Physiological brain clearance architecture revealed by neuronal protein tracing.

Physiological brain clearance architecture revealed by neuronal protein tracing - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Search Log in Primary site navigation Search Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Cell . Author manuscript; available in PMC: 2026 Jun 2. Published in final edited form as: Cell. 2026 May 29;189(15):4737–4755.e17. doi: 10.1016/j.cell.2026.04.048 Physiological brain clearance architecture revealed by neuronal protein tracing Yuichi Chayama 1, * , Nalini R Rao 1, * , Daniela Perla 1 , Zimo Zhang 1 , Madigan Reid 1 , Sophia Nelson 1, 2 , Xinlan Wen 1 , Bella Ding 1 , Jessica Blumenfeld 1, 3 , Amanda Apolonio 1 , Sahith Doddipalli 1 , Haoyue Zhou 1 , Sena Gül Turhan 4 , Pu-Yun Shih 6 , Matthias Brendel 5, 7 , Ying-Hui Fu 6, 8, 9 , Ali Ertürk 4, 7 , Zeynep Ilgin Kolabas 4, 5, 7 , Yadong Huang 1, 6 , Andrew C Yang 1, 6, 10, # Author information Article notes Copyright and License information 1 Gladstone Institute of Neurological Disease, San Francisco, CA, USA. 2 Biomedical Sciences Graduate Program, University of California, San Francisco, San Francisco, CA, USA. 3 Neuroscience Graduate Program, University of California, San Francisco, San Francisco, CA, USA. 4 Department of Intelligent Biotechnologies, Helmholtz Munich, Neuherberg, Germany 5 Department of Nuclear Medicine, LMU University Hospital, LMU Munich, Munich Germany 6 Department of Neurology, University of California San Francisco, San Francisco, CA, USA. 7 Munich Cluster for Systems Neurology (SyNergy), Munich, Germany 8 Weill Institute for Neurosciences, University of California San Francisco, San Francisco, CA. 9 Kavli Institute for Fundamental Neuroscience, University of California San Francisco, San Francisco, CA, USA. 10 Bakar Aging Research Institute, University of California San Francisco, San Francisco, CA, USA. * These authors contributed equally Author contributions A.C.Y., Y.C., and N.R.R. conceptualized the study. Y.C. and N.R.R. led experiments, including immunohistochemical stains and imaging, flow cytometry, brain border dissections, single-cell RNA-sequencing, and fluorescence quantifications, with the support of M.R., S.N., B.D., D.P., Z.Z., A.A., J.B., H.Z., S.T., S.D., and P-Y.S. N.R.R and A.C.Y. wrote the manuscript with input from all authors. A.E., Z.I.K., Y-H.F., Y.H., and A.C.Y. supervised the study. # Correspondence to andrew.yang@gladstone.ucsf.edu Issue date 2026 Jul 23. PMC Copyright notice PMCID: PMC13225612  NIHMSID: NIHMS2173840  PMID: 42214342 The publisher's version of this article is available at Cell Summary The brain must efficiently clear protein waste to maintain homeostasis, yet physiological drainage pathways remain poorly defined. Standard tracer injection approaches may not reflect endogenous efflux. Here, we develop a non-invasive genetic system to trace neuron-derived protein clearance from brain to CSF and border tissues. We identify distinct drainage routes and border hotspots missed by tracer injection, confirmed by bioorthogonal labeling of endogenous neuronal proteins. Pulse-chase kinetics reveal slow skull outflow versus rapid dural and nasal clearance. Transcriptomic analyses uncover border cells sampling neuronal antigens, including tolerogenic skull-resident B cells. Region-restricted reporter expression demonstrates compartmentalized clearance following a “nearest exit” principle, where anatomical origin dictates drainage pathway. Disease disrupts clearance through distinct mechanisms: inflammation drives vascular leakage into blood, while amyloid pathology causes parenchymal retention and border exit obstruction. These findings define brain clearance as a compartmentalized system of organized pathways and immune niches whose dysfunction may underlie regional vulnerability in neurological disease. Keywords: Brain waste clearance, Brain borders, CNS drainage, Cerebrospinal fluid, Neuroimmunology, Alzheimer’s disease Graphical Abstract In Brief Neuronal protein tracing, unlike CSF tracers, reveals that brain clearance follows a “nearest exit” principle through distinct border immune niches disrupted differently by disease. The brain’s high metabolic activity produces waste that must be continuously cleared to maintain neural function and homeostasis 1 – 3 . In the absence of a conventional lymphatic system, the central nervous system (CNS) relies on a complex network involving cerebrospinal fluid (CSF), perivascular spaces, and meningeal pathways to manage waste clearance 4 – 11 . CSF, produced primarily by the choroid plexus within the brain’s ventricles, bathes the brain and spinal cord and is the conduit for waste clearance via exchange with the brain’s interstitial fluid (ISF) 4 , 12 – 23 . Disruption of brain waste clearance has emerged as a key feature of brain aging and a common hallmark across neurodegenerative diseases, where aberrant accumulation of waste products like amyloid-beta (Aβ) drives cellular dysfunction 4 , 24 – 34 . For example, impairments in the glymphatic system 4 , 25 , 33 , 35 – 38 , which clears waste from the brain via flow-mediated CSF-ISF exchange, have been implicated in Alzheimer’s disease (AD) 24 , 25 , 27 , 39 – 44 . Once waste enters the CSF, its efficient removal relies on an elaborate network of drainage pathways. Over the past decade, tracer injections into the CSF have illuminated the anatomical landscape of viable brain waste clearance pathways and potential exit routes 4 , 5 , 12 , 15 , 27 , 45 – 54 . CSF can drain along perivascular channels before reaching one of several fates: drainage via arachnoid granulations 5 , 6 , 55 , cranial nerve sheaths 15 , 56 , meningeal lymphatics in the dorsal 5 , 6 , 27 , 57 , 58 or basal 9 , 54 , 59 aspects of the dura 5 , 9 , 12 , the skull 60 – 64 , and the cribriform plate 16 , 19 , 55 , 65 – 70 and nasal cavity 16 , 19 , 50 , 59 , 69 , 71 , 72 en route to the draining cervical lymph nodes 10 , 32 , 54 , 65 , 73 , 74 . While these approaches have mapped where CSF and solutes could anatomically flow, the relative utilization of each CNS drainage pathway for the clearance of brain-derived proteins remains an important open question. A significant consideration in interpreting these studies is that pressure-driven tracer infusion can alter intracranial pressure and fluid dynamics, potentially obscuring physiological efflux routes and forcing open anatomical paths, even at low volumes and flow rates 35 , 75 – 79 . Furthermore, technical variation in tracer type, doses, volumes, anesthesia, and experimental protocols has led to conflicting results between studies 18 , 22 , 44 , 80 – 82 . While alternative approaches tracking disease-relevant proteins such as Aβ have provided insight into pathology-specific clearance, the unique disease-predisposing biophysical properties and aggregation-prone nature of these proteins may skew their drainage distribution and sites of accumulation in ways that may not reflect the handling of the broader set of CNS proteins 4 , 40 , 42 , 44 , 48 . To complement existing approaches and elucidate the clearance routes of brain-derived proteins from their site of origin, we developed a non-invasive system to genetically express and track a fluorescent reporter secreted from neurons without altering fluid dynamics. This approach was inspired by pioneering studies tracking antigen transfer and immune surveillance in cancer immunology by using the highly stable, soluble, and biologically inert fluorescent protein ZsGreen 83 – 87 . In adopting this established system for the CNS, we reveal the physiological clearance architecture of neuron-derived proteins. To ensure these findings reflect native protein handling, we validated our reporter using endogenous biorthogonal proteome labeling and compared our results to multiple tracer injection paradigms. Using these techniques, we define the hierarchy of exit routes, identify surveilling immune cells at neuroimmune checkpoints, determine how clearance pathways are regionally organized, and investigate how they are disrupted in models of neuroinflammation and neurodegenerative disease. Results A genetic system to trace the clearance of neuron-derived proteins To identify the physiological routes through which brain-derived proteins exit the central nervous system (CNS), we sought a strategy to express and track the clearance of an endogenously expressed reporter from the brain. To do this, we adapted a Cre-dependent system to drive expression of ZsGreen 83 selectively in neurons. In this paradigm, 2–4 month old ZsGreen reporter mice (ZsG) were intravenously administered AAV-PHP.eB-hSynapsin1-Cre to induce neuron-specific expression and secretion of ZsGreen protein (Syn-ZsG mice) ( Fig. 1A ). Immunostaining, Western blot (WB), and fluorescence-based quantification of brain homogenates confirmed robust brain-wide ZsGreen expression in Syn-ZsG mice relative to PBS-injected wild-type (WT) controls two weeks after transduction ( Fig. 1B – D ). To confirm that neuron-derived ZsGreen enters the extracellular space for clearance, we biochemically fractionated 83 , 88 cortical tissue into extracellular vesicle (EV) and soluble components ( Fig. S1A ). Enrichment of Tsg101 validated the EV enrichment process ( Fig. S1B – C ) 89 . Critically, ZsGreen was significantly elevated in both the EV and soluble fractions from Syn-ZsG brains, demonstrating that it is actively secreted and available within the brain’s interstitial fluid for subsequent clearance ( Fig. 1E – F ). Figure 1. A genetic system to trace the in vivo distribution of brain-derived proteins. Open in a new tab (A) Schematic of experimental paradigm. ZsGreen reporter expression is induced in neurons via retroorbital delivery of AAV-PHP.eB-hSynapsin1-Cre into ZsGreen-Cre transgenic mice (Syn-ZsG). ZsGreen, produced in neurons is secreted in soluble form and extracellular vesicles (EV). (B) Immunohistochemistry of brain sections confirms neuronal expression of ZsGreen in Syn-ZsG mice compared to PBS-injected wild-type (WT) controls. (C-D) Western blot and fluorescent plate reader analysis of brain homogenates verifies significantly increased ZsGreen abundance in Syn-ZsG animals compared to WT controls. (E-F) Biochemical fractionation of EV and soluble fractions of Syn-ZsG brain homogenates shows ZsGreen is present in both, confirming its secretion. (F) Quantification of WB shows significantly more ZsGreen abundance in Syn-ZsG mice. (G) Immunohistochemistry reveals ZsGreen (green) within Cre-positive neurons (magenta) and also small puncta within Cre-negative astrocytes (GFAP), vascular endothelial cells (CD31), and microglia (Iba1), suggesting transfer to non-neuronal populations. (H) Flow cytometry of brain homogenates reveals significant uptake of ZsGreen by astrocytes (ACSA2+), microglia (CD45 mid CD11b+), blood endothelial cells (BECs) (CD31+), and border-associated macrophages (BAMs) (CD206+) in Syn-ZsG. Stars indicate comparisons of ZsGreen+ cells in WT versus Syn-ZsG. (I) Schematic of the CUBIC-clearing of Syn-ZsG brain and 3D imaging light sheet fluorescence microscopy (LSFM) workflow used to assess whole-brain ZsGreen distribution. (J) LSFM imaging of Syn-ZsG brains reveals widespread ZsGreen distribution, with notable signal localizing at vascular sites labeled by WGA-647 perfusion (pan blood vessels) and αSMA. White arrows indicate ZsGreen at vascular interfaces. All data are mean ± SEM with n = 3–4. *p value = < 0.05; **p value = < 0.01; ***p value = < 0.001, ****p value = < 0.0001 by Student’s t test. To visualize ZsGreen expression in the brain and investigate potential transfer to non-neuronal populations, we performed immunohistochemistry on brain sections from Syn-ZsG and WT mice. As expected, ZsGreen colocalized with Cre and the neuronal marker NeuN ( Fig. 1G ). Notably, ZsGreen puncta were also observed in non-neuronal cells lacking Cre expression, including astrocytes (GFAP + ), vascular endothelial cells (CD31 + ), and microglia (Iba1 + ), suggesting transfer of neuronal proteins across CNS cell types ( Fig. 1G , Fig. S1D – E ). Flow cytometric analysis confirmed this intercellular transfer, showing that a significant proportion of microglia, astrocytes, blood endothelial cells (BECs), and border-associated macrophages (BAMs) were ZsGreen+ in Syn-ZsG mice compared to WT controls ( Fig. 1H ). This demonstrates that neuron-derived proteins are sampled by a diverse array of cell types within the CNS parenchyma. Finally, to visualize the spatial distribution of ZsGreen at the whole-brain level, we performed light sheet fluorescence microscopy (LSFM) on CUBIC-cleared Syn-ZsG brains ( Fig. 1I ). Using WGA-647 and αSMA to label vasculature, this analysis confirmed broad ZsGreen expression in the brain with notable punctate signal at vascular interfaces ( Fig. 1J ). Together, these data establish the Syn-ZsG system as a robust platform for tracking the secretion and movement of a neuronal protein from its site of origin. We next used this system to map its precise exit routes from the brain. Drainage routes for brain-derived proteins differ from CSF-injected tracers To understand how the clearance of brain-derived proteins compares with conventionally injected tracers, we performed intracisterna magna (ICM) injections of fluorescently labeled Ovalbumin (OVA-555 or OVA-647) in Syn-ZsG mice. One-hour post-injection 4 , 52 , 53 , 73 , 90 , we collected brain border tissues or cleared whole heads for tissue clearing and 3D imaging LSFM analysis to visualize and quantify the distribution of both ZsGreen and OVA across drainage sites within the same mice ( Fig. 2A ). Immunostaining of the brain, dura, nasal cavity, and superficial cervical lymph nodes was performed to verify that brain border tissues were Cre-negative ( Fig. S2A ). LSFM imaging of the whole mouse head revealed striking bulk differences in the localization of the two fluorescent reporters. ZsGreen signal, in addition to being highly abundant in the brain, was also broadly distributed across brain borders. On the other hand, ovalbumin tracer signal was highly concentrated near the injection site ( Fig. 2B ). Figure 2. Drainage routes for brain-derived proteins differ from ICM-injected tracers. Open in a new tab (A) Schematic of the experimental workflow. For imaging experiments, mice were administered intracisterna magna (ICM) injections of Ovalbumin (OVA-647 or OVA-555) to compare exogenous tracer distribution with brain-derived protein efflux (Syn-ZsG). One-hour post-injection, whole heads were fixed and processed for tissue clearing and light sheet fluorescence microscopy (LSFM). (B) LSFM imaging of whole-head CUBIC-cleared Syn-ZsG mice reveals divergent spatial distribution patterns for neuron-derived ZsGreen and ICM-injected OVA-555. OVA-555 signal is concentrated near the injection site at the base of the brain, whereas ZsGreen signal is highly expressed in the brain and broadly dispersed. (C-D) Quantification of ZsGreen in Syn-ZsG mice (C) and OVA-555 from ICM-injected WT mice (D) intensities across brain borders and drainage tissues using fluorescence plate reader analysis. Fluorescent intensities from bulk homogenates were normalized to total protein and respective WT control tissues to enable comparison across samples of fold changed ZsGreen or OVA-555 abundance. ZsGreen was significantly enriched in the Syn-ZsG dorsal dura and skull (DSK+dura), lateral skull (LSK+dura), basal skull (BSK+dura), cribriform plate (CRP), and nasal cavity (NS), with relatively low levels in superficial and deep cervical lymph nodes (sCLNs, dCLNs). In contrast, OVA-555 showed high enrichment in sCLNs and dCLNs. (E) Data from (C-D) plotted as a heatmap of z-scored fluorescence intensities across tissues highlights three distinct drainage profiles: tissues with comparable signal for both tracers (brown), tissues with preferential OVA accumulation (pink), and tissues with elevated ZsGreen signal (green), indicating distinct clearance routes for endogenous proteins versus exogenous tracer. (F) Immunohistochemistry of dura whole mounts demonstrates both ZsGreen and OVA signal localize along the CD31+ dural sinuses, with notable enrichment in the sagittal sinus (SS/BV) bridging veins and rostral confluence of sinuses (RCS). (G) Regional anatomical analysis of tracer distribution within dura. Fluorescence intensity of ZsGreen and OVA was measured at major sinus regions (RCS, TS, SS/BV) and in non-sinus areas, then normalized to total tracer intensity across the dura. (H-I) Representative immunohistochemistry images of sCLNs and dCLNs from WT and Syn-ZsG mice. OVA signal is prominent in both lymph nodes, while ZsGreen signal is comparatively low. (J) Quantification of fluorescence intensities from (H-I), normalized to tissue area. ICM-injected OVA signal is significantly higher than neuron-derived ZsGreen in both sCLNs and dCLNs. Sagittal sinus/bridging veins = SS/BV, rostral confluence of sinus = RCCS, transverse sinus = TS, dorsal skull = DSK, lateral skull = LSK, basal skull = BSK, cribriform plate = CRP, nasal cavity = NS, superficial and deep cervical lymph nodes = sCLNs, dCLNs. All data are mean ± SEM with n = 3–4. *p value = < 0.05; **p value = < 0.01; ***p value = < 0.001, ****p value = < 0.0001 by one-way ANOVA with Tukey’s post-hoc or Student’s t test. To systematically assess the distribution of each tracer, we quantified ZsGreen and OVA-555 levels across a panel of brain-border and drainage tissues using fluorescent plate reader analyses. Fluorescence signals were normalized to both total protein and respective WT controls to enable direct comparisons between brain borders and tracers ( Fig. 2C – D ). As expected, ZsGreen was significantly enriched in the brain of Syn-ZsG mice ( Fig. 2C ). Compared to WT controls, ZsGreen signal was significantly elevated across nearly all tissues examined, with significant enrichment in the dorsal, lateral and basal skull + respective dura, nasal cavity, and cribriform plate. Notably however, the superficial and deep cervical lymph nodes (sCLNs and dCLNs) showed minimal ZsGreen. In contrast, OVA displayed a markedly different distribution profile. ICM-injected OVA levels were lower in the nasal cavity and cribriform plate but was found significantly higher in the sCLNs and dCLNs, demonstrating rapid efflux to collecting lymph nodes, likely bypassing the brain 47 ( Fig. 2D ). A heatmap of z-scored fluorescence intensities (from Fig. 2C – D ) across border tissues and lymph nodes revealed three major clusters: A heatmap of z-scored fluorescence intensities (from Fig. 2C – D ) revealed three clusters: tissues enriched for ZsGreen (nasal cavity, basal skull + dura), tissues with similar ZsGreen and OVA (dorsal/lateral skull + dura, cribriform plate), and tissues preferentially enriched for OVA (cervical lymph nodes) ( Fig. 2E ). These results demonstrate that brain-derived proteins distribute differently and likely exit through routes distinct from those used by injected tracers. Interestingly, even in tissues where both tracers were present at similar overall levels, such as the dorsal dura, we observed different within-tissue distributions comparing ZsGreen and OVA. In dural whole mounts from Syn-ZsG mice, ZsGreen and OVA both colocalized with CD31+ vasculature (perisinusal localization), but