Myeloid cells are an essential part of the glioblastoma microenvironment. However, in brain tumors the function of these immune cells is not sufficiently clarified. In our study, we investigated differential pro-angiogenic activities of resident microglia and peripheral macrophages and their impact on glioma vascularization and progression. Our data demonstrate stable accumulation of microglia/macrophages during tumor growth. These cells often interact with tumor blood vessels correlating with vascular remodeling. Here, we identified resident microglia as well as peripheral macrophages as part of the perivascular niche, primarily contacting endothelial cells. We found overexpression of a variety of pro-angiogenic molecules within freshly isolated microglia/macrophages from glioma. CXCL2, until now a poorly described chemokine, was strongly up-regulated and showed better angiogenic activity than VEGF in vitro. Blocking the CXCL2-CXCR2 signaling pathway resulted in considerably diminished glioma sizes. Additionally, the importance of microglia/macrophages in tumor angiogenesis was confirmed by depletion of these cells in vivo. Vessel density decreased by 50% leading to significantly smaller tumor volumes. Remarkably, selective reduction of resident microglia affected tumoral vessel count comparable to ablation of the whole myeloid cell fraction. These results provide evidence that resident microglia are the crucial modulatory cell population playing a central role in regulation of vascular homeostasis and angiogenesis in brain tumors. Thus, resident microglia represent an alternative source of pro-angiogenic growth factors and cytokines.

Gliomas consist of multiple cell types, including an abundant number of microglia and macrophages, whereby their impact on tumor progression is controversially discussed. To understand their unique functions and consequently manipulate either microglia or macrophages in therapeutic approaches, it is essential to discriminate between both cell populations. Because of the lack of specific markers, generally total body irradiated chimeras with labeled bone marrow cells were used to identify infiltrated cells within the brain. However, total body irradiation (TBI) affects the blood-brain barrier integrity, which in turn potentially facilitates immune cell infiltration. In this study, changes on the blood-brain barrier were avoided using headprotected irradiation (HPI). Head protection and total body irradiated chimeras exhibited similar reconstitution levels of the myeloid cell lineage in the blood, enabling the comparable analyses of brain infiltrates. We demonstrate that the HPI model impeded a massive unspecific influx of donor-derived myeloid cells into naive as well as tumor-bearing brains. Moreover, experimental artifacts such as an enlarged distribution of infiltrated cells and fourfold increased tumor volumes are prevented in head-protected chimeras. In addition, our data evidenced for the first time that microglia are able to up-regulate CD45 and represent an inherent part of the CD45 high population in the tumor context. All in all, HPI allowed for the unequivocal distinction between microglia and macrophages without alterations of tumor biology and consequently permits a detailed and realistic description of the myeloid cell composition in gliomas.

Inflammatory changes have been postulated to contribute to secondary brain injury after aneurysmal subarachnoid hemorrhage (SAH). In human specimens after SAH as well as in experimental SAH using mice, we show an intracerebral accumulation of inflammatory cells between days 4 and 28 after the bleeding. Using bone marrow chimeric mice allowing tracing of all peripherally derived immune cells, we confirm a truly CNS-intrinsic, microglial origin of these immune cells, exhibiting an inflammatory state, and rule out invasion of myeloid cells from the periphery into the brain. Furthermore, we detect secondary neuro-axonal injury throughout the time course of SAH. Since neuronal cell death and microglia accumulation follow a similar time course, we addressed whether the occurrence of activated microglia and neuro-axonal injury upon SAH are causally linked by depleting microglia in vivo. Given that the amount of neuronal cell death was significantly reduced after microglia depletion, we conclude that microglia accumulation inflicts secondary brain injury after SAH.

Group B Streptococcus (GBS) cell walls potently activate phagocytes by a largely TLR2-independent mechanism. In contrast, the cell wall component lipoteichoic acid (LTA) from diverse Gram-positive bacterial species has been shown to engage TLR2. In this study we examined the role of LTA from GBS in phagocyte activation and the requirements for TLR-LTA interaction. Using cells from knockout mice and genetic complementation in epithelial cells we found that highly pure LTA from both GBS and Staphylococcus aureus interact with TLR2 and TLR6, but not TLR1, in contrast to previous reports. Furthermore, NF-κB activation by LTA required the integrity of two putative PI3K binding domains within TLR2 and was inhibited by wortmannin, indicating an essential role for PI3K in cellular activation by LTA. However, LTA from GBS proved to be a relatively weak stimulus of phagocytes containing ∼20% of the activity observed with LTA from Staphylococcus aureus. Structural analysis by nuclear magnetic resonance spectrometry revealed important differences between LTA from GBS and S. aureus, specifically differences in glycosyl linkage, in the glycolipid anchor and a lack of N-acetylglucosamine substituents of the glycerophosphate backbone. Furthermore, GBS expressing LTA devoid of d-alanine residues, that are essential within immune activation by LTA, exhibited similar inflammatory potency as GBS with alanylated LTA. In conclusion, LTA from GBS is a TLR2/TLR6 ligand that might contribute to secreted GBS activity, but does not contribute significantly to GBS cell wall mediated macrophage activation.