Colorectal cancer cell lines show striking diversity of their O-glycome reflecting the cellular differentiation phenotype

Abstract

Alterations in protein glycosylation in colorectal cancer (CRC) have been extensively studied using cell lines as models. However, little is known about their O-glycome and the differences in glycan biosynthesis in different cell types. To provide a better understanding of the variation in O-glycosylation phenotypes and their association with other molecular features, an in-depth O-glycosylation analysis of 26 different CRC cell lines was performed. The released O-glycans were analysed on porous graphitized carbon nano-liquid chromatography system coupled to a mass spectrometer via electrospray ionization (PGC-nano-LC-ESI-MS/MS) allowing isomeric separation as well as in-depth structural characterization. Associations between the observed glycan phenotypes with previously reported cell line transcriptome signatures were examined by canonical correlation analysis. Striking differences are observed between the O-glycomes of 26 CRC cell lines. Unsupervized principal component analysis reveals a separation between well-differentiated colon-like and undifferentiated cell lines. Colon-like cell lines are characterized by a prevalence of I-branched and sialyl Lewis x/a epitope carrying glycans, while most undifferentiated cell lines show absence of Lewis epitope expression resulting in dominance of truncated α2,6-core sialylated glycans. Moreover, the expression of glycan signatures associates with the expression of glycosyltransferases that are involved in their biosynthesis, providing a deeper insight into the regulation of glycan biosynthesis in different cell types. This untargeted in-depth screening of cell line O-glycomes paves the way for future studies exploring the role of glycosylation in CRC development and drug response leading to discovery of novel targets for the development of anti-cancer antibodies.

N-Glycoproteins Have a Major Role in MGL Binding to Colorectal Cancer Cell Lines: Associations with Overall Proteome Diversity

“…I-branching (GlcNAcβ1-6Gal-R) of mucin type O -glycans is mediated by β1,6-N-acetylglucosamine transferase 3 (GCNT3). The quantitative data we obtained for this enzyme revealed higher levels in HT29 and LS174T, in comparison to HCT116 ( Figure 3 ), in accordance with transcriptomic and glycomic data from literature [ 18 , 20 ]. Another characteristic of HCT116 is the overall low level of fucosylation, associated with a more aggressive phenotype [ 17 , 18 ].…”

“…The quantitative data we obtained for this enzyme revealed higher levels in HT29 and LS174T, in comparison to HCT116 ( Figure 3 ), in accordance with transcriptomic and glycomic data from literature [ 18 , 20 ]. Another characteristic of HCT116 is the overall low level of fucosylation, associated with a more aggressive phenotype [ 17 , 18 ]. In HCT116, this feature can partially be explained by a deletion of 142 amino acids of the GDP-mannose-4,6-dehydratase (GMDS) involved in GDP-L-fucose synthesis, which may lead to misfolding and degradation of the enzyme [ 22 ].…”

“…CDX-1 is very similar to CDX-2, and these proteins share several tryptic peptides, but we did not observe a unique CDX-1 peptide in our dataset. In line with our proteomics data, higher CDX-2 (and CDX-1) mRNA levels were found in highly differentiated cells together with multi-fucosylated N - and O -glycans, e.g., LS174T [ 18 , 29 ] and mucin 2 is up-regulated by CDX-2 [ 30 ]. Hence, our data indicate that downstream targets of CDX-2 (and potentially CDX-1), especially the genes encoding proteins involved in glycosylation, could play a role in the differential binding of MGL but this warrants further investigation.…”

“…Although high MGL binding was observed to HCT116 and HT29 cells, binding to LS174T cells was negligible. The comprehensive characterization of N - [ 17 ] and O -glycans [ 18 ] expressed by the three CRC cell models used in our study, did not provide an explanation for the difference in MGL binding to these cell lines. We identified the major cell surface proteins binding to MGL in HCT116 and HT29 cells and, for some of these, found a glycopeptide with the MGL specific glycotope (e.g., LacdiNAc (on an N -glycan) and Tn antigen), but for many the glycotope remained elusive [ 16 ].…”

Structural and functional characterization of SARS-CoV-2 RBD domains produced in mammalian cells

“…Structures of detected glycans were confirmed by MS/MS in negative mode [23]. Glycan structures were assigned on the basis of the known MS/MS fragmentation patterns in negative-ion mode [24–26], elution order, and general glycobiological knowledge, with help of Glycoworkbench [27] and Glycomod software [28].…”