Catharina Steentoft

Glycosylation, the posttranslational linking of sugar molecules to proteins, is notoriously altered during tumor transformation. More specifically in carcinomas, GalNAc-type O-glycosylation, is characterized by biosynthetically immature truncated glycans present on the cancer cell surface, which profoundly impact anti-tumor immune recognition. The tumor-associated glycan pattern may thus be regarded as a biomarker of immune modulation. In epithelial ovarian cancer (EOC) there is a particular lack of specific biomarkers and molecular targets to aid early diagnosis and develop novel therapeutic interventions. The aim of this study was to investigate the ovarian cancer O-glycoproteome and identify tumor-associated glycoproteins relevant in tumor–dendritic cell (DC) interactions, mediated by macrophage galactose-like C type lectin (MGL), which recognizes the tumor-associated Tn O-glycan. Lectin weak affinity chromatography (LWAC) was employed to probe the O-glycopeptidome by MGL and Vicia villosa agglutinin (VVA) lectin using glycoengineered ovarian cancer cell lines and ovarian cancer tissues as input material. Biochemical and bioinformatics analysis gave information on the glycan arrangement recognized by MGL in tumor cells. The potential MGL binders identified were located, as expected, at the cell membrane, but also within the intracellular compartment and the matrisome, suggesting that MGL in vivo may play a complex role in sensing microenvironmental cues. The tumor glycoproteins binders for MGL may become relevant to characterize the interaction between the immune system and tumor progression and contribute to the design of glycan targeting-based strategies for EOC immunotherapeutic interventions.

Human cancer cells were eradicated by adoptive transfer of T cells transduced with a chimeric antigen receptor (CAR) made from an antibody (237Ab) that is highly specific for the murine Tn-glycosylated podoplanin (Tn-PDPN). The objectives were to determine the specificity of these CAR-transduced T (CART) cells and the mechanism for the absence of relapse. We show that although the 237Ab bound only to cell lines expressing murine Tn-PDPN, the 237Ab-derived 237CART cells lysed multiple different human and murine cancers not predicted by the 237Ab binding. Nevertheless, the 237CART cell reactivities remained cancer specific because all recognitions were dependent on the Tn glycosylation that resulted from COSMC mutations that were not present in normal tissues. While Tn was required for the recognition by 237CART, Tn alone was not sufficient for 237CART cell activation. Activation of 237CART cells required peptide backbone recognition but tolerated substitutions of up to 5 of the 7 amino acid residues in the motif recognized by 237Ab. Together, these findings demonstrate what we believe is a new principle whereby simultaneous recognition of multiple independent Tn-glycopeptide antigens on a cancer cell makes tumor escape due to antigen loss unlikely.

Complex carbohydrates serve a wide range of biological functions in cells and tissues, and their biosynthesis involves more than 200 distinct glycosyltransferases (GTfs) in human cells. The kinetic properties, cellular expression patterns and subcellular topology of the GTfs direct the glycosylation capacity of a cell. Most GTfs are ER or Golgi resident enzymes, and their specific subcellular localization is believed to be distributed in the secretory pathway according to their sequential role in the glycosylation process, although detailed knowledge for individual enzymes is still highly fragmented. Progress in quantitative transcriptome and proteome analyses has greatly advanced our understanding of the cellular expression of this class of enzymes, but availability of appropriate antibodies for in situ monitoring of expression and subcellular topology have generally been limited. We have previously used catalytically active GTfs produced as recombinant truncated secreted proteins in insect cells for generation of mouse monoclonal antibodies (mAbs) to human enzymes primarily involved in mucin-type O-glycosylation. These mAbs can be used to probe subcellular topology of active GTfs in cells and tissues as well as their presence in body fluids. Here, we present several new mAbs to human GTfs and provide a summary of our entire collection of mAbs, available to the community. Moreover, we present validation of specificity for many of our mAbs using human cell lines with CRISPR/Cas9 or zinc finger nuclease (ZFN) knockout and knockin of relevant GTfs.

Successful application of potent antibody-based T-cell engaging immunotherapeutic strategies is currently limited mainly to hematological cancers. One major reason is the lack of well-characterized antigens on solid tumors with sufficient cancer specific expression. Aberrantly O-glycosylated proteins contain promising cancer-specific O-glycopeptide epitopes suitable for immunotherapeutic applications, but currently only few examples of such antibody epitopes have been identified. We previously showed that chimeric antigen receptor T-cells directed towards aberrantly O-glycosylated MUC1 can control malignant growth in a mouse model. Here, we present a discovery platform for the generation of cancer-specific monoclonal antibodies targeting aberrant O-glycoproteins. The strategy is based on cancer cell lines engineered to homogeneously express the truncated Tn O-glycoform, the so-called SimpleCells. We used SimpleCells of different cancer origin to elicit monoclonal antibodies with selectivity for aberrant O-glycoproteins. For validation we selected and characterized one monoclonal antibody (6C5) directed to a Tn-glycopeptide in dysadherin (FXYD5), known to be upregulated in cancer and promote metastasis. While dysadherin is widely expressed also in normal cells, we demonstrated that the 6C5 epitope is specifically expressed in cancer.

