Fig 2: Purification of HisDapGalNAcT2. SDS-PAGE (A) and immunoblot analyses (B and C) of flow-through, wash and elution fractions collected during Ni2+-NTA purification (see Materials and methods) of soluble HisDapGalNAcT2. Insoluble HisDapGalNAcT2 (PC1) produced in Origami™ 2(DE3)pLysS and verified by ESI-MS or His-Em-GFP (PC2, 27 kDa) are included as controls. (A) Coomassie staining detected protein bands with an expected molecular weight (MW) of ca. 61.7 kDa in lanes 7–12. HisDapGalNAcT2 (arrows) was identified by immunoblotting using (B) specific polyclonal anti human GALNT2 antibody or (C) penta-His antibody against the N-terminal His-Tag.

Fig 3: In vitro activity of WbgU analysed by capillary electrophoresis (CE) (a) and in combination with GalNAc-T2 using a modified glycosyltransferase assay (b). Substrate conversion of the sugar substrates UDP-GlcNAc (blue squares) and UDP-GalNAc (yellow squares) using 0.15 mM (small squares) and 1.50 mM (big squares) in the presence of WbgU. UDP-GlcNAc and UDP-GalNAc were separated by CE and detected at 254 nm. The mean substrate conversion was calculated based on the relative peak area of three independent experiments and results are shown including the standard deviation (a). The mean specific activities of recombinant human GalNAc-T2 (rhGalNAc-T2) commercially produced in NS0 cells and His-tagged GalNAc-T2 (HisDapGalNAc-T2) expressed in E. coli were determined in four independent experiments as duplicates using 0.5 mM UDP-GalNAc and UDP-GlcNAc, as activated sugar donors, 0.25 mM EA2 peptide acceptor substrate and 1 µg WbgU. Samples containing WbgU were incubated for 20 min at 37 °C prior to glycosyltransferase addition and subsequently incubated for 20 min at 37 °C. Results are displayed including the standard error; inserted numbers indicate the specific activities in pmol*min−1*µg−1 (b). n.s. not significant

Fig 4: Purification of T7Muc10 using Ni–NTA affinity chromatography. Cell pellets of E. coli SHuffle® T7 Express carrying plasmid pMJS9 in combination with vector construct pET23d_galNT2_T7Muc10 or pET23d_galNT2_T7Muc10_wbgU expressing GalNAc-T2 and T7Muc10 or GalNAc-T2, T7Muc10 and WbgU were thawed and lysed. The soluble fractions were pre-purified using anion exchange chromatography (AIEX). Fractions of the flow-through containing T7Muc10 (AIEX) were loaded onto a Ni–NTA affinity chromatography spin column. Fractions of the flow-through (FT), the first washing step with equilibration buffer (W1), and the second washing step with washing buffer (W2) were collected. T7Muc10 was eluted by applying elution buffer twice (E1, E2). Aliquots of the fractions were analysed by SDS-PAGE stained with Coomassie (a) and Immunoblot treated with Anti-His.H8 antibodies (b). Molecular mass markers (MW) are in kDa

Fig 5: Co-expression of GalNAc-T2, WbgU, and T7Muc10. E. coli SHuffle® T7 Express carrying plasmid pMJS9 in combination with either construct pET23d_galNT2_T7Muc10 (+ GalNAc-T2, + T7Muc10, − WbgU), pET23d_galNT2_T7Muc10_wbgU (+ GalNAc-T2, + T7Muc10, + WbgU) or the vector control pET23d(+) (− GalNAc-T2, − T7Muc10, − WbgU) were cultured in 2 mL EnPresso® B at 30 °C. Cells were harvested 24 h after IPTG induction and lysed. Insoluble and soluble fractions were separated and analysed using Coomassie-stained SDS-Gel (a) and Immunoblot analysis with Anti-GalNAc-T2 (b), Anti-WbgU rabbit serum (c) and Anti-His.H8 antibodies (d). Purified GalNAc-T2 (HisDapGalNT2, 60 kDa) (b), WbgU (40 kDa) (c), and T7Muc10 (16.5 kDa) (d) were included as controls. Molecular mass markers (MW) are in kDa. The presence (+) or absence (−) of galNT2, T7Muc10, and wbgU in the respective expression sample is indicated. Protein bands representing GalNAc-T2 are highlighted by green, WbgU by blue, and T7Muc10 by red and orange arrows