BACKGROUND: Malignant gliomas belong to the most aggressive neoplasms in humans. The interactions and intercellular communication between malignant and non-malignant cells in the tumor microenvironment are tumor-promoting and critically to improve the understanding of the disease. N-myc downstream regulated gene 1 (NDRG1) is a stress inducible gene and key determinant of resistance towards alkylating chemotherapy in glioblastoma. NDRG1 promotes invasive growth and acts antiapoptotic in hepatocellular carcinoma. In human pancreatic cancer cells NDRG1 suppresses chemokine expression, decreases macrophage infiltration and microvascular density. Further understanding the influence of NDRG1 on the tumor microenvironment may lead to new therapeutic approaches in glioma. METHODS: To analyze the NDRG1 effects on the glioma microenvironment, we used human and murine NDRG1 knockdown (KD) cells. In orthotopic xenograft experiments control and KD microenvironments were compared for infiltrating immune cells dependent on NDRG1. Interaction and signaling mechanisms were evaluated in vitro using western blots, cytokine and migration assays and treatment with the NF-kB inhibitor JSH-23 as well as SGK1 inhibitor EMD638683 to assess the influence of NDRG1. Results: Orthotopic NDRG1 KD tumors were significantly larger compared to control tumors. Ex vivo flow cytometry analyses of the tumor microenvironment revealed a significant increase in peripheral macrophages, dendritic cells and monocytic myeloid derived suppressor cells in NDRG1 KD tumors. In vitro cytokine array analyses showed an increase of the chemokine CCL2 in NDRG1 KD cells compared to control cells. This was confirmed by ELISA. Macrophages showed an increased migration rate in the NDRG1 knockdown environment. To analyze molecular mechanisms CCL2 secretion was blocked with a NF-κB inhibitor (JSH-23) and stimulated with a SGK1 inhibitor (EMD638683). We found that NDRG1 induced NF-κB signaling attenuation is mediated through decreases in IKK α expression and IκBα phosphorylation. CONCLUSION: We identified CCL2 as a NDRG1-induced NF-κB target gene that triggers macrophage recruitment. Our findings support an immunomodulatory effect of NDRG1 on the tumor microenvironment and provide novel insights into the biological relevance of the tumor stroma and the function of NDRG1 with important implications for the development of cancer therapeutics.

Background: Dysregulation of p53 pathway is common in glioma. Frequent causes are p53 mutations or overexpression of the oncogene MDM2 in p53 wildtype tumors. Idasanutlin is a MDM2 inhibitor, which is currently under investigation for multiple myeloma and acute myeloid leukemia, but of principal interest also for other p53 mutated solid tumors. Therefore, preclinical data and molecular understanding about treatment effects and resistance mechanisms are important. Methods: Treatment effects of idasanutlin monotherapy and combination with radiotherapy were analyzed on cell viability, proliferation and clonogenicity in p53 wildtype cell lines (U87MG, A172) and glioma-initiating cells (GICs, S24, T1) at radiation doses of 2 and 4 Gray (photons). Toxicity assays were performed on freshly isolated murine astrocytes and cerebellum granule neurons. Idasanutlin resistant cells were generated by treatment of U87MG cells with increasing doses of idasanutlin for 3 months. Resistance mechanisms were analyzed by microarray analysis and confirmed via qRT-PCR, immuno blots and treatment with the nuclear factor (NF)kB inhibitor JSH-23 and transient knockdown of insulin-like growth factor binding protein (IGFBP)1 via siRNA. Results: Idasanutlin reduced cell viability and clonogenicity in p53 wildtype glioblastoma cell lines and GICs. Combined treatment with radiotherapy showed synergistic effects on clonogenicity and proliferation at clinically relevant doses without toxicity on normal brain cells. Long-term treatment with idasanutlin resulted in resistance against the drug in U87MG cells and led to a more clonogenic, invasive and proliferative phenotype of the cells. Microarray analysis revealed an activation of NFkB and ERK1/2 pathway and an upregulation of IGFBP1 expression in idasanutlin resistant cells. While radiotherapy did not overcome resistance conferred by idasanutlin, inhibition of NFkB pathway by JSH23 treatment and transient knockdown of IGFBP1 by siRNA partly restored sensitivity towards idasanutlin. Transient knockdown of IGFBP1 additionally reduced the invasiveness of idasanutlin resistant cells. Conclusions: Combination of idasanutlin treatment and radiotherapy showed synergistic effects on glioblastoma cells in a clinically relevant setting. However, long-term treatment with idasanutlin resulted in resistance against the drug mediated by activation of NFkB pathway and upregulation of IGFBP1. Targeting these pathways may provide strategies to overcome idasanutlin resistance.

