Stroke induces disease-specific myeloid cells in the brain parenchyma and pia.

Inflammation triggers secondary brain damage after stroke. The meninges and other CNS border compartments serve as invasion sites for leukocyte influx into the brain thus promoting tissue damage after stroke. However, the post-ischemic immune response of border compartments compared to brain parenchyma remains poorly characterized. Here, we deeply characterize tissue-resident leukocytes in meninges and brain parenchyma and discover that leukocytes respond differently to stroke depending on their site of residence. We thereby discover a unique phenotype of myeloid cells exclusive to the brain after stroke. These stroke-associated myeloid cells partially resemble neurodegenerative disease-associated microglia. They are mainly of resident microglial origin, partially conserved in humans and exhibit a lipid-phagocytosing phenotype. Blocking markers specific for these cells partially ameliorates stroke outcome thus providing a potential therapeutic target. The injury-response of myeloid cells in the CNS is thus compartmentalized, adjusted to the type of injury and may represent a therapeutic target.

The Involvement of Pial Microvessels in Leukocyte Invasion after Mild Traumatic Brain Injury.

The pathophysiological mechanisms underlying mild traumatic brain injury (mTBI) are not well understood, but likely involve neuroinflammation. Here the controlled cortical impact model of mTBI in rats was used to test this hypothesis. Mild TBI caused a rapid (within 6 h post-mTBI) upregulation of synthesis of TNF-α and IL-1ß in the cerebral cortex and hippocampus, followed by an increase in production of neutrophil (CXCL1-3) and monocyte (CCL2) chemoattractants. While astrocytes were not a significant source of CXC chemokines, they highly expressed CCL2. An increase in production of CXC chemokines coincided with the influx of neutrophils into the injured brain. At 6 h post-mTBI, we observed a robust influx of CCL2-expressing neutrophils across pial microvessels into the subarachnoid space (SAS) near the injury site. Mild TBI was not accompanied by any significant influx of neutrophils into the brain parenchyma until 24 h after injury. This was associated with an early induction of expression of intercellular adhesion molecule 1 on the endothelium of the ipsilateral pial, but not intraparenchymal, microvessels. At 6 h post-mTBI, we also observed a robust influx of neutrophils into the ipsilateral cistern of velum interpositum (CVI), a slit-shaped cerebrospinal fluid space located above the 3rd ventricle with highly vascularized pia mater. From SAS and CVI, neutrophils appeared to move along the perivascular spaces to enter the brain parenchyma. The monocyte influx was not observed until 24 h post-mTBI, and these inflammatory cells predominantly entered the ipsilateral SAS and CVI, with a limited invasion of brain parenchyma. These observations indicate that the endothelium of pial microvessels responds to injury differently than that of intraparenchymal microvessels, which may be associated with the lack of astrocytic ensheathment of cerebrovascular endothelium in pial microvessels. These findings also suggest that neuroinflammation represents the potential therapeutic target in mTBI.

Inflammation-inducible type 2 deiodinase expression in the leptomeninges, choroid plexus, and at brain blood vessels in male rodents.

Thyroid hormone regulates immune functions and has antiinflammatory effects. In promoter assays, the thyroid hormone-activating enzyme, type 2 deiodinase (D2), is highly inducible by the inflammatory transcription factor nuclear factor-κ B (NF-κB), but it is unknown whether D2 is induced in a similar fashion in vivo during inflammation. We first reexamined the effect of bacterial lipopolysaccharide (LPS) on D2 expression and NF-κB activation in the rat and mouse brain using in situ hybridization. In rats, LPS induced very robust D2 expression in normally non-D2-expressing cells in the leptomeninges, adjacent brain blood vessels, and the choroid plexus. These cells were vimentin-positive fibroblasts and expressed the NF-κB activation marker, inhibitor κ B-α mRNA, at 2 hours after injection, before the increase in D2 mRNA. In mice, LPS induced intense D2 expression in the choroid plexus but not in leptomeninges, with an early expression peak at 2 hours. Moderate D2 expression along numerous brain blood vessels appeared later. D2 and NF-κB activation was induced in tanycytes in both species but with a different time course. Enzymatic assays from leptomeningeal and choroid plexus samples revealed exceptionally high D2 activity in LPS-treated rats and Syrian hamsters and moderate but significant increases in mice. These data demonstrate the cell type-specific, highly inducible nature of D2 expression by inflammation, and NF-κB as a possible initiating factor, but also warrant attention for species differences. The results suggest that D2-mediated T3 production by fibroblasts regulate local inflammatory actions in the leptomeninges, choroid plexus and brain blood vessels, and perhaps also in other organs.

