EXTH-21. MECHANISMS OF SYNERGISTIC GROWTH SUPPRESSION BY RADIOTHERAPY AND C-MET INHIBITION IN EXPERIMENTAL GLIOBLASTOMA

Glioblastoma remains to be one of the most lethal solid cancers and novel therapies are urgently needed. There is increasing interest in the role of the HGF/MET pathway in the response of glioblastoma to radiotherapy. c-MET-mediated radioresistance may be partially induced via proinvasive and DNA damageresponse pathways and HGF may be involved in the regulation of immune responses. Here, we explored the role of the c-MET pathway in response to radiotherapy and investigated potential modes of action that mediate synergistic effects of MET pathway inhibition and irradiation in syngeneic murine glioma models in vitro and in vivo. Murine glioma cells express HGF and c-MET and respond with c-MET phosphorylation upon exposure to exogenous HGF. In vitro, glioma cell viability and proliferation are not affected by pharmacological or genetic c-MET pathway interference, and the c-MET inhibitor tepotinib fails to sensitize glioma cells to irradiation. Conversely, in vivo c-MET inhibition combined with focal radiotherapy synergistically prolongs survival in two syngeneic orthotopic glioma models compared with either treatment alone. Complementary studies demonstrated that synergy was lost when gliomas were established and treated in immunodeficient mice, and importantly, also when c-MET gene expression was disrupted in the tumor. Thus, synergistic suppression of experimental syngeneic glioma growth by irradiation and c-MET inhibition requires at least two mechanisms, expression of c-MET in the tumor and a functional immune system. In summary, our data suggest clinical evaluation of c-MET pathway inhibition in combination with radiotherapy in human glioblastoma.

CSIG-22. Cd95l/Cd95 GENE DELETION DECREASES MOUSE GLIOMA TUMORIGENESIS THROUGH THE INHIBITION OF NON-CANONICAL CD95L/CD95-MEDIATED CELL GROWTH AND IMMUNOSUPPRESSION

CD95 is a transmembrane receptor with potential to promote both cell death and growth. Initially described to trigger apoptosis upon ligand (CD95L) engagement, CD95 may also prompt cell proliferation, invasion and stemness. CD95 stimulation to induce cancer cell apoptosis has been proved clinically impracticable. However, in tumors expressing both CD95 and CD95L, strategically inhibiting CD95-CD95L interactions to simultaneously block cancer cell growth and apoptotic cell death in tumor microenvironment components, including CD95-expressing antitumor immune effector cells, may represent an alternative therapeutic strategy. Here we characterized the expression of CD95 and CD95L in murine glioma models in vitro and in vivo. To fully disrupt CD95-CD95L interactions, we deleted Cd95 or Cd95l by CRISPR-Cas9-mediated knockout (KO) and assessed the consequences on cell growth and tumorigenicity in immunocompetent and immunocompromised mice. CD95 expression was identified in selected murine glioma cell lines. In vitro, expression of the canonical, membrane-bound, form of CD95L was not detected but cell lines expressed a shorter non-canonical, soluble, Cd95l variant. Tumors generated upon implantation of the same cells in vivo expressed both Cd95l variants. Upon Cd95l KO, all investigated cell lines exhibited reduced growth in vitro. Cell growth reduction upon Cd95 KO in SMA-497 murine glioma cells was rescued upon Cd95 re-transfection, validating CD95 specificity of the phenotype. Cd95-overexpression in Cd95-expressing cells did not increase growth. In vivo, Cd95 or Cd95l KO cell implantation in syngeneic mice generated smaller tumors than wildtype cells, resulting in prolonged survival. While 40% Cd95l KO cell-implanted immunocompetent mice did not develop tumors, all immunodeficient mice did. Altogether, these data reveal a growth-promoting role of non-canonical CD95L-CD95 interactions in murine gliomas, which blockade through gene KO results in decreased tumorigenicity. Furthermore, our data suggest the contribution of CD95L-mediated immunosuppression to the reduction of Cd95l KO-associated tumorigenicity.

P16.05 Implementation of a novel ex-vivo brain slice model to study human glioblastoma and glioma-associated microglia

