACCTTAACCTTCCAAATGC-3' GalNAc-T12 lectin domain : T12LECFOR : 5'-CAAGGAAGCTTGGGATGTGACAGAGAGGAAG-3' T12LECREV : 5'-GCAAGCTCGAGGCGGCCGCTCATAACATGCGCTCTTTGAAGAACC-3' GaINAc-T 13 lectin domain : T13LECFOR : 5'-CAAGGAAGCTTCTGAGAAGCCAGACTGCATGG-3' T13LECREV : 5,-GCAAGCTCGAGGCGGCCGCTCATCGTTCATCCACAGCATTG-3' GalNAc-T14 lectin domain : T14LECFOR : 5'-CAAGGAAGCTTATGGAGATGTGTCAGTCAGAAAAAC-3' T14LECREV : 5'-GCAAGCTCGAGGCGGCCGCTCATGTGCCCAAGGTCATGTTCC-3' GaINAc-T15 lectin domain : T15LECFOR : 5'-CAAGGAAGCTTTCGGGAATGTTGAGAGCAGATTG-3' T15LECREV : 5'-GCAAGCTCGAGGCGGCCGCTCAAGAACTCACCATGTCCCAGTG-3' GaINAc-T16 lectin domain : T16LECFOR : 5'-CAAGGAAGCTTGCAGTGTGGCTACGCGGATAGAGCAGAGG-3' T16LECREV : 5'-GCAAGCTCGAGGCGGCCGCTCATGTGTGTGGCAACAGCTGCC-3' PCR amplifications were performed with 10 ng GalNAc-transferase plasmid DNA as template and High Fidelity PCR kit (Roche) with conditions recommended by

the manufacturer. Amplified products were digested with HindlII and XhoI and inserted into the HindIII/XhoI site of pBKS-HistagI. All constructs were fully sequenced.

Tagged lectin domain constructs were excised with NotI and sub-cloned into the Notl site of pAcGP67-A Baculo expression vector.

The exact borders of the lectin domains and the catalytic units have not been de- fined, but multiple sequence alignment analysis (Fig. 2) was used to predict the most likely borders and these were used for design of PCR primers as listed in Table II. DNA and amino acid sequences of preferred constructs of GaINAc-transferase lectin domains and their construct design include the following (Table III) : Table III: DNA and amino acid sequences of GaINAc-transferase lectin do- mains.

GaINAc-T1 lectin domain: The lectin domain polypeptide sequence comprises amino acid residues 393- 559 of GALNTI (GALNT 1 nucleotide sequence accession number is AJ505952) T1 LECTIN DNA sequence TTCCAGAAATATTCTGA-stop T1 LECTIN Amino acid sequence YGDISSRVGLRHKLQCKPFSWYLENIYPDSQIPRHYFSLGEIRNVETNQCLDNMARKENE KV GIFNCHGMGGNQVFSYTANKEIRTDDLCLDVSKLNGPVTMLKCHHLKGNQLWEYDPVKLT LQ HVNSNQCLDKATEEDSQVPSIRDCNGSRSQQWLLRNVTLPEIF* GaINAc-T2 lectin domain: The lectin domain polypeptide sequence comprises amino acid residues 408- 571 of GALNT2 (GALNT2 nucleotide sequence accession number is X85019).

T2 LECTIN DNA sequence TATCCAGAGTTAAGGGTTCCAGACCATCAGGATATAGCTTTTGGGGCCTTGCAGCAGGGA AC TAACTGCCTCGACACTTTGGGACACTTTGCTGATGGTGTGGTTGGAGTTTATGAATGTCA CA

ATGCTGGGGGAAACCAGGAATGGGCCTTGACGAAGGAGAAGTCGGTGAAGCACATGGATT TG TGCCTTACTGTGGTGGACCGGGCACCGGGCTCTCTTATAAAGCTGCAGGGCTGCCGAGAA AA TGACAGCAGACAGAAATGGGAACAGATCGAGGGCAACTCCAAGCTGAGGCACGTGGGCAG CA ACCTGTGCCTGGACAGTCGCACGGCCAAGAGCGGGGGCCTAAGCGTGGAGGTGTGTGGCC CG GCCCTTTCGCAGCAGTGGAAGTTCACGCTCAACCTGCAGCAGTAG-stop T2 LECTIN Amino acid sequence YPELRVPDHQDIAFGALQQGTNCLDTLGHFADGVVGVYECHNAGGNQEWALTKEKSVKHM DL CLTVVDRAPGSLIKLQGCRENDSRQKWEQIEGNSKLRHVGSNLCLDSRTAKSGGLSVEVC GP ALSQQWKFTLNLQQ* GaINAc-T3 lectin domain: The lectin domain polypeptide sequence comprises amino acid residues 467- 633 of GALNT3 (GALNT3 nucleotide sequence accession number is AJ505954). T3 LECTIN DNA sequence TCATTTGGTGATCTTTCAAAAAGATTTGAAATAAAACACCGTCTTCGGTGTAAAAATTTT AC ATGGTATCTGAACAACATTTATCCAGAGGTGTATGTGCCAGACCTTAATCCTGTTATATC TG GATACATTAAAAGCGTTGGTCAGCCTCTATGTCTGGATGTTGGAGAAAACAATCAAGGAG GC AAACCATTAATTATGTATACATGTCATGGACTTGGGGGAAACCAGTACTTTGAATACTCT GC TCAACATGAAATTCGGCACAACATCCAGAAGGAATTATGTCTTCATGCTGCTCAAGGTCT CG TTCAGCTGAAGGCATGTACCTACAAAGGTCACAAGACAGTTGTCACTGGAGAGCAGATAT GG GAGATCCAGAAGGATCAACTTCTATACAATCCATTCTTAAAAATGTGCCTTTCAGCAAAT GG AGAGCATCCAAGTTTAGTGTCATGCAACCCATCAGATCCACTCCAAAAATGGATACTTAG CC AAAATGATTAA-stop T3 LECTIN Amino acid sequence FGDLSKRFEIKHRLRCKNFTWYLNNIYPEVYVPDLNPVISGYIKSVGQPLCLDVGENNQG GK PLIMYTCHGLGGNQYFEYSAQHEIRHNIQKELCLHAAQGLVQLKACTYKGHKTVVTGEQI WE IQKDQLLYNPFLKMCLSANGEHPSLVSCNPSDPLQKWILSQND* GaINAc-T4 lectin domain: The lectin domain polypeptide sequence comprises amino acid residues 405- 578 of GALNT4 (GALNT4 nucleotide sequence accession number is Y08564).

T4 LECTIN DNA sequence GAGGATAGACCAGGCTGGCATGGGGCTATTCGCAGTAGAGGGATCTCGTCTGAATGTTTA GA TTATAATTCTCCTGACAACAACCCCACAGGTGCTAACCTTTCACTGTTTGGATGCCATGG TC AAGGAGGCAATCAATTCTTTGAATATACTTCAAACAAAGAAATAAGGTTTAATTCTGTGA CA GAGTTATGTGCAGAGGTACCTGAGCAAAAAAATTATGTGGGAATGCAAAATTGTCCCAAA GA TGGGTTCCCTGTACCAGCAAACATTATTTGGCATTTTAAAGAAGATGGAACTATTTTTCA CC CACACTCAGGACTGTGTCTTAGTGCTTATCGGACACCGGAGGGCCGACCTGATGTACAAA TG AGAACTTGTGATGCTCTAGATAAAAATCAAATTTGGAGTTTTGAGAAATAG-stop

T4 LECTIN Amino acid sequence AYGDISERKLLRERLRCKSFDWYLKNVFPNLHVPEDRPGWHGAIRSRGISSECLDYNSPD NN PTGANLSLFGCHGQGGNQFFEYTSNKEIRFNSVTELCAEVPEQKNYVGMQNCPKDGFPVP AN IIWHFKEDGTIFHPHSGLCLSAYRTPEGRPDVQMRTCDALDKNQIWSFEK* GaINAc-T5 lectin domain: The lectin domain polypeptide sequence comprises amino acid residues 486- 653 of GALTN5 (GALTN5 nucleotide sequence accession number is AJ505956). T5 LECTIN DNA sequence TTAGATGTTGGCAACCTCACCCAGCAAAGGGAGCTGCGAAAGAAACTGAAGTGCAAAAGT TT CAAATGGTACTTGGAGAATGTCTTTCCTGACTTAAGGGCTCCCATTGTGAGAGCTAGTGG TG TGCTTATTAATGTGGCTTTGGGTAAATGCATTTCCATTGAAAACACTACAGTCATTCTGG AA GACTGCGATGGGAGCAAAGAGCTTCAACAATTTAATTACACCTGGTTAAGACTTATTAAA TG TGGAGAATGGTGTATAGCCCCCATCCCTGATAAAGGAGCCGTAAGGCTGCACCCTTGTGA TA ACAGAAACAAAGGGCTAAAATGGCTGCATAAATCAACATCAGTCTTTCATCCAGAACTGG TG AATCACATTGTTTTTGAAAACAATCAGCAATTATTATGCTTGGAAGGAAATTTTTCTCAA AA GATCCTGAAAGTAGCTGCCTGTGACCCAGTGAAGCCATATCAAAAGTGGAAATTTGAAAA AT ATTATGAAGCCTGA-stop T5 LECTIN Amino acid sequence DVGNLTQQRELRKKLKCKSFKWYLENVFPDLRAPIVRASGVLINVALGKCISIENTTVIL ED CDGSKELQQFNYTWLRLIKCGEWCIAPIPDKGAVRLHPCDNRNKGLKWLHKSTSVFHPEL VN HIVFENNQQLLCLEGNFSQKILKVAACDPVKPYQKWKFEKYYEA* GaINAc-T6 lectin domain: The lectin polypeptide sequence comprises amino acid residues 458-622 of GALNT6 (GALNT6 nucleotide sequence accession number is AJ133523) T6 LECTIN DNA sequence TCCTTCGGTGACATTTCGGAACGACTGCAGCTGAGGGAACAACTGCACTGTCACAACTTT TC CTGGTACCTGCACAATGTCTACCCAGAGATGTTTGTTCCTGACCTGACGCCCACCTTCTA TG GTGCCATCAAGAACCTCGGCACCAACCAATGCCTGGATGTGGGTGAGAACAACCGCGGGG GG AAGCCCCTCATCATGTACTCCTGCCACGGCCTTGGCGGCAACCAGTACTTTGAGTACACA AC TCAGAGGGACCTTCGCCACAACATCGCAAAGCAGCTGTGTCTACATGTCAGCAAGGGTGC TC TGGGCCTTGGGAGCTGTCACTTCACTGGCAAGAATAGCCAGGTCCCCAAGGACGAGGAAT GG GAATTGGCCCAGGATCAGCTCATCAGGAACTCAGGATCTGGTACCTGCCTGACATCCCAG GA CAAAAAGCCAGCCATGGCCCCCTGCAATCCCAGTGACCCCCATCAGTTGTGGCTCTTTGT CT AG-stop T6 LECTIN Amino acid sequence SFGDISERLQLREQLHCHNFSWYLHNVYPEMFVPDLTPTFYGAIKNLGTNQCLDVGENNR GG KPLIMYSCHGLGGNQYFEYTTQRDLRHNIAKQLCLHVSKGALGLGSCHFTGKNSQVPKDE EW ELAQDQLIRNSGSGTCLTSQDKKPAMAPCNPSDPHQLWLFV* GaINAc-T7 lectin domain: The lectin domain polypeptide sequence comprises amino acid residues 492- 657 of GALNT7 (GALNT7 nucleotide sequence accession number is AJ505958).

