Engineering of therapeutic and diagnostic O-glycans on recombinant mucin-type immunoglobulin fusion proteins expressed in CHO cells.

Metabolic engineering of mammalian cells for optimized glycosylation is usually done to improve activity and the pharmacokinetic features of glycoprotein therapeutics. The field is mainly focused around engineering of N-glycans. We have created a platform in which recombinant mucin-type immunoglobulin fusion proteins are used as scaffolds for multivalent expression of O-glycans with diagnostic or therapeutic potential. The methods used to make stable CHO cell lines secreting a mucin-type fusion protein with blood group A or B determinants following expression of up to five different cDNAs are described.

Mucin-type fusion proteins with blood group A or B determinants on defined O-glycan core chains produced in glycoengineered Chinese hamster ovary cells and their use as immunoaffinity matrices.

Assays for quantification, and methods for removal, of anti-A and anti-B antibodies are the key for the success of ABO incompatible organ transplantation programs. In order to produce tools that can be used as substrates in tests for anti-A/anti-B quantification and specificity determination or as affinity matrices in extracorporeal immunoadsorption (IA) columns, we engineered Chinese hamster ovary (CHO) cells secreting mucin-type fusion proteins carrying blood group A or B determinants on defined O-glycan core saccharide chains. Besides the P-selectin glycoprotein ligand-1/mouse immunoglobulin G(2b) (PSGL-1/mIgG(2b)) cDNA, CHO cells were transfected with plasmids encoding core 2 (ß1,6GlcNAc-T1) or core 3 (ß1,3GlcNAc-T6 and ß1,3Gal-T5) enzymes together with α1,2Fuc-T1 or α1,2Fuc-T2 and the A or B gene-encoded α1,3GalNAcT or α1,3Gal-T, respectively. Selected clones with the correct glycophenotype were expanded and cultured in shaker flasks and Wave bioreactors. Western blotting was used to characterize purified fusion protein and liquid chromatography-mass spectrometry was used to characterize the released O-glycans. Clones producing PSGL-1/mIgG(2b) carrying O-glycans with A and B determinants on type 1 (Galß3GlcNAc), type 2 (Galß4GlcNAc) and type 3 (Galß3GalNAcα) outer core saccharide chains were established. The conversion of CHO cells from exclusive inner core 1 (Galß3GalNAc) to core 3 (GlcNAcß3GalNAc) O-glycan producers was almost complete, whereas conversion to inner core 2 (GlcNAcß6GalNAc) O-glycans was incomplete as was the α2-fucosylation of the core 1 chain. Sialylation may prevent these biosynthetic steps. The clinical utility of the blood group A and B substituted mucin-type fusion proteins as substrates in enzyme-linked immunosorbent assay or as affinity matrices in IA columns is explored.

Adsorption of chain type-specific ABO antibodies on Sepharose-linked A and B tetrasaccharides.

BACKGROUND: Antigen-specific removal of anti-A and anti-B on immunoadsorption columns carrying the blood group A and B trisaccharides is one important component of some protocols used in ABO-incompatible organ transplantation. Because ABO antibodies exist requiring parts of the core saccharide chain for binding, the anti-A and -B-binding capacity of individual and combined, Sepharose-linked Types 1 through 4 A and B tetrasaccharides with that of the A and B trisaccharides was compared. STUDY DESIGN AND METHODS: Sepharose-linked A and B tri- and tetrasaccharides were used to adsorb anti-A and -B from pooled blood group O serum. Remaining chain type-specific anti-A and -B were detected and quantified in enzyme-linked immunosorbent assays using wells coated with neoglycoproteins or recombinant mucins carrying A and B determinants on defined core saccharide chains. RESULTS: Significantly more anti-A Type 3- and 4-specific immunoglobulin (Ig)G remained after adsorption on the A trisaccharide and the A Type 1 and A Type 2 tetrasaccharide than after adsorption on the A Types 3 and 4 tetrasaccharides. Selective adsorption of chain type-specific IgG anti-B was detected on Sepharose-linked B tetrasaccharides. In contrast, there were no chain type-specific IgM anti-A or -B. A combination of the A or B tetrasaccharides adsorbed a larger fraction of the IgG anti-A and -B repertoires than the corresponding trisaccharides. CONCLUSION: There are chain type-specific anti-A and anti-B IgG, and an adsorber based on a combination of Types 1 through 4 A or B tetrasaccharides will be a more efficient adsorber than an adsorber based on the A or B trisaccharides.