only ~40% of ZsGreen signal overlapped with OVA, suggesting distinct uptake or accumulation sites within the dura ( Fig. S2B ). ZsGreen accumulated at specific sinus structures, especially the bridging veins of the sagittal sinus (SS), likely reflecting drainage through recently discovered arachnoid cuff exit (ACE) points 91 , as well as in non-sinus regions of the dura ( Fig. 2F – G ). Immunostaining of sCLNs and dCLNs confirmed significant OVA accumulation and relatively sparse ZsGreen signal ( Fig. 2H – J , Fig. S2C ). Together, these data suggest that the drainage routes of brain-derived proteins differ from those of injected tracers. To verify that our ZsGreen paradigm reflects the drainage of endogenously produced neuronal proteins, rather than an artifact of the reporter itself, we leveraged an in vivo bioorthogonal proteome-labeling system (PheRS) 92 . In this system, Cre-dependent incorporation of a BBB-permeable, non-canonical amino acid (AzF), administered intraperitoneally (i.p.), occurs into all newly synthesized proteins. Using the same Syn-Cre AAV strategy, neuronal proteins incorporate AzF, rendering them click-detectable (Syn-PheRS). Following one week of AzF administration, immunostaining and in-gel fluorescence analyses confirmed AzF-labeled neuronal proteins (AzF-647) in Syn-PheRS mice compared to controls ( Fig. S2D – F ). Consistent with our ZsGreen findings, immunostaining and in-gel fluorescence analyses from Syn-PheRS mice demonstrated that endogenous neuronal proteins mirror the anatomical distribution of ZsGreen in dura and are present at brain borders ( Fig. S2G – K ). Together, these data validate ZsGreen as a faithful reporter of the drainage of endogenous neuronal proteins. A critical question is whether the divergent distributions of ZsGreen and ICM-injected OVA was due to the site of origin (parenchyma versus CSF) or tracer duration (chronic versus acute), as ZsGreen accumulates over a two-week transduction period while OVA is assessed one-hour after injection. To address this, we chronically administered OVA into the CSF using an osmotic pump for two weeks (matching the ZsGreen production window) and quantified its distribution across brain-border tissues and draining lymph nodes. Chronically delivered OVA exhibited a distribution pattern similar to acute ICM-injection, including comparable representation in the dura, underrepresentation at skull-associated borders, and significant enrichment in both sCLNs and dCLNs ( Fig. S2L – M ). These findings demonstrate that the distinct drainage routes observed for brain-derived proteins versus injected tracers are not attributable simply to differences in tracer duration. Brain-derived proteins exit the CNS through distinct anatomical compartments with border-specific clearance kinetics To visualize the global distribution of brain-derived proteins as they exit the CNS, we performed 3D LSFM on whole, CUBIC-cleared heads from Syn-ZsG mice after ICM injection of OVA-555 tracer, revealing an unexpected ZsGreen clearance route extending laterally along the zygomatic process that was not highlighted by OVA-555 ( Fig. 3A – B ). In the nasal cavity, ZsGreen concentrated above the dorsal medial meatus (DMM) and along lateral nasal glands (LNG), with IHC confirming distributions distinct from OVA ( Fig. 3C – D , Fig. S3A – D ). This highly structured localization suggests significant efflux of brain-derived proteins to specific functional zones in the anterior head and nasal cavity. Figure 3. Distinct drainage hotspots and clearance kinetics for brain-derived proteins. Open in a new tab (A-B) LSFM of CUBIC-cleared whole anterior heads from Syn-ZsG mice following ICM-injection of OVA-555. LSFM imaging reveals a prominent ZsGreen clearance route extending laterally along the zygomatic process that is not shown by OVA (magenta), indicating selective efflux of brain-derived proteins through bone-associated pathways. (i-ii) Merged channels. (C-D) LSFM imaging of the nasal cavity identifies additional routes of ZsGreen exit. Sagittal (C) and coronal (D) views show ZsGreen concentrated in discrete hotspots above the dorsal medial meatus (DMM) and along lateral nasal glands (LNG), as well as within naso- and ethmoturbinates, with a distribution distinct from OVA. (i-ii) Insets highlight regional differences in ZsGreen and OVA localization. (E) LSFM imaging of CUBIC-cleared heads with the brain removed reveals differential distribution of ZsGreen and OVA across skull and dura compartments. ZsGreen accumulates prominently along dorsal, lateral, and basal skull-associated regions, whereas OVA shows limited representation in these compartments. (F) Sagittal IHC sections of the dorsal skull demonstrate ZsGreen localization within skull bone marrow hubs compared to OVA, indicating differential access of brain-derived proteins to skull associated cellular niches. (i-iii) Insets show zoom in of boxed regions. (G) Flow cytometric analysis of dural cell populations from Syn-PheRS mice following click labeling of AzF-incorporated neuronal proteins (AzF-647). AzF-647 signal was quantified before and after permeabilization to distinguish surface-associated and intracellular pools, revealing a substantial intracellular accumulation in both lymphatic endothelial cells (LECs) and myeloid cells. Bar plots show mean fluorescence intensity (MFI) relative to controls. LECs gated on (CD45-CD31+PDPLN+); myeloid gated on (CD45hiCD11b+). n = 2 pooled mice. (H) Immunostaining of dorsal skull sections from Syn-PheRS mice following click detection of AzF-647. Co-staining with a surface marker (WGA) highlights intracellular localization of neuron-derived proteins within resident skull cells. (I) Schematic of the Syn-PheRS pulse-chase paradigm used to assess protein flux dynamics across brain borders. After seven days of AzF labeling, AzF was withdrawn and allowed to wash out for 0, 2, or 5 days. (J-K) In-gel fluorescence analysis of AzF-647 labeled proteins across brain border tissues reveals compartment-specific clearance kinetics. AzF-647 intensities were normalized to total protein and scaled with respect to T0. Decay constants ( k ) for each border show that skull-associated borders exhibit slow outflow, whereas the dura, nasal cavity, and cribriform plate display faster clearance rates. All data are mean ± SEM with n = 3–5. *p value = < 0.05; **p value = < 0.01; ***p value = < 0.001, ****p value = < 0.0001 by one-way ANOVA with Tukey’s post-hoc. To better resolve tracer distribution in the skull, we removed the brain prior to clearing and LSFM. This revealed marked differences between ZsGreen and OVA-555 across bone-associated structures, including distinct ZsGreen accumulation in the basal and lateral skull regions ( Fig. 3E , Fig. S3E – F ). IHC analysis of dorsal skull sections revealed distinct ZsGreen signal within skull bone marrow hubs compared to ovalbumin ( Fig. 3F ). These imaging data demonstrate that brain-derived proteins drain to anatomical compartments at levels underreported by injected tracers. To determine whether brain-derived proteins are internalized by resident cells at brain borders, we leveraged the Syn-PheRS platform to click-tag endogenously produced neuronal proteins. Flow cytometric analysis of dural cells comparing AzF-647 signal before and after permeabilization revealed that a substantial fraction of AzF-labeled neuronal proteins localized intracellularly within both lymphatic endothelial cells (LECs) and myeloid cells ( Fig. 3G , Fig. S3G ). Immunostaining of cleared whole heads and dorsal skull sections followed by click detection and surface labeling confirmed intracellular accumulation of brain-derived proteins within resident cells ( Fig. 3H , Fig. S3H – I , Movie S1 ). To assess flux dynamics, we performed a Syn-PheRS pulse-chase where after seven days of labeling, AzF was withdrawn for 0, 2, or 5 days ( Fig. 3I ). In-gel AzF-647 fluorescence analyses across border tissues homogenates revealed distinct clearance (sum of efflux and degradation) kinetics across compartments ( Fig. 3J – K ). Skull-associated borders exhibited slow outflow ( k = −0.008), whereas the dura, nasal cavity, and cribriform plate showed rapid clearance ( k = 0.127–0.261). Together, these data demonstrate that brain-derived proteins are actively taken up by resident border cells and that borders exhibit divergent clearance kinetics with slow skull retention versus rapid dural and nasal turnover. Distinct brain border and skull immune cells surveil brain-derived proteins to support immune tolerance Having identified key anatomical border sites for drainage, we next sought to define the specific cell types within them that uptake and surveil neuronal proteins. We collected the dorsal dura, dorsal skull, and nasal cavity from Syn-ZsG mice and WT controls. Tissues were dissociated, and ZsGreen+ and ZsGreen− cells were quantified by flow cytometry, then sorted for scRNA-seq ( Fig. 4A ). Figure 4. Distinct brain border and skull immune cell surveil brain-derived proteins to support immune tolerance. Open in a new tab (A) Schematic of the experimental workflow. Dorsal dura, dorsal skull, and nasal cavity tissues were collected from Syn-ZsG and WT mice. Isolated endothelial, stromal, and immune cell types were analyzed by flow cytometry to quantify relative levels of ZsGreen uptake. Cells of each types major cell type were then sorted at equal proportions of ZsGreen+ and - cells for single-cell RNA sequencing (scRNA-seq). (B) Flow cytometry-based quantification of the proportion of ZsGreen+ and - cells within each immune cell population in the dorsal dura, skull, and nasal, based on flow cytometry. BECs: CD45(−)CD31+; LECs: CD45(−)CD31(−)PDPLN+; cDC1: CD45hiCD11b(−)CD11c+; cDC2: CD45hiCD11b+CD11c+MHCII+; CD206+ BAMs: CD45hiCD11b+CD206+; Monocytes: CD45hiCD11b+Ly6C+; Macrophages: CD45hiCD11b+F480+CD206(−). (C) Comparison of ZsGreen+ populations of CD206+ BAMs and CD31+ BECs across brain border tissues shows that dural resident cells take up significantly more ZsGreen compared to cells residing in both skull and nasal compartments. (D) Immunofluorescence of dorsal dura whole mounts from Syn-ZsG mice reveals colocalization of ZsGreen with CD206 + BAMs and CD31+ vasculature. ZsGreen signal shows perisinusal accumulation, including the rostral confluence of sinuses (RCS), sagittal sinus (SS), and transverse sinus (TS). (E-F) UMAP projections of scRNA-seq data from dorsal dura (E), skull (F), and nasal cavity (G), background shaded color for each cluster represents annotated cell type. Green dots represent ZsGreen+ cells, grey dots represent ZsGreen− cells. (H-J) Bar plots summarizing the number of differentially expressed genes (DEGs) between ZsGreen+ and ZsGreen− cells by cell type in dorsal dura (H), skull (I), and nasal cavity (J). The skull shows substantially greater transcriptional differences in ZsGreen+ versus ZsGreen− immune cells, compared to similar populations in the dura and nasal tissue. (K) Upregulated DEGs between ZsGreen+ versus ZsGreen− blood endothelial cells (CD31+) in the dura shows a significant enrichment of gene ontology (GO) terms related to immune cell recruitment and adhesion. (L) Upregulated DEGs between ZsGreen+ versus ZsGreen− skull B-cells reveal immunological specialization. (L.i) Upregulated DEGs were significantly enriched for GO terms such as antigen processing and UPS degradation, and MHC I antigen processing and presentation. Many of the genes in these two GO terms are key antigen processing and presentation genes. String interactome analysis confirms that the genes in this GO term have been shown to interact. (L.ii) Gene expression changes for key tolerogenic and inflammatory genes from the B cell cluster demonstrate that ZsGreen+ skull B-cells display an upregulated tolerogenic transcriptional profile. (L.iii) Pseudotime trajectory analysis comparing the expression of a tolerogenic module ( Ptpn22, Zbtb20, Cd274) across the transition from pre-B to B-cells. The expression of the tolerogenic module was significantly higher in ZsGreen+ cells compared to ZsGreen−. All data are mean ± SEM with n = 4. *p value = < 0.05; **p value = < 0.01; ***p value = < 0.001 by one-way ANOVA with Tukey’s post-hoc, Wilcoxon, or Student’s t test. Top ten to twelve GO biological processes or reactome terms were selected based on .01 FDR cut off and sorted by nGenes. Flow cytometry quantified the proportion of ZsGreen+ and - cells in each tissue. We observed significantly greater ZsGreen sampling by dural cells compared to skull and nasal across cells types such as blood endothelial cells (BECs) and conventional dendritic cells (cDC1, cDC2) ( Fig. 4B , Fig. S4A ). In the dorsal dura, for example, approximately 60% of CD206 + BAMs and 65% of BECs were ZsGreen+ compared to ~35% and ~15% in skull and nasal tissue, respectively ( Fig. 4C ). IHC analyses corroborated significant uptake of ZsGreen across CD206+ cells at dural sinuses ( Fig. 4D , Fig. S4B ). Despite similar bulk ZsGreen levels in dorsal dura and skull versus nasal tissue ( Fig. 2 ), brain-derived antigen sampling varied across borders and did not scale with total uptake, suggesting border-specific structural or cellular specialization. To assess transcriptional signatures linked to antigen sampling, we performed scRNA-seq on near-equal numbers of ZsGreen+ and ZsGreen− cells from dura, skull, and nasal cavity. UMAP embedding revealed a complex community of immune, endothelial, and stromal cells at each border 57 , 62 , 78 , 93 – 95 ( Fig. 4E – G , Fig. S5A – C , Table S1 – 2 ). Across all three borders, ZsGreen and Cre transcripts were largely undetectable, indicating ZsGreen+ cells reflect uptake rather than expression ( Fig. S5D – E ). To further investigate cell populations exhibiting distinct transcriptional profiles linked to brain-derived protein uptake, we assessed transcriptional signatures associated with ZsGreen uptake ( Fig. 4H – J ). As expected, cell types like macrophages exhibited significant ZsGreen uptake across borders. Yet, comparison of ZsGreen+ versus ZsGreen− macrophages revealed minimal transcriptional differences ( Fig. S5F – G , Table S3 – 5 ). This implies that macrophage antigen sampling is governed less by specialized transcriptional programs and more by extrinsic factors such as the anatomy of each border and their positioning relative to outflow pathways. Among endothelial and immune populations identified across brain borders, dorsal dura endothelial cells and skull-resident B cells exhibited distinctive transcriptional profiles linked to neuronal protein uptake. Differential gene expression (DEG) analysis revealed upregulated genes in ZsGreen+ cells from dorsal dura BECs which were significantly enriched in pathways associated with immune cell adhesion and recruitment, including key genes such as Vcam1, Icam1, Cxcl10, Il15 , and Plcg2 ( Fig. 4K , Table S3 ). These results suggest that select dural endothelial cells actively surveil brain-derived antigens and facilitate local immune responses 57 , 96 . Notably, skull-resident B cells demonstrated the most pronounced distinct transcriptional profile linked to ZsGreen uptake ( Fig. 4I , Table S4 ). ZsGreen+ skull B cells upregulated genes which were enriched for pathways involved in antigen processing and presentation, proteasomal degradation, and vesicular trafficking ( Fig. 4L.i ). Core genes involved in antigen processing ( Erap1 , Pdia3 , Calr , Canx , and Hspa5 ) were significantly upregulated, along with vesicle trafficking genes such as Vamp3 , Sar1b , and Sec24a/b ( Fig. 4L.ii ). Concurrently, ZsGreen+ compared to ZsGreen− skull B cells significantly upregulated numerous key tolerogenic markers, including Il10ra , Cd1d1 , and Zbtb20, Ptpn22 , which are involved in limiting immune responses and promoting immune regulation 63 , 64 , 97 – 99 . To confirm this transcriptomic signature at the protein level, we performed quantitative immunostaining for the checkpoint molecule PD-L1 (CD274). We found that ZsGreen+ skull B cells displayed significantly higher PD-L1 protein levels compared to their ZsGreen− counterparts ( Fig. S5H – I ). In contrast, classical pro-inflammatory genes such as Tnf , Il1b , and type I interferon signaling components ( Irf1 , Irf3 , Irf7 ) were either unchanged or significantly downregulated ( Fig. 4L.iii ). Trajectory analysis revealed a gradual enrichment of tolerogenic module genes ( Ptpn22, Zbtb20, Cd274 ) 99 from pre-B to B cells ( Fig. 4L.iv ). Strikingly, from this analysis, ZsGreen+ cells displayed significantly higher tolerogenic module expression compared to ZsGreen−. Expanding on and consistent with prior studies 63 , 64 , 97 , 98 , these data suggest that a distinct population of skull B cells take up and drive a tolerogenic immune response to brain-derived antigens. Brain regions preferentially utilize nearby drainage routes to clear neuron-derived proteins Having established the overall architecture of brain-wide neuronal protein clearance, we next asked whether this system is uniform or spatially compartmentalized ( Fig. 5A ). We asked whether proteins from distinct brain regions drain via common or unique pathways by restricting Cre-mediated ZsGreen production to select brain regions and comparing ZsGreen clearance dynamics. Compared to our pan-neuronal Syn-ZsG model, we generated three regionally restricted ZsG models: (1) Forebrain-ZsG (CaMKII-Cre) labeling forebrain excitatory neurons, (2) Anterior-ZsG (AAV:hDLX-4X2C-Cre) 100 labeling forebrain and olfactory bulb GABAergic interneurons, and (3) Striatal-ZsG (AAV:AiE0441h_3xC2-Cre) 101 labeling striatal medium spiny neurons. We confirmed successful, region-specific ZsGreen expression in each paradigm ( Fig. 5B ). Figure 5. Brain waste clearance is regionally compartmentalized and follows a “nearest exit” principle. Open in a new tab (A) Schematic comparing global neuronal labeling using Syn-ZsG mice (left) versus a restricted ZsGreen expression. Cartoon illustrates two possibilities for ZsGreen clearance through global non-specific (top) or region-specific pathways (bottom). (B) Representative IHC analysis of restricted ZsG expression for 3 brain regions. Forebrain-ZsG (CaMKII-positive neurons), Anterior-ZsG (forebrain and olfactory bulb GABA-ergic interneurons), and Striatal-ZsG (striatal medium spiny neurons). IHC confirms restricted ZsGreen expression for each compared to PBS::WT controls. (C) Fluorescent plate reader quantification of ZsGreen signal in the brain shows significantly different ZsGreen abundances across Syn-ZsG, Forebrain-ZsG, Anterior-ZsG, and Striatal-ZsG paradigms. (D) Heatmap of ZsGreen utilization (proportion of ZsGreen at a border/total ZsGreen across borders) from fluorescence plate reader analysis. Fluorescent intensities from bulk homogenates were normalized to total protein and respective ZsGreen abundance in the brain. Across all regionally restricted models, ZsGreen signal is enriched at borders proximal to the labeled region and reduced at more distal borders. (E) Box plots showing the proportion of ZsGreen detected at selected brain border compartments (dorsal dura, BSK+dura, and NS) across Syn-ZsG and regionally restricted labeling paradigms. Significant differences in border utilization were observed depending on the anatomical origin of ZsGreen, with enrichment at borders proximal to