BACKGROUND Glioblastoma multiforme is a highly malignant brain tumor with a devastating prognosis. Resection followed by radio-chemotherapy leads to an overall survival of only 15 months. Up to 40% of the tumor mass consist of tumor-associated microglia and macrophages (TAMs). These cells were shown to promote tumor growth and invasiveness in many murine glioma models. The interaction between TAMs and tumor cells is crucial for tumor progression and includes several known pathways. Still, murine glioma models only partially mirror the human tumor microenvironment. Several known genes, which are highly upregulated in human glioma and TAMs are only expressed in human tissue and not in mice. To further investigate some of these genes, we aimed at establishing a humanized ex-vivo brain slice model, in which human TAMs and human glioma cells can be studied in a standardized manner. MATERIAL AND METHODS We used 250 micrometer thick murine brain slices, which were depleted of intrinsic microglia by applying clodoronated liposomes. Next, we inoculated human glioma cells (originating from the cell lines mCherryU87, mCherryU251MG, mCherryLN229 and several patient derived cells lines) with or without human microglia derived from induced pluripotent stem cells (iPSCs). Slices were cultivated for 7 to 14 days. Next, we performed a detailed analysis of microglia morphology (sphericity, cell body volume, process length and branching pattern) and tumor volume. RESULTS Clodronation efficacy was high, depending on duration of treatment and length of cultivation. iPSCs and tumor cells integrated into the slice very well. The presence of tumor cells led to an increased sphericity of iPSC-dervied microglia and to an increased cell body volume. Branching pattern and process length did not differ between both conditions. Tumor volume was significantly larger when iPSC-derived microglia were present. This was found in various glioma cells lines and also in patient derived cells. CONCLUSION The newly established humanized ex-vivo brain slice system was shown to be feasible. The method successfully allows to study the interaction between human TAMs and tumor cells. Microglia foster tumor growth not only in murine glioma models, but also in a human paradigm. The humanized ex-vivo brain slice model therefore is the optimal basis to study the role human-specific genes in TAM-glioma interaction.

Microglia-Derived Small Extracellular Vesicles Reduce Glioma Growth by Modifying Tumor Cell Metabolism and Enhancing Glutamate Clearance through miR-124

Brain homeostasis needs continuous exchange of intercellular information among neurons, glial cells, and immune cells, namely microglial cells. Extracellular vesicles (EVs) are active players of this process. All the cells of the body, including the brain, release at least two subtypes of EVs, the medium/large EVs (m/lEVs) and small EVs (sEVs). sEVs released by microglia play an important role in brain patrolling in physio-pathological processes. One of the most common and malignant forms of brain cancer is glioblastoma. Altered intercellular communications constitute a base for the onset and the development of the disease. In this work, we used microglia-derived sEVs to assay their effects in vitro on murine glioma cells and in vivo in a glioma model on C57BL6/N mice. Our findings indicated that sEVs carry messages to cancer cells that modify glioma cell metabolism, reducing lactate, nitric oxide (NO), and glutamate (Glu) release. sEVs affect Glu homeostasis, increasing the expression of Glu transporter Glt-1 on astrocytes. We demonstrated that these effects are mediated by miR-124 contained in microglia-released sEVs. The in vivo benefit of microglia-derived sEVs results in a significantly reduced tumor mass and an increased survival of glioma-bearing mice, depending on miR-124.

CD137 and PD-L1 targeting with immunovirotherapy induces a potent and durable antitumor immune response in glioblastoma models

BackgroundGlioblastoma (GBM) is a devastating primary brain tumor with a highly immunosuppressive tumor microenvironment, and treatment with oncolytic viruses (OVs) has emerged as a promising strategy for these tumors. Our group constructed a new OV named Delta-24-ACT, which was based on the Delta-24-RGD platform armed with 4-1BB ligand (4-1BBL). In this study, we evaluated the antitumor effect of Delta-24-ACT alone or in combination with an immune checkpoint inhibitor (ICI) in preclinical models of glioma.MethodsThe in vitro effect of Delta-24-ACT was characterized through analyses of its infectivity, replication and cytotoxicity by flow cytometry, immunofluorescence (IF) and MTS assays, respectively. The antitumor effect and therapeutic mechanism were evaluated in vivo using several immunocompetent murine glioma models. The tumor microenvironment was studied by flow cytometry, immunohistochemistry and IF.ResultsDelta-24-ACT was able to infect and exert a cytotoxic effect on murine and human glioma cell lines. Moreover, Delta-24-ACT expressed functional 4-1BBL that was able to costimulate T lymphocytes in vitro and in vivo. Delta-24-ACT elicited a more potent antitumor effect in GBM murine models than Delta-24-RGD, as demonstrated by significant increases in median survival and the percentage of long-term survivors. Furthermore, Delta-24-ACT modulated the tumor microenvironment, which led to lymphocyte infiltration and alteration of their immune phenotype, as characterized by increases in the expression of Programmed Death 1 (PD-1) on T cells and Programmed Death-ligand 1 (PD-L1) on different myeloid cell populations. Because Delta-24-ACT did not induce an immune memory response in long-term survivors, as indicated by rechallenge experiments, we combined Delta-24-ACT with an anti-PD-L1 antibody. In GL261 tumor-bearing mice, this combination showed superior efficacy compared with either monotherapy. Specifically, this combination not only increased the median survival but also generated immune memory, which allowed long-term survival and thus tumor rejection on rechallenge.ConclusionsIn summary, our data demonstrated the efficacy of Delta-24-ACT combined with a PD-L1 inhibitor in murine glioma models. Moreover, the data underscore the potential to combine local immunovirotherapy with ICIs as an effective therapy for poorly infiltrated tumors.