The microglial/macrophagic response at the tumour-brain border of invasive meningiomas.

AIMS: Little is known about the immune response of the brain to invasive meningiomas. The present study was based upon the hypothesis that the microglial/macrophagic response towards brain-invasive meningiomas is dependent on the intactness of the pial-glial basement membrane. METHODS: We immunostained sections from 40 brain-invasive meningiomas that were graded according to World Health Organization (WHO) 2007 criteria. Thirty-three tumours were histologically WHO grade II (18, 'otherwise benign', and 15, 'otherwise atypical'), and seven, grade III. Microglial/macrophagic cells were labelled with antibodies directed against major histocompatibility complex class II, CD68, CD14 and CD163. Anti-collagen IV was used to visualize basement membranes. RESULTS: Twenty-five per cent (10/40) meningiomas (1/18 WHO grade II 'otherwise benign', 3/15 grade II 'otherwise atypical' and 6/7 WHO grade III) contained microglial/macrophagic cells at the tumour-brain border. The presence of these cells correlated with the absence of the pial-glial basement membrane (BM) and with WHO grade III. The monocytic response was of two kinds: one consisted of a dense layer of mononuclear cells at the tumour-brain border in nine cases, the other of an elevated number of microglial cells expressing CD14 or CD163 (two cases). CONCLUSIONS: The immune response at the tumour-brain interface correlates with the absence of the pial-glial BM and with malignancy grade. It remains to be established whether the mononuclear cells at the tumour-brain border are native microglia or blood-derived macrophages.

Effects of moderate hypothermia on IL-1 beta-induced leukocyte rolling and adhesion in pial microcirculation of mice and on proinflammatory gene expression in human cerebral endothelial cells.

SUMMARY: Although the neuroprotective effects of hypothermia have been known for a long time, the molecular correlates of this neuroprotection are poorly understood. In this study, the authors investigated how hypothermia affects inflammatory responses in the brain elicited by systemic injection of IL-1 beta. Leukocyte rolling and adhesion were quantified in pial venules (20 to 50 microm) of C57/Bl6 mice 4 hours after intraperitoneal injection of IL-1 beta (5 microg/kg) using an open cranial window and intravital microscopy. Animals were subjected to moderate hypothermia (32 degrees C) or normothermia (37 degrees C) for 1 or 4 hours after IL-1 beta injection. Significant increases in leukocyte rolling and adhesion were observed in IL-1 beta-injected animals as compared with sham controls. Whereas 1-hour hypothermia did not affect IL-1 beta-induced leukocyte rolling and adhesion, 4-hour hypothermia caused a reduction in both rolling and adhesion. Molecular mechanisms of hypothermic effects were investigated in cultured human cerebral endothelial cells exposed to IL-1 beta (50 U/mL) for 4 hours at 37 degrees C or 32 degrees C followed by 18 hours at 37 degrees C. Human cerebral endothelial cells exposed to IL-1 beta at 32 degrees C showed attenuated NF-kappa B activation determined by the Luciferase yellow reporter gene assay and reduced expression of IL-8 and IL-1 beta measured by reverse transcriptase-polymerase chain reaction and enzyme-linked immunosorbent assay. Intracellular adhesion molecule-1 was induced to similar levels (threefold over control) at both temperatures. The expression of CD18 on neutrophils in vitro was not affected by either IL-1 beta or hypothermia. These findings suggest that mechanisms by which hypothermia reduces leukocyte rolling and adhesion include suppression of inflammatory gene transcription in brain endothelial cells.

