Microphysiological model of the brain

Abstract

An in vitro model of the brain has been developed which mimics the parameters of normal or diseased brain, including microvasculature, lymphatics, neurons containing genetic changes associated with disease and immune cells. The model is useful for testing the effect of several factors (e.g. molecules, proteins) on blood brain barrier (BBB) function, morphology, permeability, neuronal connectivity, meningeal lymphatics function and morphology, various indicators of pathology and apoptosis. Examples demonstrate the usage of the in vitro system for modeling Alzheimer's disease and potential applications of the models by adding more cellular compartments

Description

[0001]MICROPHYSIOLOGICAL MODEL OF THE BRAIN CROSS-REFERENCE TO RELATED APPLICATIONS None. STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH The government has certain rights in the invention by virtue of Grant nos.6945261 and R01 NS121078 from the National Institutes of Health and CBET-0939511 from the National Science Foundation. FIELD OF THE INVENTION The present invention is generally in the field of in vitro models of complex tissues for use in drug screening, and in particular to a microphysiological model of the brain. BACKGROUND OF THE INVENTION As the elderly population increases, the number of people with Alzheimer's disease (AD) and other neurodegenerative diseases is rising. Key characteristics of AD include extracellular buildup of amyloid plaques composed of amyloid-β (Aβ) peptide, the formation of neurofibrillary tangles due to abnormal accumulations of tau protein, and neuroinflammation. Recently, emerging evidence has implicated cerebrovascular dysfunction, such as the breakdown of the blood-brain barrier (BBB), as another crucial factor in the development and progression of AD. The cerebral vasculature is a complex network of blood vessels that supplies oxygen and nutrients to the brain. The BBB, which comprises tightly linked endothelial cells, pericytes, astrocytes, and extracellular matrix (ECM) proteins, plays a critical role in regulating the exchange of molecules between the brain and the blood to protect the neural tissue and maintain the homeostasis of the brain microenvironment, including ion concentrations, pH, and nutrient levels. A disrupted BBB may lead to increased permeability, allowing entry of harmful blood-derived molecules, cells, and microbial agents into the brain, triggering inflammatory and immune responses, which in turn, can initiate multiple pathways of neurodegeneration. BBB breakdown can also impair the clearance of Aβ 1 45630538.1 peptides from the brain potentially contributing to the accumulation of Aβ plaques. Additionally, vascular dysfunction in AD can lead to reduced blood flow, hypoxia, and oxidative stress, all of which can further exacerbate neuronal damage and cognitive impairment. However, the mechanisms governing fluid and molecule transport across the BBB, the interaction between neurons and the BBB, and the molecular processes underlying AD onset and progression remain inadequately understood in humans. This knowledge gap is partially due to limited accessibility to human brains for research and the challenges in translating results from animal studies to clinical settings. Emerging evidence has implicated an early breakdown of the BBB in AD patients, even before cognitive decline and brain pathology. BBB breakdown has been suggested to be involved in the pathophysiology of several neurological disorders such as Alzheimer’s disease, Parkinson’s disease, Huntington’s disease, amyotrophic lateral sclerosis and multiple sclerosis. However, the mechanisms driving barrier breakdown and their link to pathophysiology of neurodegenerative diseases are not yet fully characterized. Meningeal lymphatics are responsible for clearance of Aβ peptides from the brain parenchyma, and disruption of meningeal lymphatics in AD transgenic mice accelerates Aβ deposition in the meninges, aggravates parenchymal Aβ accumulation, and induces cognitive impairment. Yet, the understanding of the role of fluid transport across the BBB, through the brain parenchyma, and exiting via the lymphatic system is poorly understood. It is apparent that models of the brain that can be used to screen for drugs and to follow the progression of neurological diseases would be extraordinarily useful. However, the brain is a complex organ, with many unique and hard to replicate features and multiple types of disparate cell types. As a result, it is still the industry standard to use genetically engineered mouse models for initial screening and then spend years in testing using human patients, with little to no opportunity to study disease in the early stages. Additionally, high failure rates in clinical trials for neurodegenerative disorders such as AD have been linked to an insufficient predictive validity of current animal-based disease models. Pre-clinical 2 45630538.1 studies in animal models poorly translate to humans and represent a major hurdle in the drug discovery pipeline, since 80% of potential treatments fail in clinical trials. This has created an increasing demand for