the labeled region. (F) Representative immunohistochemistry of dural whole mounts from Forebrain-ZsG and Striatal-ZsG mice demonstrates varied anatomical distributions of ZsGreen colocalizing with vascular structures (CD31). (G) LSFM of whole CUBIC-cleared heads from Striatal-ZsG mice with ICM-injected OVA-555. Top-down and side perspectives of the striatum-distal dorsal dura showing little ZsGreen signal in the dorsal dura of Striatal-ZsG mice. (H) Schematic depicting proposed “nearest-exit” paradigm for brain waste clearance (top). Distances from the mid-forebrain and mid-striatum to dorsal and basal brain borders illustrate how anatomical proximity may bias drainage toward closer border compartments. Sagittal sinus/bridging veins = SS/BV, rostral confluence of sinus = RCS, transverse sinus = TS. All data are mean ± SEM with n = 4. *p value = < 0.05; **p value = < 0.01; ***p value = < 0.001 by one-way ANOVA with Tukey’s post-hoc or Student’s t test. As total brain ZsGreen levels varied across paradigms, we normalized all border tissues by total brain ZsGreen and quantified the relative utilization of each border ( Fig. 5C ). Comparison of relative border utilization revealed a divergence in the drainage pathways engaged by each brain region ( Fig. 5D , Fig. S6A ). Across all three regionally restricted models, ZsGreen signal was enriched at borders proximal to the labeled region and reduced at distal borders ( Fig. 5E ). For example, dorsal dura exhibited significantly greater ZsGreen drainage in Forebrain-ZsG mice compared to Striatal-ZsG mice. Conversely, the nasal cavity showed significantly greater drainage in Striatal-ZsG mice compared to Forebrain-ZsG mice while Anterior-ZsG mice showed a strong bias towards cribriform plate drainage. Dural whole-mounts and whole-head LSFM imaging provided visual confirmation of this pattern, revealing decreased ZsGreen in the dorsal dura and increased signal in the nasal cavity and basal skull in Striatal-ZsG mice ( Fig. 5E – G , Fig. S6B – E ). Having ruled out tracer duration as the source of divergence between ZsGreen and CSF tracers, we next asked whether matching the anatomical site of origin would eliminate this difference. We compared Striatal-ZsG mice to mice receiving intraparenchymal OVA injections into the striatum (Str-Inj:OVA). IHC of brain sections confirmed OVA-647 injection at one hour, with clearance evident by 24 hours ( Fig. S6F ). Plotting the proportion of OVA and ZsGreen at each drainage border in Str-Inj:OVA and Striatal-ZsG mice respectively revealed that skull borders were underrepresented by OVA with substantially greater drainage to sCLNs ( Fig. S6G ). This indicates that bolus injection, even when delivered directly into the parenchyma, overwhelms local border uptake capacity and shunts flow to cervical lymph nodes. Together, these findings indicate that brain-derived protein clearance is regionally compartmentalized, with preferential drainage to proximal borders, supporting a “nearest exit” principle ( Fig. 5H ). Neuroinflammation and amyloid pathology differentially disrupt clearance of brain-derived proteins Having defined the architecture of physiological brain clearance, we next investigated how this system is altered under pathological conditions. We used two distinct and clinically relevant models: acute systemic inflammation induced by lipopolysaccharide (LPS), and the 5XFAD transgenic model 102 of amyloidosis and Alzheimer’s disease. First, we examined the impact of acute inflammation. We intraperitoneally (i.p.) injected Syn-ZsG mice with 3 mg/kg LPS, an endotoxin known to compromise the blood-brain barrier (BBB) and activate neuroimmune responses ( Fig. 6A ) 103 , 104 . As expected, LPS impaired BBB integrity resulting in Cadaverin-647 tracer extravasation as well as neuroinflammation confirmed by GFAP+ reactive astrogliosis ( Fig. 6B , Fig. S7A ). Next, we investigated if LPS treatment modulated ZsGreen drainage. While ZsGreen levels within the brain remained unchanged compared to WT controls, clearance to major border exit routes, including the dorsal and basal skull and nasal cavity, was significantly reduced after LPS treatment ( Fig. 6C – D ). Given the brain showed no difference in abundance and select brain borders displayed less ZsGreen, we hypothesized that the ZsGreen was inappropriately shunted through compromised vasculature into the blood ( Fig. S7B – C ). Indeed, we quantified an increase in ZsGreen fluorescence in plasma with LPS ( Fig. 6E , Fig. S7D – E ). These findings suggest that acute neuroinflammation disrupts homeostatic drainage routes through brain borders, triggering a collapse of normal efflux pathways and shunting of neuronal proteins directly into peripheral circulation through a leaky vasculature. Figure 6. Neuroinflammation and neurodegenerative pathologies induce distinct forms of drainage dysfunction. Open in a new tab (A) Schematic of the experimental paradigm for inducing acute inflammation. WT mice and Syn-ZsG mice (one week after transduction), were given LPS (3mg/kg) or PBS via intraperitoneal (i.p.) injection to induce acute neuroinflammation. After 24 hours, mice were administered intravenous Cadaverin-647 tracer to assess LPS-induced sequelae and BBB compromise. (B) Immunohistochemistry of brain sections from Syn-ZsG mice shows Cadaverin-647 extravasation in Syn-ZsG + LPS mice compared to Syn-ZsG + PBS controls. (C) Fluorescent plate reader analysis demonstrates that LPS treatment does not significantly alter ZsGreen abundance within the brain. (D) Fluorescent plate reader quantification of ZsGreen levels across brain borders and drainage sites from Syn-ZsG PBS or LPS-treated mice reveals an overall decrease in drainage to brain borders, with significant decreases across the dorsal skull, basal skull, and nasal cavity (NS) following LPS treatment. (E) Fluorescent plate reader analysis demonstrates that LPS treatment significantly increases ZsGreen abundance in the blood plasma of Syn-ZsG mice. (F) Schematic of the experimental paradigm for investigating how ZsGreen clearance is affected by amyloid pathology in the 5XFAD model. In 7 or 2-month-old 5XFAD mice, retroorbital AAV transduction of AAV::hSyn-ZsGreen-P2A-mCherry was performed. After 14 days of transduction, brain borders and drainage tissues were collected for biochemical analyses. WT::hSyn-ZsG (WT-ZsG) and 5XFAD::hSyn-ZsG (5XFAD-ZsG). (G) Immunohistochemistry of brain sections from WT-ZsG and 5XFAD-ZsG mice confirms that both genotypes have strong neuron-specific ZsGreen production, but only 5XFAD mice have amyloid pathology. Insets show ZsGreen puncta colocalized and associated with extracellular Aβ plaques. (H) Aβ 42 ELISA analysis of brain homogenates confirms that 5XFAD-ZsG mice produce significantly more Aβ 42 compared to WT-ZsG healthy controls. (I) Fluorescent plate reader analysis of ZsGreen abundance in brain homogenates reveals that 5XFAD-ZsG brains have significantly more ZsGreen compared to WT-ZsG mice. (J) Fluorescent plate reader quantification of ZsGreen levels across brain borders and drainage sites of WT-ZsG or 5XFAD-ZsG mice reveals an overall decrease in drainage to nearly all brain borders in 5XFAD mice compared to healthy controls. In 5XFAD-ZsG tissues, corresponding Aβ 42 ELISA analysis demonstrates Aβ 42 mostly accumulates at the dorsal dura, despite global impairment in ZsGreen drainage across brain borders. (K) Dural whole mounts from 7M 5XFAD confirm Aβ 42 is highly abundant in the dorsal dura at distinct bridging veins exit points. (L) Fluorescent plate reader analysis of ZsGreen abundance in blood plasma reveals 5XFAD-ZsG have significantly less ZsGreen in the blood compared to WT-ZsG mice. (M) Aβ 42 ELISA analysis of 2-month and 7-month-old 5XFAD-ZsG confirms that over time Aβ 42 accumulation is significantly increased in the brain and dura compartments, but is decreased in the plasma. (N) Correlation of ZsGreen and Aβ 42 abundance in 5XFAD-ZsG blood plasma mice reveals a significant positive correlation. (O) Summary schematic showing the change in ZsGreen abundance in the brain, borders, and blood with acute inflammation (LPS) and with amyloid pathology (5XFAD). Neuroinflammation and neurodegenerative pathologies induce distinct forms of drainage dysfunction: while LPS shunts neuron-derived ZsGreen away from homeostatic drainage routes directly into the blood likely via BBB compromise, Aβ pathology induces ZsGreen clogging and retention in the brain. Dorsal skull = DSK, lateral skull = LSK, basal skull = BSK, cribriform plate = CRP, nasal cavity = NS, superficial and deep cervical lymph nodes = sCLNs, dCLNs. All data are mean ± SEM with n = 3–4. *p value = < 0.05; **p value = < 0.01; ***p value = < 0.001, ****p value = < 0.0001 by one-way ANOVA with Tukey’s post-hoc, Spearman’s correlations, or Student’s t test. To assess how a chronic, age-related stressor impacts brain clearance, we next turned to the 5XFAD mouse model, which develops early and aggressive amyloid pathology beginning at 2 months of age 102 . While these mice do not recapitulate all aspects of human AD, they represent a strong model for investigating how ZsGreen drainage across brain borders is affected by progressing neuropathology. We leveraged AAV-based expression of ZsGreen in neurons in WT and 5XFAD mice followed by biochemical analyses ( Fig. 6F ). IHC analysis of brain sections and quantification of brain confirmed neuronal ZsGreen production and revealed that 5XFAD brains displayed substantial ZsGreen and Aβ 42 accumulation relative to controls ( Fig. 6G – I ). Given ZsGreen’s retention in the brain, we next investigated if brain borders had modulated uptake of ZsGreen. Indeed, nearly all brain borders showed significantly decreased ZsGreen levels in 5XFAD-ZsG mice compared to controls ( Fig. 6J ). ZsGreen levels across brain borders did not correlate with Aβ 42 : while ZsGreen drains across diverse brain borders, Aβ’s unique biophysical properties and regional expression may favor dorsal dura efflux pathways 27 . Immunostaining of dural whole mounts confirmed Aβ 42 at the dura with accumulation at bridging veins exit points 105 ( Fig. 6K , Fig. S7F ). This impairment in brain-to-border clearance was reflected in the periphery, where plasma ZsGreen was significantly lower in 5XFAD-ZsG mice compared to controls ( Fig. 6L ). To verify that ZsGreen production did not alter endogenous Aβ 42 distribution, we quantified Aβ 42 levels across brain borders in 5XFAD-ZsG and 5XFAD and observed comparable patterns ( Fig. S7G ). Finally, to assess how clearance impairments tracks with disease progression, we analyzed 5XFAD mice at 2 months of age ( Fig. 6M , Fig. S7H ). ZsGreen efflux from brain to blood diminished from 2 to 7-months of age and the degree of impairment correlated strongly with Aβ 42 levels ( Fig. 6N , Fig. S7I ). Together, these findings demonstrate that when disease-relevant insults disrupt CNS protein clearance, they do so through fundamentally distinct mechanisms. Acute inflammation reroutes homeostatic drainage through a destabilized vasculature directly into the bloodstream. In contrast, amyloid pathology progressively obstructs brain protein drainage by trapping proteins within the parenchyma and blocking their access to border efflux routes, resulting in reduced clearance ( Fig. 6O ). Discussion In this study, we developed and validated a genetic system to track the clearance of neuron-derived proteins as they efflux from their site of origin in the brain to exit sites across CNS borders. Using this system, we reveal that neuron-derived protein drainage differs from routes predicted by injected tracers. We establish that physiological clearance follows a “nearest exit” principle for brain regions, exhibits distinct clearance kinetics across brain borders, is surveilled by select pro-tolerogenic border immune cells, and fails via distinct mechanisms in inflammation versus neurodegeneration. These findings support a framework where brain waste clearance is spatially compartmentalized and organized to deliver neuronal antigens from specific brain regions to corresponding border niches. Our results help clarify long-standing discussion regarding the primary routes of CNS protein clearance 4 , 6 , 32 , 35 , 71 . We propose the divergence in clearance patterns observed between our ZsGreen paradigm and CSF-tracer studies reflects fundamental differences in how these approaches engage physiological pathways. We found that under physiological conditions, neuronal proteins primarily drain to dorsal and basal dura, skull, and nasal sites, with less prominent drainage to superficial and deep cervical lymph nodes (sCLNs and dCLNs) 4 , 6 , 32 , 35 . This distribution differs from many studies using intracisterna magna injected tracers, which report that the majority (typically >50–80%) of infused tracers are recovered in the CLNs 4 , 33 , 47 , 59 , 106 , 107 . This discrepancy may stem from methodology where acute bolus or chronic CSF infusions may favor rapid flow to lymph nodes and bypass the brain 47 , whereas our steady-state tracing of protein from the brain highlights the importance of immune-rich “staging areas” in the dorsal dura 5 , 6 , 9 , 12 , 27 , 57 , 58 and skull 60 – 64 upstream of cervical lymph nodes. Our findings are consistent with immunopeptidomics studies demonstrating richer repertoires of CNS antigens in dura compared to lymph nodes 108 . Critically, we demonstrate this discrepancy is not an artifact of tracer duration. Matching the chronicity of ZsGreen production with 14-day osmotic pump infusion of OVA into the CSF did not recapitulate the parenchymal drainage pattern. Instead, chronic CSF tracers still accumulated preferentially in cervical lymph nodes. This indicates that lymph node accumulation in tracer studies reflects the site of tracer introduction (CSF vs parenchyma) and dose (bolus vs continuous production). Consistently, intraparenchymal OVA injection into the striatum still substantially drained to sCLNs, unlike the drainage of Striatal-ZsG mice. By revealing neuron-derived proteins draining to both dorsal 5 , 6 , 27 , 57 , 58 and basal 9 , 54 , 59 aspects of the dura and skull, our findings suggest both routes mediate brain waste clearance and reveal an additional layer of spatial organization. Indeed, our neuron-derived tracing strategy reveals a previously unappreciated degree of spatial compartmentalization in brain clearance, indicating that “primary” routes of CNS clearance will depend on where the molecule in question originates in the brain. This clearance compartmentalization may also inform translational relevance; given evolutionary expansion of the human neocortex 109 , 110 , our “nearest exit” principle suggests a potentially greater functional role for dorsal clearance pathways in humans 111 , 112 . In addition, we found that border compartments differ in clearance kinetics, with skull-associated borders exhibiting slower outflow than dura, nasal cavity, and cribriform plate, indicating that CNS waste clearance is both spatially and temporally compartmentalized. Slower skull kinetics may facilitate prolonged antigen retention and immune surveillance. Our flow and transcriptomic analyses suggest dorsal pathways play a significant role in immune surveillance relative to other brain borders 5 , 6 , 9 , 12 , 27 , 57 , 58 , 60 – 64 , 113 , 114 . For instance, despite similar overall ZsGreen exposure, BAMs 64 , 115 , 116 in dorsal dura sampled significantly more ZsGreen than BAMs in the nasal cavity. Yet, comparison of ZsGreen+ and ZsGreen− BAMs did not reveal distinct transcriptional signatures, suggesting extrinsic, anatomical differences or BAM positioning within each border regulates their antigen uptake. In contrast, we discovered a distinct transcriptional program in ZsGreen-sampling skull-resident B cells. ZsGreen+ B cells upregulated key components of the antigen processing and presentation pathway, while simultaneously expressing tolerogenic regulators, including PD-L1 protein. This “tolerogenic-presentation” profile supports a model where antigen-sampling skull B cells maintain immune tolerance to CNS-derived proteins. This concept extends recent studies showing that the skull marrow hosts a reservoir of immature B cells capable of responding to local CNS cues 61 , 64 , 117 – 120 . The heterogeneity in clearance kinetics and CNS immune surveillance across borders, together with the “nearest exit” hypothesis, opens new hypotheses for understanding regional vulnerability in disease. For instance, this patterning could function as a biological “zip code” system, ensuring that self-antigens from a specific neuronal neighborhood are consistently delivered to an immune niche at the corresponding brain border 108 , 113 , 116 , 118 . We hypothesize that in aging and disease, these precise pathways can degrade, effectively “scrambling” these zip codes. Expanding on prior studies, we found that neuroinflammation and amyloid pathology dysregulate brain clearance via distinct mechanisms 25 , 27 , 30 , 38 , 42 , 121 – 123 . Specifically, acute inflammation shunts brain-derived proteins into the blood via vascular leakage, while amyloid pathology causes parenchyma retention and obstruction of border exits. Both mechanisms could promote dysregulated neuroimmunity, for example leakage may misroute brain antigens to the peripheral immune system rather than the proper pro-tolerogenic brain borders 57 , 64 , and retention leads to a loss of antigen drainage to and education at borders 108 . Further, this model could explain how the selective failure of clearance pathways servicing specific brain regions could lead to localized buildup of toxic proteins, initiating a cascade of pathologies characteristic of neurodegenerative disorders. Overall, by developing a genetic system to trace the in vivo distribution of brain-derived proteins, we map the brain’s physiological clearance architecture, revealing preferred drainage pathways, distinct border immune interfaces, and mechanisms of dysfunction in disease. Limitations of the study While our approach provides significant advantages over traditional tracer methods, limitations remain. First, our findings are in mice, owing to the need for genetic manipulation. Though conservation of meningeal lymphatics between rodents and humans has been shown, the precise anatomical flow patterns may differ in primates, necessitating future validation of these findings in non-human primate or human tissue studies 44 , 46 . Second, although we validated our findings with the endogenous proteome (Syn-PheRS) 92 , ZsGreen is an exogenous protein that can form tetramers. Third, while Cre immunostaining and scRNA-seq of border tissues confirmed brain-restricted Cre expression, AAV-PHP.eB can transduce extra-CNS neurons (e.g., spinal cord, dorsal root ganglia). Finally, steady-state accumulation does not equal clearance rate. We addressed this with pulse-chase kinetics, though decay rates reflect combined local degradation and efflux which cannot be fully disentangled. Complementary live imaging will be needed to capture acute temporal dynamics of CNS waste clearance. Resource availability Lead Contact Requests for further information and resources should be directed to and will be fulfilled by the lead contact, Andrew C. Yang ( andrew.yang@gladstone.ucsf.edu ) Materials Availability This study did not generate new unique reagents. Data and Code Availability Raw sequencing data are available in the NCBI Sequence Read Archive (SRA) under BioProject accession number SRP599291. This paper does not report original code. Any additional information in this paper is available from the lead contact upon request. STAR Methods EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS Animals C57BL/6J (RRID:IMSR_JAX:000664), ZsGreen (IMSR_JAX:007906), 5XFAD (RRID:MMRRC_034840-JAX), CaMKII-Cre (RRID:IMSR_JAX:005359), and PheRS (RRID:IMSR_JAX:033734) mice were obtained from Jackson Laboratory and bred in-house. For Forebrain-ZsG mice, ZsGreen mice were crossed and bred in house to CaMKII-Cre mice to achieve homozygous robust ZsGreen expression in CaMKII-positive neurons. Young adult mice of both sexes under the written age are used. All animals were fed ad libitum and housed under 12-hour light/dark cycles (7 a.m. - 7 p.m.), under the control of humidity and temperature. All experiments were approved by the Institutional Animal Care and Use Committee of the University of California, San Francisco (#AN206908–00A). AAV Production The pENN.AAV.hSyn.Cre.WPRE.hGH was a gift from James M. Wilson (Addgene plasmid # 105553, RRID:Addgene 105553) and was packaged to AAV-PHPeB in house or in VectorBuilder for bulk production. The AiP14825 : pAAV-AiE0441h_3xC2-minBG-iCre(R297T)-BGHpA (Alias: CN4825) was a gift from The Allen Institute for Brain Science & Jonathan Ting (Addgene viral prep # 214862-PHPeB; RRID:Addgene 214862) 101 . The viral prep of AiP11851 -pAAV-hDLX-minBG-iCre-4X2C-WPRE3-BGHpA (Alias: CN1851) was purchased through Addgene 100 . AAV was otherwise produced in-house using HEK293T cells as the packaging cell line. Cells were maintained in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 100 U/mL penicillin, and 100 μg/mL streptomycin at 37°C in a humidified incubator with 5% CO 2 . For transfection, cells were plated at 30–40% confluence in 15 cm dishes 24 hours prior, reaching 70–80% confluence at the time of transfection. DNA was transfected using a polyethyleneimine (PEI)-based method with modifications. Cells were incubated at 37°C for 3 days post-transfection before harvesting. They were then lysed by three freeze-thaw cycles. The lysate was filtered through a .45μm PES filter, and AAV particles were precipitated using differential-speed ultracentrifugation. Vector titers were determined by quantitative PCR (qPCR) using primers and labeled probes targeting the ITR sequence. Retroorbital Injection Male and female mice received retro-orbital injections of AAV vectors capable of crossing the blood-brain barrier. Mice were anesthetized with 2–3% isoflurane prior to the procedure. The virus was diluted to a concentration of approximately 6–8 × 10 12 genome copies (GC) per mL in sterile saline. A total volume of 100 μL (equivalent to 6–8 × 10 11 GC per mouse) was injected into the retro-orbital sinus of the right eye using a 31G needle attached to an insulin. After injection, mice were monitored and maintained for a 1–2 week expression period prior to sample collection. Intracisterna magna (ICM) injections of tracers Mice were anesthetized with 3.0–3.5% isoflurane and administered subcutaneous injections of buprenorphine and bupivacaine prior to surgery. The back of the neck was shaved and disinfected, and the mouse was positioned on a surgical stage to optimize alignment for access to the cisterna magna, as previously described. A midline incision was made, and the neck muscles were gently retracted to expose the cisterna magna. A total of 5 μL of Alexa Fluor-conjugated ovalbumin (8 mg/mL in PBS, Thermo Fischer product# O34782 ) was injected at a rate of 1 μL/min using a Hamilton syringe connected to a 33G needle via silicone tubing. Following the injection, the needle was held in place with positive pressure for 2 minutes to prevent backflow. The incision site was closed using a surgical stapler, and mice were placed on a heating pad until fully recovered. Meloxicam was administered subcutaneously immediately after surgery. Intrastriatal injection of OVA Mice were anesthetized with isoflurane before being placed in a stereotaxic injection rig. Mouse head was restrained by ear bars with placing the front teeth in the holding apparatus. Isoflurane and oxygen were provided at a low flow rate to keep the mouse anesthetized during the whole procedure. To prevent from drying, the eyeballs were covered with ophthalmic ointment. Buprenex (0.1 mg/kg) was administrated by intraperitoneal injection. 5μL of Alexa Fluor-conjugated ovalbumin (8 mg/mL in PBS) were injected in each striatum using Hamilton needles. Injection sites, ±1.8 mediolateral, 0.74 anteroposterior, and −3.5 dorsoventral. An automatic injection pump was used to control the speed at 200 nl/min. The needle was holding at injection position for 10 min before slow withdrawal. After closing the wound, meloxicam (1 mg/kg) was administrated by intraperitoneal injection. The mouse was put onto a warming pad for recovery. Chronic CSF-infusion of OVA For intracerebroventricular infusions, osmotic pumps (Alzet Model 2002, 0.50 μL/hr, Cat #0000296) were filled with OVA-647 (diluted to delivery 40μg/day) and connected to brain infusion cannulas (Alzet brain infusion kit 3, Cat. 0008851) modified with a spacer disc to enable ventricular targeting. Mice were anesthetized with isoflurane (Dechra, Cat. #200–129) and secured in a stereotaxic frame (Kopf Instruments Model 940). The scalp was shaved, and a midline incision was made to expose the skull. After identifying bregma, a stereotaxic site was drilled with a 0.5 mm microburr (Fine Science Tools, Cat. #19007–05) at coordinates X = +1.0 mm and Y = −0.25mm. Osmotic pump was implanted in a subcutaneous pocket formed on the back of the mouse, and the connected infusion cannula was inserted Z = −2.0mm through the drill site and into the right lateral ventricle. The incision was closed with nylon monofilament non-absorbable 6–0 sutures (Surgical Specialties Look, Cat. #911B). Lidocaine (Vedco) was applied subcutaneously along the incision, and buprenorphine (Henry Schein, Cat. #55175) and ketofen (Henry Schein, Cat. #005487) were administered for analgesia. Mice were monitored on a heating pad until ambulatory and then single-housed for 14 days before collection. Endogenous labeling of neuronal proteomes To generate Syn-PheRS mice, PHP.eB.hSyn.Cre.WPRE.hGH AAV was RO injected into PheRS mice and let to transduce for 8–10 days. Following transduction, 185mg/kg of g4-azido-l-phenylalanine (Vector Laboratories; 1406–5G AzF) was intraperitoneally administered to Syn-PheRS mice every day for one week 124 . For pulse-chase kinetic experiments, after Syn-PheRS received one week of labeling, wash out periods of 2 and 5 days where AzF was withdrawn. Mice were collected at T0, T2, and T5 timepoints for downstream analyses. METHOD DETAILS Cu-Click for IHC Tissue sections were processed for click chemistry labeling of azide-modified proteins following the standard immunofluorescence protocol through the blocking step. Sections were then incubated for 1 h in a CuAAC reaction containing Alexa Fluor 647-alkyne (5 mM), CuSO 4 (20 mM), THPTA (50 mM), aminoguanidine hydrochloride (50 mM), sodium l-ascorbate (50 mM), and PBS. CuSO 4 and THPTA were pre-complexed at a 1:2 ratio for 15 min prior to addition of the remaining reagents. After the reaction, sections were washed three times in TBS-T and either subjected to subsequent immunofluorescence staining or mounted and cover slipped. In-gel fluorescence detection of azide-modified proteins For detection and quantitation of azide-modified proteins across brain border homogenates from Syn-PheRS mice, we utilized DBCO-based click chemistry. Protein lysates (250–500 μg per tissue) were reduced with 10 mM dithiothreitol (DTT) for 10–15 min at room temperature, followed by alkylation with 20 mM iodoacetamide (IAA) for 20 min in the dark. Excess IAA was quenched by the addition of DTT to a final concentration of 20 mM and incubated for 15 min at room temperature. Samples were then incubated with 6 μM DBCO-Alexa Fluor 647 (Click Chemistry Tools) for 1 h at room temperature with rotation in the dark. Following labeling, samples were mixed with LDS sample buffer and boiled for 10 min. Labeled homogenates were then run by SDS-PAGE. Immunohistochemistry Mice were anesthetized via intraperitoneal injection of Avertin solution. Brains, dura, skulls, and lymph nodes were collected following transcardial perfusion with cold PBS. Brains were post-fixed in 4% paraformaldehyde (PFA) in PBS overnight at 4°C, then cryoprotected in 30% sucrose/PBS for at least two overnights. Brains were sectioned at 40μm thickness using a microtome (Leica), and sections were stored in cryoprotectant solution at −20°C until immunostaining. Before staining, brain sections were washed three times in TBS with 0.1% Tween-20 (TBST, Sigma-Aldrich) to remove the cryoprotectant, then blocked and permeabilized in blocking buffer (2% normal donkey serum, 0.2% Triton X-100 in TBS) for 30 minutes with shake at room temperature. Dura and skulls were fixed in 4% PFA/PBS for 3 days. The dorsal dura was carefully dissected from the skull cap and processed immediately without storage. Tissues were washed three times in TBST and then blocked and permeabilized in the blocking buffer for 30 minutes at room temperature. Skulls were decalcified in 0.5M EDTA for more than 5 days, cryoprotected in 30% sucrose/PBS for 1–2 days, and embedded in OCT compound. The OCT-embedded samples were stored at −80°C until sectioning. Cryosections of skull tissue (20 μm) were collected on adhesive glass slides (SuperFrost Plus, Fisher Scientific), air-dried, and stored at −80°C until immunostaining. Fresh lymph nodes were embedded in OCT compound immediately after collection from perfused animals, frozen on dry ice, and stored at −80°C. Cryosections (20 μm thickness) were placed on adhesive glass slides and stored at −80°C until immunostaining. For immunostaining, the primary antibodies were applied at a 1:100–1000 dilution in blocking buffer and incubated overnight at 4 °C After three washes with TBST, appropriate fluorophore-conjugated secondary antibodies (1:200 dilution) were applied and incubated overnight at 4 °C. Sections were then washed again with TBST and mounted on glass slides using VectaShield mounting medium (H1800 or H1700). Images were acquired using confocal microscopes (LSM880 and LSM700). Tissue Clearing and Light Sheet Fluorescence Imaging Mice were anesthetized via intraperitoneal injection of Avertin solution, euthanized by severing the right atrium to allow exsanguination, and then intracardially perfused with 50–80 mL of cold PBS to flush out blood, followed by 35 mL of Alexa Fluor–conjugated wheat germ agglutinin (10 μg/mL in PBS) to label blood vessels, and finally 40 mL of 4% paraformaldehyde (PFA) in PBS for fixation. For whole-brain clearing, the brain was carefully extracted post-perfusion and incubated in 4% PFA/PBS at 4 °C overnight, followed by the CUBIC tissue clearing protocol. For head clearing, the entire head was collected after perfusion, trimmed to remove skin and eyeballs while preserving the skull and surrounding tissues, including muscles, salivary glands, lymph nodes, and ligaments. The trimmed head was incubated in 4% PFA/PBS overnight at 4 °C, then decalcified in 0.5 M EDTA at 37 °C for 5–7 days. Following three PBS washes, the head was coronally bisected at the level of the bregma into anterior and posterior halves prior to CUBIC clearing. For both brain and head samples, tissues were incubated in CUBIC-L (TCI, cat #T3740) at 37 °C for 5–7 days with agitation for delipidation. The CUBIC-L solution was refreshed periodically when it turned yellow during the process. After delipidation, tissues were washed three times in PBS, then sequentially immersed in 50% and 100% CUBIC-R+(N) (TCI, cat# T3983) for refractive index (RI) matching. For skull imaging (head samples without brain), the head was sagittally halved after EDTA decalcification, and the brain was gently removed. The skull was delipidated in CUBIC-L at 37 °C for 5–7 days with agitation, washed three times in PBS, and blocked/permeabilized overnight at room temperature in blocking buffer (2% normal donkey serum, 0.2% Triton X-100, 0.2% NaN 3 in TBS). Following three additional washes with TBST and three rinses with PBS, the samples were immersed in CUBIC-R+(N) for refractive index (RI) matching. Fluorescence 3D imaging was conducted using a light sheet fluorescence microscope (UltraMicroscope Blaze, Miltenyi) with EasyIndex Matched Immersion Oil (RI:1.52) (life canvas technologies, cat# Oil-1.52). Imaging was performed using 405, 488, 588, and 647 nm lasers, with a 1× objective lens and 1× zoom. Each optical plane was illuminated bilaterally from both the left and right sides. Tile scans were acquired and raw image files were converted to IMS format, stitched, and rendered into 3D reconstructions using Imaris software (Oxford Instrument, version10.2.1) for visualization and analysis. Tissue homogenization and plate reader quantification Mice were anesthetized via intraperitoneal injection of Avertin. Blood was collected from the right ventricle using an EDTA-coated syringe (25G, Fisher Scientific), transferred to a 1.5 mL protein low-binding tube, and centrifuged at 2,000 × g for 15 min at 4 °C to isolate plasma. Following blood collection, mice were perfused intracardially with cold PBS via the left ventricle. Tissues were dissected to remove muscle and connective tissue, segmented into defined anatomical regions, and transferred into individual tubes. Dorsal durae were carefully peeled from the dorsal skull. The basal skull and nasal cavity were further divided into subregions, including the nasal cavity, cribriform plate, lateral skull, and basal skull. All collected tissues were snap-frozen on dry ice and stored at −80 °C until further processing for protein extraction. Hard tissues (dorsal, lateral, and basal skulls, nasal cavity bone, and cribriform plate), along with the brain, were homogenized in RIPA lysis buffer supplemented with Halt protease inhibitor (Thermo Scientific), using metal beads (Revity) and a tissue homogenizer. Soft tissues (dura, sCLNs, and dCLNs) were homogenized using a probe sonicator. Homogenates were centrifuged at 15,000 × g for 15 min at 4 °C, and the supernatants were collected. Total protein concentrations were determined using the BCA Protein Assay Kit (Thermo Scientific). ZsGreen fluorescence was measured using 96-well fluorescence assay plates and a plate reader (Molecular Devices). The relative intensity values were calculated by normalizing fluorescence intensity to total protein concentration. To determine the fold enrichment of respective tracers at each border tissues were normalized to WT controls. For proportion of tracer at borders analyses, the border tracer levels were divided by the sum of tracer amounts across borders. Western blot For western blot analysis, protein concentrations in each sample were measured with BCA protein Assay Kit (Thermo Scientific). Equal quantities of protein samples were boiled for ten minutes in LDS buffer. Samples were then loaded onto the NU-PAGE precast gels and electrophoresed for high-resolution separation of proteins based on the size. Transfer onto 0.45-micron size nitrocellulose membrane was achieved in a Bio-Rad semi-dry quick transfer apparatus. Membrane blocking was done with 5% milk for 60 min before primary antibody incubation overnight at 4 °C. The next day, following four washes in TBST, five minutes each with shaking, membranes were probed with HRP-conjugated secondary antibodies obtained from the same host. Flow cytometry The brain, dura, skull, nasal cavity, and superficial and deep cervical lymph nodes (sCLNs and dCLNs, respectively) were collected in cold PBS or RPMI medium from mice, following intracardiac perfusion with cold PBS. Muscles and connective tissues were trimmed from the dorsal skull and nasal cavity bones. Brain cell suspensions were prepared using the Neural Tissue Dissociation Kit (Miltenyi Biotec, cat# 130-092-628), followed by myelin removal via centrifugation in 0.9 M sucrose/PBS. Dural cell suspensions were prepared according to a previously published protocol 125 : dorsal dura were peeled from the skull, pooled, and digested in an enzyme mix (2.5 mg/mL Collagenase D and 0.1 mg/mL DNase I in RPMI) at 37 °C for 30 minutes, followed by mechanical dissociation with 20–30 pipette strokes using a P1000 tip. Skull and nasal cavity cell suspensions were prepared following a previously described method 62 . Pooled tissues were minced in RPMI medium, vortexed at maximum speed for 30 seconds, and filtered through a 100 μm cell strainer. This process was repeated twice with fresh RPMI, and the resulting filtrates were pooled to generate a single cell suspension. Dissociated single-cell pellets were resuspended in cold PBS and passed through a 35 μm cell strainer of a FACS tube. From surface versus intracellular assays, dissociated cells were split in half and incubated in buffer with or without detergents. After centrifugation, cells were stained with Zombie NIR viability dye (BioLegend, cat#423106; 1:1000 dilution in PBS) on ice for 20 minutes. Following a wash with PBS, cells were blocked with Fc block (BD Biosciences, 1:100) and incubated on ice for 20 minutes with antibodies. After staining, cells were washed with FACS buffer, permeabilized (if needed) in IC buffer, fixed for 20 minutes, and stored in FACS buffer at 4 °C until analysis. For AzF detection, cells after FACs buffer and/or permeabilization were alkylated with 20mM IAA for 20 min in the dark before 1 hour incubation with 6μM DBCO-647. Flow cytometry was performed using a BD FACSAria Fusion (BD Biosciences). Acquired data was analyzed with FlowJo. Single-Cell RNA Sequencing Two weeks after AAV injection, the brain, dura, skull, and nasal cavity bone were collected in cold PBS or RPMI medium from Syn-ZsG mice (both sexes) following intracardiac perfusion with cold PBS. Single-cell suspensions from each tissue were prepared and stained using the same protocol as described for the flow cytometry experiments. To obtain sufficient cell numbers for sorting, tissues were pooled as follows: brain (n = 4), dorsal dura (n = 8), skull (n = 8), and nasal cavity bone (n = 4). Following antibody staining, cells were fixed overnight at 4 °C in Fixation Buffer from the GEM-X Flex Sample Preparation v2 Kit (10x Genomics, #PN-1000781). The next day, fixed cells were treated with Additive C (10x Genomics, #PN-1000781), centrifuged, and resuspended in PBS containing 1% nuclease-free BSA and 0.2 U/μL RNase inhibitor (Roche). Flow cytometry gating was performed to exclude debris (SSC-A vs. FSC-A), followed by selection of singlet cells (FSC-W vs. FSC-A), and gating of ZsGreen+ and - populations for sorting. Doublets and multiplets were excluded. Up to 50,000 to 100,000 cells were sorted per sample. Sorted cells were centrifuged and stored overnight at 4 °C prior to probe hybridization and GEM generation. Sorted cells were hybridized with probes (10x Genomics) according to the manufacturer’s instructions. Custom probes targeting Cre recombinase and ZsGreen were designed following 10x Genomics guidelines. After hybridization, cells were loaded into the 10x Genomics Chromium system, and library preparation was performed using the GEM-X FLEX kit, following the manufacturer’s protocol. Sequencing was conducted on an Illumina NovaSeq platform using paired-end reads in accordance with 10x Genomics recommendations. Blood-Brain Barrier Permeability Assay Mice were retro-orbitally injected under anesthesia using an insulin syringe (BD, 328421; 0.33 mm × 12.7 mm) with cadaverine-Alexa Fluor 647 (Invitrogen, A30677) at a concentration of 0.5 mg/mL, 100 μL per 20 g body weight. After a 2-hour circulation period, mice were perfused with 0.01 M PBS for 5 minutes, and brains were harvested for analysis. One hemisphere was weighed and homogenized in 1% Triton X-100 in 0.01 M PBS. The homogenate was centrifuged at 21,000 × g for 20 minutes at 4 °C, and the supernatant was collected. LPS-induced inflammation ZsGreen expression was allowed to develop for one-week. Mice then received an intraperitoneal injection of LPS (3 mg/kg) or PBS as a control. One day after LPS administration, tissues were collected for further analysis. Sample processing for protein quantification followed the same procedures described in the previous section. 