T7 LECTIN DNA sequence TATGGGGATATATCGGAGCTGAAAAAATTTCGAGAAGATCACAACTGCCAAAGTTTTAAG TG GTTCATGGAAGAAATAGCTTATGATATCACCTCACACTACCCTTTGCCACCCAAAAATGT TG ACTGGGGAGAAATCAGAGGCTTCGAAACTGCTTACTGCATTGATAGCATGGGAAAAACAA AT GGAGGCTTTGTTGAACTAGGACCCTGCCACAGGATGGGAGGGAATCAGCTTTTCAGAATC AA TGAAGCAAATCAACTCATGCAGTATGACCAGTGTTTGACAAAGGGAGCTGATGGATCAAA AG TTATGATTACACACTGTAATCTAAATGAATTTAAGGAATGGCAGTACTTCAAGAACCTGC AC AGATTTACTCATATTCCTTCAGGAAAGTGTTTAGATCGCTCAGAGGTCCTGCATCAAGTA TT CATCTCCAATTGTGACTCCAGTAAAACGACTCAAAAATGGGAAATGAATAACATCCATAG TG TTTAG-stop T7 LECTIN Amino acid sequence YGDISELKKFREDHNCQSFKWFMEEIAYDITSHYPLPPKNVDWGEIRGFETAYCIDS MGKTNGGFVELGPCHRMGGNQLFRINEANQLMQYDQCLTKGADGSKVMITHCNL NEFKEWQYFKNLHRFTHIPSGKCLDRSEVLHQVFISNCDSSKTTQKWEMNNIHSV* GaINAc-T8 lectin domain : The lectin domain polypeptide sequence comprises amino acid residues 459- 637 of GALNT8 (GALNT8 nucleotide sequence accession number is AJ505959). T8 LECTIN DNA sequence GACGTTTCTTCCAGAATGGCACTCCGGGAAAAACTGAAATGTAAAACTTTTGAC TGGTACCTGAAAAATGTTTATCCACTCTTGAAGCCACTCCACACCATCGTGGGCT ATGGAAGAATGAAAAACCTATTGGATGAAAATGTCTGCTTGGATCAGGGACCCG TTCCAGGCAACACCCCCATCATGTATTACTGCCATGAATTCAGCTCACAGAATGT CTACTATCACCTAACTGGGGAGCTCTATGTGGGACAACTGATTGCAGAGGCCAG TGCTAGTGATCGCTGCCTGACAGACCCTGGCAAGGCGGAGAAGCCCACCTTAGA ACCATGCTCCAAGGCAGCTAAGAATAGACTGCATATATATTGGGATTTTAAACC GGGAGGAGCTGTCATAAACAGAGATACCAAGCGGTGTCTGGAGATGAAGAAGG ATCTTTTGGGTAGCCACGTGCTTGTGCTCCAGACCTGTAGCACGCAAGTGTGGGA AATCCAGCACACTGTCAGAGACTGGGGTCAGACCAACAGCCAGTGA// T8 LECTIN Amino acid sequence FGDVSSRMALREKLKCKTFDWYLKNVYPLLKPLHTIVGYGRMKNLLDENVCLDQG PVPGNTPIMYYCHEFSSQNVYYHLTGELYVGQLIAEASASDRCLTDPGKAEKPTLEP <BR> <BR> CSKAAKNRLHIYWDFKPGGAVINRDTKRCLEMKKDLLGSHVLVLQTCSTQVWEIQ<B R> <BR> <BR> HTVRDWGQTNSQ// GaINAc-T9 lectin domain: The lectin domain polypeptide sequence comprises amino acid residues 427- 603 of GALNT9 (GALNT9 nucleotide sequence accession number is AJ505960). T9 LECTIN DNA sequence TTCGGGGACGTGTCTGAGAGGCTGGCCCTGCGTCAGAGGCTGAAGTGTCGCAGCTTCAAG TG GTACCTGGAGAACGTGTACCCGGAGATGAGGGTCTACAACAACACCCTCACGTACGGAGA GG TGAGAAACAGCAAAGCCAGTGCCTACTGTCTGGACCAGGGAGCGGAGGACGGCGACCGGG CG ATCCTCTACCCCTGCCACGGGATGTCCTCCCAGCTGGTGCGGTACAGCGCTGACGGCCTG CT GCAGCTGGGGCCTCTGGGCTCCACAGCCTTCTTGCCTGACTCCAAGTGTCTGGTGGATGA CG GCACGGGCCGCATGCCCACCCTGAAGAGGTGTGAGGATGTGGCGCGGCCAACACAGCGGC TG TGGGACTTCACCCAGAGTGGCCCCATTGTGAGCCGGGCCACGGGCCGCTGCCTGGAGGTG GA

GATGTCCAAAGATGCCAACTTTGGGCTCCGGCTGGTGGTACAGAGGTGCTCGGGGCAGAA GT GGATGATCAGAAACTGGATCAAACACGCACGGCACTGA-stop T9 LECTIN Amino acid sequence FGDVSERLALRQRLKCRSFKWYLENVYPEMRVYNNTLTYGEVRNSKASAYCLDQG AEDGDRAILYPCHGMSSQLVRYSADGLLQLGPLGSTAFLPDSKCLVDDGTGRMPTL KRCEDVARPTQRLWDFTQSGPIVSRATGRCLEVEMSKDANFGLRLVVQRCSGQKW MIRNWIKHARH* GatNAc-TIO lectin domain: The lectin domain polypeptide sequence comprises amino acid residues 417- 603 of GALNTIO (GALNT10 nucleotide sequence accession number is AJ505950). T10 LECTIN DNA sequence GCTGGGGATGTCGCAGTCCAGAAAAAGCTCCGCAGCTCCCTTAACTGCAAGAGTTTCAAG TGGTTTAT GACGAAGATAGCCTGGGACCTGCCCAAATTCTACCCACCCGTGGAGCCCCCGGCTGCAGC TTGGGGGG AGATCCGAAATGTGGGCACAGGGCTGTGTGCAGACACAAAGCACGGGGCCTTGGGCTCCC CACTAAGG CTAGAGGGCTGCGTCCGAGGCCGTGGGGAGGCTGCCTGGAACAACATGCAGGTATTCACC TTCACCTG GAGAGAGGACATCCGGCCTGGAGACCCCCAGCACACCAAGAAGTTCTGCTTTGATGCCAT TTCCCACA CCAGCCCTGTCACGCTGTACGACTGCCACAGCATGAAGGGCAACCAGCTGTGGAAATACC GCAAAGAC AAGACCCTGTACCACCCTGTCAGTGGCAGCTGCATGGACTGCAGTGAAAGTGACCATAGG ATCTTCAT GAACACCTGCAACCCATCCTCTCTCACCCAGCAGTGGCTGTTTGAACACACCAACTCAAC AGTCTTGG AAAAATTCAATAGGAACTGA T10 LECTIN Amino acid sequence AGDVAVQKKLRSSLNCKSFKWFMTKIAWDLPKFYPPVEPPAAAWGEIRNVGTGLCADTKH G ALGSPLRLEGCVRGRGEAAWNNMQVFTFTWREDIRPGDPQHTKKFCFDAISHTSPVTLYD CH SMKGNQLWKYRKDKTLYHPVSGSCMDCSESDHRIFMNTCNPSSLTQQWLFEHTNSTVLEK F NRN* GaINAc-Tll lectin domain : The lectin domain polypeptide sequence comprises amino acid residues 492- 608 of GALNTll (GALNT11 nucleotide sequence accession number is Y12434). Tll LECTIN DNA sequence TGCAATATCAGTGAGCGTGTGGAACTGAGAAAGAAGTTGGGCTGTAAATCATTTAAATGG TA TTTGGATAATGTATACCCAGAGATGCAGATATCTGGGTCCCACGCCAAACCCCAACAACC CA TTTTTGTCAATAGAGGGCCAAAACGACCCAAAGTCCTTCAACGTGGAAGGCTCTATCACC TC CAGACCAACAAATGCCTGGTGGCCCAGGGCCGCCCAAGTCAGAAGGGAGGTCTCGTGGTG CT TAAGGCCTGTGACTACAGTGACCCAAATCAGATCTGGATCTATAATGAAGAGCATGAATT GG TTTTAAATAGTCTCCTTTGTCTAGATATGTCAGAGACTCGCTCATCAGACCCGCCACGGC TC ATGAAATGCCACGGGTCAGGAGGATCCCAGCAGTGGACCTTTGGGAAAAACAATCGGCTA TA CCAGGTGTCGGTTGGACAGTGCCTGAGAGCAGTGGATCCCCTGGGTCAGAAGGGCTCTGT CG CCATGGCGATCTGCGATGGCTCCTCTTCACAGCAGTGGCATTTGGAAGGTTAA Tll LECTIN Amino acid sequence NISERVELRKKLGCKSFKWYLDNVYPEMQISGSHAKPQQPIFVNRGPKRPKVLQRG RLYHLQTNKCLVAQGRPSQKGGLVVLKACDYSDPNQIWIYNEEHELVLNSLLCLD MSETRSSDPPRLMKCHGSGGSQQWTFGKNNRLYQVSVGQCLRAVDPLGQKGSVA MAICDGSSSQQWHLEG*

GaINAc-T12 lectin domain: The lectin domain polypeptide sequence comprises amino acid residues 428- 581 of GALNT12 (GALNT12 nucleotide sequence accession number is AJ505963). T12 LECTIN DNA sequence TGGGATGTGACAGAGAGGAAGCAGCTCCGGGACAAGCTCCAGTGTAAAGACTTCAAGTGG TT CTTGGAGACTGTGTATCCAGAACTGCATGTGCCTGAGGACAGGCCTGGCTTCTTCGGGAT GC TCCAGAACAAAGGACTAACAGACTACTGCTTTGACTATAACCCTCCCGATGAAAACCAGA TT GTGGGACACCAGGTCATTCTGTACCTCTGTCATGGGATGGGCCAGAATCAGTTTTTCGAG TA CACGTCCCAGAAAGAAATACGCTATAACACCCACCAGCCTGAGGGCTGCATTGCTGTGGA AG CAGGAATGGATACCCTTATCATGCATCTCTGCGAAGAAACTGCCCCAGAGAATCAGAAGT TC ATCTTGCAGGAGGATGGATCTTTATTTCACGAACAGTCCAAGAAATGTGTCCAGGCTGCG AG GAAGGAGTCGAGTGACAGTTTCGTTCCACTCTTACGAGACTGCACCAACTCGGATCATCA GA AATGGTTCTTCAAAGAGCGCATGTTATGA T12 LECTIN Amino acid sequence DVTERKQLRDKLQCKDFKWFLETVYPELHVPEDRPGFFGMLQNKGLTDYCFDYNPPDENQ IV GHQVILYLCHGMGQNQFFEYTSQKEIRYNTHQPEGCIAVEAGMDTLIMHLCEETAPENQK FI LQEDGSLFHEQSKKCVQAARKESSDSFVPLLRDCTNSDHQKWFFKERML* GaINAc-T13 lectin domain: The lectin domain polypeptide sequence comprises amino acid residues 466- 639 of GALNT13 (GALNT13 nucleotide sequence accession number is AJ505964). T13 LECTIN DNA sequence TCTGAGAAGCCAGACTGCATGGAACGCTTGCAGCTGCAAAGGAGACTGGGTTGTCGGACA TT CCACTGGTTTCTGGCTAATGTCTACCCTGAGCTGTACCCATCTGAACCCAGGCCCAGTTT CT CTGGAAAGCTCCACAACACTGGACTTGGGCTCTGTGCAGACTGCCAGGCAGAAGGGGACA TC CTGGGCTGTCCCATGGTGTTGGCTCCTTGCAGTGACAGCCGGCAGCAACAGTACCTGCAG CA CACCAGCAGGAAGGAGATTCACTTTGGCAGCCCACAGCACCTGTGCTTTGCTGTCAGGCA GG AGCAGGTGATTCTTCAGAACTGCACGGAGGAAGGCCTGGCCATCCACCAGCAGCACTGGG AC TTCCAGGAGAATGGGATGATTGTCCACATTCTTTCTGGGAAATGCATGGAAGCTGTGGTG CA AGAAAACAATAAAGATTTGTACCTGCGTCCGTGTGATGGAAAAGCCCGCCAGCAGTGGCG TT TTGACCAGATCAATGCTGTGGATGAACGATGA T13 LECTIN Amino acid sequence EKPDCMERLQLQRRLGCRTFHWFLANVYPELYPSEPRPSFSGKLHNTGLGLCADCQAEGD IL GCPMVLAPCSDSRQQQYLQHTSRKEIHFGSPQHLCFAVRQEQVILQNCTEEGLAIHQQHW DF QENGMIVHILSGKCMEAVVQENNKDLYLRPCDGKARQQWRFDQINAVDER* GaINAc-T14 lectin domain: The lectin domain polypeptide sequence comprises amino acid residues 352- 516 of of GALNT14 (GALNT14 nucleotide sequence accession number is AJ505991). T14 LECTIN DNA sequence TATGGAGATGTGTCAGTCAGAAAAACACTAAGAGAAAATCTGAAGTGTAAGCCCTTTTCT TGGTACCT AGAAAACATCTATCCGGACTCCCAGATCCCAAGACGTTATTACTCACTTGGTGAGATAAG AAATGTTG AAACCAATCAGTGTTTAGACAACATGGGCCGCAAGGAAAATGAAAAAGTGGGTATATTCA ACTGTCAT GGTATGGGAGGAAATCAGGTATTTTCTTACACTGCTGACAAAGAAATCCGAACCGATGAC TTGTGCTT