alternative, human-based models capable of emulating key pathological phenotypes in vitro. There are no physiologically accurate models for neurodegenerative diseases that allow for rapid screening of drugs. There is a growing interest in the development of in vitro human disease models to better understand and study the physiological and pathological mechanisms associated with AD and other neurological diseases and degenerative disorders, and for more rapid screening of potential therapeutic agents than the current human clinical trials that take years and a great deal of money to conduct. In vitro models are widely used to study cellular interactions and function in many organ-specific contexts. This is made possible through technological advances in bioengineered microphysiological systems (MPS) that support 3D cellular self-organization, such as microfluidics or 3D patterning in combination with biocompatible hydrogels. While these individual technological advances have progressed incrementally over decades, their combination was only achieved recently. Resulting MPS now facilitate the formation, maturation and maintenance of complex organotypic structures in well-defined and customizable microenvironments. MPS can recapitulate tissue organization and function in many organ-specific contexts, providing physiological relevance, potential for customization to different population groups, and scalability for drug screening purposes. Recent BBB in vitro models using human cells of the neurovascular unit (NVU) show physiological morphology and permeability values, and reproduce brain neurological disorder phenotypes. Considering the implications of barrier breakdown and vascular leakiness in neurodegenerative diseases and the advances in MPS, there is an urgent need to create more physiologically relevant MPS to elucidate the pathophysiological mechanisms of vascular instability and accelerate therapeutic interventions. MPS are especially important to model the NVU, where close interactions between endothelial cells (ECs), pericytes and astrocytes are crucial for proper function. 3 45630538.1 Various in vitro models and assays are being developed to investigate BBB function and pathogenesis. ECs in mono-culture or in co-culture together with pericytes and astrocytes have been grown in different configurations as 2D monolayers and more recently in 3D MPS to interrogate barrier properties in standard culture or in response to treatments disrupting or maintaining vascular permeability. Recently, the assembly of endothelial and perivascular cells was accomplished in a manner that resembles NVUs found in CNS micro-vasculatures, by self-organization of cells into microvascular networks (MVNs) in MPS. Deeper characterization of vascular models using primary cell sources alone, or in combination with induced pluripotent stem cell (iPSC)-derived ECs, demonstrated acquisition of tissue-specific cellular identities representative of physiological human blood-neural barriers. Such systems are essential to study human-based advanced cellular models customized to specific patient groups / populations, which are scalable for drug discovery purposes, show physiologically relevant morphology and permeability values, and reproduce pathophysiological phenotypes such as vascular leakage and neurodegeneration. It is therefore an object of the present invention to provide a three- dimensional (3D) in vitro model by integrating BBB, meningeal lymphatics, immune cells and stem-cell-derived AD cell culture to closely mimic the AD brain environment. It is a further object of the present invention to to identify AD- specific mechanisms underlying pathophysiological changes in the BBB, meningeal lymphatics, and Aβ clearance. It is still another object of the present invention to provide a 3D AD model not only for studying the mechanisms of AD progression but also for drug discovery in a human brain-like environment. It is a further object of the present invention to provide an MPS which includes not only the neural cells and structures, the vasculature including the BBB, but also the lymphatics and the immune cells which are present in the brain, both normal and diseased. 4 45630538.1 SUMMARY OF THE INVENTION An in vitro microphysiological system (MPS) model of the brain has been developed which mimics the parameters of normal or diseased brain, providing a mechanism for screening and testing of compounds for the effect on blood brain barrier function, permeability, neuronal activity, various indicators of pathology and apoptosis for neurodegenerative diseases. The model can include cells (stem cell derived or primary) of such diverse origins as neural cells (neurons, astrocytes), vascular endothelial cells, mural cells (pericytes), and lymphatic cells, all integrated to mimic a normal, diseased, or degenerating brain. Cells can be obtained from normal or abnormal brains, or cell lines, and de-differentiated then differentiate to produce cells and structures incorporating the cells that recapitulate the naturally occurring characteristic of the brains of origin. The MPS includes: 1. a blood brain barrier (“BBB”) chamber (device design 1, no central