5XFAD ZsGreen Induction Neuronal ZsGreen expression was induced in 2- and 7-month-old 5XFAD and age-matched WT mice with retroorbital administration of 6 × 10 11 GC PHPeB-hSyn-ZsGreen-P2A-mCherry-NLS-SV40. Transduction was allowed for 14 days before tissue collection and downstream analyses. Aβ 42 ELISA Aβ 42 ELISA analysis was performed as previously described 126 , in short homogenized tissues from brain, brain border tissues, and blood were diluted in the standard diluent buffer. 50μL of blank solution, standards, and samples were loaded into antibody-coated wells and incubated with detection antibody for 3hr at RT. After multiple washes, HRP-conjugated antibody was added for 30 min. After washes, the samples were incubated with stabilized chromogen for 30 min, and the reaction was stopped with an acid-based stop solution. OD was measured at 450 nm with plate reader (Molecular Devices) and compared to a standard curve to determine the final concentration. QUANTIFICATION AND STATISTICAL ANALYSIS scRNA-seq data preprocessing Raw sequencing reads were aligned to a custom mm10 reference genome containing ZsGreen and Cre transgene sequences using Cell Ranger, and samples were demultiplexed using barcode information from the 10x Genomics Feature Barcoding Multiplexing kit. Filtered gene-barcode matrices were imported into R for downstream analysis with Seurat 127 – 129 (v5.2.1). Sample metadata, including tissue source (dura, skull, nasal) and ZsGreen status, were derived from sample identifiers and added to the object. Cells were filtered based on standard quality metrics: those with fewer than 200 detected genes, more than 6,000 genes, or over 10% mitochondrial transcript content were removed. Additional thresholds for total UMI counts and gene complexity were used to refine filtering. Doublets were identified and excluded using DoubletFinder 130 , with the expected doublet rate set to 7.5% and neighborhood size optimized per tissue. Each tissue was processed separately for normalization, feature selection, and dimensionality reduction. PCA was performed on the top variable genes, and Harmony was applied for sample-level integration within each tissue type to account for technical variation. Clustering and cell annotation Each tissue dataset (dura, skull, nasal) was independently preprocessed, including normalization, scaling, and principal component analysis (PCA), followed by batch correction using Harmony 131 . The top 30 PCs were used to construct a shared nearest neighbor graph and perform unsupervised clustering using the Leiden algorithm, with resolution parameters adjusted (0.2–0.4) based on tissue complexity. UMAP was applied for two-dimensional visualization. Clusters were reviewed post hoc and iteratively filtered to remove low-quality or artifactual populations, including those with elevated mitochondrial content, ambiguous transcriptomic profiles, or contaminant gene signatures (e.g., skeletal muscle). Following removal, re-clustering was performed and new markers based on final objects were used for cell type annotation. In the nasal dataset, initial clustering identified neuronal clusters based on strong expression of Snap25, Rbfox3, and Tubb3. Neuronal and non-neuronal compartments were then independently subset and reprocessed to optimize subtype resolution. Cluster-specific marker genes were identified via differential expression testing. Final annotations were guided by reference to published adult mouse single-nucleus RNA-seq atlases. Annotated objects were saved for downstream tissue-specific and integrative analyses. Differential expression analysis To identify transcriptional differences associated with ZsGreen expression, differential expression analysis was performed using Memento 132 (v0.1.5), a model-aware framework optimized for single-cell RNA-seq count data. Analyses were conducted independently for each tissue (brain, dura, skull, and nasal) and within each annotated cell type. From each tissue-specific Seurat object, both normalized (data slot) and raw UMI count matrices (counts slot) were exported to AnnData format to preserve full expression and metadata information. For differential testing, only raw counts were used, consistent with Memento’s statistical model. Within each annotated cell type, cells were filtered to include only those labeled as ZsGreen+ or ZsGreen− in the ZsGreen metadata column. A binary label was created (0 = negative, 1 = positive) for use in the model. Genes not detected in at least 10% of cells in either group were excluded. Cell types with fewer than 10 cells in both ZsGreen+ and ZsGreen− groups were skipped. Differential expression was performed using Memento’s binary_test_1d function, with a fixed capture rate of 0.25 and 5,000 bootstrap iterations. Sparse count matrices were converted to compressed sparse row (CSR) format for memory efficiency. The test produced mean expression estimates and raw p-values, which were corrected for multiple testing using the Benjamini-Hochberg procedure. Genes with adjusted p-values < 0.05 were considered significantly differentially expressed. Monocle Trajectory and Module Analysis Trajectory analysis was conducted using Monocle3 133 – 135 on skull-derived B-lineage cells to model the transition from pre-B cells to B cells. Cells were embedded using UMAP, clustered, and ordered in pseudotime, with the trajectory rooted in the Pre B cell population. Pseudotime values were transferred back to the Seurat object. To assess immune regulation along the trajectory, a tolerogenic module score was computed based on expression of Ptpn22, Zbtb20, Cd274 genes 99 . Scores were visualized across pseudotime and compared between ZsGreen+ and - cells using statistical tests, including Wilcox. Statistical analysis Statistical analysis and data visualization were performed using GraphPad Prism or R. The specific statistical tests used, including corrections for multiple comparisons, are indicated in the respective figure legends. The number of biological replicates for each experiment is provided in the respective figure legends. In box plots, the horizontal bar shows the median. Error bars throughout represent standard error of the mean. Hierarchical clustering was performed in to identify clustering with the number of clusters (k) selected based on optimal silhouette score. For Gene Ontology (GO) analysis, minimum 10 or 20 group size and maximum of 3000 was used. Statistical significance is indicated in each figure legend. Supplementary Material 1 Movie S1. Whole head clearing and click-staining of Syn-PheRS animals confirms the presence of endogenous neuronal proteins in dorsal skull bone marrows niches. Download video file (83.9MB, mp4) 2 Figure S1. Validation of extracellular vesicle isolation and neuron-specific ZsGreen expression in brain tissue. Figure S1. Extracellular vesicle isolation paradigm and ZsGreen distribution in brain sections. (A) Schematic of biochemical strategy for extracellular vesicle (EV) and soluble fraction isolations from brain homogenates. (B-C) Full western blot of EV protein fractions from brain homogenates shows enrichment of EV marker Tsg101 in the EV fraction and ZsGreen detection in Syn-ZsG with no detection of ZsGreen in the WT controls. (C) Quantification of Tsg101 shows significant enrichment of EVs in EV fraction. (D-E) IHC analysis of brain sections from WT and Syn-ZsG mice shows DAPI, Cre, ZsG, Iba1, CD206, CD31, and GFAP staining confirming exclusive ZsGreen and Cre expression in Syn-ZsG. All data are mean ± SEM with n = 3–4. *p value = < 0.05; **p value = < 0.01; ***p value = < 0.001, ****p value = < 0.0001 by Student’s t test. NIHMS2173840-supplement-2.pdf (1.1MB, pdf) 3 Figure S2. Validation of Cre specificity across brain borders and of endogenous neuronal proteome-labeling paradigm. Figure S2. ZsGreen paradigm validated by tracing endogenous neuronal protein clearance. (A) IHC of brain, dura, nasal cavity, and superficial cervical lymph nodes (sCLNs) from Syn-ZsG mice confirms Cre-dependent ZsGreen expression restricted to neuronal compartments, with brain border tissues remaining Cre-negative. (B) Quantification of ZsGreen and OVA-647 colocalization with CD31 + vasculature in dural whole mounts using Mander’s correlation coefficients. While both tracers localize perisinusally, only ~40% of ZsGreen signal overlaps with OVA-647, indicating distinct distribution and uptake patterns. (C) Quantification of IHC signal intensity in dura, deep cervical lymph nodes (dCLNs), and nasal cavity (NS) demonstrates significantly elevated ZsGreen abundance in Syn-ZsG mice compared to WT controls. (D-E) Validation of the Syn-PheRS paradigm using click detection (AzF-647) of AzF-incorporated proteins in brain sections (D) and corresponding in-gel fluorescence analysis (E) of brain homogenates confirming significant labeling of Syn-PheRS brains. (F) Quantification of AzF-647 in-gel fluorescence intensity demonstrates significantly increased AzF-labeled protein abundance in Syn-PheRS mice compared to WT controls. (G) Dural whole mounts from WT and Syn-PheRS mice demonstrates the distribution of AzF-labeled neuronal proteins in Syn-PheRS mice paralleling anatomical patterns observed in the Syn-ZsG paradigm. (H) High-magnification views of dural whole mounts from Syn-PheRS mice show AzF-labeled neuronal proteins colocalized with CD206, CD31, and LYVE-1. (I) In-gel fluorescence analysis of dura homogenates confirms the presence of AzF-labeled neuronal proteins in Syn-PheRS compared to WT. (J-K) Representative in-gel fluorescence assay and quantification of AzF-647 labeled proteins from brain border homogenates confirms the detection of endogenously produced neuronal proteins at each brain border. (L) Experimental schematic illustrating chronic CSF-infusion of OVA-647 via osmotic pump and subsequent tissue collection. (M) Quantification comparing the proportion of ZsGreen or OVA at each border from Syn-ZsG versus chronically CSF-infused OVA-647, respectively. Plotting this data showed that chronic tracer delivery recapitulates a similar distribution observed after acute ICM-injection ( Fig. 2D ), with significantly more OVA at sCLNs and dCLNS. Sagittal sinus/bridging veins = SS/BV, caudal confluence of sinus = COS, rostral confluence of sinus = ROS, transverse sinus = TS. All data are mean ± SEM with n = 3–4. *p value = < 0.05; **p value = < 0.01; ***p value = < 0.001, ****p value = < 0.0001 by one-way ANOVA with Tukey’s post-hoc or Student’s t test. NIHMS2173840-supplement-3.pdf (2.1MB, pdf) 4 Figure S3. Spatial segregation and uptake of brain-derived proteins at brain border compartments. Figure S3. LSFM and IHC analyses reveals pattern of ZsGreen and ICM-injected Ovalbumin tracer. (A-C) IHC on coronal and sagittal nasal cavity sections highlights that despite both ZsGreen and OVA-647 being detected, the signals are not concentrated in similar anatomical regions. (i-iii) Zoom in of boxed regions. DMM = dorsal medial meatus, ET = Ethmoturbinates, LNG = lateral nasal glands, S = Septum. (D) Quantification of proportion of area that is exclusively ZsGreen positive across nasal cavity sections shows that ~85% of the total ZsGreen positive area does not colocalize with OVA-647. (E-F) LSFM of whole heads with the brain removed allows visualization of protein distribution across dura and skull compartments. Panel (E) shows ZsGreen drainage along the lateral and basal dura and skull, regions with distinct spatial differences compared to OVA-555. (F) LSFM additionally highlighted the retroorbital vein. (G) Flow cytometric analysis of all dural cell populations from WT and Syn-PheRS mice following click labeling of AzF-incorporated neuronal proteins (AzF-647). AzF-647 signal was quantified before (surface) and after permeabilization (intracellular + surface), revealing increased intracellular accumulation of neuron-derived proteins across dural cell populations in Syn-PheRS mice compared to WT. (H-I) Immunostaining of dorsal skull sections and in-gel fluorescence assay from WT and Syn-PheRS mice following click detection of AzF-labeled neuronal proteins (AzF-647). NIHMS2173840-supplement-4.pdf (2.3MB, pdf) 5 Figure S4. Representative flow cytometry gating strategy for dural cell populations. Figure S4. Representative flow cytometry gating of dural cell populations. (A) Representative flow cytometry-based gating of the cell types population in the dorsal dura. BECs: CD45(−)CD31+; LECs: CD45(−)CD31(−)PDPLN+; cDC1: CD45hiCD11b(−)CD11c+; cDC2: CD45hiCD11b+CD11c+MHCII+; CD206+ BAMs: CD45hiCD11b+CD206+; Monocytes: CD45hiCD11b+Ly6C+; Macs: CD45hiCD11b+F480+CD206(−). (B) IHC of dura whole mounts from Syn-ZsG mice highlighting CD206 colocalization with ZsGreen (Merges in Fig. 4D ). Sagittal sinus/bridging veins = SS, rostral confluence of sinus = RCS, transverse sinus = TS NIHMS2173840-supplement-5.pdf (354.6KB, pdf) 6 Figure S5. Single-cell transcriptomic profiling across dura, skull, and nasal cavity borders. Figure S5. Flow cytometry and scRNA-seq of ZsGreen distribution across dura, skull, and nasal borders. (A-C) UMAP of cell type clusters in (B) dorsal dura, (C) dorsal skull, and (D) nasal cavity. Cell types in each tissue are annotated by unique colors. (D) ZsGreen and Cre transcripts are not detectable across all cell types in dorsal dura, skull, and nasal cavity compared to housekeeping genes. Mean expression for genes is plotted. (E) Proportion of ZsGreen+ and ZsGreen− cells within each resident cell population in the dorsal dura, skull, and nasal based on scRNA-seq analysis. (F-G) Merged UMAP of macrophage annotated cells across dorsal dura, skull, and nasal borders showing distribution of ZsGreen+ and ZsGreen− cells. (H-I) Quantification of CD274 intensity at ZsGreen+ B cells compared to ZsGreen− B cells in dorsal skull section. Three sections each from n = 2. All data are mean ± SEM with n = 2–4. *p value = < 0.05; **p value = < 0.01; ***p value = < 0.001, ****p value = < 0.0001 by Student’s t test. NIHMS2173840-supplement-6.pdf (584.5KB, pdf) 7 Figure S6. Region-specific ZsGreen-labeling reveals spatially compartmentalized clearance and distinct routing compared to intraparenchymal tracer injection. Figure S6. Brain region-specific ZsGreen labeling reveals spatially compartmentalized protein clearance and distinct routing compared to intraparenchymal tracer injection. (A) Distance matrix of normalized ZsGreen proportion across brain border and drainage tissues comparing pan-neuronal (Syn-ZsG) and regionally restricted labeling paradigms (Forebrain-ZsG, Anterior-ZsG, and Striatal-ZsG). (B-E) LSFM imaging of CUBIC-cleared whole heads from Striatal-ZsG mice showing ZsGreen distribution, including lateral skull + dura (LSK + dura), and basal skull + dura (BSK + dura), and nasal cavity. (F) Representative IHC of brain sections following intraparenchymal injection of OVA-647 into the striatum (Str-Inj:OVA). Images show tracer localization in the striatum one hour post-injection and substantial clearance by 24 hours, confirming tracer clearance from the injection site. After one-hour, brain OVA-647 levels were quantified and compared to WT controls. (G) Heatmap comparing the proportion of ZsGreen (Str-ZsG) versus intraparenchymal OVA-647 (Str-Inj:OVA) detected across brain border and drainage tissues. In contrast to ZsGreen, intraparenchymal OVA shows relative underrepresentation at skull-associated borders and increased accumulation in superficial cervical lymph nodes (sCLNs). All data are mean ± SEM with n = 3–4. *p value = < 0.05; **p value = < 0.01; ***p value = < 0.001, ****p value = < 0.0001 by Student’s t test. NIHMS2173840-supplement-7.pdf (351.9KB, pdf) 8 Figure S7. Acute inflammation and amyloid pathology differentially redistribute brain-derived proteins across vascular, border, and blood compartments. Figure S7. LPS and 5XFAD alter brain-derived protein drainage (A) Quantification of Cadaverin-647 intensity from IHC images from Fig. 6B , confirms LPS-induced acute inflammation results in a comprised blood brain barrier as shown with the significantly elevated levels of Cadaverin-647 tracer in the brain after treatment compared to controls. (B-C) IHC analysis and quantification of brain sections from Syn-ZsG PBS and LPS treated mice shows significantly increased abundance of ZsGreen at CD31+ vessels. (D) WB analysis of ZsGreen in plasma EV and soluble fractions shows significantly increased ZsGreen signal in LPS-treated Syn-ZsG mice compared to controls. (E) Quantification of (C). ZsGreen from plasma EV and soluble fractions confirms significantly elevated ZsGreen levels after LPS. Total protein normalization was used for all comparisons. (F) IHC of dura whole mounts from WT and 5XFAD mice highlight the anatomical distribution of Aβ 42 . Zoom in of boxes shown in Fig. 6K . (G) Aβ 42 ELISA from brain border extracts from WT, 5XFAD-ZsG, and 5XFAD mice confirm that ZsGreen expression does not perturb the normal drainage routing of endogenous Aβ 42 . (H) 2-month-old WT-ZsG and 5XFAD-ZsG clearance of ZsGreen across brain borders quantified by fluorescent plate reader and normalized to respective brain ZsGreen. (I) ZsGreen abundance in blood plasma of 2-month-old 5XFAD mice was significantly increased compared to WT-ZsG healthy controls All data are mean ± SEM with n = 3–5. *p value = < 0.05; **p value = < 0.01; ***p value = < 0.001, ****p value = < 0.0001 by one-way ANOVA with Tukey’s post-hoc or Student’s t test. NIHMS2173840-supplement-8.pdf (1MB, pdf) 9 Table S1. List of Top 10 Markers used to annotate cell types for each cluster in dura, skull, and nasal. Related to Figure 4 . NIHMS2173840-supplement-9.xlsx (60.4KB, xlsx) 10 Table S2. Number of cells for each cell type in dura, skull, and nasal. Number of ZsGreen+ and – cells collected after flow cytometry. Related to Figure 4 . NIHMS2173840-supplement-10.xlsx (12.7KB, xlsx) 11 Table S3. Differentially expressed genes for each cell type in dura. Full dataset shown for all cell types which showed significantly modulated genes. Related to Figure 4 . NIHMS2173840-supplement-11.xlsx (902.6KB, xlsx) 12 Table S4. Differentially