GGATGTTTCTAGACTCAATGGACCTGTAATCATGTTAAAATGCCACCATATGAGAGGAAA TCAGTTAT GGGAATATGATGCTGAGAGACTCACGTTGCGACATGTTAACAGTAACCAATGTCTCGATG AACCTTCT GAAGAAGACAAAATGGTGCCTACAATGCAGGACTGTAGTGGAAGCAGATCCCAACAGTGG CTGCTAAG GAACATGACCTTGGGCACATGA T14 LECTIN Amino acid sequence YGDVSVRKTLRENLKCKPFSWYLENIYPDSQIPRRYYSLGEIRNVETNQCLDNMGR KENEKVGIFNCHGMGGNQVFSYTADKEIRTDDLCLDVSRLNGPVIMLKCHHMRGN QLWEYDAERLTLRHVNSNQCLDEPSEEDKMVPTMQDCSGSRSQQWLLRNMTLGT* GalNAc-T15 lectin domain: The lectin domain polypeptide sequence comprises amino acid residues 382- 552 of GALNT15 (GALNT15 nucleotide sequence accession number is AJ505966). T15 LECTIN DNA sequence TCGGGAATGTTGAGAGCAGATTGGACCTGAGGAAGAATCTGCGCTGCCAGAGCTTCAAGT GG TACCTGGAGAATATCTACCCTGAACTCAGCATCCCCAAGGAGTCCTCCATCCAGAAGGGC AA TATCCGACAGAGACAGAAGTGCCTGGAATCTCAAAGGCAGAACAACCAAGAAACCCCAAA CC TAAAGTTGAGCCCCTGTGCCAAGGTCAAAGGCGAAGATGCAAAGTCCCAGGTATGGGCCT TC ACATACACCCAGAAGATCCTCCAGGAGGAGCTGTGCCTGTCAGTCATCACCTTGTTCCCT GG CGCCCCAGTGGTTCTTGTCCTTTGCAAGAATGGAGATGACCGACAGCAATGGACCAAAAC TG GTTCCCACATCGAGCACATAGCATCCCACCTCTGCCTCGATACAGATATGTTCGGTGATG GC ACCGAGAACGGCAAGGAAATCGGCGTCAACCCATGTGAGTCCTCACTCATGAGCCAGCAC TG GGACATGGTGAGTTCTTGAG T15 LECTIN Amino acid sequence FGNVESRLDLRKNLRCQSFKWYLENIYPELSIPKESSIQKGNIRQRQKCLESQRQNNQET PN LKLSPCAKVKGEDAKSQVWAFTYTQKILQEELCLSVITLFPGAPVVLVLCKNGDDRQQWT KT GSHIEHIASHLCLDTDMFGDGTENGKEIGVNPCESSLMSQHWDMVSS* GaINAc-T16 lectin domain: The lectin domain polypeptide sequence comprises amino acid residues 396- 558 of GALNT16 (GALNT16 nucleotide sequence accession number is AJ505951). T16 LECTIN DNA sequence AGTGTGGCTACGCGGATAGAGCAGAGGAAGAAGATGAACTGCAAGTCCTTCCGCTGGTA CCTGGAGAACGTCTACCCAGAGCTCACGGTCCCCGTGAAGGAAGCACTCCCCGGCATCAT TAAGCAGGGGGTGAACTGCTTAGAATCTCAGGGCCAGAACACAGCTGGTGACTTCCTGCT TGGAATGGGGATCTGCAGAGGGTCTGCCAAGAACCCGCAGCCCGCCCAGGCATGGCTGT TCAGTGACCACCTCATCCAGCAGCAGGGGAAGTGCCTGGCTGCCACCTCCACCTTAATGT CCTCCCCTGGATCCCCAGTCATACTGCAGATGTGCAACCCTAGAGAAGGCAAGCAGAAA TGGAGGAGAAAAGGATCTTTCATCCAGCATTCAGTCAGTGGCCTCTGCCTGGAGACAAA GCCTGCCCAGCTGGTGACCAGCAAGTGTCAGGCTGACGCCCAGGCCCAGCAGTGGCAGC TGTTGCCACACACATGA T16 LECTIN Amino acid sequence SVATRIEQRKKMNCKSFRWYLENVYPELTVPVKEALPGIIKQGVNCLESQGQNTAGDF LLGMGICRGSAKNPQPAQAWLFSDHLIQQQGKCLAATSTLMSSPGSPVILQMCNPREGKQ KWR RKGSFIQHSVSGLCLETKPAQLVTSKCQADAQAQQWQLLPHT*

In this Example we have defined minimal sequences of functional lectin do- mains based on multiple sequence alignments. It is clear that changes in the length of sequences used may not affect functionality of the lectins. Such changes could consti- tute, for example, plus or minus 10-20 amino acid residues of the GalNAc-transferase sequence at their amino or carboxy termini.. For example, the GaINAc-Tl lectin do- main may comprise 10-20 fewer amino acid residues at its carboxy and/or amino ter- mini than shown in Table III's T1 lectin domain sequence; i. e. the T1 lectin domain could, for example, stretch from amino acids 403-549 of the GALNT1 sequence, or, for example, from amino acids 413-539 of the GALNT1 sequence. Additionally, the Gal- NAc-T1 lectin domain may comprise 10-20 more amino acid residues at its carboxy and/or amino termini than shown in Table III's Tl lection domain sequence; i. e. the T1 lectin domain could, for example, stretch from amino acids 383-569 of the GALNT1 sequence, or, for example, from amino acids 373-579 of the GALNT1 sequence.

Sf9 cells were co-transfected with pACGP67-GaINAc-transferase soluble ex- pression constructs and Baculo-GoldT" DNA (Pharmingen) as previously described 33.

Briefly, 0.4 zig DNA was mixed with 0.1 pg Baculo-Gold DNA and co-transfected in Sf9 cells in 24-well plates. Ninety-six hours post-transfection recombinant virus was amplified in 6-well plates at dilutions of 1: 10 and 1: 50. Titer of amplified virus was es- timated by titration in 24-well plates. For large scale production and purification of re- combinant secreted enzymes and lectins the amplified vira were used to infect High Five cells grown in serum free medium (Invitrogen) in upright roller bottles shaking at 140 rpm in 27 °C waterbaths. Recombinant proteins were purified by nickel NiTA chromatography using nickel agarose (Qiagen) as recommended by the manufacturer or by consecutive chromatographies on Amberlite, S-sepharose and Mono-S as previously described 39.

2. Direct binding assay for determination of carbohydrate specificity of polypeptide GaINAc-tranferase lectins using soluble GatNAc-transferase enzymes: GalNAc-transferase lectins were previously shown to direct GaINAc- glycopeptide substrate specificities of some GalNAc-transferase 2. The mechanism by which the lectin domains mediate this specificity is unknown but the finding that the monosaccharide GaINAc selectively inhibits GaINAc-glycopeptide specificity of some

isoforms suggested that the putative lectin domains were involved in an interaction with the substrate at least partly through the GaINAc-residue. Nevertheless, it has not been possible in the past despite many different attempts to demonstrate direct binding of the enzyme protein or fragments hereof to glycopeptides or saccharides 2 (PCT WO 01/85215 A2). In this Example a binding assay using HIS-tagged affinity purified and biotinylated secreted enzyme was developed. HIS-tagged secreted human GaINAc-T2 and-T4 were prepared from pAC-GP67-T2-sol and pAC-GP67-T4-sol cDNA 10 32 by PCR as described in Example 1.

Secreted GaINAc-T2 and-T4 and variant proteins were obtained from infected High Five cells grown in serum-free medium (Invitrogen) in upright roller bottles shaken 140 rpm in waterbaths at 27 °C. Purification of the recombinant proteins were performed by iminodiacetic acid metal affinity chromatography (IMAC) Ni2+-charged (QIAGEN). Elution was achieved with 250 mM imidazole in 50 mM sodium phosphate (pH 8.0) and 500 mM NaCI. In some cases, recombinant proteins were purified by con- secutive ionexchange chromatographies as developed and described previously 39, be- fore final purification by Ni2+-chromatography. Proteins eluted were dialyzed three times against PBS (10 mM sodium phosphate (pH 7.4) 150 mM NaCl) and concen- trated by centrifugal filter device (Millipore; 10,000 kDa cut off). Purity was analyzed by SDS-PAGE under reducing conditions, and stained for proteins with Coomassie Blue R 250.

Protein biotinylation was made as previously reported 49. The pH of 1 ml puri- fied protein (0.3 mg/ml) in PBS was adjusted to pH 9 with 1 M NaOH and 40 pI N- hydroxy-succinimidobiotin (Sigma) dissolved in DMF (10 mg/ml) was added. The so- lution was mixed end-over-end for 2 hours at room temperature, and dialyzed three times against PBS and an equal volume of glycerol was added. The biotinylated pro- teins were stored at-20°C in 50% glycerol until use.

Glycosylation of MUC1 peptides (0.1 mM) was made in 20 mM cacodylate buffer (pH 8.0), 10 mM MnCI2, 10 mM UDP-GaINAc, and 20 pg purified polypeptide GaINAc-T1 or-T2 with or without subsequent glycosylation with GaINAc-T4 at 37°C during overnight. Glycopeptides were purified by C-18 reverse phase HPLC. Peptides were custom synthesized by Neosystems (Strasbourg). Biotinylated Helix Pomatia

lectin (HPA) was from KemEnTec (Denmark). Anti-MUC1 HMFG2 monoclonal anti- body was a generous gift from Joyce Taylor-Papadimitriou. Anti-MUCI 5E5 mono- clonal antibody was developed by immunizing Balb/c mice with 60-mer MUC1 tandem repeat peptide glycosylated with 5 moles GaINAc per repeat. Monoclonal antibodies to the lectin domains of human GalNAc-T2 and-T4 were developed as previously de- scribed i050.