hole); 2. a neuronal tissue chamber within (device design 2, with central hole); 3. a neuronal tissue media and BBB media compartments; 4. iPSC-derived endothelial cells, pericytes, astrocytes; 5. neurospheroids or organoids (as referred to as cerebral organoid or spheroid); 6. lymphatic microvasculature (device design 3); 7. immune cells. The MPS is typically formed of an inert polymer such as polydimethylsiloxane (PDMS) elastomer but can be formed of other polymers such as cyclic olefin polymers (COP) or copolymers (COC). This includes at least three separate but communicating compartments: a BBB vasculature compartment typically in the form of a central circular chamber containing the BBB vasculature (device design 1); a neuronal cell culture compartment in the form of a circular chamber within the vasculature compartment in which neurons and astrocytes are embedded in an extracellular material such as MATRIGEL®, the neuronal cell culture compartment is open and accessible from the top for separate media introduction and extraction; and at least two media channels for controlled media administration to the culture compartments (FIG.1A). If a lymphatic 5 45630538.1 compartment is added in this system, the MPS has one more region where lymphatics are formed against the BBB region (FIG.2A). This compartment has a separate medium channel. Two or four fluid-holding syringes are connected to the medium chambers to create interstitial flow (IF) or vascular flow (VF) (four for circulating medium). Alternatively, tubing may be connected to the chamber openings in order to be connected to a pump controlling fluid flow rate to push fluid through the tissue (IF) or vascular networks (VF), to promote BBB formation and provide long-term device culture. Alternatively, a pumping system may be incorporated into the device to produce one integrated system (FIG 3A). The main microfluidic device has three designs: one with only brain vasculature, one with brain vasculature and neurons and one with brain vasculature, neurons and lymphatics. The pump can fit to all these designs. The pump can be plugged into the devices directly. FIG.4A-4C are prospective and cross-sectional schematics of the pump used with the microfluidic devices. FIG.4A is a cross-sectional view of the pump. FIG.4B is a cross-sectional side view of the pump of FIG. 4A. FIG.4C is a top view of the pump. FIG.5A-5C are prospective views of the pump and microfluidic device for culturing BBB with neurons and lymphatics systems. FIG.5A shows the wells and channels with the culture media chambers. FIG.5C is a mold for manufacture of the device for culturing the BBB with neurons and lymphatics systems. . The model allows for the engineering of perfusable BBB microvascular networks within days and accommodates long-term, at least four weeks, culture requirements necessary for neuronal cells to generate elevated levels of Aβ. The co-culture system exhibits 3D vascular endothelial networks in direct contact with pericytes and astrocytic endfeet, closely mimicking in vivo multicellular interactions crucial for BBB functionality, and induces the deposition of amyloid plaques, faithfully recapitulating key aspects of AD pathology. The lymphatic microvascular platform and incorporation of inflammatory cells such as dendritic cells, microglial cells, macrophages and lymphocytes allows even greater functionality in mimicking normal brain for use in testing and diagnostics. 6 45630538.1 Studies demonstrated alterations in the function and features of the BBB and Aβ accumulation during AD progression. The BBB permeability increases over time in AD conditions, potentially caused by dysregulation in key endothelial markers. Quantification of morphological alterations of the vasculature and visualization of Aβ plaque formation within the model replicates key AD phenotypes. Quantitative analysis of soluble Aβ levels in distinct compartments of the microfluidic devices highlights the relevance of the model in replicating AD-related dynamics. Additionally, this model can be easily adapted to study other neurodegenerative diseases, such as Parkinson’s disease, Huntington’s disease, and amyotrophic lateral sclerosis, as well as for other brain conditions and for screening of compounds for use in treatment thereof. Examples further demonstrate the replacement of the suspended neurons with neurospheres for reducing co-culture time, the effect of compounds in the model, an integrated pump in the system and the addition of the lymphatic compartment and immune cells.. BRIEF DESCRIPTION OF THE DRAWINGS FIG.1A is a perspective view of the microfluidic platform 10 used to create the BBB in a circular compartment 12a that surrounds the neural sphere compartment 12b (ID 12a top chamber 0.6 mm; ID 12b lower chamber 0.2 mm) allowing direct communication throughout the device. The loading ports 16a, 16b (ID 0.15 mm) are connected via loading channels 16c, 16d to the neurosphere medium compartment 14. Medium reservoirs are shown at 18a, 18b, 18c, 18d. FIG.1B is a 3D rendering of the platform 10 in FIG.1A. The media channels 20a, 20b, 20c, 20d are used to perfuse the BBB by generating a pressure difference