Professional phagocytic cells such as macrophages are a central part of innate immune defence. They ingest microorganisms into membrane-bound compartments (phagosomes), which acidify and eventually fuse with lysosomes, exposing their contents to a microbicidal environment. Gram-positive Rhodococcus equi can cause pneumonia in young foals and in immunocompromised humans. The possession of a virulence plasmid allows them to subvert host defence mechanisms and to multiply in macrophages. Here, we show that the plasmid-encoded and secreted virulenceassociated protein A (VapA) participates in exclusion of the proton-pumping vacuolar-ATPase complex from phagosomes and causes membrane permeabilisation, thus contributing to a pH-neutral phagosome lumen. Using fluorescence and electron microscopy, we show that VapA is also transferred from phagosomes to lysosomes where it permeabilises the limiting membranes for small ions such as protons. This permeabilisation process is different from that of known membrane pore formers as revealed by experiments with artificial lipid bilayers. We demonstrate that, at 24 hr of infection, virulent R. equi is contained in a vacuole, which is enriched in lysosome material, yet possesses a pH of 7.2 whereas phagosomes containing a vapA deletion mutant have a pH of 5.8 and those with virulence plasmid-less sister strains have a pH of 5.2. Experimentally neutralising the macrophage endocytic system allows avirulent R. equi to multiply. This observation is mirrored in the fact that virulent and avirulent R. equi multiply well in extracts of purified lysosomes at pH 7.2 but not at pH 5.1. Together these data indicate that the major function of VapA is to generate a pH-neutral and hence growth-promoting intracellular niche. VapA represents a new type of Gram-positive virulence factor by trafficking from one subcellular compartment to another, affecting membrane permeability, excluding proton-pumping ATPase, and consequently disarming host defences.

Central nervous system germinomas are characterized by a massive immune cell infiltrate. We systematically characterized these immune cells in 28 germinomas by immunophenotyping and image analysis. mRNA expression was analyzed by Nanostring technology and in situ RNA hybridization. Tumor infiltrating lymphocytes (TILs) were composed of 61.8% ± 3.1% (mean ± SE) CD3-positive T cells, including 45.2% ± 3.5% of CD4-positive T-helper cells, 23.4% ± 1.5% of CD8-positive cytotoxic T cells, 5.5% ± 0.9% of FoxP3-positive regulatory T cells, and 11.9% ±1.3% PD-1-positive TILs. B cells accounted for 35.8% ± 2.9% of TILs and plasma cells for 9.3% ± 1.6%. Tumor-associated macrophages consisted of clusters of activated PD-L1-positive macrophages and interspersed anti-inflammatory macrophages expressing CD163. Germinoma cells did not express PD-L1. Expression of genes encoding immune cell markers and cytokines was high and comparable to mRNA levels in lymph node tissue. IFNG and IL10 mRNA was detected in subfractions of TILs and in PD-L1-positive macrophages. Taken together, the strong immune reaction observed in germinomas involves inflammatory as well as various suppressive mechanisms. Expression of PD-1 and PD-L1 and infiltration of cytotoxic T cells are biomarkers predictive of response to anti-PD-1/PD-L1 therapies, constituting a rationale for possible novel treatment approaches.

No abstract

Approximately 20-30% of diffusely infiltrative gliomas in adults contain a point mutation in isocitrate dehydrogenase 1 (IDH1mut), which leads to production of D-2-hydroxyglutarate (D2HG). Production of D2HG is so efficient that D2HG concentrations often reach 30 mM within IDH1mut gliomas. Yet, while up to 100 µM D2HG can be detected in the circulating cerebrospinal fluid of IDH1mut glioma patients, the amount of D2HG to which nonneoplastic cells are exposed within and surrounding an IDH1mut glioma is unknown and difficult to measure directly in patients or animal models. Thus, to date, the majority of research on D2HG has focused on its effects inside tumor cells, with very little known about its role in the tumor microenvironment. Liquid chromatography-mass spectrometry (LC-MS) analysis indicates that patient-derived IDH1mut glioma cells release approximately 3.7-6.7 pg D2HG per cell per week in culture. Extrapolating this to an averagesized tumor (30 ml glioma volume and 1x108 cells/ml tumor, the latter based on a previously published estimate of tumor cell density), the rate of D2HG release by an IDH1mut glioma (SA) is calculated to be 3.2-5.8x10-12 moles/ml/sec. Mathematical models of D2HG production and diffusion around an IDH1mut glioma were independently generated based on intracerebral fluid dynamics and previously reported D2HG concentrations in the cerebrospinal fluid. These models estimate an SA of 2.9-17.0x10-12 moles/ml/sec, in good agreement with the LC-MS data generated from in vitro models. In even the most conservative of these diffusion models, the extracellular concentration of D2HG exceeds 3 mM within a 2 cm radius from the center of an IDH1mut glioma. This has profound implications for understanding how D2HG may alter the microenvironmental landscape in IDH1mut gliomas.