Protein O-glycosylation has shown to be critical for a wide range of biological processes, resulting in an increased interest in studying the alterations in O-glycosylation patterns of biological samples as disease biomarkers as well as for patient stratification and personalized medicine. Given the complexity of O-glycans, often a large number of samples have to be analysed in order to obtain conclusive results. However, most of the O-glycan analysis work done so far has been performed using glycoanalytical technologies that would not be suitable for the analysis of large sample sets, mainly due to limitations in sample throughput and affordability of the methods. Here we report a largely automated system for O-glycan analysis. We adapted reductive β-elimination release of O-glycans to a 96-well plate system and transferred the protocol onto a liquid handling robot. The workflow includes O-glycan release, purification and derivatization through permethylation followed by MALDI-TOF-MS. The method has been validated according to the ICH Q2 (R1) guidelines for the validation of analytical procedures. The semi-automated reductive β-elimination system enabled for the characterization and relative quantitation of O-glycans from commercially available standards. Results of the semi-automated method were in good agreement with the conventional manual in-solution method while even outperforming it in terms of repeatability. Release of O-glycans for 96 samples was achieved within 2.5 hours, and the automated data acquisition on MALDI-TOF-MS took less than 1 minute per sample. This largely automated workflow for O-glycosylation analysis showed to produce rapid, accurate and reliable data, and has the potential to be applied for O-glycan characterization of biological samples, biopharmaceuticals as well as for biomarker discovery.

Aberrant fucosylation in cancer cells is considered as a signature of malignant cell transformation and it is associated with tumor progression, metastasis and resistance to chemotherapy. Specifically, in colorectal cancer cells, increased levels of the fucosylated Lewisx antigen are attributed to the deregulated expression of pertinent fucosyltransferases, like fucosyltransferase 4 (FUT4) and fucosyltransferase 9 (FUT9). However, the lack of experimental models closely mimicking cancer-specific regulation of fucosyltransferase gene expression has, so far, limited our knowledge regarding the substrate specificity of these enzymes and the impact of Lewisx synthesis on the glycome of colorectal cancer cells. Therefore, we sought to transcriptionally activate the Fut4 and Fut9 genes in the well-known murine colorectal cancer cell line, MC38, which lacks expression of the FUT4 and FUT9 enzymes. For this purpose, we utilized a physiologically relevant, guide RNA-based model of de novo gene expression, namely the CRISPR-dCas9-VPR system. Induction of the Fut4 and Fut9 genes in MC38 cells using CRISPR-dCas9-VPR resulted in specific neo-expression of functional Lewisx antigen on the cell surface. Interestingly, Lewisx was mainly carried by N-linked glycans in both MC38-FUT4 and MC38-FUT9 cells, despite pronounced differences in the biosynthetic properties and the expression stability of the induced enzymes. Moreover, Lewisx expression was found to influence core-fucosylation, sialylation, antennarity and the subtypes of N-glycans in the MC38-glycovariants. In conclusion, exploiting the CRISPR-dCas9-VPR system to augment glycosyltransferase expression is a promising method of transcriptional gene activation with broad application possibilities in glycobiology and oncology research.

No abstract