Direct binding ELISA assay was developed as follows: Polystyrene microtiter plates (Maxisorb, Nunc, Denmark) were coated with peptides or enzymatically glycosy- lated glycopeptides in PBS overnight at 4°C. Plates were washed and blocked with 0. 1 % Tween20 and 0.2% BSA in PBS for 1 h at room temperature, followed by incuba- tion with biotinylated proteins in PBS with 0.05% Tween20 for 2 h at room tempera- ture. After four washes with PBS, plates were incubated with 1 : 2,000 dilution strepta- vidin-HRP (Sigma) in PBS for 30 min at room temperature and washed four times with PBS. Development was performed with 0.5 mg/ml o-phenylenediamine and 0.02% H202 at room temperature for 15 min, and reaction stopped by adding 100 ul/well of 0.5 N H2SO4. Competitive inhibition assays were done at end-point titers of GalNAc- transferase proteins with one hour preincubations with inhibitor.

In initial binding experiments it was determined that secreted GaINAc-T2 and- T4 could bind their peptides substrates in the presence of 5 mM UDP and Mon++ (may be substituted with other divalent cation), whereas no binding was observed in the absence or when 10 mM EDTA was included. This correlates with our previous findings that GalNAc-T2 can be purified on an acceptor substrate peptide in the presence of UDP and Mn++, and can be eluted by removing UDP in EDTA 32. This binding is mediated by the catalytic unit of the enzyme, which exposes the acceptor-binding pocket only in the presence of UDP and Mon4, predicted by the ordered catalytic reaction.

In order to selectively evaluate the binding characteristics of the putative lectin domain, assays were carried out in the absence of UDP and Mon++. Significant binding to GaINAc-peptides was found for both GaINAc-T2 and-T4 (Figure 4).

We have previously shown that GalNAc-T4 with a single amino acid change in the lectin domain selectively impairs the GaINAc-glycopeptide substrate specificity of this enzyme. In agreement with this the lectin mutated enzyme protein did not bind

GaINAc-glycopeptides. The binding to GaINAc-glycopeptides were unaffected by Ca++ and EDTA further confirming that the catalytic units of the enzyme proteins are not involve in binding.

GaINAc and GalNAca-benzyl have previously been shown to inhibit the Gal- NAc-glycopeptide substrate specificity of GaINAc-T4 2 (PCT WO 01/85215 A2).

Lectins and antibodies to carbohydrates usually recognize the anomeric configuration of the sugar structures they bind. However, surprisingly, both GaINAc-T2 and-T4 exhibit equal inhibition with GaINAca-benzyl (Sigma) and GaINAcp-benzyl (NuRx, Alberta Research Council (Fig. 5). 11, 59 Similar results were obtained with other aryl derivatives.

The methods described in this Example utilize recombinant polypeptide Gal- NAc-transferases in binding assays and excludes potential binding activity through the catalytic unit. It is clear that recombinant polypeptide GalNAc-transferases with muta- tions that inactivates the binding activity of the catalytic unit can be used, as well as any truncation of the enzyme protein that eliminate the binding activity of the catalytic unit.

While this binding assay establishes a method for screening for inhibitors of lec- tion mediated binding mediated through the lectins of human GaINAc-T2 and GaINAc- T4, it is clear that the same method with modifications can be applied to all animal and mammalian polypeptide GalNAc-transferases with a functional lectin domain. The ligand target used in this Example is a GalNAc-MIJC 1 glycopeptide produced enzy- matically from synthetic peptides. It is clear that GaINAc-glycopeptides based on any number of peptides with GaINAc attached can be used as target for the binding assay. It is also clear that the assay developed can be modified to accommodate high through put screening by any assay method available in the prior art that can detect and quantify binding between the polypeptide GaINAc-tranferase mediated through its lectin domain and a suitable ligand.

3. Direct binding assay for determination of carbohydrate specificity of poly- peptide GaINAc-tranferase lectins using truncated GaINAc-transferase lectin domains: In Example 2 it was established that secreted soluble enzymatically active pro- teins of GaINAc-T2 and-T4 bind GaINAc-MUC 1 glycopeptides, and that GalNAc could inhibit the binding. The catalytic unit of polypeptide GalNAc-transferase can in-

teract and bind acceptor substrate peptides and possible glycopeptides 38, however, binding studies without donor substrates (UDP), and in the presence of EDTA to che- late Mn, suggested that the binding was not mediated through the catalytic unit. In this Example direct binding to GaINAc-glycopeptides through the lectin domains of poly- peptide GaINAc-transferases GaINAc-T2 and-T4 was established. Attempts to express C-terminal truncated GalNAc-transferase proteins failed due to low secretion rate pre- sumable related to folding problems and intracellular degradation. Similar phenomenon has recently been reported for GalNAc-TI 14. Numerous attempts to express isolated lectin domains in insect cells and P. pastoris have failed due to low expression and ap- parent degradation. As described in Example 1 successful expression was finally achieved with constructs truncated as described in Table II using an expression vector with a N-terminal HIS tag and thrombin cleavage site as well as a T7 tag. HIS-tagged truncated GaINAc-T2 and-T4 lectins were expressed and purified as described in Ex- ample 2. Lectins were used in binding studies directly or after biotinylation as described in Example 2. In binding studies using lectins without biotinylation antibodies to the HIS-tag and the T7-tag, or in some experiments antibodies raised to GalNAc-T2 and- T4 enzymes were used to detect binding.

Inhibition experiments were used to further define the binding specificity of GaINAc-T2 and-T4 secreted soluble enzymes as well as lectin domains compared to Helix Pomatia lectin (Table IV). GalNAca-and GalNAcß-aryl structures inhibited binding of both enzymes and isolated lectins at comparable levels. In contrast, Helix Pomatia showed strong preference for GalNAca-derivatives. Gal and other sugars had no inhibitory effect. Interestingly, UDP-GaINAc was not a significant inhibitor of the GalNAc-transferase lectin binding, but a strong inhibitor of Helix Pomatia binding.

Table IV. Inhibition of carbohydrate-binding proteins by using saccharides and related structures.

Carbohydrates T21d sT2 T41d sT4 HPA Glc >100a >100 >100 >100 >100 GlcNAc >100 >100 >100 >100 8 BzlaGlcNAc >100 >100 >100 >100 2 BzlßGlcNAc >100 >100 >100 >100 >20 Gal >100 >100 >100 50 >100 MeαGal >100 >100 >100 50 >100 Me (3Gal >100 >100 >100 40 >100 GalNAc 37 15 5 1 2 BzlaGalNAc 20 15 5 1 0.5 PhlaGalNAc 15 5 5 1 1 oNPaGaINAc 12 7 5 0.5 1.5 oNPßGalNAc >12 10 8 1 >20 pNPaGaINAc >10 >10 10 1 2 pNPßGalNAc >10 >10 8 0.8 >20 UDPaGalNAc >100 65 50 30 2 UDP >100 >100 >100 >100 >100 Lactose >100 >100 >100 >100 ND EDTA >10 >10 >10 >10 >10 a Concentration (mM) required for 50% inhibition (IC50).

ND, not determined.

While this binding assay establishes a method for screening for inhibitors of iso- lated human GaINAc-T2 and GaINAc-T4 lections, it is clear that the same method with modifications can be applied to all animal and mammalian polypeptide GalNAc- transferase lectins. The ligand target used in this Example is a GalNAc-MUC1 glyco-

peptide produced enzymatically from synthetic peptides. It is clear that GalNAc- glycopeptides based on any number of peptides with GaINAc attached can be used as target for the binding assay. It is also clear that the assay developed can be modified to accommodate high through put screening by any assay method available in the prior art that can detect and quantify binding between the isolated lectin and a suitable ligand.

The methods described in this Example utilize recombinant GalNAc-transferase lectins in binding assays which excludes potential binding activity through the catalytic unit. It is clear that recombinant polypeptide GaINAc-transferases with mutations that inactivates the binding activity of the catalytic unit can be used, as well as any trunca- tion of the enzyme protein that eliminate the binding activity of the catalytic unit.

4. Establishment of cell tine model systems for cell surface expression of mucin and secreted mucin-Stably transfected CHO and CHO Idled cells : Cell lines and expression constructs : Wild type Chinese Hamster Ovary cells (CHO) and the glycosylation deficient mutant cell line CHO ldlD 40 were stably trans- fected with a full coding MUC1 construct (MUC1F, supplied by M. A. Hollingsworth, Nebraska, USA) containing 32 tandem repeats using the pCDNA3 vector (Invitrogen).

A secreted MUC1 construct (MUCl-IgGHIS) was generate by insertion of mouse IgGy2a domain fused to 6x histidine tag at the BsU36I site downstream of the tandem repeat region of MUC1F Cells were generally grown in Hams F12 containing 10% Fetal Bovine Serum at 37° C at 5% CO2, and plated 12-24 hours prior to transfection in 6 well plates and grown to approximately 50% confluency. One hour before transfec- tion cells were washed in serum free medium Optimem (Invitrogen) and cells were transfected with 1-2 sug DNA using the Lipofectamine plus reagent (Invitrogen) in a to- tal volume of 1 mL as recommended by the supplier. Three hours after the transfection one mL of Hams F 12 containing 10% Fetal Bovine Serum was added and cells grown 24-48 hours before medium was replaced with 2 mL Hams F12 containing 10% Fetal Bovine Serum. Two to three days after transfection cells were trypzinized and plated in 75 mL T-flasks or in 24/96 well microtiter plates in the same medium containing the appropriate selection agent (1 mg/mL G418 or 0.4 mg/mL Zeocin). Selection medium was changed twice weekly until clones appeared. The medium used for CHO ldlD cells included 1 mM GaINAe and 0.1 mM Gal. Transfectant clones were selected by immu-

nocytology with anti-MUC 1 antibodies and SDS-PAGE western blot analysis to dem- onstrate cell surface expression and secretion of MUC 1.

Immunocytology : Two different procedures were applied: i) For general screen- ing purposes, cells grown in plates or flasks were trypsinized, washed in saline, and air- dried on multiwell coverslides. Slides were fixed in ice-cold acetone and stained with monoclonal antibodies and FITC-conjugated rabbit anti-mouse Ig as previously de- scribed 50. ii) For analysis of cell surface expression, cells were seeded in 6 well plates and grown for 6 hours in Hams F12 medium with serum until approximately 30-50 % subconfluent. Medium was hereafter replaced with Optimem supplemented with 1.0 mM GaINAc and 0.1 mM Gal and cells grown for 18 to 42 hours. Cells were washed once in PBS (phosphate buffered saline without Calcium and Magnesium) and subse- quently fixed in 2 ml 3% paraformaldehyde at 25°C for 20 min followed by three washes with PBS. Free aldehyde groups were quenched by incubating in 2 mL 50 mM Ammonium Chloride in PBS for 10 min, followed by three washes with PBS and three washes with 5 min incubations each with PBS containing 0.2 % Fish Skin Gelatin (Sigma). Immunostaining of cells was performed by incubation with monoclonal anti- bodies for 40 min at 25°C, followed by three washes with PBS and three washes of 5 min each with PBS containing 0.2 % Fish Skin Gelatin. Subsequently, cells were incu- bated with FITC-conjugated rabbit anti-mouse Ig (Dako, F261) diluted 1: 150 in PBS containing 0.2 % Fish Skin Gelatin) for 20 min at 25°C, followed by the same washing procedure, after which wells were cut out of plates and mounted with glycerol as for glass slides.