Pichia pastoris-produced mucin-type fusion proteins with multivalent O-glycan substitution as targeting molecules for mannose-specific receptors of the immune system.

Mannose-binding proteins like the macrophage mannose receptor (MR), the dendritic cell-specific intercellular adhesion molecule-3 grabbing non-integrin (DC-SIGN) and mannose-binding lectin (MBL) play crucial roles in both innate and adaptive immune responses. Immunoglobulin fusion proteins of the P-selectin glycoprotein ligand-1 (PSGL-1/mIgG(2b)) carrying mostly O-glycans and, as a control, the α1-acid glycoprotein (AGP/mIgG(2b)) carrying mainly N-linked glycans were stably expressed in the yeast Pichia pastoris. Pichia pastoris-produced PSGL-1/mIgG(2b) was shown to carry O-glycans that mediated strong binding to mannose-specific lectins in a lectin array and were susceptible to cleavage by α-mannosidases including an α1,2- but not an α1,6-mannosidase. Electrospray ionization ion-trap mass spectrometry confirmed the presence of O-glycans containing up to nine hexoses with the penta- and hexasaccharides being the predominant ones. α1,2- and α1,3-linked, but not α1,6-linked, mannose residues were detected by (1)H-nuclear magnetic resonance spectroscopy confirming the results of the mannosidase cleavage. The apparent equilibrium dissociation constants for binding of PNGase F-treated mannosylated PSGL-1/mIgG(2b) to MR, DC-SIGN and MBL were shown by surface plasmon resonance to be 126, 56 and 16 nM, respectively. In conclusion, PSGL-1/mIgG(2b) expressed in P. pastoris carried O-glycans mainly comprised of α-linked mannoses and with up to nine residues. It bound mannose-specific receptors with high apparent affinity and may become a potent targeting molecule for these receptors in vivo.

Is there a clinical need for a diagnostic test allowing detection of chain type-specific anti-A and anti-B?

BACKGROUND: Hemagglutination for detection and semiquantification of ABO antibodies is associated with large center-to-center variations and poor reproducibility. Because acceptance for transplantation and diagnosis of rejection in ABO-incompatible transplantation rely on the levels and specificity of ABO antibodies, reproducible tests that allow their detection and specificity determination are required. STUDY DESIGN AND METHODS: The level of chain type-specific anti-A and anti-B were analyzed in the sera of 44 healthy individuals of known ABO blood group using an enzyme-linked immunosorbent assay (ELISA) with polyacrylamide (PAA) conjugates of blood group A and B trisaccharides or Type 2 chain A and B tetrasaccharides. Selected sera were further analyzed by hemagglutination and in an ELISA with Types 1 to 4 chain A or B neoglycolipids (NGL) as antigens. RESULTS: Immunoglobulin (Ig)G anti-A and anti-B levels were higher (p ≤ 0.05) in blood group O than in B and A individuals. More IgM anti-A and anti-B cross-reactivity was detected in AB serum on PAA-conjugated A and B trisaccharides than on the tetrasaccharides. One of 11 blood group B and two of 12 A individuals had IgG antibodies binding the tetrasaccharide despite lack of, or very low reactivity with, the trisaccharides. IgG antibodies preferring the A and B Type 2 tetrasaccharides were of the IgG2 subclass. The NGL ELISA further supported the presence of chain type-specific anti-A and -B antibodies among nonsensitized, healthy individuals. CONCLUSION: An ELISA with structurally defined ABH antigens will allow the antibody class and fine specificity of ABO antibodies to be determined, which may improve risk assessment in ABO-incompatible transplantation.