across the central compartments 12a, 12b. FIG.1C shows the experimental seeding protocol and device culture protocol for maintaining the model in co-culture for 30 days. Suspended cells are injected through the BBB loading ports 16a, 16b. Neurospheres are seeded at day 7 (D-21) (suspended neurons at day 0 after BBB injection) into the central neurosphere compartment 12 and neuronal media added and BBB media added to the media channels 20a-20d flanking the central chamber 12. Suspended cells are injected through the BBB loading port 16a, 16b, chips loaded with primary human astrocytes, primary human microvascular 7 45630538.1 endothelial cells, primary human microvascular pericytes and optionally immune and lymphatic cells. Fluid flow is initiated two days later, and permeability and morphological studies performed seven days after loading. FIG.1D shows the confocal projected z-stack of one portion of the full device and FIG.1E shows immunofluorescence staining after 30 days of co-culture. Green: GFP ECs; Red: mCherry ReN AD; Texas Red: PDGFRα + PDGFRβ antibody-PCs; White: S-100 antibody-ACs; DAPI: nucleus staining. Scale bars are 200 μm. E) Neurospheres (instead of suspended neurons) in the neuronal compartment. FIGs.2A-2B. FIG.2A is a schematic of the microfluidic system for seeding and culturing a lymphatic endothelial monolayer next to the BBB in a separate lymphatic chamber with the neuronal compartment. FIG.2B is a confocal z-projection of the full device. Neurons that are surrounded by the brain vasculature (BBB) on the left and lymphatic endothelial cells that have been formed in the right gel area of the microfluidic device. FIG.3A is a cross-sectional schematic of a microfluidic pumping system that fits on top of the microfluidic devices shown in FIG.1A-1D and / or FIG.2A to create circulating flow. FIG.3B is a photograph of a perspective view of the device shown in FIG.3A. FIG.3C are graphs of the interstitial fluid flow pressure: pressure input profile and fluid flow rate in vessels. FIG.4A-4 are prospective and cross-sectional schematics of the pump used with the microfluidic devices. FIG.4A is a cross-sectional view of the pump. FIG.4B is a cross-sectional side view of the pump of FIG. 4A. FIG.4C is a top view of the pump. FIG.5A-5C are prospective views of the pump and microfluidic device for culturing BBB with neurons and lymphatics systems . FIG.5A shows the wells and channels with the culture media chambers. FIG.5C is a mold for manufacture of the BBB with neurons and lymphatics culture device. FIG.6A-6D are confocal fluorescence micrographs of the cells in the microfluidic device. FIG.6A shows the meningeal lymphatics over the blood brain barrier; FIG.6B shows the neurons over the blood brain barrier over the meningeal lymphatics. FIG.6C shows the meningeal lymphatic 8 45630538.1 cells. FIG.6D shows the endothelial cells forming the blood brain barrier, using dextran as vascular perfusability proof. DETAILED DESCRIPTION OF THE INVENTION I. Definitions The blood-brain barrier (BBB), first described by Paul Ehlrich (1885), is a highly selective semi-permeable membrane between the blood and brain interstitium. This unique barrier allows cerebral blood vessels to regulate the movement of molecules and ions between the blood and the brain. “Microvascular” refers to the part of the circulatory system made up of minute vessels (such as venules or capillaries) that average less than 0.3 millimeters in diameter. A microvascular network (MVN) is defined as a perfusable microvasculature network developed in a self-organizing manner by embedded endothelial cells, or endothelial cells and corresponding supporting cells, within a hydrogel containing one or more extracellular matrix components such as fibrin, usually by co-culturing the endothelial cells with stromal cells such as fibroblasts and pericytes. A neurosphere is a culture system composed of free-floating clusters of neural stem cells. Organoids are an artificially grown mass of cells or tissue that resembles an organ. An organoid is a miniaturized and simplified version of an organ produced in vitro in three dimensions that mimics the key functional, structural and biological complexity of that organ. Neurodegenerative diseases occur when nerve cells in the brain or peripheral nervous system lose structure and function over time and ultimately die. In some cases, these include motor neurons. Neurodegenerative diseases affect millions of people worldwide. Alzheimer’s disease and Parkinson’s disease are the most common neurodegenerative diseases. In the United States, as many as 6.2 million people may have Alzheimer’s disease, according to a report from the Alzheimer's Disease Association in 2022. Nearly a million Americans are living with Parkinson’s disease, according to the Parkinson’s Foundation. Other examples include amyotrophic lateral sclerosis and motor neuron 9 45630538.1 disease. In addition to Alzheimer's disease, frontotemporal dementia, chronic traumatic encephalopathy (CTE), Lewy body dementia and limbic predominant age-related TDP-43 encephalopathy (LATE) are more likely to lead to death. II. Microfluidic Devices and Cells Microfluidic