Matteo Bocci

Glioblastoma (GBM) is characterized by fast progression, infiltrative growth pattern, and a high relapse rate. A defining feature of GBM is the existence of spatially and functionally distinct cellular niches, where malignant cells engage in paracrine crosstalk with cell types comprising the tumor microenvironment. Here, we identify pericytes as the most active paracrine signaling hub within the tumor parenchyma. Their depletion through genetic engineering results in accelerated tumor progression and shortened survival. Mechanistic studies reveal that pericyte deficiency remodels the endothelium and impacts the immune cell landscape, exacerbating tumor cell invasion and immune suppression. Specifically, the pericyte-deprived endothelium recruits perivascular, tumor-associated macrophages polarized towards an immune-suppressive phenotype. The recruited macrophages express Hepatocyte Growth Factor, which reinforces activation of its receptor tyrosine kinase MET on GBM cells harboring a pronounced mesenchymal subtype driven by the key phenotypic regulator Fosl1 . Indeed, orthotopic implantation of MET-expressing GBM cells corroborates their superior tumor-initiating and invasive capabilities. Thus, pericytes represent critical modulators of GBM development by orchestrating a tumor-suppressive microenvironment, highlighting the importance of their preservation in therapy.

Descriptive statistics of the distances from each tumor core (TC) region of interest and the nearest immune infiltrate, measured using QuPath on the digital scan of the microscope slide. All distances are given in micrometers.

S5. Comprehensive cell counts from the multiplex immunohistochemistry analysis.

S1. Prognostic Effect of CD45+ Infiltration in Lung Cancer Brain Metastases.

S4. Immune signatures in the lung cancer brain metastasis cohort.

Predicted ligand-target interactions between immune and MTC networks across all networks (first sheet), adjacent networks (second sheet), immune networks including those not spatially adjacent (third sheet) and interactions plotted in Fig. 4c-d (fourth sheet).

S9. Immune-MTC cellular Networks in Lung Cancer Brain Metastasis.

Differentially expressed genes between all TI ROIs and TC ROIs with log2-fold change, p-value and q-value, for comparison with all patients (first sheet) as well as comparison per patient (second sheet).

Differentially expressed genes (first sheet) and cytokine genes (second sheet) from each tumor cell network with log2-fold changes, p-values, q-values calculated within each network and q-values calculated across all comparisons.

S3. Analysis of the immune landscape in human lung cancer brain metastases using the second reference dataset.