Contemporary Mouse Models in Glioma Research

Despite advances in understanding of the molecular pathogenesis of glioma, outcomes remain dismal. Developing successful treatments for glioma requires faithful in vivo disease modeling and rigorous preclinical testing. Murine models, including xenograft, syngeneic, and genetically engineered models, are used to study glioma-genesis, identify methods of tumor progression, and test novel treatment strategies. Since the discovery of highly recurrent isocitrate dehydrogenase (IDH) mutations in lower-grade gliomas, there is increasing emphasis on effective modeling of IDH mutant brain tumors. Improvements in preclinical models that capture the phenotypic and molecular heterogeneity of gliomas are critical for the development of effective new therapies. Herein, we explore the current status, advancements, and challenges with contemporary murine glioma models.

Contemporary Mouse Models in Glioma Research

Despite advances in understanding of the molecular pathogenesis of glioma, outcomes remain dismal. Developing successful treatments for glioma requires faithful in vivo disease modeling and rigorous preclinical testing. Murine models, including xenograft, syngeneic, and genetically engineered models, are used to study gliomagenesis, identify methods of tumor progression, and test novel treatment strategies. Since the discovery of highly recurrent isocitrate dehydrogenase (IDH) mutations in lower-grade gliomas, there is increasing emphasis on effective modeling of IDH mutant brain tumors. Improvements in preclinical models that capture the phenotypic and molecular heterogeneity of gliomas are critical for the development of effective new therapies. Herein, we explore the current status, advancements, and challenges with contemporary murine glioma models.

NFB-01. FUNCTIONAL CHARACTERIZATION OF ATRX LOSS IN NF1-ASSOCIATED GLIOMA AND MPNST

To identify the biologic relevance of ATRX loss in NF1-associated gliomagenesis, we studied the effects of Atrx loss using four previously characterized Nf1+/-Trp53+/- murine glioma lines. Lines 130G#3 and 158D#8 (corresponding to grade IV and III gliomas, respectively) displayed preserved ATRX protein expression compared to NIH-3T3 cells. We studied the effects of Atrx knockdown in these two lines in the presence and absence of the TERT inhibitor, BIRBR1532. Using a telomere-specific FISH assay, we identified increased signal intensity after Atrx knockdown, only in the presence of the TERT inhibitor. These features are reminiscent of ALT, although there were no significant alterations in cell growth. Next, we studied the effect of ATRX loss in MPNST lines ST88-14, NF90-8, STS-26T. These cell lines all expressed ATRX and DAXX. However, STS-26T contained a TERT promoter mutation and ST88-14 had a known SNP in the TERT promoter, while NF90-8 had no alterations. ATRX siRNA knockdown showed no significant effects in cell proliferation or apoptosis. However, ATRX knockdown resulted in rare ultra-bright foci, indicative of ALT. Next, we studied the in vitro effect of the ATR inhibitor VE-821 in MPNST cell lines. Only NF90-8 (lacking TERT alterations) demonstrated a decrease in growth after ATRX knockdown and VE-821 treatment. However, ATRX knockdown alone did not affect sensitivity to carboplatin. Our findings further support a role for ATRX loss with subsequent ALT activation in a biologic subset of NF1-associated malignancies, thereby opening an opportunity for therapeutic targeting of these aggressive tumors using specific classes of drugs.

IMMU-39. TIM-3 APTAMER IN COMBINATION WITH RADIOTHERAPY RESULTS IN ENHANCED SURVIVAL IN DIPG MODELS

Pediatric diffuse midline gliomas-H3-K27M-mutant are aggressive brain tumours that arise during childhood. Despite new advances in genomic knowledge and the significant number of clinical trial testing new targeted therapies, patient outcome is still insufficient. Cancer immunotherapy is opening new therapeutic options representing a hope for this orphan disease. Aptamers are single-stranded nucleic acid ligands design to achieve a remarkable affinity and specificity to their targets, comparable to antibodies. TIM-3, is a potential immune checkpoint target, typically involved in T-cell exhaustion. Recent studies showed that TIM-3 is also expressed in tumour and glial cells and it plays an important role in brain tumour responses mediated by myeloid cells. In this work, we examined the anti-tumour effect of an aptamer against TIM-3 alone or in combination with radiotherapy. Of importance, we tested TIM-3 aptamer in a murine glioma and DIPG model, where we not observed any toxicity. TIM-3 administration increased overall survival but was unable to control the disease. Of importance, TIM-3 combination with radiotherapy improved the overall survival of treated mice when compared with single treatments leading to 50% of long-term survivors. TIM-3 aptamer administration increase T-infiltration in the tumour site compared to non-treated or library control. Mechanistic studies performed on day 16 showed an increase in CD8 effector cells, a decrease in T-regulators Foxp3+ cells and an increase in IFN-gamma expression suggesting the triggering of an antitumor-immune response. Rechallenge experiments demonstrated immune memory in the long-term responders that led to reject tumour re-implantation, confirming that TIM-3 aptamer treatment in combination with RT elicits specific antitumor immunity in mouse glioma models. These results suggest that immuno-therapies approaches in combination with radiotherapy would be worth exploring in the treatment of deadly DMG-H3K27-Mutant tumours.