Devices In the simplest embodiment, polymeric microfluidic devices, preferably formed from polydimethylsiloxane (PDMS), are molded from a silicon wafer fabricated by soft photolithography, but can also be made from other types of plastics such as COC or COP. Appropriate cell types such as human endothelial cells (ECs), pericytes (PCs) and astrocytes (ACs) are embedded in a fibrin hydrogel or MATRIGEL® and loaded inside the microfluidic chip (FIG.1A-1D), as shown in FIG.1E. The protocol is depicted in FIG.1C. As a result of cell self-organization, microvascular networks (MVNs) form after a few days of culture within the chips. Additionally, neuronal and lymphatic compartments can be added using a modified version of the same microfluidic device. See FIG.2A. The entire system is user friendly and lends itself to construction in hard plastic with no need for internal membranes or PDMS fabrication and can readily be translated to a high-throughput platform as needed by the pharma industry. Based on the design that allows co-culturing using one (BBB only) or two different hydrogels (multi-chambers for cells), microfluidic chips are fabricated from polydimethylsiloxane (PDMS) elastomer, which facilitates the co-culture of BBB and neurons / astrocytes to model AD in vitro (FIG. 1A-1E). The microfluidic chip offers versatility through the integration of three separate but communicating compartments, each serving a specific purpose within the co-culture system: i) BBB vasculature compartment: a central circular chamber 12 containing the BBB vasculature, which allows accurate reproduction of both the structural and functional characteristics of the BBB, ii) neuronal cell culture compartment: within the BBB vasculature compartment lies another circular chamber 14 in which neurons and astrocytes (suspended or in neurospheres, FIG.1D) are embedded. This chamber is open and accessible from the top, which allows for separate 10 45630538.1 media introduction and extraction, facilitating precise sampling of the neuronal culture, and iii) two media channels or media ports 16c, 16d: two additional channels or ports 18a,b and 18c,d, flank the central compartment, specifically designed for controlled media administration to the co-culture system. A BBB + lymphatic microvasculature model has been developed that mimics the morphological and functional properties of in vivo lymphatic capillaries (FIG.2A). The formation of lymphatic vasculature in this microfluidic platform 20 is achieved by first seeding a confluent monolayer of human lymphatic endothelial cells at the media-gel interface 22 and, subsequently, introducing a pro-angiogenic stimulus (growth factors or interstitial flow) at the opposite media channel 24 to promote lymphangiogenesis into the central 3D gel channel 26. By continuously exposing the lymphatic endothelial cells to interstitial flow with or without pro-angiogenic factors towards them over 4-6 days, the resultant lymphatic network 2B that invades the 3D gel region recapitulates key physiological characteristics, including vessel density, diameter, and dead-ended. The microfluidic design consists of a vascular component with three channels (a central channel for 3D BBB 26 and neuron co-culture 28 and two side channels 30a,b and 30c,d for BBB medium that run parallel to it) and a lymphatic component 32, composed of one channel 34 that forms a T- junction with the central channel 26 where media can come into contact with the gel in the BBB compartment. The BBB compartment 26 incorporates a neuronal culture islet with an open-top well for seeding brain spheroids to come into contact with three media channels (FIG.2A 30a,b 30c,d and 36a,36b). FIG.2B shows the neurons, BBB and lymphatic endothelial cells in culture in the device of FIG.2C. Inflammatory cells can be added to further replicate the conditions of the normal and diseased brain. Representative cells include microglial cells, macrophages, lymphocytes, dendritic cells, and macrophages. FIG.3A is a schematic of the microfluidic device 40 including closed ports 42 which can be connected by pumps 44 to control fluid flow through the microfluidic device 40. The pump 50 controls flow rates and 11 45630538.1 pressure using one way valves 52 (in), 54 (out) and capacitors 56, 58. FIG. 3B is a photograph of the device 40 and pump 50. FIG.3C is a graph of pressure input profile and fluid flow rate in vessels. Cells The brain includes different cell types, as well as subclasses of the different cell types. Neurons are the main signaling units, communicating with each other via synapses. Among the non-neuronal cells are the endothelial cells lining blood vessels, pericytes, astrocytes and glial cells. Glial cells are oligodendrocytes (insulating neuronal axons for faster signal transmission), microglia are immune cells (the brain macrophage with a hematopoietic origin) and astrocytes and pericytes are involved in numerous functions such as blood brain barrier, homeostasis, neuronal growth and neurotransmitter recycling. There are also cells forming meningeal lymphatics and other types of immune cells. The microfluidic device can include any of the various cell classes ( endothelial, neural or glial) and subpopulations of cells (microglia, pericytes, oligodendrocytes and astrocytes) as well promote and support the subcellular locations (axons, dendrites, synapse and glial endfeet which interact with the vascular tree in the CNS). To form the most accurate model of the brain, the model must include cells that form a vasculature having the unique properties of the blood brain barrier, neural cells, astrocytes, pericytes, immune cells, and cells forming the meningeal lymphatics. The cells may be normal or abnormal, derived by genetic modification, by dedifferentiation and re-differentiation of cells from patients, who may have a disease or disorder such as Alzheimers Disease, Parkinsons, Lou Gehrig disease, aged, or having suffered from a stroke, for example, where the cells are from an individual or cell line from an individual, the individual having a neurodegenerative disease or disorder, having had a stroke or other interruption in flow of blood to the brain, a human that is over the age of 40, 50, 60 or 70, or a combination thereof. Cells can be obtained by biopsy, surgery, postmortem, or be from cell lines, for example, the parental line KOLF2.1J; AD line PSEN1 E276A SNV / WT and control PSEN1 E276A Rev / WT; and PD line SNCA A53T SNV-WT and control SNCA A53T REV-WT. 12 45630538.1 Alzheimer’s Disease, Neurospheres and Microvasculature A three-dimensional Alzheimer’s disease model was developed using a compartmentalized microfluidic device that combines a self- assembled microvascular network of the human blood-brain barrier with neurospheres derived from Alzheimer’s disease-specific neural progenitor cells. To shorten microfluidic co-culture times, neurospheres are pre-differentiated for 21 days to express Alzheimer’s disease-specific pathological phenotypes prior to the introduction into the microfluidic device. In agreement with post-mortem studies and Alzheimer’s disease in vivo models, after 7 days of co-culture with pre- differentiated Alzheimer’s disease-specific neurospheres, the three- dimensional blood-brain barrier network exhibited significant changes in barrier permeability and morphology. Furthermore, vascular networks in co- culture with Alzheimer’s disease-specific microtissues displayed localized β- amyloid deposition. By interconnecting a microvascular network of the blood-brain barrier with pre- differentiated neurospheres, the model can replicate key neurovascular phenotypes of neurodegenerative disorders in vitro. By adding lymphatic microvasculature and immune cells, one is able to more completely assess the functional characteristics of abnormal, diseased or genetically compromised brains, and the effects of potential therapeutic and diagnostic agents thereon. Alterations in the function and features of the BBB reveal that the BBB permeability increases over time in AD conditions, potentially caused by dysregulation in key endothelial markers. Additionally, morphological alterations of the vasculature and visualization of Aβ plaque formation within the model replicate key AD phenotypes. Neurospheres are seeded in the compartment on the BBB network. Support for the BBB cells is provided using an extracellular matrix (ECM) protein such as fibrinogen; the neurospheres are supported using a media such as MATRIGEL®. Neurospheres are formed from iPSCs, either of normal or abnormal neural cell origin. ECs and neurospheres can be of normal origin or from a patient with a defect and / or disease and / or old age, all of which can have different genetics and functions. Cells can be obtained from patients, for example, such as Alzheimer’s disease (“AD”) patients, Parkinson’s disease patients (“PD”), people with dementia due to age, people 13 45630538.1 with blood or cancer disorders, or people having inflammation of the brain. iPSC-derived cells are used to recapitulate the functioning structures of the BBB in vitro, which can be used for measurement of vascular permeabilities, the evaluation of cellular architecture, and the analysis of gene and protein expression, offering a powerful tool to study the mechanisms underlying cognitive impairment. The model allows for the engineering of perfusable BBB microvascular networks within days and accommodates long-term (at least four weeks) culture requirements necessary for neuronal cells to generate elevated levels of Aβ. The co-culture system exhibits 3D vascular networks in direct contact with pericytes and astrocytic endfeet, closely mimicking in vivo multicellular interactions crucial for BBB functionality, and induces the deposition of amyloid plaques, faithfully recapitulating key aspects of AD pathology. Alterations in the function and features of the BBB and Aβ accumulation during AD progression were observed. The BBB permeability increases over time in AD conditions, potentially caused by dysregulation in key endothelial markers. Quantification of morphological alterations of the vasculature and visualization of Aβ plaque formation within the model replicates key AD phenotypes. Quantitative analysis of soluble Aβ levels in distinct compartments of the microfluidic devices highlights the relevance of the model in