In the absence of conventional lymphatics, drainage of interstitial fluid and solutes from the brain parenchyma to cervical lymph nodes is along basement membranes in the walls of cerebral capillaries and tunica media of arteries. Perivascular pathways are also involved in the entry of CSF into the brain by the convective influx/glymphatic system. The objective of this study is to differentiate the cerebral vascular basement membrane pathways by which fluid passes out of the brain from the pathway by which CSF enters the brain. Experiment 1: 0.5 µl of soluble biotinylated or fluorescent Aβ, or 1 µl 15 nm gold nanoparticles was injected into the mouse hippocampus and their distributions determined at 5 min by transmission electron microscopy. Aβ was distributed within the extracellular spaces of the hippocampus and within basement membranes of capillaries and tunica media of arteries. Nanoparticles did not enter capillary basement membranes from the extracellular spaces. Experiment 2: 2 µl of 15 nm nanoparticles were injected into mouse CSF. Within 5min, groups of nanoparticles were present in the pial-glial basement membrane on the outer aspect of cortical arteries between the investing layer of pia mater and the glia limitans. The results of this study and previous research suggest that cerebral vascular basement membranes form the pathways by which fluid passes into and out of the brain but that different basement membrane layers are involved. The significance of these findings for neuroimmunology, Alzheimer’s disease, drug delivery to the brain and the concept of the Virchow–Robin space are discussed.

Tracers injected into CSF pass into the brain alongside arteries and out again. This has been recently termed the “glymphatic system” that proposes tracers enter the brain along periarterial “spaces” and leave the brain along the walls of veins. The object of the present study is to test the hypothesis that: (1) tracers from the CSF enter the cerebral cortex along pial-glial basement membranes as there are no perivascular “spaces” around cortical arteries, (2) tracers leave the brain along smooth muscle cell basement membranes that form the Intramural Peri-Arterial Drainage (IPAD) pathways for the elimination of interstitial fluid and solutes from the brain. 2 μL of 100 μM soluble, fluorescent fixable amyloid β (Aβ) were injected into the CSF of the cisterna magna of 6–10 and 24–30 month-old male mice and their brains were examined 5 and 30 min later. At 5 min, immunocytochemistry and confocal microscopy revealed Aβ on the outer aspects of cortical arteries colocalized with α-2 laminin in the pial-glial basement membranes. At 30 min, Aβ was colocalised with collagen IV in smooth muscle cell basement membranes in the walls of cortical arteries corresponding to the IPAD pathways. No evidence for drainage along the walls of veins was found. Measurements of the depth of penetration of tracer were taken from 11 regions of the brain. Maximum depths of penetration of tracer into the brain were achieved in the pons and caudoputamen. Conclusions drawn from the present study are that tracers injected into the CSF enter and leave the brain along separate periarterial basement membrane pathways. The exit route is along IPAD pathways in which Aβ accumulates in cerebral amyloid angiopathy (CAA) in Alzheimer’s disease. Results from this study suggest that CSF may be a suitable route for delivery of therapies for neurological diseases, including CAA.Electronic supplementary materialThe online version of this article (10.1007/s00401-018-1862-7) contains supplementary material, which is available to authorized users.

CMG were employees of Invicro, LLC when the studies were performed. JMS is currently an employee of Biogen Inc. NJA, LK, and ROC are employees of the University of Southampton. MB is an employee of 3D Imaging, and AV was an employee of Biogen Inc. at the time that this work was performed.

The accumulation of proteinaceous deposits comprised largely of the α-synuclein protein is one of the main hallmarks of Parkinson's disease (PD) and related synucleinopathies. Their progressive development coincides with site-specific phosphorylation, oxidative stress and eventually, compromised neuronal function. However, modeling protein aggregate formation in animal or in vitro models has proven notably difficult. Here, we took advantage of a preclinical organotypic brain slice culture model to study α-synuclein aggregate formation ex vivo. We monitored the progressive and gradual changes induced by α-synuclein such as cellular toxicity, autophagy activation, mitochondrial dysfunction, cellular death as well as α-synuclein modification including site-specific phosphorylation. Our results demonstrate that organotypic brain slice cultures can be cultured for long periods of time and when cultured in the presence of aggregated α-synuclein, the molecular features of PD are recapitulated. Taken together, this ex vivo model allows for detailed modeling of the molecular features of PD, thus enabling studies on the cumulative effects of α-synuclein in a complex environment. This provides a platform to screen potential disease-modifying therapeutic candidates aimed at impeding α-synuclein aggregation and/or cellular transmission. Moreover, this model provides a robust replacement for in vivo studies that do not include behavioral experiments, thus providing a way to reduce the number of animals used in an accelerated timescale.