My name is Jennifer Sokoloski, and I'm a resident at the University of Virginia. Today I'm going to show that arachnoid granulations seem to be immune tissue poised for immune surveillance of the brain. How does the immune system sense CNS in dream pathology? My data suggests that meninges and arachnoid granulations are a hub for immune responses. Specifically, arachnoid granulations seem to have machinery for immune surveillance of the brain. Most of what we know about the immune system is gleaned from studies done in the periphery, so we know that immune cell traffic through the blood can extravasate into tissue to survey the tissue, and this occurs under normal physiologic circumstances and can be increased in response to injury. Immune cells, as well as lymph fluid, then travel through lymphatic vessels to lymph nodes, where an immune response can be initiated. In the CNS, we don't see immune cells such as B and T cells surveying the parenchyma, and that's largely because of the blood-brain barrier. It limits extravasation of immune cells. We really only see B cells and T cells in the parenchyma when the endothelium is activated, which, for example, occurs in the context of injury. So under baseline conditions, immune surveillance of the parenchyma by B cells and T cells really does not happen. How does immune surveillance occur in the CNS? The CNS has a specialized system for clearance of waste. Current dogma suggests that it's through the contribution of glymphatic drainage and meningeal lymphatics that immune surveillance would occur. Glymphatic drainage refers to the concept that CSF produced by the choroid plexus flows convectively through the parenchyma, carrying with it macromolecules, debris, metabolites. CSF could then drain through the meningeal lymphatics downstream to the cervical lymph nodes where an immune response would occur. Most of these studies have been done in real rodents, however, and so our understanding of how this translates to humans is relatively limited. We do know that the human meninges, specifically the dura, contain lymphatic vessels. So this is some staining done in the dura of humans, looking at lymphatic vessels with DAB staining here or in this green immunofluorescence, and these lymphatic vessels are found adjacent to the superior sagittal sinus. So current dogma is really that meningeal lymphatics are the key structure. They are the highway for antigens and immune cells to travel through to induce an immune response. However, again, these studies were really done in rodents, and rodents do not have the same meningeal structure as humans. Specifically, rodents do not have readily identifiable arachnoid granulations, and studies so far have really failed to understand how the meningeal lymphatics and arachnoid granulations would interrelate. Models as depicted now really depict that CSF flowing through arachnoid granulations is basically just dumped in the superior sagittal sinus. But does it really make sense that CSF flowing through arachnoid granulations would bypass immune surveillance? Additionally, does it make sense that the first initiation of an adaptive immune response would be all the way down in the cervical lymph nodes? Instead of thinking that it's just the meningeal lymphatics that are of interest, I would suggest that it's the arachnoid granulations as well that are interesting as a potential hub for immune responses. Arachnoid granulations are definitively distinct from these dural lymphatics, but it turns out that they also contain lymphatic endothelial cells and immune cells. They seem to be capable of antigen presentation, and they seem to be responding to injury, notably the density of immune cells increases after hemorrhagic injury. I'm going to start by orienting you to arachnoid granulations, showing you some gross anatomy histology. Here are some cadaveric dissections done by Kan-Yag Merlu, and you can see here he's unroofed the superior sagittal sinus, taking away that top layer of dura, and you can see the arachnoid granulations adjacent to the sinus. In a coronal view depicted here in an H&E section, you can see that most of the arachnoid granulations, the protrusions shown here, are really adjacent to the sinus. While some protrude directly into the sinus, most of them actually seem to be protruding into the parasagittal venous lacunae. When we zoom in on one of these and we do some immunostaining, looking at markers of lymphatic endothelium, podoplanin and LyB1, it appears that these arachnoid granulations contain lymph channels. Here we've added a marker for lymphatic endothelial cells, another one, CD31. CD31 is also well known for highlighting blood vessels, but it also highlights lymphatic endothelium. And you can see nice colocalization of these markers in these cells