replicating AD-related dynamics. Additionally, this model can be easily adapted to study other neurodegenerative diseases, such as Parkinson’s disease, Huntington’s disease, and other forms of dementia. Menigeal Lymphatics In one embodiment, the cells are introduced into the central compartment of the MPS, for example, in a concentration of between 0.5 and 6 x 106cells / ml, media and growth factors as appropriate, MATRIGEL® (10 microliters) added, then media added (50 microliters). To create a more in vivo-like system and better capture the complexity of Aβ clearance mechanisms in the brain, meningeal lymphatics can be incorporated. A neurovascular AD model with lymphatic system provides a more comprehensive representation of the major Aβ clearance pathways and elucidate the interplay between the BBB, meningeal lymphatics, and other Aβ clearance pathways, offering a deeper understanding of Aβ dynamics in 14 45630538.1 AD. Similarly, adding microglia, the resident immune cells of the brain, into the system further improves the physiological relevance of the model and provides insights into the neuroinflammatory aspects of AD and the interaction between immune responses and the BBB. The assembly of endothelial and perivascular cells was accomplished in a manner that resembles NVUs found in CNS micro-vasculatures, by self- organization of cells into microvascular networks (MVNs) in MPS. Deeper characterization of vascular models using primary cell sources alone, or in combination with induced pluripotent stem cell (iPSC)-derived ECs, demonstrated acquisition of tissue-specific cellular identities representative of physiological human blood-neural barriers. Such systems are essential to study human-based advanced cellular models customized to specific patient groups / populations, which are scalable for drug discovery purposes. Examples demonstrate that the BBB permeability increases and perfusability is impaired in neurovascular co- cultures with exogenously added Aβ42. Examples of disease modeling show how PD can be created using a triplication mutation of the SYN gene, familial PD, increased accumulation of SYN, and early onset of PD, showing pre-onset of PD phenotype onset of PD, pronounced PD phenotypes. PD models show that MPS emulates vascular phenotypes associated with PD including impaired network morphology, reduced vessel area fraction, length and lower segment counts when co-cultured with 2xSNCA, and significantly impaired barrier permeability of D65 of human midbrain organoid (hMO) differentiation. Meningeal Lymphatics Genetic, pathology and clinical studies have shown that many neurological diseases have an essential immunological component. In addition to multiple sclerosis (MS), a prototypic neuroinflammatory disease caused by autoreactive T cells that attack the central nervous system (CNS), several lines of evidence indicate that Alzheimer disease and many other neurodegenerative diseases, feature a remarkable activation of CNS-resident macrophages, which are commonly known as microglia. Moreover, recent findings on the meningeal lymphatics have highlighted an additional layer of immune circuitry that may participate in neurological diseases. Thus, 15 45630538.1 understanding the dynamics of neuro-immune interactions within the CNS has become the focal point of many studies. After an acute stroke, loss of blood brain barrier (BBB) integrity leads to a dramatic infiltration of peripheral immune cells into the brain, further contributing to neuroinflammation. The composition of these infiltrating cells in animal models differs between young and aged brains, as does their inflammatory potential, as assessed in a functional flowing lymphatic system in the CNS using specific lymphatic endothelial markers. Immune Cells After finding lymphatic elements in the human brain, the presence and distribution of T lymphocytes was assessed to determine whether lymphatic endothelial cells and lymphocytes in the CNS are associated with one another. Immune cells are known to monitor the CNS for danger signals and play a role in restoration of CNS homeostasis. Peripherally derived T cells, macrophages, and dendritic cells all appear to participate in this. Since 80% of the CNS sentinel cells are T cells, CD3 was used to look for these cells. T cells can enter the CNS in any of three ways. They can get into the CSF through the choroid plexus, through the Virchow–Robin space, or through postcapillary venules. These perivascular lymphocytes are thought to screen the CSF for potentially harmful molecules. Infusing an antigen into the CSF space will activate the peripheral immune system, suggesting a direct connection between the SAS and the periphery. Few lymphocytes are present in the healthy brain, and most of these are effector and central memory T cells. The T cells that have been examined are always in close proximity to cells that express lymphatic markers and adhesion molecules. T cells may move in these spaces just as they “roll” on endothelial cells and are likely to engage in surveillance of the CNS environment. Their movement can be orchestrated by the synchronized expression of the adhesion molecule by the lymphoid epithelial cells and its cognate ligand LFA1, produced by the immune cells. This might result in a slow “flow” of immune cells along the nerves in the perivascular space (PVS). The cells may be gently pushed along by the movement of interstitial fluid / CSF around the vasculature or by endoneurial fluid flowing among the fibers in cranial / peripheral nerves. 