Characterization of wild type and MUCI stable transfectant CHO cells : Several representative clones expressing the full coding or secreted MUC1 construct were se- lected and characterized for expression of MUC 1 as well as 0-glycosylation. Wild type CHO/MUClF-clonel expressed MUC1 at the cell surface as detected by anti-MUCl monoclonal antibodies on non-permeabilized cells, while wild type CHO/MUClsol- cloneC4 only was labeled weakly at the surface. Staining with a panel of anti-MUCI antibodies of permeabilized cells showed intracellular accumulation of MUC1 detected by HMFG2 (general anti-MUC1 reactive, 52), SM3 (reactive with cancer-associated MUC1, 52), VU-4H5 (reactive with low density O-glycosylated MUC1, 53), VU-2G7 (reactive with high density 0-glycosylated MUC1, 54), and a novel antibody 5E5 reac-

tive exclusively with STn/Tn-glycosylated MUC1 glycoforms. In contrast, staining of non-permeabilized cells were only reactive with the anti-MUC1 antibodies HMFG2, SM3 and weakly VU-2G7. Analysis of O-glycosylation using a panel of anti- carbohydrate monoclonal antibodies revealed that wild type CHO cells label very weakly with anti-T antibodies (HH8, 3C9, 55 at the surface after neuraminidase treat- ment, while untreated cells are negative indicating that wild type CHO cells express very little 0-glycoproteins and the glycosylation is mainly of sialylated core 1 structure (ST) (Fig. 6). Antibodies to Tn (1E3, SF4, 55) were weakly reactive without and with neuraminidase treatment and antibodies to STn (TKH2, 3F1, 55) were negative. Staining with the lectins PNA (T), HPA (Tn), SNA (a2, 6sialic acid) and MAA (a2, 3sialic acid) were in agreement except the finding of weak reactivity with SNA indicating some presence of a2, 6 linked sialic acids which may be derived from N-linked or 0-linked glycans. These results demonstrate that the main form of O-glycosylation found on MUC1 expressed in wild type CHO is the sialyl-T structure as found for other recombi- nant glycoproteins Staining of permeabilized wild type CHO/MUClsol-clone-C4 with anti-MUCl antibodies revealed strong intracellular expression of MUC1 with HMFG2, SM3, vu- 4H5, and vu-2G7 (Fig. 6). Staining with anti-carbohydrate antibodies revealed strong intracellular staining with anti-T after neuraminidase only, while anti-Tn only labeled weakly. These results indicate that the main glycosylation of MUC1 in CHO wild type is ST similar to untransfected cells.

In order to characterize the secreted MUCK product SDS-PAGE western blot analysis of harvested culture medium of confluent cultures were performed. Ten to twenty uL culture supernatant was analysed directly or treated with 0.1 U/mL neura- minidase (C. Perfringes VI, Sigma) for 30-60 min at 37°C. Samples were mixed with SDS sample buffer, reduced with DTT, and run on precast 4-20 % gels (Biorad). As shown in Figure 7 anti-MUCI antibodies detected two forms of MUC1 in the medium; a low molecular weight form migrating as 130-140 kd corresponding to unglycosylated product, and a high molecular weight form migrating above 250 kd. The high molecular weight form was sensitive to neuraminidase treatment as evidenced by a marked shift and retardation in migration. It is known that sialylated glyoproteins migrate aberrantly

and often desialylation results in slower migration by SDS-PAGE analysis regardless of the mass. Interestingly, the antibody VLJ-4H5 reacted mainly with the unglycosylated form and only a very weak band was found in the high molecular weight forms after neuraminidase treatment. This result indicates that the PDTR region is O-glycosylated as the VU-4H5 antibody was previously found to tolerate O-glycosylation most posi- tions in the tandem repeat except the PDTR region 53. In agreement with this the anti- body VU-2G7 raised against a MUC1 GalNAc-glycopeptide with only one GaINAc per repeat attached in the PDTR region reacted strongly with the secreted MUC1. Further- more, reactivity with the anti-T antibody after neuraminidase treatment showed that the main type of O-glycosylation on secreted MUC 1 was sialylated-T. Anti-Tn and STn produced no staining.

Characterization of mutant and MUCI stable transfectant CHO ldlD cells : CHO Idled cells stably transfected with full coding MUC1, e. g. CHOIdID/MUClF- clone2, expressed MUC1 at the cell surface as detected by anti-MUCl antibodies when cells were grown in GalNAc and Gal (Fig. 8). Cells were seeded at approximately 30- 50 % confluency (approx. 0.2 X 106 per 6 well plate) in Hams F 12 medium supple- mented with 10% Fetal Bovine serum and grown for 6 hours. Medium was replaced with Optimem with or without 1.0 mM GalNAc and/or 0.1 mM Gal, and cells grown for 18-36 hours after which cells were trypsinised and washed in saline and processed as described for immunocytology. CHOldlD/MUCIF-clone2 cells grown in the absence of sugars and analysed after permeabilization produced very little MUC 1 detectable by HMFG2 but not by 5E5. In contrast, cells grown in the presence of only GalNAc strongly expressed MUC1 as evaluated by HMFG2 and 5E5, specifically reactive with GalNAc-MUCl. In agreement with reactivity with 5E5 these cells also labeled strongly with anti-Tn antibodies, 5F4 and 1E3, while anti-T antibodies, HH8 and 3C9, did not label the cells. Very weak or no staining with anti-STn antibodies (3F 1 and TKH2) in- dicates that a2, 6 sialylation to form STn does not occur in CHO IdID cells.

CH01dlD/MUClF-clone2 cells grown in the presence of both GalNAc and Gal show reactivity at the surface with anti-T antibodies (HH8 and 3C9) only after neuraminidase pretreatment, confirming previous results that the predominiant glycoform in CHO cells is sialyl-T (Fig 8). No staining with anti-Tn or STn antibodies was detected with cells grown in both Gal and GaINAc. CHOIdID/MUClF-clone2 cells grown in the absence

of GaINAc and Gal or in the presence of GaINAc alone showed no reactivity with anti- Tn and T antibodies or lectins (DBA, HPA, VVA, PNA, not shown) were detected in- dicating complete lack of O-glycosylation (Fig 8). Cell surface expression of MUC1 was detected in CHOldlD/MUC I F-clone2 cells grown in the presence of GaINAc, while cells grown without GaINAc showed no or only weak expression of MUC I at the surface (Fig 8). Surface expression of MUC 1 was detected with HMFG2 in cells grown in GaINAc as well as cells grown in both Gal and GaINAc, however, expression ana- lysed with the Tn/STn-MUC1 glycoform specific antibody 5E5 revealed surface ex- pression only with cells grown in the presence of GalNAc (Fig 8). This latter finding is in agreement with the O-glycosylation pattern determined above in these cells.

CHOIdID/MUCsol-cloneD5 secretes MUC 1 to the culture medium, and perme- abilized cells immunostain with antibodies to MUC1 in the cytoplasm. Cells were grown in Hams F12 medium supplemented with 10% Fetal Bovine serum and seeded at a density of 0.2 x100 in 6 well plates. Following growth for 6 hours, the medium was replaced with Optimem supplemented with 1 mM GalNAc, 0.1 mM Gal, or 1 mM Gal- NAc and 0. 1 mM Gal, and cells grown for 18-72 hours. Secretion of MUC1 was monitored by immunochemical assays of culture supernatants at differing time points.

SDS-PAGE western blot analysis was performed with 5 pl culture supernatant mixed with 5 p1 of 2x SDS sample buffer containing 1 mM DTT. Samples were heated to 100°C for 2 min and loaded on a precast 4-20% gradient gel and run at 125 V for 75 min. Transfer to nitrocellulose membrane was performed by elecroblotting using Biorad Mini Trans Blot apparatus at 350 mA for 1 hour. Membranes were blocked with 15% skimmed milk prepared in dH2O for 2 hours and stained with anti-MUCI and anti- carbohydrate monoclonal antibodies for 18 hours at 4 °C, followed by washing with Tris buffered saline (TBS) (10 mM Tris pH 8,0 with 8.5% NaCI) 5 times for 5 min, and incubation with with biotinylated rabbit anti-mouse IgG subclass specific antibodies (1: 1000 dilution in TBS) for 1 hour at 25°C. Following 5 washes for 5 min each in TBS, membranes were incubated in HRP conjugated Streptavidin (1: 3000 dilution in TBS) for 30 min at 25°C. After 5 washings of 5 min each in TBS the blot was developed in b 0.04 % 4-Chloro-1-Naphthol prepared in 50 mM Tris-HCI (pH 7,4) containing 0.025 % H202. Similar to the findings with full coding MUC1 expressed at the cell surface of

CHO Idled cells, glycosylation of the secreted MUC1 was dependent on Gal and Gal- NAc sugars in culture medium. Cells grown without sugars produced and secreted low amounts of a low molecular weight MUC1 molecule of apparent mw of 120-130 kd without glycosylation detectable by HMFG2 but not 5E5 or anti-Tn and anti-T antibod- ies (Fig. 9). In contrast, cells grown in 1 mM GalNAc secreted MUC1 glycosylated with GaINAc (Tn) as evidenced by reactivity with both HMFG2 and 5E5 as well as anti-Tn antibodies (Fig. 9). The apparent molecular weight of secreted Tn-MUCl was 250-300 kd and no significant shift in migration was observed with pretreatment with neuraminidase (0.1 U/ml for 30 min at 37°C), suggesting lack of a2, 6sialylation (STn).

This was confirmed by lack of staining with anti-STn antibodies. Cells grown in both 0.1 mM Gal and 1 mM GalNAc produced and secreted MUC1 with sialylated core 1 (T) glycoforms reactive with HMFG2 but not 5E5 (Fig. 9). Pretreatment with neura- minidase resulted in a significant shift in migration and reactivity with anti-T antibodies as well as the lectin PNA. Two novel anti-MUC1 antibodies described recently have been suggested to react with the MUC1 tandem repeat sequence without (Mab VU- 4H5) or with (Mab VU-2G7) 0-glycans attached in the central immunodominant epi- tope-PDTR-. Analysis of secreted MUCI produced in CHO Idled cells grown without GalNAc show reactivity with unglycosylated MUC1 migrating at mw 120-130 kd with VU-4H5, while no or only weak reactivity was observed when grown in GalNAc with or without Gal (Fig. 10). In contrast, the Mab VU-2G7 reacted strongly with MUC1 migrating at 250-300 kd secreted from cells grown in GalNAc with or without Gal (Fig.

10). Although, Mab VU-2G7 reacted weakly with unglycosylated MUC1 the combined results suggest that MUC1 produced in CHO Idled cells carry 0-glycans on all five sites of the, tandem repeat.

5. The inhibitor GaINAca-benzyl inhibits MUC1 expression independently of 0-glycan processing.

As shown in Example 4 CH01dID/MUClF-clone2 cells grown in the presence of GalNAc but not Gal have limited O-glycosylation capacity, only produce the Tn gly- coform of MUC1, but expresses comparable levels of MUC1 at the cell surface as in wild type CHO cells or in CHO Idled cells grown in both GalNAc and Gal. This sug- gested that cell surface expression was not related to O-glycosylation and particular

glycoforms as previously proposed (for a review see Huet). We therefore investigated the effect of treatment with GalNAca-benzyl of CHOldlD/MUC I F-clone2 cells grown in the presence of GalNAc. CHOldlD/MUClF-clone2 cells were seeded in 6 well plates at a density of 0.2 x 106 per well and were grown for 6 hours in Hams F 12 medium with serum until approximately 30% subconfluent. Medium was hereafter replaced with Op- timem supplemented with 1.0 mM GaINAc or 1.0 mM GaINAc and 0. 1 mM Gal with or without the inhibitors GalNAca-benzyl, GaINAcp-benzyl or the control GIcNAca- benzyl. After 18 hours, the medium was replaced with fresh Optimem containing the sugars and benzyl derivatives as above and grown for 12-48 hours. Initially we ana- lyzed surface expression of MUCl by immunocytology. Cells were washed once in PBS-CMF (phosphate buffered saline Calcium and Magnesium free) after carefully re- moving the medium from the wells and subsequently fixed in 2 ml 3% paraformalde- hyde at 25°C for 20 min followed by three washes with PBS-CMF 3 times. Free alde- hyde groups were quenched by incubating in 2 mL 50 mM Ammonium Chloride in PBS-CMF for 10 min, followed by three washes with PBS-CMF and three washes with 5 min incubations each with PBS-CMF containing 0.2 % Fish Skin Gelatin (Sigma).

