Bypassing the barrier: new routes for delivery of macromolecules to the central nervous system
Year: 2017
Type: article
Abstract: The blood–brain barriers of the central nervous system (CNS) provide a great deal of protection to the brain and spinal cord by blocking penetration of harmful molecules from the peripheral bloodstream. This protection is afforded by barriers at the level of the CNS blood vasculature (the blood–brain barrier proper), and the choroid plexus and arachnoid epithelia (the blood–cerebrospinal fluid (CSF) barrier). While important for the controlled delivery of peripherally produced molecules during early brain development (i.e. for normal brain growth and maturation), in the adult the penetration of therapeutic molecules is hampered by these protective measures (Saunders et al. 2016). As a result, there remains a huge unmet need for proper delivery of drugs into the CNS, and as such, there is increased interest and advantage in a semi-invasive CSF administration that delivers drugs behind these barriers. However, several unresolved issues remain about the sites and mechanism of entry from the CSF into the CNS parenchyma. A lot of work has focused on the route of entry for macromolecules from the blood into the CSF, through the choroid plexus epithelium, but very little has focused on the penetration of these molecules from the CSF further into the brain parenchyma or spinal cord. A long-standing belief has described penetration of molecules by size-dependent diffusional transport and gross fluid convection (Davson & Segal, 1995). Though the exact route into the brain has remained largely unidentified, physiological evidence indicates that in the smaller extracellular spaces passive diffusion is the major factor (Wolak & Thorne, 2013), while bulk-flow convection forces are more important for distribution in the CSF-filled ventricular and subarachnoid spaces (Davson & Segal, 1995). More recently, a route was proposed through the perivascular spaces (Virchow Robin spaces) between the blood vessel wall and the glial limitans of astrocyte endfeet encapsulating these vessels (Iliff et al. 2012). This ‘glymphatic’ hypothesis provided evidence of bulk fluid flow to remove metabolites from the brain parenchyma, but it did not address where and how CSF-borne molecules might enter the perivascular space. In this issue of The Journal of Physiology, Pizzo et al. (2018) study a new route of entry for macromolecules, such as antibodies, from the cerebrospinal fluid into the brain. Their careful descriptions of stomata present on the CSF-facing leptomeningeal cells in the subarachnoid space suggest this may be a unique route for molecules to enter the perivascular spaces. Using whole and fragmented antibodies across a broad size range (∼15–150 kDa), this work highlights a possible delivery route for most major endogenous biomolecules, as well as larger therapeutic molecules, to the brain. Pizzo and colleagues used rodents to investigate brain penetration following intrathecal infusion of whole IgG antibodies, and smaller truncated single-domain antibodies. By using magnetic resonance imaging and fluorescently-tagged antibodies, the authors show clearance of the antibody from the subarachnoid space to the blood stream (presumably via the arachnoid villi), followed by substantial delivery to the perivascular spaces around most major vessels of the brain. They report diffusion at the brain surface and widespread distribution to deeper brain regions along these perivascular spaces. Consistent with a size exclusion model, the smaller single-domain antibody fragments were more penetrant than the whole antibody molecules. Next the authors asked if they could manipulate the degree of penetration of these antibodies into the brain parenchyma. By co-infusing the osmolyte mannitol, they hypothesized that they might increase the access of the larger full IgG, by drawing water out of the leptomeningeal cells on the adventitial surface vessels of the subarachnoid space. In line with their hypothesis, they report a dose-dependent increase in the penetration of IgG antibodies in all brain regions. But how would these molecules have been moving from the CSF spaces to the perivascular spaces around the blood vessels? To address this question, the exact route of entry for the two antibodies was investigated using scanning electron microscopy of fixed tissue. Looking at the meningeal vessels on the surface of the brain of untreated animals, these preparations revealed many stomata (micrometre-sized pores) on the CSF-facing outer surfaces of the leptomeningeal vessels. The authors speculate that these unique fenestrations are the route of entry of macromolecules to the perivascular spaces, before further penetration in the brain parenchyma. It seems likely that the route described here by Pizzo et al. (2018) elucidates how macromolecules in the CSF access perivascular convective flow pathways – such as the rapid perivascular transport described in prior work, including the ‘glymphatic’ system suggested by Maiken Nedergaard and colleagues (Iliff et al. 2012). The transport of macromolecules in the CSF by diffusion into extracellular spaces at the brain surface (through newly described stomata on the leptomeningeal vessels), followed by convection in the perivascular spaces of the cerebral blood vessels to deeper brain regions, describes new details for what are now becoming appreciated as important new pathways for entry of macromolecules to the brain. Combined, the results of this new study by Pizzo and colleagues demonstrate how the distinctive architecture of the CNS produces the unique size-dependent distribution of antibodies from the CSF to the brain. Furthermore, they show that they can improve delivery of these antibodies by osmotic manipulation – increasing the distribution into the perivascular spaces of the brain parenchyma. This study reveals a previously undescribed entry mechanism for endogenous macromolecules into the central nervous system. What remains is to determine if this route will prove beneficial for the improved administration of drugs to the brain parenchyma via intrathecal delivery. Ultimately the study by Pizzo et al. (2018) provide much-needed information on an additional exploitable route for drug delivery to the CNS, and the prospect of using the intrathecal route of administration for pre-clinical and translational drug trials is highly appealing. None None received for this work.
Source: The Journal of Physiology
DOI: 10.1113/jp275499
Author: Shane Liddelow
Institution: Stanford University
Language: English
Subfield: Neurology
Field: Neuroscience
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W2796013484