Transient Glyco-Engineering to Produce Recombinant IgA1 with Defined N- and O-Glycans in Plants

Abstract

The production of therapeutic antibodies to combat pathogens and treat diseases, such as cancer is of great interest for the biotechnology industry. The recent development of plant-based expression systems has demonstrated that plants are well-suited for the production of recombinant monoclonal antibodies with defined glycosylation. Compared to immunoglobulin G (IgG), less effort has been undertaken to express immunoglobulin A (IgA), which is the most prevalent antibody class at mucosal sites and a promising candidate for novel recombinant biopharmaceuticals with enhanced anti-tumor activity. Here, we transiently expressed recombinant human IgA1 against the VP8* rotavirus antigen in glyco-engineered ΔXT/FT Nicotiana benthamiana plants. Mass spectrometric analysis of IgA1 glycopeptides revealed the presence of complex biantennary N-glycans with terminal N-acetylglucosamine present on the N-glycosylation site of the CH2 domain in the IgA1 alpha chain. Analysis of the peptide carrying nine potential O-glycosylation sites in the IgA1 alpha chain hinge region showed the presence of plant-specific modifications including hydroxyproline formation and the attachment of pentoses. By co-expression of enzymes required for initiation and elongation of human O-glycosylation it was possible to generate disialylated mucin-type core 1 O-glycans on plant-produced IgA1. Our data demonstrate that ΔXT/FT N. benthamiana plants can be engineered toward the production of recombinant IgA1 with defined human-type N- and O-linked glycans.

Keywords: N-glycosylation; O-glycosylation; antibody; glyco-engineering; monomeric IgA; plant-made pharmaceuticals; protein glycosylation; recombinant glycoprotein.

FIGURE 1

Schematic overview of constructs for expression and glyco-engineering. (A) A monomeric IgA1 with its different domains is indicated. The N-glycosylation sites in the CH2 domain and in the tailpiece of the alpha chain are marked using the symbols according to the nomenclature from the Consortium for Functional Glycomics (http://www.functionalglycomics.org/). The O-glycans in the hinge region are depicted by yellow squares (GalNAc), yellow cycles (galactose) and purple diamonds (sialic acid). (B) Illustration of the multigene expression vector for sIgA1. 35S, cauliflower mosaic virus 35S promoter; SC, secretory component; t, terminator sequence; JC, joining chain; αC, alpha chain; λC, lambda light chain. (C) GnTII: human N-acetylglucosaminyltransferase II used for N-glycan engineering. (D) Enzymes for sialylated core 1 formation: GalNAc-T2, human polypeptide GalNAc-transferase 2; C1GalT1, Drosophila melanogaster core 1 β1,3-galactosyltransferase; ST6GalNAc, Mus musculus α2,6-sialyltransferase III/IV; ST3Gal-I, human α2,3-sialyltransferase I.

FIGURE 2

Analysis of sIgA1 expression. (A) SDS-PAGE and immunoblotting of crude protein extracts (Ex), SSL7-purified sIgA1 (P), and intercellular fluid (IF) from N. benthamiana wild-type (wt) or ΔXT/FT (ΔXF) infiltrated plants with either anti-alpha chain (anti-αC), anti-secretory component (anti-SC), or anti-lambda light chain (anti-λC) antibodies. (B) SDS-PAGE under non-reducing conditions followed by immunoblotting. (C) SDS-PAGE under reducing conditions and Coomassie staining of SSL-purified sIgA1 from plants. Human serum IgA was loaded for comparison.

FIGURE 3

Enzymatic deglycosylation of expressed sIgA1 chains. (A) Crude protein extracts were digested with Endo H or PNGase F. Proteins were separated by SDS-PAGE and analyzed by immunoblotting with antibodies against the alpha chain (anti-αC). (B) SSL7-purified samples were Endo H and PNGase F digested, respectively, subjected to SDS-PAGE followed by immunoblotting with antibodies against the alpha chain. (C) IF and crude extracts were digested with PNGase F and analyzed by SDS-PAGE and immunoblotting with antibodies against the secretory component (anti-SC).

FIGURE 4

N-glycan analysis of sIgA1. (A) Mass spectra of the tryptic glycopeptide from the CH2 domain of sIgA1 expressed in N. benthamiana wild-type (wt) or ΔXT/FT (ΔXF). The amino acid sequence of the identified peptide is highlighted. The N-glycosylation site is underlined. (B) sIgA1 was transiently co-expressed with human GnTII and analyzed as mentioned in (A). (C) The corresponding Glu-C/trypsin digested glycopeptide from human serum IgA. (D) The spectrum of the tailpiece glycopeptide of the alpha chain from human serum IgA. (E) Spectra from two glycopeptides of the SC derived from the IF of wild-type plants. A detailed explanation of the used N-glycan abbreviations can be found at the ProGlycAn homepage (http://www.proglycan.com/index.php?page=pga_nomenclature). The graphical depictions of glycan-structures follow the style of the Consortium for Functional Glycomics (http://www.functionalglycomics.org/static/consortium/Nomenclature.shtml).

FIGURE 5

MS analysis showing the presence of plant specific O-glycan modifications. The analyzed peptide sequence with potential O-glycosylation sites (highlighted in bold) is shown. (A) The glycopeptide corresponding to the alpha chain hinge region contains different amounts of hydroxyproline (Hyp) residues. (B) Peaks corresponding to Hyp + pentoses (Pent) were detected in wild-type and ΔXT/FT plants.

FIGURE 6

Co-expression of GalNAc-T2 and C1GalT1 results in the modification of the hinge region peptide. (A) Co-expression of GalNAc-T2 leads to the incorporation of HexNAc residues. (B) Co-expression of GalNAc-T2 and C1GalT1 results in the generation of core 1 structures: HexNAc + hexose (Hex). (C) Incubation with jacalin agarose (+) reveals binding of the sIgA1 alpha chain when modified in planta with mucin-type O-glycan biosynthesis enzymes. The amount of sIgA1 before binding to jacalin is also shown (-). Human serum IgA was used as a positive control and sIgA1 alone or co-expressed with the N-glycan modifying β1,4-galactosyltransferase (GalT) was used as negative control.

FIGURE 7

In planta generation of sialylated core 1 structures on the alpha chain hinge region. (A) Mass-spectrum of trypsin-digested sIgA1 expressed in N. benthamiana ΔXT/FT (ΔXT) line is shown. sIgA1 was coexpressed with GalNAc-T2, C1GalT1, ST6GalNAc-III/IV, ST3Gal-I and the required proteins for CMP-sialic acid synthesis and Golgi-transport to generate disialyl core 1 structures. The inset shows peaks corresponding to modifications of four O-glycosylation sites with sialic acid (4–4–8: 4x hexose – 4x HexNAc – 8x N-acetylneuraminic acid). (B) For comparison, the O-glycan structures derived from human serum IgA are indicated.