β-Hexosaminidases Along the Secretory Pathway of Nicotiana benthamiana Have Distinct Specificities Toward Engineered Helminth N-Glycans on Recombinant Glycoproteins

Abstract

Secretions of parasitic worms (helminths) contain a wide collection of immunomodulatory glycoproteins with the potential to treat inflammatory disorders, like autoimmune diseases. Yet, the identification of single molecules that can be developed into novel biopharmaceuticals is hampered by the limited availability of native parasite-derived proteins. Recently, pioneering work has shown that helminth glycoproteins can be produced transiently in Nicotiana benthamiana plants while simultaneously mimicking their native helminth N-glycan composition by co-expression of desired glycosyltransferases. However, efficient "helminthization" of N-glycans in plants by glyco-engineering seems to be hampered by the undesired truncation of complex N-glycans by β-N-acetyl-hexosaminidases, in particular when aiming for the synthesis of N-glycans with antennary GalNAcβ1-4GlcNAc (LacdiNAc or LDN). In this study, we cloned novel β-hexosaminidase open reading frames from N. benthamiana and characterized the biochemical activity of these enzymes. We identified HEXO2 and HEXO3 as enzymes responsible for the cleavage of antennary GalNAc residues of N-glycans on the model helminth glycoprotein kappa-5. Furthermore, we reveal that each member of the HEXO family has a distinct specificity for N-glycan substrates, where HEXO2 has strict β-galactosaminidase activity, whereas HEXO3 cleaves both GlcNAc and GalNAc. The identification of HEXO2 and HEXO3 as major targets for LDN cleavage will enable a targeted genome editing approach to reduce undesired processing of these N-glycans. Effective knockout of these enzymes could allow the production of therapeutically relevant glycoproteins with tailor-made helminth N-glycans in plants.

Keywords: LacdiNAc; N-glycosylation; Nicotiana benthamiana; glyco-engineering; β-hexosaminidases.

FIGURE 1

(A) Schematic illustration of proposed plant HEXO activity on complex LDN-carrying N-glycans. (B) Schematic overview of the expression vectors used in this study. LB, left border; Pnos, nopaline synthase gene promoter; Hyg, hygromycin B phosphotransferase gene; Kan, neomycin phosphotransferase 2 gene; Tnos, nopaline synthase gene terminator; CaMV, cauliflower mosaic virus; d35S, dual cauliflower mosaic virus 35S gene promoter; P35S, cauliflower mosaic virus 35S gene promoter; HEXO, N. benthamiana HEXO ORF; GalNAcT, β1,4-N-acetyl-galactosaminyltransferase from Caenorhabditis elegans; GFP, green fluorescent protein; RFP, red fluorescent protein; XYLT, Arabidopsis thaliana b1,2-xylosyltransferase ORF; S: HEXO RNAi sequence in sense orientation; XTI2, intron 2 from A. thaliana XYLT; AS, HEXO RNAi sequence in antisense orientation; g7T, Agrobacterium gene 7 terminator; and RB, right border.

FIGURE 2

Expression of Nicotiana benthamiana beta-hexosaminidases (HEXOs). GFP- or RFP-tagged HEXOs or IL-22-GFP (used as apoplast control protein) were transiently expressed in N. benthamiana leaves. As a negative control (–) p19 was used. Leaves were harvested 3 days post infiltration and analyzed by western blot. (A) anti-GFP western blot under reducing conditions of 10 μg total soluble protein of apoplast fluids or intracellular fraction upon expression of IL-22-GFP (22), HEXO2A (2A), HEXO2B (2B), or HEXO3B (3B). (B) anti-RFP western blot under reducing conditions of 10 μg total soluble protein of apoplast fluids or intracellular fraction upon expression of HEXO1A (1A) or HEXO3A (3A).

FIGURE 3

Subcellular localization of Nicotiana benthamiana HEXOs. GFP- or RFP-tagged HEXOs or formin-mCherry (used as plasma membrane marker) were transiently expressed in N. benthamiana leaves. Whole mount confocal microscopy pictures were taken at 3 days post infiltration. For each construct overview pictures are given (scale bar of 20 μm) and a picture at higher magnification underneath (scale bar 5 μm).

FIGURE 4

Cleavage of terminal GlcNAc residues by N. benthamiana HEXOs. Kappa-5 was co-expressed with p19 and different HEXO genes in wild-type plants. After isolation of the apoplast fluid (at 6 dpi) the glycan composition of kappa-5 was analyzed. (A) Griffonia simplicifolia lectin II (GSL-II) binding assay for detection of terminal GlcNAc. Lectin binding is measured as absorbance at 450 nm. (B–D) Kappa-5 was purified from the apoplast fluid and the N-glycan composition was analyzed by MALDI-TOF-MS. MS profiles are given for N-glycans of wild-type kappa-5 (B) or upon co-expression of HEXO1A (C) or HEXO3A (D). m/z refers to mass-to-charge ratio.

FIGURE 5

Cleavage of engineered LDN motifs by N. benthamiana HEXOs. Kappa-5 was co-expressed with p19, CeGalNAcT and different HEXO genes in wild-type plants. After isolation of the apoplast fluid (at 6 dpi) the presence of the LDN motif on kappa-5 N-glycans was analyzed. (A) Soybean agglutinin (SBA) binding assay for detection of terminal GalNAc. Lectin binding is measured as absorbance at 450 nm. (B–D) Kappa-5 was purified from the apoplast fluid and the N-glycan composition was analyzed by MALDI-TOF-MS. MS profiles are given for N-glycans of kappa-5 upon co-expression of GalNAcT only (B), or combined with co-expression of HEXO2B (C), or HEXO3A (D). When a MS peak represents multiple N-glycan structures of identical mass, the possible positions of these sugar residues on the N-glycan are indicated between brackets. m/z refers to mass-to-charge ratio.

FIGURE 6

Transient silencing of HEXO2 and HEXO3. Kappa-5 was co-expressed with CeGalNAcT and constructs targeting HEXO2 (Δ2), HEXO3 (Δ3), or both (Δ2/3). The samples are grouped in three blocks depending on whether a HEXO variant (2B or 3A) was co-expressed. Control samples where a particular construct is not co-infiltrated is indicated as well (–). After isolation of the apoplast fluid (at 5 dpi) the presence of the LDN motif on kappa-5 N-glycans was analyzed. (A) Soybean agglutinin (SBA) binding assay for detection of terminal GalNAc. Lectin binding is measured as absorbance at 450 nm. (B,C) Kappa-5 was purified from the apoplast fluid and the N-glycan composition was analyzed by MALDI-TOF-MS. MS profiles are given for glucosaminidase treated N-glycans of kappa-5 upon co-expression of GalNAcT only (B) or combined with the hairpin construct targeting HEXO2 and HEXO3 (Δ2/Δ3; C). m/z refers to mass-to-charge ratio.