lining these sinus type structures. These are pretty reminiscent of structures that you would see on lymph nodes. Notably, these are distinct from those lymphatic vessels that I previously talked about that run in the dura. So this is an example of a zoomed in picture of a lymphatic vessel shown here. You can see that the caliber of this vessel is minuscule compared to the size of the sinus. This is really in contrast to rodents, where the lymphatic vessels running adjacent to the sinus are at least on the same order of magnitude as the sinus. And it seems reasonable that in rodents, they may indeed be responsible for transporting a significant proportion of CSF. However, in humans, the minuscule nature of these makes it seem like they would be unlikely to be able to carry the amount of CSF needed for drainage. I would propose that that's why arachnoid granulations have evolved. However, instead of thinking of arachnoid granulations as just simply a way to dump CSF, I would propose that CSF may indeed be sampled as it is draining through arachnoid granulations. So here we're looking at, for example, you've got brain and then arachnoid, and then the arachnoid membrane interdigitating into the dura. Here we're looking at markers for lymphatic endothelium, and LIV1 does also highlight a subset of macrophages as well. But here we're looking at specifically just the CD45 cells, and you can see that there are a large number of immune cells in this membrane. Coming in again, showing podoplanin, the lymphatic endothelial marker, and CD45 in red, the immune cell marker, you can see that even just under baseline physiologic conditions, there are a large number of immune cells in arachnoid granulations. Next I'm going to show you that the lymphatic channels and immune cells are capable of antigen presentation. The MHC molecules, known as HLA in humans, are the molecules or the complexes responsible for antigen presentation. You can imagine CSF flowing by would contain debris and antigens that could be endocytosed by a cell. These proteins are then broken up into smaller peptides that are bound in a binding pocket on MHC, and then this is presented on the surface of the cell to T cells. The MHC peptide complex can either induce activation of an immune cell or energy, depending on the co-stimulatory molecules that are present. I'm going to show that lymphatic endothelium presents or expresses MHC2. This is not completely novel, as there has been a paper that showed this before. This is standing specifically in arachnoid granulations, showing that the lymphatic endothelium there is expressing MHC2. So in the top row, we're looking at co-localization of the lymphatic marker LiV1 and MHC2. In the middle row, we're showing co-localization of some of the immune cells with MHC2. And in the bottom row, we're looking after a hemorrhagic injury, and it seems subjectively that the intensity of MHC2 increases. We do know that the number of immune cells increases after hemorrhage. So we've quantified the number of immune cells, both under baseline conditions shown in these black bars here, and we can see that arachnoid granulations are enriched for immune cells even at baseline. That's compared to distal arachnoid tissue. The number of immune cells or the density of immune cells further increases in the context of hemorrhage. But this increase is specific to arachnoid granulations, as we don't see an increase in distal arachnoid tissue. What types of immune cells are present? Looking under baseline conditions versus after hemorrhage, we've stained for dendritic cells in green, B cells in red, and T cells in red here. And we see dendritic cells take on a more activated morphology, suggesting of a response to injury, and there seem to be more B cells. There also seem to be more T cells, specifically labeling for CD4 cells versus CD8 cells. It seems like there are mostly CD8 cells that are in the arachnoid granulations after injury. For some future directions, we'd like to continue to examine the changes in immune cells in arachnoid granulations after injury, comparing to the parenchyma and understanding the time course. What happens first, increases in the parenchyma versus increases in the meninges in arachnoid? Importantly, we'd like to further characterize the immune cell profile through single cell sequencing, and we'd like to understand what are the injury signals or the peptides that are presented on the MHC by looking at the MHC peptidyl. First I'd like to thank the NREF for a grant that funded this research, and I'd like to thank everybody on this list, but especially Dr. Min Park, Dr. Deryasha Kalani, and Dr. Tonya Agmarloo for their help with this research, as well as Dr. Peter Tverdyk.

arachnoid granulations

immune surveillance

brain

meninges

lymphatic channels