Identification of β-galactosidases along the secretory pathway of Nicotiana benthamiana that collectively hamper engineering of galactose-extended glycans on recombinant glycoproteins.

Glycosylation is an important aspect for many biopharmaceuticals, including vaccines against parasitic helminths. Plants, especially Nicotiana benthamiana, have proven to be excellent production hosts for biopharmaceuticals with tailor-made glycosylation. If desired, galactosylation can be introduced on biopharmaceuticals through co-expression of the appropriate glycosyltransferase. However, achieving homogenous glycoforms with terminal galactose residues remains difficult as native N. benthamiana β-galactosidases (NbBGALs) truncate these glycans. Recently, the first NbBGAL has been identified, but a knockout line was insufficient to achieve near complete galactosylation, suggesting that other enzymes could have similar activity. In this study, we selected 10 NbBGALs for further investigation into subcellular localization, in vitro and in vivo activity against β1,4-linked galactose on N-glycans and β1,3-linked galactose on O-glycans. We show that NbBGAL3B is localized in the apoplast and has similar specificity for β1,4-linked galactose on N-glycans as the previously identified NbBGAL1. In contrast, none of the selected NbBGALs cleaved β1,3-linked galactose from O-glycans besides BGAL1. In addition, we provide a novel strategy to achieve near complete galactosylation on galactosidase-prone glycoproteins by using the protective capacity of the Lewis X motif and subsequent removal of the antennary fucose residues. Taken together, our results provide a broad view of the ability of NbBGALs to cleave galactoses and have identified NbBGAL3B as the second major contributor of undesired β-galactosidase activity while engineering N-glycans. This work lays the foundation for generating knockout lines that are devoid of undesired NbBGALs and therefore do not hamper the production of recombinant glycoproteins with galactose-extended glycans.