Novel insights into the glycosylation signature of myeloid-derived suppressor cells in health and cancer

ADA G. BLIDNER; GABRIEL A RABINOVICH

Medellin

Congreso; IMMUNOCOLOMBIA2015 - 11th Congress of the Latin American Association of Immunology - 10o. Congreso de la Asociación Colombiana de Alergia, Asma e Inmunología; 2015

IUIS, SAI

Myeloid derived cells (MDSCs, macrophages, dendritic cells and neutrophils) comprise a heterogenous cell population that is origined from a single myeloid progenitor cell in the bone marrow, and are exported to peripheral tissues as immature cells to complete their maturation process. Macrophages, dendritic cells and neutrophils represent the first barrier of defense against pathogens; however, they are crucial to shape the adaptive response. In the tumor microenvironment, these cells have gained relevance because of their dual roles and plasticity displaying either pro- or anti-tumor activities. The glycome, defined as the entire set of glycans displayed by cells, tissues and the extracellular matrix encodes critical information that is required for cellular activation, differentiation, survival and homeostasis. The glycosylation signature of a given cell type can bring important information about their selective interactions with endogenous glycan-binding proteins present in inflammatory or tumor microenvironments. Here we examined the glycosylation profile of different myeloid cells with the major goal of investigating lectin-driven immunoregulatory circuits driven by these cells in inflammatory and tumor settings.As MDSCs are heterogeneous population of immature myeloid cells and, in normal conditions, give rise to DCs and macrophages, we examined whether the differentiation process influenced their cell surface glycosylation pattern. We differentiated MDSCs, dendritic cells (DCs) and macrophages (Ms) from the bone marrow of wild type C57BL/6 mice. We obtained bone marrow progenitors and cultured them with 20ng/ml mouse recombinant GM-CSF for 4 or 8 days to obtain MDSCs and DCs respectively. To obtain macrophages we used 3% M-CSF for 8 days. We analyzed the cell surface glycosylation profiles by flow cytometry, using biotinylated lectins and FITC-associated streptavidin. We stained MDSCs with the surface markers CD11b and Gr1, along with the different biotinylated lectins. We found that Gr1highCD11b+ and Gr1low/int CD11b+ populations presented differential glycosylation profiles: the Gr1highCD11b+ cells stained positively for PHA-L and LEL, reflecting abundant beta 1-6 N-glycan branching and poly-LacNAc extension and a particular subpopulation of the cells was positive for PNA which recognizes asialo-core 1-O-glycans, whereas these cells did not show binding for SNA, a lectin that specifically recognizes alfa 2,6-linked sialic acid . On the other hand, Gr1low/int CD11b+ cells showed positive binding for PHA-L, PNA and LEL and were also positive for SNA; although a Gr1int sub-subset did not show SNA staining, suggesting spatio-temporal regulation of alfa 2,6-sialyltransferase (ST6Gal1) during conversion from granulocytic toward monocytic MDSCsOn the other hand, both conventional bone marrow-derived DCs and monocytes showed similar cell surface glycosylation profiles with positive staining for PNA, LEL and PHA-L, and subpopulatons that were either positive or negative for SNA.Next, we examined the glycosylation status of MDSCs present in different hematopoietic organs in normal conditions and whether this phenotype could be influenced by tumor growth We inoculated C57BL/6 with 5x105 Lewis Lung Carcinoma cells in the right flank and left a control group without tumor inoculation. After tumors reached a volume of 800-1000 mm3, we harvested spleens, draining lymph nodes (dLNs), bone marrow and tumors and stained the cell suspension for Gr1, CD11b and biotinylated lectins. We found profound differences in the glycosylation signature of MDSCs present in different tissues. Binding of SNA to MDSCs from LN of normal mice revealed the presence of alfa 2,6-linked sialic acid; however this glycosydic structure was not present in splenic MDSCs from the same mice.And in different subsets within the same MDSCs population (e.g. alfa 2,6-linked sialic acid was abundant on the surface of Gr1low/int splenic MDSCs from normal mice but was almost absent in the Gr1high cluster, and this cluster exhibited higher expression of tri- and tetra-antennary complex N-glycans than the Gr1low/int cluster. Moreover, we found differences in MDSCs from different tissues from tumor bearing mice (e.g. lack of SNA binding to splenic MDSCs and positive SNA reactivity in the case of tumor MDSCs. Moreover, differences were also observed when compared to control MDSCs (e.g. positive SNA staining in MDSCs isolated from the bone marrow of normal but not tumor-bearing mice).Thus, the plasticity of MDSCs in different tissues and in response to different physiologic or pathologic conditions is also reflected by their cell surface glycosylation profiles. These results may imply selective exposure or masking of galectin-specific binding sites by different myeloid-derived populatoins (particularly GR1high and GR1low clusters), suggesting tissue-specific regulatory responses of these clusters under normal conditions or during tumor growth (Blidner et al, Journal of Immunology, 2015).Our group has been intensively studying the role of galectin-1, a glycan binding protein, in cancer and inflammation. We previously found that galectin-1 binds to DCs, through their glycosidic ligands rendering them tolerogenic (Ilarregui et al, Nature Immunology, 2009), on the other hand, galectin-1 is unable to bind glycan structures that include syalic acid in the alfa 2-6 position (Toscano et al, Nature Immunology, 2007). Therefore, the balance of the cognate and the inhibitory ligands for galectin-1, brings information about the interaction of myeloid cells with this lectin in the different microenvironments.