Immune surveillance of mouse brain perivascular spaces by blood-borne macrophages.

Virchow-Robin's perivascular spaces lie between the basement membrane around pericytes and the basement membrane at the surface of the glia limitans of the brain vessels. They are directly connected to the subpial space and harbour a population of cells distinct from pericytes, perivascular microglia and other cells within perivascular spaces (e.g. T cells and mast cells) in their ability to quickly phagocytose particles from the cerebrospinal fluid (CSF). Morphology, function, and cell surface proteins of these perivascular cells suggest an origin from the monocyte/macrophage lineage. It is currently unclear to what extent these brain perivascular cells represent a resident population of histiocytes or undergo continuous supplementation from blood monocytes. Using transplants of green-fluorescent-protein (GFP)-transfected bone marrow cells, we therefore investigated the replacement of perivascular cells by blood-borne macrophages in adult mice. GFP-positive cells in the perivascular spaces were found as early as 2 weeks post transplantation. The substitution of host perivascular cells by donor-derived macrophages was then evaluated using immunocytochemistry and intraventricular injection of hydrophilic rhodamine-fluorescent tracers. Such tracers diffuse along perivascular spaces and are subsequently phagocytosed by perivascular cells leading to stable phagocytosis-dependent labelling. Thus, the population of newly immigrated macrophages could be related to the total number of perivascular macrophages. This approach revealed a continuous increase of donor-derived perivascular cells. At 14 weeks post transplantation, all perivascular cells were donor-derived. These data show that brain perivascular cells are a population of migratory macrophages and not resident histiocytes.

Identification of cell types producing RANTES, MIP-1 alpha and MIP-1 beta in rat experimental autoimmune encephalomyelitis by in situ hybridization.

The chemokines RANTES, macrophage inflammatory protein (MIP)-1 alpha and MIP-1 beta are members of the beta-family of chemokines and potent chemoattractants for lymphocytes and monocytes. To investigate the factors which regulate lymphocyte traffic in experimental autoimmune encephalomyelitis (EAE), we studied, by in situ hybridization analysis, the kinetics of mRNA expression and the potent cellular sources of RANTES, MIP-1 alpha and MIP-1 beta in the central nervous system (CNS) during the course of EAE. RANTES-positive cells appeared in the subarachnoid space and infiltrated the subpial region at around day 10, increased to a peak at days 12-13 and then decreased following the resolution of the acute phase of EAE, though elevated RANTES message expressions still remained on chronic subclinical stage. Most of RANTES positive cells were identified as T-lymphocytes located mainly around blood vessels, by combined studies of in situ hybridization and immunohistochemistry. The remainder of the RANTES-positive cells were astrocytes and macrophages/microglia. MIP-1 alpha and MIP-1 beta mRNA-positive cells appeared around day 10, increased further on days 12-13 and then gradually decreased. Most of the MIP-1 alpha- and MIP-1 beta-positive mononuclear cells were located around blood vessels. The kinetics of RANTES, MIP-1 alpha and MIP-1 beta expression paralleled those of the recruitment of infiltrating inflammatory cells and disease severity. Our observations support the possibility that chemokine production by T-cells, macrophages and astrocytes lead to the infiltration of inflammatory cells into the CNS parenchyma during the acute phase of EAE.

Substance P-, calcitonin gene-related peptide, growth-associated protein-43, and neurotrophin receptor-like immunoreactivity associated with unmyelinated axons in feline ventral roots and pia mater.