expressed genes for each cell type in skull. Full dataset shown for all cell types which showed significantly modulated genes. Related to Figure 4 . NIHMS2173840-supplement-12.xlsx (4.7MB, xlsx) 13 Table S5. Differentially expressed genes for each cell type in nasal. Full dataset shown for all cell types which showed significantly modulated genes. Related to Figure 4 . NIHMS2173840-supplement-13.xlsx (2.1MB, xlsx) Key Resources Table REAGENT or RESOURCE SOURCE IDENTIFIER Antibodies NeuN Abcam Cat#ab177487; RRID:AB_2532109 Cre recombinase Cell Systems Technologies Cat#15036S; RRID:AB_2798694 GFAP Thermo Fisher Scientific Cat#50-173-4108 Iba1 Wako Cat#019-19741; RRID:AB_839504 CD31 (PECAM-1) R&D Systems Cat#AF3628; RRID:AB_2161028 CD206 Thermo Fisher Scientific Cat#MA5-16868; RRID:AB_2538346 ACSA-2 Miltenyi Biotec Cat#130-116-243 TSG101 Santa Cruz Biotechnology Cat#sc-7964; RRID:AB_671392 Aβ42 Santa Cruz Biotechnology Cat#sc-28365; RRID:AB_626669 CD274 (PD-L1) Santa Cruz Biotechnology Cat#sc-28365; RRID:AB_467781 CD11b BioLegend Cat#101257; RRID:AB_2561373 Ly6C BioLegend Cat#128026; RRID:AB_10640120 LYVE-1 AngioBio Cat#11-034; RRID:AB_2813732 CD81 Santa Cruz Biotechnology Cat#sc-166029; RRID:AB_2275892 B220 Thermo Fisher Scientific Cat#19-0452-82; RRID:AB_467254 Bacterial and virus strains AAV-PHP.eB-hSyn-Cre.WPRE.hGH James M. Wilson; Addgene Addgene plasmid #105553; RRID:Addgene_105553 AiP14825 : pAAV-AiE0441h_3xC2-minBG-iCre(R297T)-BGHpA (Alias: CN4825), AAV-PHP.eB viral prep Allen Institute for Brain Science & Jonathan Ting; Addgene Addgene viral prep #214862-PHPeB; RRID:Addgene_214862 AiP11851 : pAAV-hDLX-minBG-iCre-4X2C-WPRE3-BGHpA (Alias: CN1851) viral prep Addgene Addgene viral prep #164450-PHPeB; RRID:Addgene_164450 AAV-PHP.eB-hSyn-ZsGreen-P2A-mCherry-NLS-SV40 This paper Chemicals, peptides, and recombinant proteins Alexa Fluor-conjugated ovalbumin Thermo Fisher Scientific Cat# O34784 ; O34782 g4-azido-L-phenylalanine (AzF) Vector Laboratories Cat#1406-5G DBCO-Alexa Fluor 647 Vector Laboratories CCT-1302-1 Cadaverine-Alexa Fluor 647 Invitrogen Cat#A30677 CUBIC-L TCI Cat#T3740 CUBIC-R+(N) TCI Cat#T3983 EasyIndex Matched Immersion Oil (RI:1.52) LifeCanvas Technologies Cat#Oil-1.52 Lipopolysaccharide (LPS) Sigma-Aldrich L2880A WGA-647 Fisher Scientific Cat# W32466 Copper (II) sulfate (CuSO 4 ) Sigma-Aldrich Cat# 451657-10G THPTA Vector Laboratories Cat# 1010-500 Aminoguanidine hydrochloride Sigma-Aldrich Cat# 396494-25G Sodium ascorbate (sodium L-ascorbate) TCI Cat# A0539-25G Iodoacetamide (IAA) Sigma-Aldrich Cat# I1149-25G LDS Sample Buffer Invitrogen Cat# NP0008 Critical commercial assays BCA Protein Assay Kit Thermo Scientific Cat#23225 Neural Tissue Dissociation Kit Miltenyi Biotec Cat#130-092-628 GEM-X Flex Sample Preparation v2 Kit 10x Genomics Cat#PN-1000781 Feature Barcoding Multiplexing Kit 10x Genomics N/A Experimental models: Organisms/strains C57BL/6J The Jackson Laboratory RRID:IMSR_JAX:000664 ZsGreen The Jackson Laboratory RRID:IMSR_JAX:007906 5XFAD The Jackson Laboratory RRID:MMRRC_034840-JAX CaMKII-Cre The Jackson Laboratory RRID:IMSR_JAX:005359 PheRS The Jackson Laboratory RRID:IMSR_JAX:033734 Forebrain-ZsG: CaMKII-Cre x ZsGreen This paper N/A Oligonucleotides Primers and labeled probes targeting AAV ITR sequence for qPCR titering This paper N/A Custom 10x Genomics probe set targeting Cre recombinase This paper N/A Custom 10x Genomics probe set targeting ZsGreen This paper N/A Software and algorithms Imaris v10.2.1 Oxford Instruments https://imaris.oxinst.com/ FlowJo FlowJo, LLC / BD Biosciences https://www.flowjo.com/ Cell Ranger 10x Genomics https://www.10xgenomics.com/support/software/cell-ranger/latest R (v4.5.1) R Foundation for Statistical Computing https://www.r-project.org/ (RRID:SCR_001905) Seurat v5.2.1 Satija Lab https://satijalab.org/seurat/ (RRID:SCR_016341) DoubletFinder (v 2.0.4) McGinnis & Gartner Lab https://github.com/chris-mcginnis-ucsf/DoubletFinder (RRID:SCR_018771) Harmony (v 1.2.3) Raychaudhuri Lab https://github.com/immunogenomics/harmony (RRID:SCR_022206) Memento v0.1.5 Ye Lab https://memento.readthedocs.io/ Monocle3 (v1.4.26) Cole Trapnell Lab https://cole-trapnell-lab.github.io/monocle3/ (RRID:SCR_018685) Python (v 3.12.0) Python Software Foundation https://www.python.org/downloads/source/ (RRID:SCR_008394) Scanpy (v 1.10.4) Theis Lab https://scanpy.readthedocs.io/en/stable/ (RRID:SCR_018139) Numpy (v 1.26.4) NumPy Community https://numpy.org/ (RRID:SCR_008633) Pandas (v 2.3.3) pandas Development Team https://pandas.pydata.org/ (RRID:SCR_018214) Other Stereotaxic frame (Model 940) Kopf Instruments Model 940 Hamilton syringe Hamilton Company 33G/31G needle VWR Insulin syringe Fisher Scientific Cat# 22-024-306 Osmotic pump (Model 2002; 0.50 μL/hr) Alzet Cat#0000296 Brain infusion cannula kit 3 (for ICV infusion) Alzet Cat#0008851 0.5 mm microburr Fine Science Tools Cat#19007-05 Open in a new tab Highlights. Neuronal protein tracing maps brain border clearance routes distinct from CSF tracers Skull borders clear slowly and harbor tolerogenic B cells that sample brain antigens Clearance is compartmentalized and follows a “nearest exit” principle by brain region Neuroinflammation and amyloid pathology disrupt clearance via distinct mechanisms Acknowledgements This research was supported by the National Institutes of Health (DP5OD033381), National Institute of Neurological Disorders and Stroke (1R01NS128909, 1RF1NS139975), Alzheimer’s Association (ADSF-24-1345199-C), Burroughs Wellcome Fund Career Awards at the Scientific Interface, Ludwig Family Foundation, Longevity Impetus Grant from Norn Group, UCSF Sandler Program for Breakthrough Biomedical Research New Frontier Research Award, and the Dolby Family. Y.C. is employed by Daiichi Sankyo Co., Ltd. and was a visiting scholar. M.B., A.E. and Z.I.K. were funded by the Deutsche Forschungsgemeinschaft (DFG) under Germany’s Excellence Strategy within Munich Cluster for Systems Neurology (EXC 2145 SyNergy - ID 390857198). We thank Mylinh Bernardi, Felicia Miller, and Horng-Ru Lin of the Gladstone Genomics Core for their assistance. We also thank participants of the 2024 ‘Brainwashing: Where Do We Stand on It?’ meeting and 2025 CSHL Brain Barriers meeting and members of the Yang laboratory for their feedback and support. Footnotes Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. Declaration of interests The authors declare no competing interests. References 1. Murphy JB, and Sturm E (1923). Conditions Determining the Transplantability of Tissues in the Brain. J Exp Med 38, 183–197. 10.1084/jem.38.2.183. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 2. Attwell D, and Laughlin SB (2001). An energy budget for signaling in the grey matter of the brain. J Cereb Blood Flow Metab 21, 1133–1145. 10.1097/00004647-200110000-00001. [ DOI ] [ PubMed ] [ Google Scholar ] 3. Harris JJ, Jolivet R, and Attwell D (2012). Synaptic energy use and supply. Neuron 75, 762–777. 10.1016/j.neuron.2012.08.019. [ DOI ] [ PubMed ] [ Google Scholar ] 4. Iliff JJ, Wang M, Liao Y, Plogg BA, Peng W, Gundersen GA, Benveniste H, Vates GE, Deane R, Goldman SA, et al. (2012). A paravascular pathway facilitates CSF flow through the brain parenchyma and the clearance of interstitial solutes, including amyloid beta. Sci Transl Med 4, 147ra111. 10.1126/scitranslmed.3003748. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 5. Aspelund A, Antila S, Proulx ST, Karlsen TV, Karaman S, Detmar M, Wiig H, and Alitalo K (2015). A dural lymphatic vascular system that drains brain interstitial fluid and macromolecules. J Exp Med 212, 991–999. 10.1084/jem.20142290. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 6. Louveau A, Smirnov I, Keyes TJ, Eccles JD, Rouhani SJ, Peske JD, Derecki NC, Castle D, Mandell JW, Lee KS, et al. (2015). Structural and functional features of central nervous system lymphatic vessels. Nature 523, 337–341. 10.1038/nature14432. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 7. Smets NG, van der Panne SA, Strijkers GJ, and Bakker E (2024). Perivascular spaces around arteries exceed perivenous spaces in the mouse brain. Sci Rep 14, 17655. 10.1038/s41598-024-67885-y. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 8. Bojarskaite L, Nafari S, Ravnanger AK, Frey MM, Skauli N, Abjorsbraten KS, Roth LC, Amiry-Moghaddam M, Nagelhus EA, Ottersen OP, et al. (2024). Role of aquaporin-4 polarization in extracellular solute clearance. Fluids Barriers CNS 21, 28. 10.1186/s12987-024-00527-7. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 9. Ahn JH, Cho H, Kim JH, Kim SH, Ham JS, Park I, Suh SH, Hong SP, Song JH, Hong YK, et al. (2019). Meningeal lymphatic vessels at the skull base drain cerebrospinal fluid. Nature 572, 62–66. 10.1038/s41586-019-1419-5. [ DOI ] [ PubMed ] [ Google Scholar ] 10. Oliver G, Kipnis J, Randolph GJ, and Harvey NL (2020). The Lymphatic Vasculature in the 21(st) Century: Novel Functional Roles in Homeostasis and Disease. Cell 182, 270–296. 10.1016/j.cell.2020.06.039. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 11. Hannocks MJ, Pizzo ME, Huppert J, Deshpande T, Abbott NJ, Thorne RG, and Sorokin L (2018). Molecular characterization of perivascular drainage pathways in the murine brain. J Cereb Blood Flow Metab 38, 669–686. 10.1177/0271678X17749689. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 12. Zhang Q, Niu Y, Li Y, Xia C, Chen Z, Chen Y, and Feng H (2025). Meningeal lymphatic drainage: novel insights into central nervous system disease. Signal Transduct Target Ther 10, 142. 10.1038/s41392-025-02177-z. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 13. Lun MP, Monuki ES, and Lehtinen MK (2015). Development and functions of the choroid plexus-cerebrospinal fluid system. Nat Rev Neurosci 16, 445–457. 10.1038/nrn3921. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 14. Engelhardt B, and Coisne C (2011). Fluids and barriers of the CNS establish immune privilege by confining immune surveillance to a two-walled castle moat surrounding the CNS castle. Fluids Barriers CNS 8, 4. 10.1186/2045-8118-8-4. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 15. Koh L, Zakharov A, and Johnston M (2005). Integration of the subarachnoid space and lymphatics: is it time to embrace a new concept of cerebrospinal fluid absorption? Cerebrospinal Fluid Res 2, 6. 10.1186/1743-8454-2-6. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 16. Kida S, Pantazis A, and Weller RO (1993). CSF drains directly from the subarachnoid space into nasal lymphatics in the rat. Anatomy, histology and immunological significance. Neuropathol Appl Neurobiol 19, 480–488. 10.1111/j.1365-2990.1993.tb00476.x. [ DOI ] [ PubMed ] [ Google Scholar ] 17. Upton ML, and Weller RO (1985). The morphology of cerebrospinal fluid drainage pathways in human arachnoid granulations. J Neurosurg 63, 867–875. 10.3171/jns.1985.63.6.0867. [ DOI ] [ PubMed ] [ Google Scholar ] 18. Rennels ML, Gregory TF, Blaumanis OR, Fujimoto K, and Grady PA (1985). Evidence for a ‘paravascular’ fluid circulation in the mammalian central nervous system, provided by the rapid distribution of tracer protein throughout the brain from the subarachnoid space. Brain Res 326, 47–63. 10.1016/0006-8993(85)91383-6. [ DOI ] [ PubMed ] [ Google Scholar ] 19. Bradbury MW, and Westrop RJ (1983). Factors influencing exit of substances from cerebrospinal fluid into deep cervical lymph of the rabbit. J Physiol 339, 519–534. 10.1113/jphysiol.1983.sp014731. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 20. Dani N, Herbst RH, Habib N, Head J, Dionne D, Nguyen L, McCabe C, Cui J, Shipley FB, Jang A, et al. (2019). A cellular and spatial map of the choroid plexus across brain ventricles and ages. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 21. Lehtinen MK, Bjornsson CS, Dymecki SM, Gilbertson RJ, Holtzman DM, and Monuki ES (2013). The choroid plexus and cerebrospinal fluid: emerging roles in development, disease, and therapy. J Neurosci 33, 17553–17559. 10.1523/JNEUROSCI.3258-13.2013. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 22. Pizzo ME, Wolak DJ, Kumar NN, Brunette E, Brunnquell CL, Hannocks MJ, Abbott NJ, Meyerand ME, Sorokin L, Stanimirovic DB, and Thorne RG (2018). Intrathecal antibody distribution in the rat brain: surface diffusion, perivascular transport and osmotic enhancement of delivery. J Physiol 596, 445–475. 10.1113/JP275105. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 23. Pollock H, Hutchings M, Weller RO, and Zhang ET (1997). Perivascular spaces in the basal ganglia of the human brain: their relationship to lacunes. J Anat 191 (Pt 3), 337–346. 10.1046/j.1469-7580.1997.19130337.x. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 24. Murdock MH, Yang CY, Sun N, Pao PC, Blanco-Duque C, Kahn MC, Kim T, Lavoie NS, Victor MB, Islam MR, et al. (2024). Multisensory gamma stimulation promotes glymphatic clearance of amyloid. Nature 627, 149–156. 10.1038/s41586-024-07132-6. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 25. Nedergaard M, and Goldman SA (2020). Glymphatic failure as a final common pathway to dementia. Science 370, 50–56. 10.1126/science.abb8739. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 26. Kress BT, Iliff JJ, Xia M, Wang M, Wei HS, Zeppenfeld D, Xie L, Kang H, Xu Q, Liew JA, et al. (2014). Impairment of paravascular clearance pathways in the aging brain. Ann Neurol 76, 845–861. 10.1002/ana.24271. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 27. Da Mesquita S, Louveau A, Vaccari A, Smirnov I, Cornelison RC, Kingsmore KM, Contarino C, Onengut-Gumuscu S, Farber E, Raper D, et al. (2018). Functional aspects of meningeal lymphatics in ageing and Alzheimer’s disease. Nature 560, 185–191. 10.1038/s41586-018-0368-8. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 28. Ding XB, Wang XX, Xia DH, Liu H, Tian HY, Fu Y, Chen YK, Qin C, Wang JQ, Xiang Z, et al. (2021). Impaired meningeal lymphatic drainage in patients with idiopathic Parkinson’s disease. Nat Med 27, 411–418. 10.1038/s41591-020-01198-1. [ DOI ] [ PubMed ] [ Google Scholar ] 29. Wen YR, Yang JH, Wang X, and Yao ZB (2018). Induced dural lymphangiogenesis facilities soluble amyloid-beta clearance from brain in a transgenic mouse model of Alzheimer’s disease. Neural Regen Res 13, 709–716. 10.4103/1673-5374.230299. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 30. Zamani A, Walker AK, Rollo B, Ayers KL, Farah R, O’Brien TJ, and Wright DK (2022). Impaired glymphatic function in the early stages of disease in a TDP-43 mouse model of amyotrophic lateral sclerosis. Transl Neurodegener 11, 17. 10.1186/s40035-022-00291-4. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 31. Wang L, Zhang Y, Zhao Y, Marshall C, Wu T, and Xiao M (2019). Deep cervical lymph node ligation aggravates AD-like pathology of APP/PS1 mice. Brain Pathol 29, 176–192. 10.1111/bpa.12656. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 32. Ma Q, Ineichen BV, Detmar M, and Proulx ST (2017). Outflow of cerebrospinal fluid is predominantly through lymphatic vessels and is reduced in aged mice. Nat Commun 8, 1434. 10.1038/s41467-017-01484-6. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 33. Benveniste H, Liu X, Koundal S, Sanggaard S, Lee H, and Wardlaw J (2019). The Glymphatic System and Waste Clearance with Brain Aging: A Review. Gerontology 65, 106–119. 10.1159/000490349. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 34. Voumvourakis KI, Sideri E, Papadimitropoulos GN, Tsantzali I, Hewlett P, Kitsos D, Stefanou M, Bonakis A, Giannopoulos S, Tsivgoulis G, and Paraskevas GP (2023). The Dynamic Relationship between the Glymphatic System, Aging, Memory, and Sleep. Biomedicines 11. 10.3390/biomedicines11082092. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 35. Mestre H, Mori Y, and Nedergaard M (2020). The Brain’s Glymphatic System: Current Controversies. Trends Neurosci 43, 458–466. 10.1016/j.tins.2020.04.003. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 36. Hablitz LM, and Nedergaard M (2021). The glymphatic system: A novel component of fundamental neurobiology. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 37. Louveau A, Plog BA, Antila S, Alitalo K, Nedergaard M, and Kipnis J (2017). Understanding the functions and relationships of the glymphatic system and meningeal lymphatics. Journal of Clinical Investigation 127, 3210–3219. 10.1172/JCI90603. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 38. Hauglund NL, Andersen M, Tokarska K, Radovanovic T, Kjaerby C, Sorensen FL, Bojarowska Z, Untiet V, Ballestero SB, Kolmos MG, et al. (2025). Norepinephrine-mediated slow vasomotion drives glymphatic clearance during sleep. Cell 188, 606–622 e617. 10.1016/j.cell.2024.11.027. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 39. Harrison IF, Ismail O, Machhada A, Colgan N, Ohene Y, Nahavandi P, Ahmed Z, Fisher A, Meftah S, Murray TK, et al. (2020). Impaired glymphatic function and clearance of tau in an Alzheimer’s disease model. Brain 143, 2576–2593. 10.1093/brain/awaa179. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 40. Peng W, Achariyar TM, Li B, Liao Y, Mestre H, Hitomi E, Regan S, Kasper T, Peng S, Ding F, et al. (2016). Suppression of glymphatic fluid transport in a mouse model of Alzheimer’s disease. Neurobiol Dis 93, 215–225. 10.1016/j.nbd.2016.05.015. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 41. Chen X, Liu X, Koundal S, Elkin R, Zhu X, Monte B, Xu F, Dai F, Pedram M, Lee H, et al. (2022). Cerebral amyloid angiopathy is associated with glymphatic transport reduction and time-delayed solute drainage along the neck arteries. Nat Aging 2, 214–223. 10.1038/s43587-022-00181-4. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 42. Xie L, Kang H, Xu Q, Chen MJ, Liao Y, Thiyagarajan M, O’Donnell J, Christensen DJ, Nicholson C, Iliff JJ, et al. (2013). Sleep drives metabolite clearance from the adult brain. Science 342, 373–377. 10.1126/science.1241224. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 43. Da Mesquita S, Papadopoulos Z, Dykstra T, Brase L, Farias FG, Wall M, Jiang H, Kodira CD, de Lima KA, Herz J, et al. (2021). Meningeal lymphatics affect microglia responses and anti-Abeta immunotherapy. Nature 593, 255–260. 10.1038/s41586-021-03489-0. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 44. Keil SA, Jansson D, Braun M, and Iliff JJ (2025). Glymphatic dysfunction in Alzheimer’s disease: A critical appraisal. Science 389, eadv8269. 10.1126/science.adv8269. [ DOI ] [ PubMed ] [ Google Scholar ] 45. Smyth LCD, and Kipnis J (2025). Redefining CNS immune privilege. Nat Rev Immunol. 10.1038/s41577-025-01175-0. [ DOI ] [ PubMed ] [ Google Scholar ] 46. Jacob L, de Brito Neto J, Lenck S, Corcy C, Benbelkacem F, Geraldo LH, Xu Y, Thomas JM, El Kamouh MR, Spajer M, et al. (2022). Conserved meningeal lymphatic drainage circuits in mice and humans. J Exp Med 219. 10.1084/jem.20220035. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 47. Ma Q, Ries M, Decker Y, Muller A, Riner C, Bucker A, Fassbender K, Detmar M, and Proulx ST (2019). Rapid lymphatic efflux limits cerebrospinal fluid flow to the brain. Acta Neuropathol 137, 151–165. 10.1007/s00401-018-1916-x. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 48. Albargothy NJ, Johnston DA, MacGregor-Sharp M, Weller RO, Verma A, Hawkes CA, and Carare RO (2018). Convective influx/glymphatic system: tracers injected into the CSF enter and leave the brain along separate periarterial basement membrane pathways. Acta Neuropathol 136, 139–152. 