Primary central nervous system lymphoma reveals a pro-tumoral brain-specific myeloid landscape.

Abstract Background: Primary Central Nervous System Lymphoma (PCNSL) ranks among the most aggressive MCD-ABC-DLBCLs driven by MYD88L265P and CD79B mutations resulting in activation of NF-κB signaling. PCNSL is associated with poor clinical outcomes, and patients often exhibit resistance to treatment. Previous clinical trials aimed at targeting NF-κB dependency in CNSL have achieved limited success. Within the CNS, local gliosis and the infiltration of inflammatory cells, particularly macrophages, may suppress antitumor immunity and enhance NF-κB signaling in MCD-DLBCL cells, contributing to treatment resistance. We propose that brain-resident myeloid cells (bTAMs) are reprogrammed by MCD-DLBCL cells into pro-tumoral states, driving aggressive phenotypes in the CNS. Disrupting the reprogramming of tumor-responsive inflammatory myeloid cells could mitigate the pro-tumorigenic effect and enhance the effectiveness of existing therapies. Methods: To model PCNSL in vivo, we implanted mouse DLBCL cells harboring Myd88L252P, Cd79b mutations, and BCL2 overexpression unilaterally into the striatum of C57BL/6 mice. Single-cell RNA-sequencing of tumor-implanted (TS) and contralateral (NTS) brain regions was conducted to investigate the lymphoma microenvironment (LME) and define bTAMs. We identified candidate mediators of lymphoma cells and bTAM crosstalk using CellChat. Candidates were prioritized for validation according to presence in human disease and potential for pharmacological targeting. Human MCD-PCNSL (MYD88L265P, CD79B) was investigated by spatial transcriptomic profiling using CosMx (6K transcripts, 64 proteins) and multiparametric imaging. This allowed us to identify transcriptionally distinct spatial communities and assess the organization of bTAMs and lymphoma cells. Finally, we used pharmacological intervention in a co-culture system to target bTAMs to test antitumor activity. Results: Single-cell spatial transcriptomics/proteomics of human MCD-PCNSL samples (totalizing >43,000 cells) identified multiple spatial communities, with some enriched for bTAM-associated markers (e.g., CD68, CXCR4, ITGB2) and cytokine/immune-related genes (e.g., CXCL13, IL32, APOE), indicating active microglial signaling within the LME. In the PCNSL murine model, immunofluorescence staining confirmed localized microglial activation in response to the implanted MCD-DLBCL cells. Single-cell RNA-seq revealed 10 myeloid sub-clusters in TS and NTS sites, with four showing increased abundance in TS. These clusters reflected distinct immune functions, including interferon signaling, complement activation (Ifitm3, Ly6a, C1qa, C1qb, C1qc, Tyrobp) antigen presentation (H2-Aa, H2-Ab1, H2-Eb1, H2-Q7), pro-inflammatory cytokine production and chemotaxis (Cxcl13, Cxcl10, Ccl5, Ccl8, Il12b, Pim1, Il1b, Trps1, Ccl2, Ccl12). PPI network analysis (via MCODE) of TS-enriched clusters highlighted key pathways in pro-inflammatory activation and antigen presentation (Tlr4, Stat4, Il12rb1, Tnf, Cd28), tissue remodeling (Thbs1, Sdc4, Itgb1, Itga5), reactive oxidative stress, and the development of a tolerogenic state (classical disease-associated microglial signature of Apoe, App, Psen1, Psen2), corroborating human data. Based on interaction analysis, we identified communication between lymphoma cells and bTAMs through the MIF–CD74 pathway, the Nampt–Itga5/Itgb1 axis, PPIA–BSG, and ApoE–TREM2/TYROBP complexes. Recent studies have implicated TREM2 activation and the SPP1–integrin axis in tumor progression and immune evasion. We found that SPP1 expression was increased in TS myeloid clusters and thus designed a therapeutic strategy to (i) inhibit TREM2–Syk signaling and (ii) block integrins to disrupt bTAM–lymphoma crosstalk and determine its effects on MCD-DLBCL proliferation. Conclusion: We established a novel immunocompetent murine model of PCNSL and uncovered distinct brain bTAM populations that likely participate in lymphoma progression through specific ligand–receptor interactions, laying the groundwork for the development of bTAM-directed therapies to overcome immune evasion in this aggressive lymphoma subtype.