Immunostaining of cells was performed by incubation with monoclonal antibodies for 40 min at 25°C, followed by three washes with PBS-CMF and three washes of 5 min each with PBS-CMF containing 0.2 % Fish Skin Gelatin. Subsequently, cells were in- cubated with FITC-conjugated rabbit anti-mouse Ig (Dako F261) diluted 1: 150 in PBS- CMF containing 0.2 % Fish Skin Gelatin) for 20 min at 25°C, followed by the same washing procedure, after which wells were cut out of plates and mounted with glycerol as for glass slides. Figure 11 shows that treatment with 1 mM GalNAca-benzyl, pro- duced strong inhibition of cell surface expression of MUC1, while treatment with a similar control benzyl derivative showed no inhibition. GIcNAca-benzyl was chosen as a control because this sugar does not serve as a substrate for mammalian glycosyltrans- ferases and hence was not expected to interfere with O-glycosylation in CHO cells.

Most anti-MUC1 antibodies reacted with cells grown in GaINAc including VU-2G7 and 5E5, and only VU-4H5 did not react. Reactivity with VU-2G7 indicates that the- PDTR-region is O-glycosylated, while reactivity with 5E5 confirms that the glyco- forms of surface MUC1 is mainly or exclusively Tn.

We next analysed the expression of MUC I produced by CHOIdID/MUC I F- clone2 cells by SDS-PAGE western analysis. Cells were grown for 24 hours or 48 hours in the presence of 1 mM GaINAc or 1 mM GalNAc and 0.1 mM Gal to limit core O-glycosylation to GalNAcal-O-Ser/Thr and Galßl-3GalNAcal-O-Ser/Thr, respec- tively. Cells were further treated with 2 mM GalNAca-benzyl, GlcNAca-benzyl or no inhibitor. Cells were washed and lysed at 24 or 48 hours and the lysates subjected to immunoprecipitation with monoclonal antibody HMFG2, which broadly recognize MUC1 glycoforms. Immunoprecipitates were analysed by SDS-PAGE and western blot using HMFG2 antibody to detect MUC1 expression. As shown in Figure 12 the MUC1 glycoforms at 24 hours expressed by cells grown in GalNAc or Gal and GaINAc mi- grated similarly with only little high molecular weight forms in both, indicating that synthesis of sialylated core 1 O-glycans were time limited. At 48 hours, MUC1 glyco- forms migrating as higher molecular weight species were expressed more pronounced and selectively by cells grown in Gal and GaINAc. Treatment with GIcNAca-benzyl produced the same glycoforms at similar intensity as cells without treatment. In striking contrast, treatment with GalNAca-benzyl had significant effect after 48 hours. A sig- nificant reduction in MUC1 expression was found in cells grown in GalNAc as well as in Gal and GalNAc. In the latter case a significant shift in migration into two bands fur- ther confirmed that GalNAca-benzyl also serves as an inhibitor of O-glycan extension and reduces O-glycosylation to GalNAcal-O-Ser/Thr. Analysis of the same blots with the anti-MUC1 antibody 5E5 (Figure 13) produced essentially the same results except that the antibody only labeled the lower migrating band of the two bands labeled by HMFG2 in cells grown in Gal and GalNAc and treated with GalNAca-benzyl. This in- dicates some heterogeneity in glycosylation.

These results show for the first time that the effect of GalNAca-benzyl on mucin transport and surface expression is independent of its effects on O-glycosylation in striking contrast to the prevailing hypothesis 26. Because cells grown in the presence of GalNAc do not produce core 1 (Galpl-3GalNAcal-0-Ser/Thr) 0-glycosylation, GalNAca-benzyl cannot serve as a competitive substrate for the core 1 , B3galactosyltransferase and subsequently for sialyltransferases. GalNAca-benzyl must therefore exert its function on mucin transport by another unknown mechanism.

The in vivo cell line model system developed is one example of a method to screen for inhibitors effects of one or more compounds on transport of mucins and O- linked glycoproteins in cells. The Example utilizes MUC1 but any mucin or 0-linked glycoprotein could be used with appropriate expression constructs, antibodies and re- agents. The developed cell model and modifications hereof can be used for high throughput screens of inhibitors in combination with or as a second screen after the binding assays disclosed in Examples 2 and 3.

6. Identification of a novel selective inhibitor, GalNAcß-benzyl, of polypeptide GaINAc-transferase lectins that inhibits MUC1 expression without affecting O- glycosylation.

As shown in Examples 2 and 3, polypeptide GaINAc-transferases contain lectin domains with binding properties for GaINAc-peptides including GaINAc-MUC1 pep- tides. Since GalNAca-benzyl was found to inhibit the binding properties of GalNAc- transferase lectins, we tested the possibility that the independent effect on mucin ex- pression this O-glycosylation inhibitor has, could be related to an inhibitory effect on GaINAc-transferase lectins. In Examples 2 and 3 we found surprisingly that the lectin domains of several GaINAc-transferases in addition to GalNAca-benzyl, which mimics the GaINAc-glycopeptide targets of the lectins, also were inhibited by ßGalNAc deriva- tives. Initial tests with commercially available GaINAc (3 and GalNAca derivatives, p- nitrophenyl and umbrelliferyl did not produce significant effects in our model system.

GaINAcp-benzyl, the (3-anomeric configuration of GaINAc-benzyl (there is a ß linkage between the N-acetylgalactosamine and the benzyl ring), was custom synthesized by Alberta Research Council (Canada), and its structure was confirmed by mass spec- trometry and'H-NMR. CHOldID/MUC lF-clone2 cells were grown for 12 hours in the presence of 1 mM GalNAc or 1 mM GalNAc and 0.1 mM Gal to limit core O- glycosylation to GalNAcal-O-Ser/Thr and Galpl-3GalNAcal-0-Ser/Thr, respec- tively. Cells were then treated with 2 mM GalNAcα-benzyl, GalNAcß-benzyl, or GlcNAca-benzyl as control (GlcNAca-benzyl was shown in Example 4 to have no ef- fect). Cells were washed and lysed as described above after 36 hours and the lysates subjected to immunoprecipitation with anti-MUC1 monoclonal antibodies HMFG2 or

5E5. Figure 14 illustrates that treatment with GaINAcp-benzyl produced the same or better reduction in MUC 1 expression as treatment with GalNAca-benzyl in cells grown in GalNAc as well as in Gal and GaINAc. In cells grown in Gal and GalNAc MUCI expression was reduced with GalNAcß-benzyl treatment, but in contrast to cells treated with GalNAca-benzyl, GaINAcp-benzyl produced no change in the migration of MUC I demonstrating that this inhibitor does not affect the O-glycan processing. The lack of immunoprecipitation of MUC1 by antibody 5E5 in cells grown in Gal and Gal- NAc indicates that MUC1, is glycosylated with more complex structures than Gal- NAcal-O-Ser/Thr as recognized by this antibody. As shown in Figure 16 the main O- glycan phenotype of CHO IdlD cells grown in Gal and GalNAc is sialylated-T, and 5E5 does not react with MUC1, with T or sialylated T glycoforms of MUC1. Figure 15 il- lustrates the same experiment as in Figure 14 except that the detection antibody is 5E5 and only Tn and STn MUCI glycoforms are visualised. This experiment confirms the strong inhibition of MUC1 expression in GalNAca-benzyl and GalNAcß-benzyl treated cells.

The finding that GalNAca-benzyl and GaINAcp-benzyl exhibit the same inhibi- tory effect on GalNAc-transferase lectin binding, and that they have similar effects on inhibition of MUC1 expression, clearly indicate that the effects these compounds have on mucin expression and secretion are directed through interaction with the lectin do- mains of polypeptide GalNAc-transferases. The effect GalNAca-benzyl has on O- glycan processing is a separate phenomenon directed by its ability to serve as a com- petitive substrate for the core 1 ß1, 3galactosyltransferase.

GaINAcp-benzyl is the first identified selective inhibitor of polypeptide Gal- NAc-transferase lectins and their roles in transport and secretion, which does not modu- late O-glycosylation in cells (i. e. does not serve a substrate for mainly core 1 ß3galactosyltransferase activities, a2, 6sialyltransferase activities, and core 3 ß3GlcNAc-transferase activities). Inhibitors structurally related to GaINAcp-benzyl with the same properties, as identifiable by the binding assays disclosed in Examples 2 and 3, may be designed and syntesized to obtain higher affinity binders. Such inhibitors may be based on carbohydrates such as the monosaccharide GalNAc or modifications thereof, inhibitors may be based on structural and functional mimetics such as polypep-

tides, glycopeptides, DNA, RNA, antibodies, and antibody fragments including phage antibodies, and inhibitors may be natural or synthetic organic or inorganic compounds.

One common feature for such preferred inhibitors is the ability to inhibit the binding of one or more polypeptide GaINAc-transferase lectins to its binding ligand, such as Gal- NAc-glycopeptides and mucins as exemplified in Examples 2 and 3. Another feature of the novel inhibitor GaINAcp-benzyl is its ability to enter living cells and reach the Golgi apparatus for in vivo binding to polypeptide GalNAc-transferase. The hydrophic benzyl aglycone is one example of an aryl compound suitable for the (3-anomeric con- figuration of GaINAc-R, but other aryl substituents include, without limitation, p- nitrophenyl, umbrelliferyl, and naphtalenmethanol. Any pharmaceutical carrier known in the art may be used to achieve the same effect. The appropriate carrier will be evident to those skilled in the art and will depend in large part upon the route of administration.

7. Inhibition of secretion of mucins: Because GalNAca-benzyl exerts separate effects on O-glycan processing and mucin expression, the use of the novel selective inhibitor of mucin expression, Gal- NAc (3-benzyl, allow analysis of mucin expression and secretion in different cell line models. Examples of human cell lines (available from ATCC, USA) expressing and se- creting mucins are without limitations LS 174T, HT29, Colo205, CALU, MCF7, T47D, NCI-H292, and A549. Most human adenocarcinoma cell lines express and secret mucins and analysis with antibodies to detect protein expression or probes to detect mRNA can reveal the types and quantities of mucins. The human colon carcinoma cell line LS 174T was previously shown to exhibit reduced secretion of mucin following treatment with GalNAca-benzyl 57. This and other cell lines can be used to treat with 2 mM GaINAcp-benzyl. In this Example we used wild type CHO/MUClsol-cloneC4 and western blot analysis of medium of cells treated with 1-2 mM GalNAca-benzyl, Gal- NAc (3-benzyl, or the control GlcNAca-benzyl. Treatment with both GalNAca-benzyl and GaINAcp-benzyl showed inhibition of secreted MUC1 compared to control treated or non-treated cells. The inhibitory effect on mucin secretion can be quantified by a number of assays known to the skilled in the art including western blot, ELISA, gelfil- tration, immunocapture, and other assays.

The in vivo cell line model system developed is one example of a method to screen for inhibitors effects of one or more compounds on secretion of mucins and O- linked glycoproteins in cells. The Example utilizes MUC 1 but any mucin or 0-linked glycoprotein could be used with appropriate expression constructs, antibodies and re- agents. The developed cell model and modifications hereof can be used for high throughput screens of inhibitors in combination with or as a second screen after the binding assays disclosed in Examples 2 and 3.

