An in vitro one pot synthetic biology approach to emulating Golgi O-glycosylation divergence in MUC1 and tumor associated MUC1

Peptide O-glycosylation is a non-template driven process dependent on an orchestrated collaboration of glycosyltransferases (GTs) in the ER and golgi apparatus. An in vitro modelling of this requires an understanding of GT specificities, kinetics and their spatial distribution along the ER-Golgi axis. This study explores the specificity and kinetics of O-GalNAc glyco-sylation on a designed tandem repeat of MUC1 (23mer), a protein significantly implicated in cancer progression. Using an in vitro one-pot synthetic biology approach, MUC1 glycopeptides were produced, and GT specificity for cancer-associated anti-gens Tn, T, and their sialylated forms were assessed. The impact of GT re-localization from the ER to Golgi on glycosyla-tion patterns was modelled and analyzed. The model suggests that presenting MUC1 to GALNTs in isolation from other GTs, namely C1GALT1 and ST6GALNAC1 significantly increases the extent of site occupancy by Tn a tumour epitope. This mimics the ER localization of GALNTs associated with cancer. In contrast when MUC1 is presented to GALNTs in combination with C1GALT1 there is a decline in observed occupancy. Our modelling shows this to be the result of a com-petition between the secondary glycosylation of the GalNac sites by C1GALT1 and the GALNTs’ lectins that need to bind to the GalNac sites to further proceed with the primary glycosylation of adjacent free sites. In the case of normal MUC1 C1GALT1 shows greater specificity to the under-occupied (GalNAc3-23mer) form of Tn. In the cancer case ST6GALNAC1 shows greater specificity to the fully occupied form (GalNAc5-23mer). These results suggest a mechanism of regulation of tumour associated MUC1 antigens independent of GTs expression level.