10.1007/s00401-018-1862-7. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 49. Ringstad G, Valnes LM, Dale AM, Pripp AH, Vatnehol SS, Emblem KE, Mardal KA, and Eide PK (2018). Brain-wide glymphatic enhancement and clearance in humans assessed with MRI. JCI Insight 3. 10.1172/jci.insight.121537. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 50. Melin E, Eide PK, and Ringstad G (2020). In vivo assessment of cerebrospinal fluid efflux to nasal mucosa in humans. Sci Rep 10, 14974. 10.1038/s41598-020-72031-5. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 51. van Osch MJP, Wahlin A, Scheyhing P, Mossige I, Hirschler L, Eklund A, Mogensen K, Gomolka R, Radbruch A, Qvarlander S, et al. (2024). Human brain clearance imaging: Pathways taken by magnetic resonance imaging contrast agents after administration in cerebrospinal fluid and blood. NMR Biomed 37, e5159. 10.1002/nbm.5159. [ DOI ] [ PubMed ] [ Google Scholar ] 52. Kwon S, Janssen CF, Velasquez FC, and Sevick-Muraca EM (2017). Fluorescence imaging of lymphatic outflow of cerebrospinal fluid in mice. J Immunol Methods 449, 37–43. 10.1016/j.jim.2017.06.010. [ DOI ] [ PubMed ] [ Google Scholar ] 53. Jiang-Xie LF, Drieu A, Bhasiin K, Quintero D, Smirnov I, and Kipnis J (2024). Neuronal dynamics direct cerebrospinal fluid perfusion and brain clearance. Nature 627, 157–164. 10.1038/s41586-024-07108-6. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 54. Jin H, Yoon JH, Hong SP, Hwang YS, Yang MJ, Choi J, Kang HJ, Baek SE, Jin C, Jung J, et al. (2025). Increased CSF drainage by non-invasive manipulation of cervical lymphatics. Nature. 10.1038/s41586-025-09052-5. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 55. Proulx ST (2021). Cerebrospinal fluid outflow: a review of the historical and contemporary evidence for arachnoid villi, perineural routes, and dural lymphatics. Cell Mol Life Sci 78, 2429–2457. 10.1007/s00018-020-03706-5. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 56. Zhou Y, Ran W, Luo Z, Wang J, Fang M, Wei K, Sun J, and Lou M (2022). Impaired peri-olfactory cerebrospinal fluid clearance is associated with ageing, cognitive decline and dyssomnia. EBioMedicine 86, 104381. 10.1016/j.ebiom.2022.104381. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 57. Rustenhoven J, Drieu A, Mamuladze T, de Lima KA, Dykstra T, Wall M, Papadopoulos Z, Kanamori M, Salvador AF, Baker W, et al. (2021). Functional characterization of the dural sinuses as a neuroimmune interface. Cell. 10.1016/j.cell.2020.12.040. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 58. Fitzpatrick Z, Ghabdan Zanluqui N, Rosenblum JS, Tuong ZK, Lee CYC, Chandrashekhar V, Negro-Demontel ML, Stewart AP, Posner DA, Buckley M, et al. (2024). Venous-plexus-associated lymphoid hubs support meningeal humoral immunity. Nature. 10.1038/s41586-024-07202-9. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 59. Yoon J-H, Jin H, Kim HJ, Hong SP, Yang MJ, Ahn JH, Kim Y-C, Seo J, Lee Y, McDonald DM, et al. (2024). Nasopharyngeal lymphatic plexus is a hub for cerebrospinal fluid drainage. Nature 625, 768–777. 10.1038/s41586-023-06899-4. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 60. Herisson F, Frodermann V, Courties G, Rohde D, Sun Y, Vandoorne K, Wojtkiewicz GR, Masson GS, Vinegoni C, Kim J, et al. (2018). Direct vascular channels connect skull bone marrow and the brain surface enabling myeloid cell migration. Nature Neuroscience. 10.1038/s41593-018-0213-2. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 61. Mazzitelli JA, Smyth LCD, Cross KA, Dykstra T, Sun J, Du S, Mamuladze T, Smirnov I, Rustenhoven J, and Kipnis J (2022). Cerebrospinal fluid regulates skull bone marrow niches via direct access through dural channels. Nat Neurosci 25, 555–560. 10.1038/s41593-022-01029-1. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 62. Kolabas ZI, Kuemmerle LB, Perneczky R, Förstera B, Ulukaya S, Ali M, Kapoor S, Bartos LM, Büttner M, Caliskan OS, et al. (2023). Distinct molecular profiles of skull bone marrow in health and neurological disorders. Cell 186, 3706–3725.e3729. 10.1016/j.cell.2023.07.009. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 63. Brioschi S, Wang W-L, Peng V, Wang M, Shchukina I, Greenberg ZJ, Bando JK, Jaeger N, Czepielewski RS, Swain A, et al. (2021). Heterogeneity of meningeal B cells reveals a lymphopoietic niche at the CNS borders. Science. 10.1126/science.abf9277. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 64. Cugurra A, Mamuladze T, Rustenhoven J, Dykstra T, Beroshvili G, Greenberg ZJ, Baker W, Papadopoulos Z, Drieu A, Blackburn S, et al. (2021). Skull and vertebral bone marrow are myeloid cell reservoirs for the meninges and CNS parenchyma. Science. 10.1126/science.abf7844. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 65. Norwood JN, Zhang Q, Card D, Craine A, Ryan TM, and Drew PJ (2019). Anatomical basis and physiological role of cerebrospinal fluid transport through the murine cribriform plate. Elife 8. 10.7554/eLife.44278. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 66. Weller RO, Sharp MM, Christodoulides M, Carare RO, and Mollgård K (2018). The meninges as barriers and facilitators for the movement of fluid, cells and pathogens related to the rodent and human CNS. Acta Neuropathologica 135, 363–385. 10.1007/s00401-018-1809-z. [ DOI ] [ PubMed ] [ Google Scholar ] 67. Laman JD, and Weller RO (2013). Drainage of cells and soluble antigen from the CNS to regional lymph nodes. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 68. Koh L, Zakharov A, Nagra G, Armstrong D, Friendship R, and Johnston M (2006). Development of cerebrospinal fluid absorption sites in the pig and rat: connections between the subarachnoid space and lymphatic vessels in the olfactory turbinates. Anat Embryol 211, 335–344. 10.1007/s00429-006-0085-1. [ DOI ] [ PubMed ] [ Google Scholar ] 69. Johnston M, Zakharov A, Koh L, and Armstrong D (2005). Subarachnoid injection of Microfil reveals connections between cerebrospinal fluid and nasal lymphatics in the non-human primate. Neuropath Appl Neuro 31, 632–640. 10.1111/j.1365-2990.2005.00679.x. [ DOI ] [ PubMed ] [ Google Scholar ] 70. Mollanji R, Bozanovic-Sosic R, Zakharov A, Makarian L, and Johnston MG (2002). Blocking cerebrospinal fluid absorption through the cribriform plate increases resting intracranial pressure. Am J Physiol-Reg I 282, R1593–R1599. 10.1152/ajpregu.00695.2001. [ DOI ] [ PubMed ] [ Google Scholar ] 71. Walter BA, Valera VA, Takahashi S, and Ushiki T (2006). The olfactory route for cerebrospinal fluid drainage into the peripheral lymphatic system. Neuropathol Appl Neurobiol 32, 388–396. 10.1111/j.1365-2990.2006.00737.x. [ DOI ] [ PubMed ] [ Google Scholar ] 72. Ma QL, Ineichen BV, Detmar M, and Proulx ST (2017). Outflow of cerebrospinal fluid is predominantly through lymphatic vessels and is reduced in aged mice. Nature Communications 8. ARTN 1434 10.1038/s41467-017-01484-6. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 73. Yoon JH, Jin H, Kim HJ, Hong SP, Yang MJ, Ahn JH, Kim YC, Seo J, Lee Y, McDonald DM, et al. (2024). Nasopharyngeal lymphatic plexus is a hub for cerebrospinal fluid drainage. Nature 625, 768–777. 10.1038/s41586-023-06899-4. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 74. Decker Y, Kramer J, Xin L, Muller A, Scheller A, Fassbender K, and Proulx ST (2022). Magnetic resonance imaging of cerebrospinal fluid outflow after low-rate lateral ventricle infusion in mice. JCI Insight 7. 10.1172/jci.insight.150881. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 75. Raghunandan A, Ladron-de-Guevara A, Tithof J, Mestre H, Du T, Nedergaard M, Thomas JH, and Kelley DH (2021). Bulk flow of cerebrospinal fluid observed in periarterial spaces is not an artifact of injection. Elife 10. 10.7554/eLife.65958. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 76. Gahn-Martinez D, Giannetto M, Chang E, Beam N, Tobin P, Pla V, and Nedergaard M (2025). Chronic Intraventricular Cannulation for the Study of Glymphatic 1473 Transport. eNeuro 12. 10.1523/ENEURO.0537-24.2025. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 77. Kiviniemi V, Wang X, Korhonen V, Keinanen T, Tuovinen T, Autio J, LeVan P, Keilholz S, Zang YF, Hennig J, and Nedergaard M (2016). Ultra-fast magnetic resonance encephalography of physiological brain activity - Glymphatic pulsation mechanisms? J Cereb Blood Flow Metab 36, 1033–1045. 10.1177/0271678X15622047. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 78. Betsholtz C, Engelhardt B, Koh GY, McDonald DM, Proulx ST, and Siegenthaler J (2024). Advances and controversies in meningeal biology. Nat Neurosci 27, 2056–2072. 10.1038/s41593-024-01701-8. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 79. Papadopoulos Z, Smyth LCD, Smirnov I, Gibson DA, Herz J, and Kipnis J (2025). Differential impact of lymphatic outflow pathways on cerebrospinal fluid homeostasis. J Exp Med 222. 10.1084/jem.20241752. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 80. Ayyappan K, Unger L, Kitchen P, Bill RM, and Salman MM (2026). Measuring glymphatic function: Assessing the toolkit. Neural Regen Res 21, 534–541. 10.4103/NRR.NRR-D-24-01013. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 81. Yang L, Kress BT, Weber HJ, Thiyagarajan M, Wang B, Deane R, Benveniste H, Iliff JJ, and Nedergaard M (2013). Evaluating glymphatic pathway function utilizing clinically relevant intrathecal infusion of CSF tracer. J Transl Med 11, 107. 10.1186/1479-5876-11-107. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 82. Hablitz LM, Vinitsky HS, Sun Q, Staeger FF, Sigurdsson B, Mortensen KN, Lilius TO, and Nedergaard M (2019). Increased glymphatic influx is correlated with high EEG delta power and low heart rate in mice under anesthesia. Sci Adv 5, eaav5447. 10.1126/sciadv.aav5447. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 83. Ruhland MK, Roberts EW, Cai E, Mujal AM, Marchuk K, Beppler C, Nam D, Serwas NK, Binnewies M, and Krummel MF (2020). Visualizing Synaptic Transfer of Tumor Antigens among Dendritic Cells. Cancer Cell 37, 786–799 e785. 10.1016/j.ccell.2020.05.002. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 84. Nakamura Y, Ishii J, and Kondo A (2013). Bright fluorescence monitoring system utilizing Zoanthus sp. green fluorescent protein (ZsGreen) for human G-protein-coupled receptor signaling in microbial yeast cells. PLoS One 8, e82237. 10.1371/journal.pone.0082237. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 85. Roberts EW, Broz ML, Binnewies M, Headley MB, Nelson AE, Wolf DM, Kaisho T, Bogunovic D, Bhardwaj N, and Krummel MF (2016). Critical Role for CD103(+)/CD141(+) Dendritic Cells Bearing CCR7 for Tumor Antigen Trafficking and Priming of T Cell Immunity in Melanoma. Cancer Cell 30, 324–336. 10.1016/j.ccell.2016.06.003. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 86. Kersten K, You R, Liang S, Tharp KM, Pollack J, Weaver VM, Krummel MF, and Headley MB (2023). Uptake of tumor-derived microparticles induces metabolic reprogramming of macrophages in the early metastatic lung. Cell Rep 42, 112582. 10.1016/j.celrep.2023.112582. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 87. Yi R, Chen E, Roberts EW, Krummel MF, and Serwas NK (2022). Impact of protein identity on tumor-associated antigen uptake into infiltrating immune cells: A comparison of different fluorescent proteins as model antigens. PLoS One 17, e0272857. 10.1371/journal.pone.0272857. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 88. Szatanek R, Baran J, Siedlar M, and Baj-Krzyworzeka M (2015). Isolation of extracellular vesicles: Determining the correct approach (Review). Int J Mol Med 36, 11–17. 10.3892/ijmm.2015.2194. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 89. Matamoros-Angles A, Karadjuzovic E, Mohammadi B, Song F, Brenna S, Meister SC, Siebels B, Voss H, Seuring C, Ferrer I, et al. (2024). Efficient enzyme-free isolation of brain-derived extracellular vesicles. J Extracell Vesicles 13, e70011. 10.1002/jev2.70011. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 90. Mestre H, Hablitz LM, Xavier AL, Feng W, Zou W, Pu T, Monai H, Murlidharan G, Castellanos Rivera RM, Simon MJ, et al. (2018). Aquaporin-4-dependent glymphatic solute transport in the rodent brain. Elife 7. 10.7554/eLife.40070. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 91. Smyth LCD, Xu D, Okar SV, Dykstra T, Rustenhoven J, Papadopoulos Z, Bhasiin K, Kim MW, Drieu A, Mamuladze T, et al. (2024). Identification of direct connections between the dura and the brain. Nature. 10.1038/s41586-023-06993-7. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 92. Yang AC, du Bois H, Olsson N, Gate D, Lehallier B, Berdnik D, Brewer KD, Bertozzi CR, Elias JE, and Wyss-Coray T (2018). Multiple Click-Selective tRNA Synthetases Expand Mammalian Cell-Specific Proteomics. J Am Chem Soc 140, 7046–7051. 10.1021/jacs.8b03074. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 93. Pietila R, Del Gaudio F, He L, Vazquez-Liebanas E, Vanlandewijck M, Muhl L, Mocci G, Bjornholm KD, Lindblad C, Fletcher-Sandersjoo A, et al. (2023). Molecular anatomy of adult mouse leptomeninges. Neuron 111, 3745–3764 e3747. 10.1016/j.neuron.2023.09.002. [ DOI ] [ PubMed ] [ Google Scholar ] 94. DeSisto J, O’Rourke R, Jones HE, Pawlikowski B, Malek AD, Bonney S, Guimiot F, Jones KL, and Siegenthaler JA (2020). Single-Cell Transcriptomic Analyses of the Developing Meninges Reveal Meningeal Fibroblast Diversity and Function. Developmental Cell. 10.1016/j.devcel.2020.06.009. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 95. Van Hove H, Martens L, Scheyltjens I, De Vlaminck K, Pombo Antunes AR, De Prijck S, Vandamme N, De Schepper S, Van Isterdael G, Scott CL, et al. (2019). A single-cell atlas of mouse brain macrophages reveals unique transcriptional identities shaped by ontogeny and tissue environment. Nature Neuroscience. 10.1038/s41593-019-0393-4. [ DOI ] [ PubMed ] [ Google Scholar ] 96. Rustenhoven J, and Kipnis J (2022). Brain borders at the central stage of neuroimmunology. Nature 612, 417–429. 10.1038/s41586-022-05474-7. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 97. Wang Y, Chen D, Xu D, Huang C, Xing R, He D, and Xu H (2021). Early developing B cells undergo negative selection by central nervous system-specific antigens in the meninges. Immunity 54, 2784–2794 e2786. 10.1016/j.immuni.2021.09.016. [ DOI ] [ PubMed ] [ Google Scholar ] 98. Schafflick D, Wolbert J, Heming M, Thomas C, Hartlehnert M, Borsch AL, Ricci A, Martin-Salamanca S, Li X, Lu IN, et al. (2021). Single-cell profiling of CNS border compartment leukocytes reveals that B cells and their progenitors reside in non-diseased meninges. Nat Neurosci 24, 1225–1234. 10.1038/s41593-021-00880-y. [ DOI ] [ PubMed ] [ Google Scholar ] 99. Yang SY, Long J, Huang MX, Luo PY, Bian ZH, Xu YF, Wang CB, Yang SH, Li L, Selmi C, et al. (2021). Characterization of Organ-Specific Regulatory B Cells Using Single-Cell RNA Sequencing. Front Immunol 12, 711980. 10.3389/fimmu.2021.711980. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 100. Graybuck LT, Daigle TL, Sedeno-Cortes AE, Walker M, Kalmbach B, Lenz GH, Morin E, Nguyen TN, Garren E, Bendrick JL, et al. (2021). Enhancer viruses for combinatorial cell-subclass-specific labeling. Neuron 109, 1449–1464 e1413. 10.1016/j.neuron.2021.03.011. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 101. Hunker AC, Wirthlin ME, Gill G, Johansen NJ, Hooper M, Omstead V, Vargas S, Lerma MN, Taskin N, Weed N, et al. (2025). Enhancer AAV toolbox for accessing and perturbing striatal cell types and circuits. Neuron 113, 1507–1524 e1517. 10.1016/j.neuron.2025.04.035. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 102. Oakley H, Cole SL, Logan S, Maus E, Shao P, Craft J, Guillozet-Bongaarts A, Ohno M, Disterhoft J, Van Eldik L, et al. (2006). Intraneuronal beta-amyloid aggregates, neurodegeneration, and neuron loss in transgenic mice with five familial Alzheimer’s disease mutations: potential factors in amyloid plaque formation. J Neurosci 26, 10129–10140. 10.1523/JNEUROSCI.1202-06.2006. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 103. Banks WA, Gray AM, Erickson MA, Salameh TS, Damodarasamy M, Sheibani N, Meabon JS, Wing EE, Morofuji Y, Cook DG, and Reed MJ (2015). Lipopolysaccharide-induced blood-brain barrier disruption: Roles of cyclooxygenase, oxidative stress, neuroinflammation, and elements of the neurovascular unit. Journal of Neuroinflammation. 10.1186/s12974-015-0434-1. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 104. Zhao J, Bi W, Xiao S, Lan X, Cheng X, Zhang J, Lu D, Wei W, Wang Y, Li H, et al. (2019). Neuroinflammation induced by lipopolysaccharide causes cognitive impairment in mice. Sci Rep 9, 5790. 10.1038/s41598-019-42286-8. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 105. Smyth LCD, Verhaege D, Standen-Bloom E, Wu Y, Gan Y, Storck SE, Boskovic P, Plog BA, Mamuladze T, Mazzitelli JA, et al. (2026). Amyloidosis of bridging veins is a pathologic feature of Alzheimer’s disease. J Exp Med 223. 10.1084/jem.20251860. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 106. Xue Y, Liu X, Koundal S, Constantinou S, Dai F, Santambrogio L, Lee H, and Benveniste H (2020). In vivo T1 mapping for quantifying glymphatic system transport and cervical lymph node drainage. Sci Rep 10, 14592. 10.1038/s41598-020-71582-x. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 107. Sarker A, Suh M, Choi Y, Park JY, Lee YS, and Lee DS (2023). Intrathecal [(64)Cu]Cu-albumin PET reveals age-related decline of lymphatic drainage of cerebrospinal fluid. Sci Rep 13, 12930. 10.1038/s41598-023-39903-y. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 108. Kim MW, Gao W, Lichti CF, Gu X, Dykstra T, Cao J, Smirnov I, Boskovic P, Kleverov D, Salvador AFM, et al. (2025). Endogenous self-peptides guard immune privilege of the central nervous system. Nature 637, 176–183. 10.1038/s41586-024-08279-y. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 109. Rakic P (2009). Evolution of the neocortex: a perspective from developmental biology. Nat Rev Neurosci 10, 724–735. 10.1038/nrn2719. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 110. Buckner RL, and Krienen FM (2013). The evolution of distributed association networks in the human brain. Trends Cogn Sci 17, 648–665. 10.1016/j.tics.2013.09.017. [ DOI ] [ PubMed ] [ Google Scholar ] 111. Ringstad G, and Eide PK (2020). Cerebrospinal fluid tracer efflux to parasagittal dura in humans. Nat Commun 11, 354. 10.1038/s41467-019-14195-x. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 112. Absinta M, Ha SK, Nair G, Sati P, Luciano NJ, Palisoc M, Louveau A, Zaghloul KA, Pittaluga S, Kipnis J, and Reich DS (2017). Human and nonhuman primate meninges harbor lymphatic vessels that can be visualized noninvasively by MRI. Elife 6. 10.7554/eLife.29738. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 113. Mrdjen D, Pavlovic A, Hartmann FJ, Schreiner B, Utz SG, Leung BP, Lelios I, Heppner FL, Kipnis J, Merkler D, et al. (2018). High-Dimensional Single-Cell Mapping of Central Nervous System Immune Cells Reveals Distinct Myeloid Subsets in Health, Aging, and Disease. Immunity 48, 599. 10.1016/j.immuni.2018.02.014. [ DOI ] [ PubMed ] [ Google Scholar ] 114. Vara-Perez M, and Movahedi K (2025). Border-associated macrophages as gatekeepers of brain homeostasis and immunity. Immunity 58, 1085–1100. 