INTRODUCTION:The treatment of glioblastoma multiforme (GBM) is a major challenge of neuro-oncology.Despite aggressive therapeutic strategies including surgery and radio-chemotherapy, median survival remains extremely low.This points to the urgent need for alternative treatment strategies such as immunotherapy.Immunological escape mechanisms, involving the Natural Killer Group 2, member D (NKG2D) receptor-ligand system, play a major role in tumor progression.Cell-bound NKG2D-ligands such as MHC class I related molecule A and B (MICA and MICB), and the UL-16 binding protein family (ULBP1-6) are recognized by the NKG2D-receptor (NKG2Dr) and trigger cytotoxic effector functions in NK-cells and T-cell subsets.By releasing soluble NKG2DL (sNKG2DL), which then bind to NKG2Dr, tumor cells inhibit the killing potential of the effector cells.Numerous studies documented the importance of NKG2D-system in vitro GBM-model systems.Here, we aimed to analyze NKG2D-system in GBM-patients ex vivo.MATERIALS AND METHODS: Until now, 37 GBM patients and 19 healthy controls (HCs) have been included in the study.A total of 24 GBM patients were on medication with dexamethasone prior to surgery.We analyzed serum levels of sNKG2DLs in HCs and GBM patients before and 3 months after surgery and radio-chemotherapy via Luminex-based multiplex assay.Absolute cell numbers of distinct immune cell subsets and the expression of NKG2Dr were analyzed by flow cytometry.NKG2DLexpression on GBM primary cell cultures was studied via flow cytometry and western-blotting.RESULTS: Our data show that in comparison to healthy controls GBM patients show reduced numbers of leukocytes and cytotoxic effector cells in peripheral blood, but elevated numbers of immunosuppressive cell subsets.Whereas NKG2Dr is not detectable in primary cell cultures, the tumor cells moderately express NKG2DL and release sNKG2DL into the culture medium.Serum levels of sNKG2DL are differentially modulated in GBM patients when comparing pre-to 3 months postop.This effect is independent of medication with dexamethasone.CONCLUSION: GBM patients seem to have impaired cytotoxic immune response compared to HCs.As NKG2DL are released by the tumor cells, this effect might also be generated via sNKG2DL.Nevertheless, serum levels of NKG2DL do not decline after treatment in most patients.A better understanding of the mechanisms regulating the NKG2D-system in GBM patients will be crucial for the development of new therapeutic strategies targeting the NKG2D system.

11587 Background: Malignant glioma belongs to the most aggressive neoplasms in humans. They are highly invasive and the cellular and genetic inter- and intratumor heterogeneity contributes to treatment resistance. The interactions and intercellular communications between malignant and non-malignant cells in the tumor microenvironment are deemed tumor-promoting and critically to improve the understanding of the disease. N-myc downstream regulated gene 1 (NDRG1) is a stress inducible gene and key determinant of resistance towards alkylating chemotherapy in glioblastoma. Methods: To analyze the NDRG1 effects on the brain tumor microenvironment, we used a human NDRG1 knockdown (KD) glioma model system. In orthotopic xenograft experiments control and KD microenvironments were compared for angiogenesis and infiltrating immune cells dependent on NDRG1. Clodronate liposomes were used to deplete macrophages. In vitro and in vivo angiogenesis assays were used to assess neovascularization. Results: Orthotopic tumors showed significant volume differences between control and NDRG1 KD implanted cells and demonstrated a markedly increased vessel density in the NDRG1 KD microenvironment. Ex vivo flow cytometry analyses of the tumor stromal cells revealed a significant increase in peripheral macrophages in NDRG1 KD microenvironment compared to the control microenvironment. The same was shown for dendritic cells as well as for monocytic myeloid derived suppressor cells (MDSCs). Depletion of macrophages however did not alter tumor growth of NDRG1 KD tumors. Cytokine array analysis on U87MG NDRG1 KD and control supernatants showed a marked increase in CCL2 secretion in the NDRG1 KD cells. Macrophages showed an increased migration rate towards NDRG1 knockdown environment in vitro. Conclusions: Glioma NDRG1 shapes the tumor microenvironment by regulating angiogenesis and influences macrophage recruitment in vivo and in vitro. CCL2 was identified as a relevant cytokine dependent of the NDRG1 status in this human glioma model. These data could help to better understand the relationship between NDRG1 and the tumor microenvironment.

Brain metastases (BM) are an increasing challenge. Insight in the pathology of the brain metastatic cascade, and in particular in the characteristics of the BM initiating cells can help to identify new treatment targets. PHK26 membrane dye was used in stably GFP expressing human breast cancer to differentiate slow from fast cycling cells by membrane signal intensity changes using multiphoton laser scanning microscopy, both in vitro and in vivo. The heterogeneous population of fast and slow cycling cells as well as resting cells was injected intracardially and followed with in vivo repetitive multiphoton laser microcopy via a chronic intracranial window. Here, slow cycling cells, representing only 16% of the entire cell population, were the only ones that mastered all steps of the brain metastatic cascade (0% marcometastasis formation after intravascular arrest of fast cycling cells vs. 6.15% marcometastasis formation after intravascular arrest of slow cycling cells; p<0.001), namely intravascular arrest, extravasation, perivascular survival, and marcometastasis outgrowth. These slow cycling cells showed a high overlap with established markers for tumor stem-like cells, like Oct4/Sox2, Notch and WNT, and also (but less) with low 26S proteasome activity. Illumina gene expression profiling of slow versus fast cycling JIMT1 breast cancer cells revealed up-regulation of N-Myc down regulated gene (NDGR1). Knock down of NDRG1 resulted in complete inhibition of BM formation by preventing successful colonization of the perivascular niche. In conclusion, slow cycling cells resemble the population of BM initiating cells. Increased NDRG1 expression is a characteristic of slow cycling cells, and is a pivotal molecular precondition for successful BM formation that might serve as a potential target for BM prevention or treatment.