Glycosaminoglycans (GAGs) are essential polysaccharides in normal physiology and disease. However, understanding of the contribution of specific GAG structures to specific biological functions is limited, largely because of the great structural heterogeneity among GAGs themselves, as well as technical limitations in the structural characterization and chemical synthesis of GAGs. Here we describe a cell-based method to produce and display distinct GAGs with a broad repertoire of modifications, a library we refer to as the GAGOme. By using precise gene editing, we engineered a large panel of Chinese hamster ovary cells with knockout or knock-in of the genes encoding most of the enzymes involved in GAG biosynthesis, to generate a library of isogenic cell lines that differentially display distinct GAG features. We show that this library can be used for cell-based binding assays, recombinant expression of proteoglycans with distinct GAG structures, and production of distinct GAG chains on metabolic primers that may be used for the assembly of GAG glycan microarrays.

Optimal research results rely on the selection of cellular models capable of recapitulating the characteristics of primary tumours from which they originate. The expression of mucins (MUC16 and MUC1) and truncated O-glycans (Tn, STn and T) represents a characteristic footprint of serous ovarian carcinomas (SOCs). Therefore, selecting ovarian cancer (OVCA) cell lines that reflect this phenotype is crucial to explore the putative biological role of these biomarkers in the SOC setting. Here, we investigated a panel of OVCA cell lines commonly used as SOC models, and tested whether, when cultured in 2D and 3D conditions, these recapitulate the mucin and O-glycan expression profiles of SOCs. We further explored the role of truncating the O-glycosylation capacity in OVCAR3 cells through knockout of the COSMC chaperone, using in vitro and in vivo assays. We found that the majority of OVCA cell lines of serous origin do not share the mucin and truncated O-glycan footprint of SOCs, although 3D cultures showed a higher resemblance. We also found that genetic truncation of the O-glycosylation capacity of OVCAR3 cells did not enhance oncogenic features either in vitro or in vivo. This study underscores the importance of well-characterized cellular models to study specific features of ovarian cancer.

Cancer immunotherapy is rapidly advancing in the treatment of a variety of hematopoietic cancers, including pediatric acute lymphoblastic leukemia and diffuse large B cell lymphoma, with chimeric antigen receptor (CAR)-T cells. CARs are genetically encoded artificial T cell receptors that combine the antigen specificity of an antibody with the machinery of T cell activation. However, implementation of CAR technology in the treatment of solid tumors has been progressing much slower. Solid tumors are characterized by a number of challenges that need to be overcome, including cellular heterogeneity, immunosuppressive tumor microenvironment (TME), and, in particular, few known cancer-specific targets. Post-translational modifications that differentially occur in malignant cells generate valid cell surface, cancer-specific targets for CAR-T cells. We previously demonstrated that CAR-T cells targeting an aberrant O-glycosylation of MUC1, a common cancer marker associated with changes in cell adhesion, tumor growth and poor prognosis, could control malignant growth in mouse models. Here, we discuss the field of glycan-directed CAR-T cells and review the different classes of antibodies specific for glycan-targeting, including the generation of high affinity O-glycopeptide antibodies. Finally, we discuss historic and recently investigated glycan targets for CAR-T cells and provide our perspective on how targeting the tumor glycoproteome and/or glycome will improve CAR-T immunotherapy.

The GlycoDomainViewer is a bioinformatic tool to aid in the mining of glycoproteomic datasets from different sources and facilitate incorporation of glycosylation into studies of protein structure and function. We present a version 2.0 of GlycoDomainViewer incorporating a number of advanced features, which enhances visibility and accessibility of the wealth of glycoproteomic data being generated. The GlycoDomainViewer enables visual exploration of glycoproteomic data, incorporating information from recent N- and O-glycoproteome studies on human and animal cell lines and some organs and body fluids. The initial data comprises sites of glycosylation for N-linked, O-GalNAc, O-Fucose, O-Xyl, O-Mannose (in both human and yeast) and cytosolic O-GlcNAc type. The data made available via this tool will be regularly updated to improve the coverage of known glycosylation sites and datasets, reflecting the advances currently being made in characterization of glycoproteomes. The tool is available at https://glycodomain.glycomics.ku.dk.

Cholera toxin (CT) enters and intoxicates host cells after binding cell surface receptors using its B subunit (CTB). The ganglioside (glycolipid) GM1 is thought to be the sole CT receptor; however, the mechanism by which CTB binding to GM1 mediates internalization of CT remains enigmatic. Here we report that CTB binds cell surface glycoproteins. Relative contributions of gangliosides and glycoproteins to CTB binding depend on cell type, and CTB binds primarily to glycoproteins in colonic epithelial cell lines. Using a metabolically incorporated photocrosslinking sugar, we identified one CTB-binding glycoprotein and demonstrated that the glycan portion of the molecule, not the protein, provides the CTB interaction motif. We further show that fucosylated structures promote CTB entry into a colonic epithelial cell line and subsequent host cell intoxication. CTB-binding fucosylated glycoproteins are present in normal human intestinal epithelia and could play a role in cholera.