Network analysis to dissect glioblastoma immunosuppression complexity: identification of C-type lectins as potential therapeutic candidates and biomarkers

Introduction: Glioblastoma (GBM) is one of the most threatening tumors in adults. The failure of treatments, GBM aggressiveness and the poorness of patients’ quality of life underline the unmet need to identify novel therapeutic targets. The strong immunosuppression, hypoxia and altered vasculature make account for GBM resistance. Lectins are immunoregulatory receptors, that in tumors promote immunosuppression. Lectins expression on distinct cell type (i.e.myeloid, lymphoid and cancer cells) and the distribution of their respective ligands generate a complex and intricate network of interactions that support immunosuppressive mechanisms. So far, little is known about lectins contribution to immunosuppression and progression in GBM. Aims: to investigate the lectin profile relevant in glioblastoma immunosuppression exploiting network-based approaches and to investigate their possible role as immune biomarkers and target Results: Using three public databases (TCGA and CGGA for GBM; GTEx for healthy brain tissue), we selected a panel of 67 lectins from the Siglec, galectin, and C-type lectin families. Setting rigorous filtering parameters based on lectins expression levels, a final set of 39 lectins was employed to construct co-expression and differential co-expression networks. The analysis revealed significant correlations between lectin profiles across TCGA and CGGA. C-type lectins, that have never been investigated in GBM, were the most represented (ASGR2, CLEC12A, CLEC10A, CLEC4A, CLEC7A, CLEC2B, CD209) followed by Siglec7, Siglec9 and galectin-9, that have been already described in GBM progression and immunosuppression. The results obtained by computational analysis were validated on patients-matched tumor samples and liquid biopsies. CLEC10A was expressed in GBM-infiltrating immune cells and particularly its high expression on infiltrating macrophages correlated with poor survival. Distinct CLEC10A+ myeloid cells subsets correlate with different T cell infiltrate. Indeed, CLEC10A+ macrophages were positively associated with activated CD4+ T cells, that are likely to be regulatory T cells, while CLEC10A+ LOX1+ PMN-MDSCs negatively correlate with effector and antitumor CD8+ T cells. These results underline that distinct CLEC10A+ myeloid cells may be involved in different immunomodulatory mechanisms in GBM. CLEC10A ligands Tn and STn were expressed in GBM tissues. Tn is exclusively expressed on cancer cells in GBM in correspondence of macrophages, while STn was expressed also in low-grade gliomas in vascular area. These results highlight a possible role of CLEC10A-Tn axis in GBM TME and immunosuppression, as already described in other cancer histytopes. However, GBM immunosuppression is complex and likely several actors may contribute, thus the intricate network of interactions is not associated to a single lectin and a single cell type. The novel C-type lectins cluster (ASGR2, CLEC12A, CLEC4A, CLEC7A) identified by network analysis was investigated in GBM patients. ASGR2 and CLEC12A emerged as strongest hub of both GBM differential co- 1 expression networks and low levels of both lectins were associated with improved prognosis. Combining ASGR2 and CLEC12A expression with lectins of other families already known (siglec7, siglec9, and galectin9), we identified a lectin profile that better discriminated survival of GBM patients. So far, no evidence of these C-type lectins is available in cancer. We found that all C-type lectins are expressed on GBM-infiltrating immune cells. ASGR2 levels were significantly higher on MDMs compared to resident microglia (MG) and it was undetectable on circulating immune cells, suggesting a specific role in the GBM TME. CLEC7A expression was higher in macrophagic subpopulations compared to MDSCs. CLEC4A and CLEC12A were highly expressed on all circulating immune cell subsets, particularly those of monocytic origin. Their expression was significantly higher on circulating cells than in infiltrating cells. Similar lectin expression patterns were observed in patients with other cancers (NSCLC, head and neck cancer, HER2+ breast cancer), indicating that CLEC4A and CLEC12A may be relevant across different tumor types. Conclusions: We described for the first time C-type lectins cluster (ASGR2, CLEC4A, CLEC7A, CLEC12A, CLEC10A) that have never been described in GBM by exploiting network-based approaches. CLEC10A-Tn axis emerged as possible mechanism that supports GBM immunosuppression, underlining an interplay between macrophages and cancer cells. Moreover, ASGR2 expression appeared to be restricted to GBM TME and in association to macrophages and could represent a lectin that is induced under pathological context and may be involved in immunosuppressive mechanisms. The C-type lectins profile appears to be associated to distinct immunosuppressive myeloid cell subsets that are found not only in GBM TME but also peripheral blood, underlining their possible role as circulating immune biomarkers. Further analysis is required to investigate the metabolic and transcriptomic profile of lectins-expressing myeloid cells. Moreover, identifying the lectins ligands and understanding the impact of their engagement on the myeloid cells may be relevant to identify novel possible target to be exploited in GBM treatment. Investigating changes in lectin expression post-surgery or treatment, along with their relationship to immune features and clinical outcomes, could provide valuable insights into their roles in tumor progression and response to therapies.