The spinal pia mater receives a rich innervation of small sensory axons via the ventral roots. Other sensory axons enter the ventral roots but end blindly or turn abruptly in hairpin loop-like formations and continue in a distal direction. In the present study, the content of substance P (SP)-, calcitonin gene-related peptide (CGRP)-, growth-associated protein (GAP-43)-, and low-affinity neurotrophin receptor protein (p75NGFr)-like immunoreactivity (-LI) associated with these different types of sensory axons was assessed with light and electron microscopic immunohistochemical techniques. In addition, the binding of antibodies against synthetic peptides representing unique sequences of residues in the products of the trk and trkB protooncogenes was analyzed. These genes encode membrane spanning proteins, which have been shown to constitute specific high affinity binding sites for several members of the nerve growth factor family of neurotrophic factors. The results of the present study imply that the ventral root afferents comprise several different types of sensory axons, which all contain SP-, CGRP-, GAP-43-, and p75NGFr-like immunoreactivities. In addition, at least some of the presumed sensory fiber bundles in ventral roots and the pia mater were immunoreactive for the trkB gene product. Moreover, leptomeningeal cells and nonneuronal cells of the ventral roots were shown to bind antibodies to both the trk and trkB gene products. The ventral root afferents seem to share their immunohistochemical pattern with pain-transducing axons at some other locations, such as the tooth pulp. The contents of SP- and CGRP-LI in sensory axons that reach the central nervous system (CNS) through the ventral root indicate that ventral root afferents may be involved in sensory mechanisms, such as the ventral root pain reaction, as well as in the control of the pial blood vessels. The demonstration of GAP-43 and neurotrophin receptor-immunoreactivities associated with unmyelinated fibers in ventral roots and the pia mater is discussed in relation to previous reports on postnatal plasticity in these axonal populations.

Surgical treatment of "hindbrain related" syringomyelia: new data for pathogenesis.

52 patients with "hindbrain related" syringomyelia underwent surgical treatment. All patients underwent primary reconstructive surgery at the craniovertebral junction. Terminal ventriculostomy was performed as the secondary operation in 2 cases. The surgical treatment arrested progression of signs in 33 (63.5%), stabilized disease in 9 (17%) cases. Postoperative deterioration occurred in 8 (15%) cases. Mortality was 4% (2 patients). Percutaneous or intra-operative injection of myodil and gas into the syrinx, as well as CT, revealed the existence of communication with the 4th ventricle in 14 patients. Investigation of cerebrospinal and syrinx fluid revealed increased level of IgG, IgM or IgA in the syrinx fluid in 16 out of 22 patients. Immunohistological examination of pia mater revealed specific staining for IgG. Thus, syrinx formation may be due to synergic action of hydrodynamic and immunopathological mechanisms. Results indicate that early surgical treatment is preferable to patients with hindbrain anomalies and hydromyelia. We consider primary reconstructive operation at the posterior fossa as the preferred surgical management of "hindbrain related" syringomyelia.

Antibody against the C-terminal portion of dystrophin crossreacts with the 400 kDa protein in the pia mater of dystrophin-deficient mdx mouse brain.

The mdx mouse is an animal model for X-linked Duchenne muscular dystrophy. A polyclonal antibody against a synthetic peptide IV equivalent to the C-terminal portion (amino acids 3495-3544) of dystrophin crossreacted with a 400 kDa protein in the brain and the spinal cord of mdx mouse, as well as in the control B10 mouse. However, the protein did not crossreact with the polyclonal antibody raised against the N-terminal portion of dystrophin peptide I (amino acids 215-264). Immunofluorescent micrography revealed that the outside of the small arteries and the pia mater of the brain strongly reacted with the anti-peptide IV antibody. These results strongly suggest the presence of a crossreactive protein other than dystrophin, possibly a dystrophin-related autosomal gene product, in the pia mater.

Scanning electron microscopy of the subarachnoid space in the dog: evidence for a non-hematogenous origin of subarachnoid macrophages.

Injection of viable BCG into the subarachnoid space of immunized and non-immunized dogs produced a 10-fold increase in the populations of pial free cells. In immunized animals injected three days previously with BCG, stereoscopic SEM revealed that many pial cells had rounded up and were protruding into the subarachnoid space. With continued rounding these cells took on amoeboid characteristics, with shapes that suggested a capacity for cell movement. Internally, these pial cells possessed an increased volume of perinuclear cytoplasm and organelles. Reactive pial cells could be distinguished from macrophages of presumed hematogenous origin on the basis of their surface morphology. These findings suggested that pial cells had the ability to alter their normal structural and behavioral characteristics and to become macrophage-like under these conditions of secondary challenge by BCG.