8. GaINAcp-benzyl inhibits mucin glycosylation and MUC5AC production in HT29MTX cells.

In order to evaluate GaINAcp-benzyl as an inhibitor of mucin glycosylation and secretion we used the cultured colon carcinoma cells of mucin secreting phenotypes: HT-29 metotrexate (MTX) cells selected from the HT-29 cell line by culture in the presence of MTX (kind gift from G. Huet, France). The HT-29 MTX cells were se- lected to yield a goblet cell like phenotype with constitutive production of MUC5AC. In order to ensure proper goblet cell differentiation the HT29MTX cell line were continu- ously grown up to 21 days in the presence of 5 mM GalNAca-benzyl, 5 mM GalNAcß- benzyl, 5 mM GlcNAca-benzyl and without inhibitor using culture conditions as de- scribed in Hennebicq-Reig et al. (Permanent exposure of mucin-secreting HT-29 cells to benzyl-N-acetyl-alpha-D-galactosaminide induces abnormal O-glycosylation of mucins and inhibits constitutive and stimulated MUC5AC secretion. Biochem J.

1998; 334: 283-95). At 7,14 and 21 days of continuous culture, media was collected and cells were fixed in 3% paraformaldehyde and stained for the presence of MUC5AC and specific carbohydrate structures using immuno-histochemistry with the following monoclonal antibodies: CLH2 recognizing MUC5AC (Reis C. et al. Immunohisto- chemical study of MUC5AC expression in human gastric carcinomas using a novel monoclonal antibody. Int. J. Cancer. 1997; 74: 112-21), 1E3 recognizing Tn (GalNAca linked to serine or threonine), 3C9 recognizing T (Galp1-3GalNAca linked to serine or threonine, and 3F1 recognizing Sialyl-Tn (NeuAca2-6GalNAca linked to serine or threonine. In order to evaluate the amount of MUC5AC secreted by HT-29 MTX cells with and without GaINAcp-benzyl treatment, media were subjected to SDS-PAGE (Tris-NuPAGE Gels 14-21 %) followed by Western blotting using polyclonal rabbit

anti human MUC5AC antibody LUM5-1 (Carlstedt et al. MUC5AC, but not MUC2, is a prominent mucin in respiratory secretions. Glycoconj J. 1996; 13: 839-47). Further- more the viscosity of the media from cells with and without inhibitor was evaluated us- ing viscosity micro-measurement.

Treatment of cells with GaINAcp-benzyl and GalNAca-benzyl diminished se- cretion of MUC5AC as evaluated by western blot compared with GlcNAcß-benzyl and cells treated with media alone (Fig. 17). Furthermore in accordance with earlier finding (Hennebicq-Reig et al. Permanent exposure of mucin-secreting HT-29 cells to benzyl- N-acetyl-alpha-D-galactosaminide induces abnormal O-glycosylation of mucins and inhibits constitutive and stimulated MUC5AC secretion. Biochem J. 1998 ; 334: 283-95) treatment of cells with GalNAca-benzyl (Fig. 18, Panel A) induced a storage phenotype with increased intracellular staining of MUC5AC. In contrast cells treated with Gal- NAc (3-benzyl (Fig. 18, Panel B) did not lead to any storage disease but diminished in- tracellular staining of MUC5AC compared with and GIcNAcp-benzyl (Fig 18, panel C) and cells treated with media alone (Fig 18, panel D). In addition the 0-linked mucin carbohydrate structures T, Tn and sialyl-Tn were upregulated in cells treated with GlcNAcp-benzyl compared to controls. Media from GaINAcp-benzyl treated cells were less viscous compared with control media (data not shown). In conclusion the novel in- hibitor GalNAcß-benzyl inhibits both 0-linked mucin glycosylation and MUC5AC se- cretion.

Reference List 1. Paulson, J. C. and Colley, K. J. Glycosyltransferases. Structure, localization, and control of cell type-specific glycosylation. J. Biol. Chem. , 264 : 17615-17618,1989.

2. Hassan, H. , Reis, C. A. , Bennett, E. P. , Mirgorodskaya, E. , Roepstorff, P., Hollingsworth, M. <BR> <BR> <BR> <BR> <BR> <P> A. , Burchell, J. , Taylor-Papadimitriou, J. , and Clausen, H. The lectin domain of UDP-N-acetyl- D-galactosamine : polypeptide N-acetylgalactosaminyltransferase-T4 directs its glycopeptide specificities. J Biol Chem, 275 : 38197-38205,2000.

3. Schwientek, T. , Bennett, E. P., Flores, C. , Thacker, J., Hollmann, M. , Reis, C. A. , Behrens, J., Mandel, U. , Keck, B. , Schafer, M. A., Haselmann, K. , Zubarev, R. , Roepstorff, P., Burchell, J.

M. , Taylor-Papadimitriou, J. , Hollingsworth, M. A. , and Clausen, H. Functional conservation of subfamilies of putative UDP-N-acetylgalactosamine : polypeptide N- acetylgalactosaminyltransferases in Drosophila, Caenorhabditis elegans, and mammals. One subfamily composed of 1 (2) 35Aa is essentail in Drosophila. J. Biol. Chem., 277: 22623-22638, 2002.

4. Hagen, F. K. , Hazes, B. , Raffo, R., deSa, D. , and Tabak, L. A. Structure-Function Analysis of the UDP-N-acetyl-D-galactosamine : Polypeptide N-acetylgalactosaminyltransferase. Essential residues lie in a predicted active site cleft resembling a lactose repressor fold. J Biol Chem, 274 : 6797-6803, 1999.

5. Hazes, B. The (QxW) 3 domain: a flexible lectin scaffold. Protein Science, 5 : 1490-1501,1996.

6. Imberty A., Piller V. , Piller F. , and Breton C. Fold recognition and molecular modeling of a lectin-like domain in UDP-GaINac : polypeptide N-acetylgalactosaminyltransferases. Protein Eng., 10 : 1353-1356,1997.

7. BRETON, C. and Imberty, A. Structure/function studies of glycosyltransferases.

Curr. Opin. Struct. Biol, 9: 563-571,1999.

8. Bennett, E. P. , Hassan, H. , Hollingsworth, M. A. , and Clausen, H. A novel human UDP-N- acetyl-D-Galactosamine : polypeptide N-acetylgalactosaminyltransferase, GaINAc-T7, with specificity for partial GaINAc-glycosylated acceptor substrates. FEBS Letters, 460 : 226-230, 1999.

9. Ten Hagen, K. G. , Bedi, G. S. , Tetaert, D. , Kingsley, P. D. , Hagen, F. , BALYS, M. M., BERES, T. M. , Degand, P. , and Tabak, L. A. Cloning and Characterization of a Ninth Member of the UDP-GalNAc : Polypeptide N-Acetylgalactosaminyltransferase Family, ppGaNTase-T9. J Biol Chem, 276 : 17395-17404, 2001.

10. Bennett, E. P., Hassan, H. , Mandel, U. , Mirgorodskaya, E. , Roepstorff, P., Burchell, J. , Taylor- Papadamitriou, J. , Hollingsworth, M. A. , Merkx, G. , Geurts van Kessel, A. , Eiberg, H., Steffen- sen, R. , and Clausen, H. Cloning of a human UDP-N-acetyl-a-D-galactosamine : polypeptide N- acetylgalactosaminyltransferase that complements other GalNAc-transferases in complete O- glycosylation of the MUC I tandem repeat. J. Biol. Chem. , 273 : 30472-30481,1998.

11. Ten Hagen, K. G. , Tetaert, D. , Hagen, F. K. , Richet, C. , BERES, T. M. , Gagnon, J. , BALYS, M. M. , VanWuyckhuyse, B. , Bedi, G. S. , Degand, P. , and Tabak, L. A. Characterization of a UDP-GaINAc : polypeptide N-acetylgalactosaminyltransferase that displays glycopeptide N- acetylgalactosaminyltransferase activity. J. Biol. Chem. , 274 : 27867-27874,1999.

12. Muller, S. , Alving, K. , Peter-Katalinic, J. , Zachara, N. , Gooley, A. A. , and Hanisch, F. G. High density O-glycosylation on tandem repeat peptide from secretory MUC1 of T47D breast cancer cells. J. Biol. Chem. , 274 : 18165-18172,1999.

13. Muller, S. , Goletz, S. , Packer, N. , Gooley, A. A. , Lawson, A. M. , and Hanisch, F. G. Localiza- tion of O-glycosylation sites on glycopeptide fragments from lactation-associated MUC1.

J. Biol. Chem. , 272: 24780-24793,1997.

14. Tenno, M., Saeki, A. , Kezdy, F. J., Elhammer, A. P. , and Kurosaka, A. The lectin domain of UDP-GaINAc : Polypeptide N-acetylgalactosaminyltransferase 1 (GaINAc-T1) is involved in O-glycosylation of a polypeptide with multiple acceptor sites. J. Biol. Chem. , 2002.

15. Jentoft, N. Why are proteins O-glycosylated ? Trends. Biochem Sci 1990. Aug., 15 : 291-294.

16. Tabak, L. A. In defense of the oral cavity: structure, biosynthesis, and function of salivary mucins. Annual Review of Physiology, 57 : 547-564,1995.

17. Van den Steen, P. , Rudd, P. M. , Dwek, R. A. , and Opdenakker, G. Concepts and principles of O-linked glycosylation. Crit. Rev. Biochem Mol. Biol 1998., 33 : 151-208.

18. Taylor-Papadimitriou, J. and Epenetos, A. A. Exploiting altered glycosylation patterns in can- cer: progress and challenges in diagnosis and therapy. Trends In Biotechnology, 12 : 227-233, 1994.

19. Taylor-Papadimitriou, J. and Finn, O. J. Biology, biochemistry and immunology of carcinoma- associated mucins. Immunology Today, 18 : 105-107,1997.

20. Scharfman, A. , Lamblin, G. , and Roussel, P. Interactions between human respiratory mucins and pathogens. Biochemical Society Transactions, 23: 836-839,1995.

21. Rose, M. C. Mucins: structure, function, and role in pulmonary diseases. Am. J Physio. 1992. 0ct., 263 : L413-L429.

22. Thomsson, K. A. , Carlstedt, I., Karlsson, N. G. , Karlsson, H. , and Hansson, G. C. Different O- glycosylation of respiratory mucin glycopeptides from a patient with cystic fibrosis. Glyco- conj. J 1998. Aug., 15 : 823-833.

23. Kuan, S. F. , Byrd, J. C. , Basbaum, C. , and Kim, Y. S. Inhibition of mucin glycosylation by aryl-N-acetyl-alpha-galactosaminides in human colon cancer cells. J. Biol. Chem., 264 : 19271- 19277,1989.

24. Huet, G. , Hennebicq-Reig, S. , de Bolos, C. , Ulloa, F. , Lesuffleur, T. , Barbat, A., Carrier, V. , Kim, I., Real, F. X. , Delannoy, P. , and Zweibaum, A. GaINAc-alpha-O-benzyl inhibits NeuA- alpha2-3 glycosylation and blocks the intracellular transport of apical glycoproteins and mu- cus in differentiated HT-29 cells. J. Cell. Biol., 141 : 1311-1322,1998.

25. Byrd, J. C. , Dahiya, R. , Huang, J. , and Kim, Y. S. Inhibition of mucin synthesis by benzyl- alpha-GaINAc in KATO III gastric cancer and Caco-2 colon cancer cells. Eur J Cancer 1995., 31A : 1498-1505.