10.1016/j.immuni.2025.04.005. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 115. Kierdorf K, Masuda T, Jordao MJC, and Prinz M (2019). Macrophages at CNS interfaces: ontogeny and function in health and disease. Nat Rev Neurosci 20, 547–562. 10.1038/s41583-019-0201-x. [ DOI ] [ PubMed ] [ Google Scholar ] 116. Min H, O’Neil SM, Xu L, Moseman EA, Kurtzberg J, and Filiano AJ (2024). Mural cells interact with macrophages in the dura mater to regulate CNS immune surveillance. J Exp Med 221. 10.1084/jem.20230326. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 117. Yang M, Rui K, Wang S, and Lu L (2013). Regulatory B cells in autoimmune diseases. Cell Mol Immunol 10, 122–132. 10.1038/cmi.2012.60. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 118. Brioschi S, Wang WL, Peng V, Wang M, Shchukina I, Greenberg ZJ, Bando JK, Jaeger N, Czepielewski RS, Swain A, et al. (2021). Heterogeneity of meningeal B cells reveals a lymphopoietic niche at the CNS borders. Science 373. 10.1126/science.abf9277. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 119. Liu L, Zhang X, Chai Y, Zhang J, Deng Q, and Chen X (2025). Skull bone marrow and skull meninges channels: redefining the landscape of central nervous system immune surveillance. Cell Death Dis 16, 53. 10.1038/s41419-025-07336-2. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 120. Eme-Scolan E, Gomes M, Bani A, Romano A, Kassem O, Roussel-Queval A, Casel B, Slimani L, Lawrence T, and Rua R (2026). Remodeling of skull bone channels regulates immune infiltration into the meninges during neuroinflammation. Immunity 59, 116–128 e117. 10.1016/j.immuni.2025.11.019. [ DOI ] [ PubMed ] [ Google Scholar ] 121. Rustenhoven J, Pavlou G, Storck SE, Dykstra T, Du S, Wan Z, Quintero D, Scallan JP, Smirnov I, Kamm RD, and Kipnis J (2023). Age-related alterations in meningeal immunity drive impaired CNS lymphatic drainage. J Exp Med 220. 10.1084/jem.20221929. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 122. Plog BA, Dashnaw ML, Hitomi E, Peng W, Liao Y, Lou N, Deane R, and Nedergaard M (2015). Biomarkers of traumatic injury are transported from brain to blood via the glymphatic system. J Neurosci 35, 518–526. 10.1523/JNEUROSCI.3742-14.2015. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 123. Arbel-Ornath M, Hudry E, Eikermann-Haerter K, Hou S, Gregory JL, Zhao L, Betensky RA, Frosch MP, Greenberg SM, and Bacskai BJ (2013). Interstitial fluid drainage is impaired in ischemic stroke and Alzheimer’s disease mouse models. Acta Neuropathol 126, 353–364. 10.1007/s00401-013-1145-2. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 124. Guldner IH, Wagner VP, Moran-Losada P, Shi SM, Golub SW, Hevler JF, Chen K, Meese BT, Ghoochani A, Pulido E, et al. (2026). Ageing promotes microglial accumulation of slow-degrading synaptic proteins. Nature. 10.1038/s41586-025-09987-9. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 125. Roussel-Queval A, Rebejac J, Eme-Scolan E, Paroutaud LA, and Rua R (2023). Flow cytometry and immunohistochemistry of the mouse dural meninges for immunological and virological assessments. STAR Protoc 4, 102119. 10.1016/j.xpro.2023.102119. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 126. Upadhyay A, Chhangani D, Rao NR, Kofler J, Vassar R, Rincon-Limas DE, and Savas JN (2023). Amyloid fibril proteomics of AD brains reveals modifiers of aggregation and toxicity. Mol Neurodegener 18, 61. 10.1186/s13024-023-00654-z. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 127. Stuart T, Butler A, Hoffman P, Hafemeister C, Papalexi E, Mauck WM 3rd, Hao Y, Stoeckius M, Smibert P, and Satija R (2019). Comprehensive Integration of Single-Cell Data. Cell 177, 1888–1902 e1821. 10.1016/j.cell.2019.05.031. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 128. Satija R, Farrell JA, Gennert D, Schier AF, and Regev A (2015). Spatial reconstruction of single-cell gene expression data. Nat Biotechnol 33, 495–502. 10.1038/nbt.3192. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 129. Hao Y, Stuart T, Kowalski MH, Choudhary S, Hoffman P, Hartman A, Srivastava A, Molla G, Madad S, Fernandez-Granda C, and Satija R (2024). Dictionary learning for integrative, multimodal and scalable single-cell analysis. Nat Biotechnol 42, 293–304. 10.1038/s41587-023-01767-y. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 130. McGinnis CS, Murrow LM, and Gartner ZJ (2019). DoubletFinder: Doublet Detection in Single-Cell RNA Sequencing Data Using Artificial Nearest Neighbors. Cell Syst 8, 329–337 e324. 10.1016/j.cels.2019.03.003. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 131. Korsunsky I, Millard N, Fan J, Slowikowski K, Zhang F, Wei K, Baglaenko Y, Brenner M, Loh PR, and Raychaudhuri S (2019). Fast, sensitive and accurate integration of single-cell data with Harmony. Nat Methods 16, 1289–1296. 10.1038/s41592-019-0619-0. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 132. Kim MC, Gate R, Lee DS, Tolopko A, Lu A, Gordon E, Shifrut E, Garcia-Nieto PE, Marson A, Ntranos V, and Ye CJ (2024). Method of moments framework for differential expression analysis of single-cell RNA sequencing data. Cell 187, 6393–6410 e6316. 10.1016/j.cell.2024.09.044. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 133. Levine JH, Simonds EF, Bendall SC, Davis KL, Amir el AD, Tadmor MD, Litvin O, Fienberg HG, Jager A, Zunder ER, et al. (2015). Data-Driven Phenotypic Dissection of AML Reveals Progenitor-like Cells that Correlate with Prognosis. Cell 162, 184–197. 10.1016/j.cell.2015.05.047. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 134. Cao J, Packer JS, Ramani V, Cusanovich DA, Huynh C, Daza R, Qiu X, Lee C, Furlan SN, Steemers FJ, et al. (2017). Comprehensive single-cell transcriptional profiling of a multicellular organism. Science 357, 661–667. 10.1126/science.aam8940. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 135. Trapnell C, Cacchiarelli D, Grimsby J, Pokharel P, Li S, Morse M, Lennon NJ, Livak KJ, Mikkelsen TS, and Rinn JL (2014). The dynamics and regulators of cell fate decisions are revealed by pseudotemporal ordering of single cells. Nat Biotechnol 32, 381–386. 10.1038/nbt.2859. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials 1 Movie S1. Whole head clearing and click-staining of Syn-PheRS animals confirms the presence of endogenous neuronal proteins in dorsal skull bone marrows niches. Download video file (83.9MB, mp4) 2 Figure S1. Validation of extracellular vesicle isolation and neuron-specific ZsGreen expression in brain tissue. Figure S1. Extracellular vesicle isolation paradigm and ZsGreen distribution in brain sections. (A) Schematic of biochemical strategy for extracellular vesicle (EV) and soluble fraction isolations from brain homogenates. (B-C) Full western blot of EV protein fractions from brain homogenates shows enrichment of EV marker Tsg101 in the EV fraction and ZsGreen detection in Syn-ZsG with no detection of ZsGreen in the WT controls. (C) Quantification of Tsg101 shows significant enrichment of EVs in EV fraction. (D-E) IHC analysis of brain sections from WT and Syn-ZsG mice shows DAPI, Cre, ZsG, Iba1, CD206, CD31, and GFAP staining confirming exclusive ZsGreen and Cre expression in Syn-ZsG. All data are mean ± SEM with n = 3–4. *p value = < 0.05; **p value = < 0.01; ***p value = < 0.001, ****p value = < 0.0001 by Student’s t test. NIHMS2173840-supplement-2.pdf (1.1MB, pdf) 3 Figure S2. Validation of Cre specificity across brain borders and of endogenous neuronal proteome-labeling paradigm. Figure S2. ZsGreen paradigm validated by tracing endogenous neuronal protein clearance. (A) IHC of brain, dura, nasal cavity, and superficial cervical lymph nodes (sCLNs) from Syn-ZsG mice confirms Cre-dependent ZsGreen expression restricted to neuronal compartments, with brain border tissues remaining Cre-negative. (B) Quantification of ZsGreen and OVA-647 colocalization with CD31 + vasculature in dural whole mounts using Mander’s correlation coefficients. While both tracers localize perisinusally, only ~40% of ZsGreen signal overlaps with OVA-647, indicating distinct distribution and uptake patterns. (C) Quantification of IHC signal intensity in dura, deep cervical lymph nodes (dCLNs), and nasal cavity (NS) demonstrates significantly elevated ZsGreen abundance in Syn-ZsG mice compared to WT controls. (D-E) Validation of the Syn-PheRS paradigm using click detection (AzF-647) of AzF-incorporated proteins in brain sections (D) and corresponding in-gel fluorescence analysis (E) of brain homogenates confirming significant labeling of Syn-PheRS brains. (F) Quantification of AzF-647 in-gel fluorescence intensity demonstrates significantly increased AzF-labeled protein abundance in Syn-PheRS mice compared to WT controls. (G) Dural whole mounts from WT and Syn-PheRS mice demonstrates the distribution of AzF-labeled neuronal proteins in Syn-PheRS mice paralleling anatomical patterns observed in the Syn-ZsG paradigm. (H) High-magnification views of dural whole mounts from Syn-PheRS mice show AzF-labeled neuronal proteins colocalized with CD206, CD31, and LYVE-1. (I) In-gel fluorescence analysis of dura homogenates confirms the presence of AzF-labeled neuronal proteins in Syn-PheRS compared to WT. (J-K) Representative in-gel fluorescence assay and quantification of AzF-647 labeled proteins from brain border homogenates confirms the detection of endogenously produced neuronal proteins at each brain border. (L) Experimental schematic illustrating chronic CSF-infusion of OVA-647 via osmotic pump and subsequent tissue collection. (M) Quantification comparing the proportion of ZsGreen or OVA at each border from Syn-ZsG versus chronically CSF-infused OVA-647, respectively. Plotting this data showed that chronic tracer delivery recapitulates a similar distribution observed after acute ICM-injection ( Fig. 2D ), with significantly more OVA at sCLNs and dCLNS. Sagittal sinus/bridging veins = SS/BV, caudal confluence of sinus = COS, rostral confluence of sinus = ROS, transverse sinus = TS. All data are mean ± SEM with n = 3–4. *p value = < 0.05; **p value = < 0.01; ***p value = < 0.001, ****p value = < 0.0001 by one-way ANOVA with Tukey’s post-hoc or Student’s t test. NIHMS2173840-supplement-3.pdf (2.1MB, pdf) 4 Figure S3. Spatial segregation and uptake of brain-derived proteins at brain border compartments. Figure S3. LSFM and IHC analyses reveals pattern of ZsGreen and ICM-injected Ovalbumin tracer. (A-C) IHC on coronal and sagittal nasal cavity sections highlights that despite both ZsGreen and OVA-647 being detected, the signals are not concentrated in similar anatomical regions. (i-iii) Zoom in of boxed regions. DMM = dorsal medial meatus, ET = Ethmoturbinates, LNG = lateral nasal glands, S = Septum. (D) Quantification of proportion of area that is exclusively ZsGreen positive across nasal cavity sections shows that ~85% of the total ZsGreen positive area does not colocalize with OVA-647. (E-F) LSFM of whole heads with the brain removed allows visualization of protein distribution across dura and skull compartments. Panel (E) shows ZsGreen drainage along the lateral and basal dura and skull, regions with distinct spatial differences compared to OVA-555. (F) LSFM additionally highlighted the retroorbital vein. (G) Flow cytometric analysis of all dural cell populations from WT and Syn-PheRS mice following click labeling of AzF-incorporated neuronal proteins (AzF-647). AzF-647 signal was quantified before (surface) and after permeabilization (intracellular + surface), revealing increased intracellular accumulation of neuron-derived proteins across dural cell populations in Syn-PheRS mice compared to WT. (H-I) Immunostaining of dorsal skull sections and in-gel fluorescence assay from WT and Syn-PheRS mice following click detection of AzF-labeled neuronal proteins (AzF-647). NIHMS2173840-supplement-4.pdf (2.3MB, pdf) 5 Figure S4. Representative flow cytometry gating strategy for dural cell populations. Figure S4. Representative flow cytometry gating of dural cell populations. (A) Representative flow cytometry-based gating of the cell types population in the dorsal dura. BECs: CD45(−)CD31+; LECs: CD45(−)CD31(−)PDPLN+; cDC1: CD45hiCD11b(−)CD11c+; cDC2: CD45hiCD11b+CD11c+MHCII+; CD206+ BAMs: CD45hiCD11b+CD206+; Monocytes: CD45hiCD11b+Ly6C+; Macs: CD45hiCD11b+F480+CD206(−). (B) IHC of dura whole mounts from Syn-ZsG mice highlighting CD206 colocalization with ZsGreen (Merges in Fig. 4D ). Sagittal sinus/bridging veins = SS, rostral confluence of sinus = RCS, transverse sinus = TS NIHMS2173840-supplement-5.pdf (354.6KB, pdf) 6 Figure S5. Single-cell transcriptomic profiling across dura, skull, and nasal cavity borders. Figure S5. Flow cytometry and scRNA-seq of ZsGreen distribution across dura, skull, and nasal borders. (A-C) UMAP of cell type clusters in (B) dorsal dura, (C) dorsal skull, and (D) nasal cavity. Cell types in each tissue are annotated by unique colors. (D) ZsGreen and Cre transcripts are not detectable across all cell types in dorsal dura, skull, and nasal cavity compared to housekeeping genes. Mean expression for genes is plotted. (E) Proportion of ZsGreen+ and ZsGreen− cells within each resident cell population in the dorsal dura, skull, and nasal based on scRNA-seq analysis. (F-G) Merged UMAP of macrophage annotated cells across dorsal dura, skull, and nasal borders showing distribution of ZsGreen+ and ZsGreen− cells. (H-I) Quantification of CD274 intensity at ZsGreen+ B cells compared to ZsGreen− B cells in dorsal skull section. Three sections each from n = 2. All data are mean ± SEM with n = 2–4. *p value = < 0.05; **p value = < 0.01; ***p value = < 0.001, ****p value = < 0.0001 by Student’s t test. NIHMS2173840-supplement-6.pdf (584.5KB, pdf) 7 Figure S6. Region-specific ZsGreen-labeling reveals spatially compartmentalized clearance and distinct routing compared to intraparenchymal tracer injection. Figure S6. Brain region-specific ZsGreen labeling reveals spatially compartmentalized protein clearance and distinct routing compared to intraparenchymal tracer injection. (A) Distance matrix of normalized ZsGreen proportion across brain border and drainage tissues comparing pan-neuronal (Syn-ZsG) and regionally restricted labeling paradigms (Forebrain-ZsG, Anterior-ZsG, and Striatal-ZsG). (B-E) LSFM imaging of CUBIC-cleared whole heads from Striatal-ZsG mice showing ZsGreen distribution, including lateral skull + dura (LSK + dura), and basal skull + dura (BSK + dura), and nasal cavity. (F) Representative IHC of brain sections following intraparenchymal injection of OVA-647 into the striatum (Str-Inj:OVA). Images show tracer localization in the striatum one hour post-injection and substantial clearance by 24 hours, confirming tracer clearance from the injection site. After one-hour, brain OVA-647 levels were quantified and compared to WT controls. (G) Heatmap comparing the proportion of ZsGreen (Str-ZsG) versus intraparenchymal OVA-647 (Str-Inj:OVA) detected across brain border and drainage tissues. In contrast to ZsGreen, intraparenchymal OVA shows relative underrepresentation at skull-associated borders and increased accumulation in superficial cervical lymph nodes (sCLNs). All data are mean ± SEM with n = 3–4. *p value = < 0.05; **p value = < 0.01; ***p value = < 0.001, ****p value = < 0.0001 by Student’s t test. NIHMS2173840-supplement-7.pdf (351.9KB, pdf) 8 Figure S7. Acute inflammation and amyloid pathology differentially redistribute brain-derived proteins across vascular, border, and blood compartments. Figure S7. LPS and 5XFAD alter brain-derived protein drainage (A) Quantification of Cadaverin-647 intensity from IHC images from Fig. 6B , confirms LPS-induced acute inflammation results in a comprised blood brain barrier as shown with the significantly elevated levels of Cadaverin-647 tracer in the brain after treatment compared to controls. (B-C) IHC analysis and quantification of brain sections from Syn-ZsG PBS and LPS treated mice shows significantly increased abundance of ZsGreen at CD31+ vessels. (D) WB analysis of ZsGreen in plasma EV and soluble fractions shows significantly increased ZsGreen signal in LPS-treated Syn-ZsG mice compared to controls. (E) Quantification of (C). ZsGreen from plasma EV and soluble fractions confirms significantly elevated ZsGreen levels after LPS. Total protein normalization was used for all comparisons. (F) IHC of dura whole mounts from WT and 5XFAD mice highlight the anatomical distribution of Aβ 42 . Zoom in of boxes shown in Fig. 6K . (G) Aβ 42 ELISA from brain border extracts from WT, 5XFAD-ZsG, and 5XFAD mice confirm that ZsGreen expression does not perturb the normal drainage routing of endogenous Aβ 42 . (H) 2-month-old WT-ZsG and 5XFAD-ZsG clearance of ZsGreen across brain borders quantified by fluorescent plate reader and normalized to respective brain ZsGreen. (I) ZsGreen abundance in blood plasma of 2-month-old 5XFAD mice was significantly increased compared to WT-ZsG healthy controls All data are mean ± SEM with n = 3–5. *p value = < 0.05; **p value = < 0.01; ***p value = < 0.001, ****p value = < 0.0001 by one-way ANOVA with Tukey’s post-hoc or Student’s t test. NIHMS2173840-supplement-8.pdf (1MB, pdf) 9 Table S1. List of Top 10 Markers used to annotate cell types for each cluster in dura, skull, and nasal. Related to Figure 4 . NIHMS2173840-supplement-9.xlsx (60.4KB, xlsx) 10 Table S2. Number of cells for each cell type in dura, skull, and nasal. Number of ZsGreen+ and – cells collected after flow cytometry. Related to Figure 4 . NIHMS2173840-supplement-10.xlsx (12.7KB, xlsx) 11 Table S3. Differentially expressed genes for each cell type in dura. Full dataset shown for all cell types which showed significantly modulated genes. Related to Figure 4 . NIHMS2173840-supplement-11.xlsx (902.6KB, xlsx) 12 Table S4. Differentially expressed genes for each cell type in skull. Full dataset shown for all cell types which showed significantly modulated genes. Related to Figure 4 . NIHMS2173840-supplement-12.xlsx (4.7MB, xlsx) 13 Table S5. Differentially expressed genes for each cell type in nasal. Full dataset shown for all cell types which showed significantly modulated genes. Related to Figure 4 . NIHMS2173840-supplement-13.xlsx (2.1MB, xlsx) Data Availability Statement Raw sequencing data are available in the NCBI Sequence Read Archive (SRA) under BioProject accession number SRP599291. This paper does not report original code. Any additional information in this paper is available from the lead contact upon request. ACTIONS View on publisher site PDF (4.2 MB) Cite Collections Permalink RESOURCES Similar articles Cited by other articles Links to NCBI Databases Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top

Related documents

• Multimodal neuroimaging of brain microstructure and hippocampal morphology alterations in type 2 diabetes mellitus.#242171
• Supplementary Material from How brain pulsations drive solute transport in the cranial subarachnoid space: insights from a toy model#658833
• Precision HD-tACS for Enhancing Sensory Experience in Schizophrenia#838811
• Spatio-temporal coupling of the electric and hemodynamic brain responses in humans#401938
• Modeling interactions between neurons and astrocytes : Application to brain information processing#246719