26. Gouyer, V. , Leteurtre, E. , Zanetta, J. P., Lesuffleur, T. , Delannoy, P. , and Huet, G. Inhibition of the glycosylation and alteration in the intracellular trafficking of mucins and other glycopro- teins by GalNAcalpha-O-bn in mucosal cell lines: an effect mediated through the intracellular synthesis of complex GalNAcalpha-O-bn oligosaccharides. Front Biosci. , 6: D1235-D1244, 2001.

27. Alfalah, M. , Jacob, R. , Preuss, U. , Zimmer, K. P., Naim, H. , and Naim, H. Y. 0-linked glycans mediate apical sorting of human intestinal sucrase-isomaltase through association with lipid rafts. Curr. Biol, 9: 593-596,1999.

28. Yeaman, C. , Le Gall, A. H. , Baldwin, A. N. , Monlauzeur, L. , Le Bivic, A. , and Rodriguez- Boulan, E. The O-glycosylated stalk domain is required for apical sorting of neurotrophin re- ceptors in polarized MDCK cells. J Cell Biol 1997. Nov. 17., 139 : 929-940.

29. Ulloa, F. , Franci, C. , and Real, F. X. GaINAc-a-O-Benzyl inhibits sialylation of de novo syn- thesized apical, but not basolateral, sialoglycoproteins and blocks lysosomal enzyme processing in a post-TGN compartment. J Biol Chem 2000. Apr. 5..

30. Muller, S. and Hanisch, F. G. Recombinant MUC1 probe authentically reflects cell-specific O- glycosylation profiles of endogenous breast cancer mucin: High-density and prevalent core2- based glycosylation. J. Biol. Chem. , 2002.

31. Hassan, H. , Bennett, E. P. , Mandel, U., Hollingsworth, M. A. , and Clausen, H. Control of Mucin-Type O-Glycosylation : O-Glycan Occupancy is Directed by Substrate Specificities of Polypeptide GaINAc-Transferases. In B. Ernst, G. W. Hart, and P. Sinay (eds.), Carbohydrates in Chemistry and Biology, vol. 3,1 ed, pp. 273-292. New York: Wiley-VCH, 2000.

32. White, T. , Bennett, E. P. , Takio, K. , Sorensen, T. , Bonding, N. , and Clausen, H. Purification and cDNA cloning of a human UDP-N-acetyl-alpha-D-galactosamine : polypeptide N- acetylgalactosaminyltransferase. J. Biol. Chem. , 270: 24156-24165,1995.

33. Bennett, E. P. , Hassan, H. , and Clausen, H. cDNA cloning and expression of a novel human UDP-N-acetyl-alpha-D-galactosamine. Polypeptide N-acetylgalactosaminyltransferase, Gal- NAc-T3. J. Biol. Chem. , 271 : 17006-17012,1996.

34. Bennett, E. P. , Hassan, H. , Mandel, U., Hollingsworth, M. A. , Akisawa, N. , Ikematsu, Y., Merkx, G. , Geurts van Kessel, A. , Olofsson, S. , and Clausen, H. Cloning and Characterization of a Close Homologue of Human UDP-N-acetyl-a-D-galactosamine : Polypeptide N- Acetylgalactosaminyltransferase T3, designated GaINAc-T6 : Evidence for Genetic but not Functional Redundancy. J. Biol. Chem. , 274 : 25362-25370,1999.

35. White, K. E. , Lorenz, B. , Evans, W. E. , Meitinger, T. , Strom, T. M. , and Econs, M. J. Molecu- lar cloning of a novel human UDP-GaINAc : polypeptide N-acetylgalactosaminyltransferase, GaINAc-T8, and analysis as a candidate autosomal dominant hypophosphatemic rickets (ADHR) gene. Gene, 246 : 347-356,2000.

36. Toba, S. , Tenno, M. , Konishi, M. , Mikami, T. , Itoh, N. , and Kurosaka, A. Brain-specific ex- pression of a novel human UDP-GaINAc : polypeptide N-acetylgalactosaminyltransferase (Gal- NAc-T9). Biochim. Biophys. Acta, 1493 : 264-268,2000.

37. Homa, F. L., Hollander, T. , Lehman, D. J., Thomsen, D. R. , and Elhammer, A. P. Isolation and expression of a cDNA clone encoding a bovine UDP-GaINAc : polypeptide N- acetylgalactosaminyltransferase. J. Biol. Chem. , 268 : 12609-12616, 1993.

38. Sorensen, T. , White, T., Wandall, H. H., Kristensen, A. K. , Roepstorff, P. , and Clausen, H.

UDP-N-acetyl-alpha-D-galactosamine : polypeptide N-acetylgalactosaminyltransferase. Identi- fication and separation of two distinct transferase activities. J. Biol. Chem. , 270: 24166-24173, 1995.

39. Wandall, H. H. , Hassan, H. , Mirgorodskaya, E. , Kristensen, A. K. , Roepstorff, P. , Bennett, E.

P. , Nielsen, P. A., Hollingsworth, M. A. , Burchell, J., Taylor-Papadimitriou, J. , and Clausen, H.

Substrate specificities of three members of the human UDP-N-acetyl-alpha-D- galactosamine : Polypeptide N-acetylgalactosaminyltransferase family, GaINAc-Tl,-T2, and- T3. J. Biol. Chem., 272 : 23503-23514, 1997.

40. Kingsley, D. M. , Kozarsky, K. F., Segal, M. , and Krieger, M. Three types of low density lipo- protein receptor-deficient mutant have pleiotropic defects in the synthesis of N-linked, O- linked, and lipid-linked carbohydrate chains. J. Cell Biol., 102 : 1576-1585,1986.

41. Altschuler, Y. , Kinlough, C. L. , Poland, P. A. , Bruns, J. B. , Apodaca, G. , Weisz, O. A. , and Hughey, R. P. Clathrin-mediated endocytosis of MUC I is modulated by its glycosylation state.

Mol. Biol. Cell, 11 : 819-831, 2000.

42. Zanetta, J. P. , Gouyer, V. , Maes, E. , Pons, A. , Hemon, B. , Zweibaum, A. , Delannoy, P. , and Huet, G. Massive in vitro synthesis of tagged oligosaccharides in 1-benzyl-2-acetamido-2- deoxy-alpha-D-galactopyranoside treated HT-29 cells. Glycobiology, 10 : 565-575,2000.

43. Trombetta, E. S. and Helenius, A. Lectins as chaperones in glycoprotein folding.

Curr. Opin. Struct. Biol, 8 : 587-592, 1998.

44. Amado, M., Almeida, R. , Carneiro, F. , Levery, S. B., Holmes, E. H. , Nomoto, M., Hollingsworth, M. A. , Hassan, H. , Schwientek, T. , Nielsen, P. A. , Bennett, E. P. , and Clausen, H. A family of human beta3-galactosyltransferases. Characterization of four members of a UDP-galactose : beta-N-acetyl-glucosamine/beta-nacetyl-galactosamine beta-1,3- galactosyltransferase family. J. Biol. Chem. , 273: 12770-12778,1998.

45. Amado, M. , Almeida, R. , Schwientek, T. , and Clausen, H. Identification and characterization of large galactosyltransferase gene families: galactosyltransferases for all functions. Bio- chim. Biophys. Acta, 1473 : 35-53,1999.

46. Dohi, T. , Yuyama, Y. , Natori, Y. , SMITH, P. L. , Lowe, J. B. , and Oshima, M. Detection of N- acetylgalactosaminyltransferase mRNA which determines expression of Sda blood group car- bohydrate structure in human gastrointestinal mucosa and cancer. Int. J. Cancer, 67: 626-631, 1996.

47. Colley, K. J. Golgi localization of glycosyltransferases : more questions than answers. Glycobi- ology, 7: 1-13,1997.

48. Schwientek, T. J. , Bennett, E. P., Flores, C. , Thacker, J. , Hollman, M. , Reis, C. A. , Behrens, J., Mandel, U. , Keck, B. , Schafer, M. A., Hazelmann, K. , Zubarev, R. , Roepstorff, P., Hollingsworth, M. A. , and Clausen, H. Functional conservation of subfamilies of putative UDP-N-acetylgalactosamine: Polypeptide N-acetylgalactosaminyltransferases in drosophila, C. elegans and mammals: One subfamily comprised of 1 (2) 35Aa is essential in drosophila.

J. Biol. Chem. , 2002.

49. Jorgensen, C. S. , Heegaard, N. H. , Holm, A. , Hojrup, P. , and Houen, G. Polypeptide binding properties of the chaperone calreticulin. Eur. J. Biochem. , 267: 2945-2954,2000.

50. Mandel, U. , Hassan, H. , Therkildsen, M. H. , Rygaard, J. , Jacobsen, M. , Juhl, B. R. , Dabelsteen, E. , and Clausen, H. Expression of polypeptide GaINAc-transferases in stratified epithelia and squamous cell carcinomas: immunohistological evaluation using monoclonal antibodies to three members of the GalNAc-transferase family. Glycobiology, 9 : 43-52,1999.

51. Burdick, M. D. , Harris, A. , Reid, C. J., Iwamura, T. , and Hollingsworth, M. A. Oligosaccha- rides expressed on MUC1 produced by pancreatic and colon tumor cell lines. J. Biol. Chem. , 272: 24198-24202,1997.

52. Girling, A. , Bartkova, J. , Burchell, J. , Gendler, S., Gillett, C. , and Taylor-Papadimitriou, J. A core protein epitope of the polymorphic epithelial mucin detected by the monoclonal antibody SM-3 is selectively exposed in a range of primary carcinomas. International Journal of Cancer, 43 : 1072-1076, 1989.

53. Reis, C. A. , Hassan, H. , Bennett, E. P. , and Clausen, H. Characterization of a panel of mono- clonal antibodies using GaINAc glycosylated peptides and recombinant MUGI. Tumour Biol- ogy, 19steppl 1 : 127-133,1998.

54. Ryuko, K. , Schol, D. J. , Snijdewint, F. G., Mensdorff-Pouilly, S. , Poort-Keesom, R. J. , Ka- runtu-Wanamarta, Y. A. , Verstraeten, R. A. , Miyazaki, K. , Kenemans, P. , and Hilgers, J. Char- acterization of a new MUC1 monoclonal antibody (VU-2-G7) directed to the glycosylated PDTR sequence of MUC1. Tumour. Biol., 21 : 197-210,2000.

55. Mandel, U. , Petersen, O. W. , Sorensen, H. , Vedtofte, P. , Hakomori, S. , Clausen, H. , and Dabel- steen, E. Simple mucin-type carbohydrates in oral stratified squamous and salivary gland epi- thelia. Journal of Investigative Dermatology, 97 : 713-721, 1991.

56. Sasaki, H. , Bothner, B. , Dell, A. , and Fukuda, M. Carbohydrate structure of erythropoietin ex- pressed in Chinese hamster ovary cells by a human erythropoietin cDNA. J. Biol. Chem., 262 : 12059-12076,1987.

57. Kuan, S. F. , Byrd, J. C. , Basbaum, C. , and Kim, Y. S. Inhibition of mucin glycosylation by aryl-N-acetyl-alpha-galactosaminides in human colon cancer cells. J. Biol. Chem. , 264 : 19271- 19277,1989.

58. Ramakrishnan, B. , Balaji, P. V. , and Qasba, P. K. Crystal Structure of B-1, 4- Galactosyltransferase Complex with UDP-Gal Reveals an Oligosaccharide Acceptor Binding Site. J. Mol Biol. , 318: 491-502,2002.

59. Ramakrishnan, B. , and Qasba, P. K. Structure-based Design of ß4Gal-TI with Equally Efficient N-Acetylgalactosaminyltransferase Activity. J. Bio. Chem. , 277,23, 20833-20839,2002.