16 45630538.1 Accumulation of senescent cells during aging promotes chronic inflammation and tissue dysfunction and is an essential contributor to the progression of age-associated diseases (i.e., AD and atherosclerosis). Endothelial, epithelial, and stromal cells can express the SASP, leading to the recruitment of immune cells and an increased pro-inflammatory milieu. Glial cells also undergo senescence both in vitro and in vivo, contributing to age-related neuroinflammation and vascular dysfunction. Enhancing the clearance of senescent microglia and astrocytes using genetic or pharmacological approaches reduced tau aggregation and led to the preservation of cognitive function in murine models. This suggests that there is potential to reverse some of the detrimental immune responses seen with aging. Aging results in a chronic increase in systemic inflammation, termed “inflammaging” which can be caused by aging, termed sterile inflammation (indicating no detectable pathogens). Inflammaging and inappropriate immune activation contribute to the pathogenesis of many age-related diseases, including diabetes, atherosclerosis, and AD. Sporadic AD development has been linked to enriched risk genes that are present in aged microglia, the primary resident immune cell of the brain. This implicates brain innate immunity in neurodegeneration. This concept is further supported by the presence of microglia with a senescent (dystrophic) phenotype in post-mortem AD brains. The adaptive immune system and clonally expanded senescent T-cells also contribute to AD pathogenesis. The risk of ischemic stroke and vascular dementia is increased with aging, in part due to increased reactive oxygen species (ROS) and enhanced coagulation induced by inflammation. However, the mechanisms that drive age-related chronic inflammation in the brain and cerebral vasculature are not fully understood. Both immunosenescence and inflammaging alter the microenvironment of the central nervous system (CNS), primarily by actions on microglia. Normally microglia are in a homeostatic state maintained by cellular signals and interactions with ligands that inhibit microglial activation. Ligands such as CD200, CXCL1, and CD47 are expressed by neurons and bind to corresponding receptors on microglia. As neurons are 17 45630538.1 damaged with age or vascular insults, these inhibitory ligand-receptor interactions with microglia are disrupted. In addition, misfolded proteins, such as amyloid-beta (Aβ), accumulate during normal aging and lead to an increase in the levels of microglial pro-inflammatory cytokines. With age, there is also an increase in the expression of specific cytokines, such as transforming growth factor-β (TGFβ). Chronic exposure of microglia to TGFβ impairs their capacity to secrete anti-inflammatory cytokines and leads to the downregulation of interferon regulatory factor-7, an important factor in switching microglia from a pro-inflammatory to an anti-inflammatory phenotype. These can be modeled with the system disclosed herein to study the effects of senescence as well as the effect of compounds thereon. For example, microglia cells can be introduced in the neuronal cell compartments of the model on top of a layer of MATRIGEL®or other hydrogel to determine if they are activated and recruited to the neuronal cells in the 3D AD µVN-ML. The human immortalized microglial cell lines (SV- 40) and human iPSC-derived microglia-like cells (iMGL) can be used. Immunofluorescence staining can be used to quantify the recruitment of microglial cells, by visualizing microglia-specific markers, such as Iba1, CD11b, and CD33, using confocal microscopy. To characterize the migration, individual microglial cells can be monitored using time-lapse confocal z-stack imaging. CD33 is particularly crucial because, in microglia, CD33 may play a role in the phagocytosis of Aβ. The interaction of microglia cells with the neurovascular and meningeal lymphatic endothelial cells can be assessed along with any changes in this interaction with disruption to the neurovascular or meningeal lymphatic endothelial cells, for example, by screening for the secretion of cytokines that may facilitate communication between these cell types. Furthermore, the study after the addition of immune cells in the meningeal compartment can further dissect the interaction between different cell types during pathological conditions. An anti-CCL2 antibody can be used to modulate microglia recruitment and the crosstalk with BBB and meningeal lymphatics. Neutralization of CCL2 chemokine signaling using an anti- CCL2 antibody reduces microglia migration in response to neuron and astrocyte co-culture. 18 45630538.1 III. Methods of Making the Model In the simplest embodiment, polymeric microfluidic devices, preferably formed from polydimethylsiloxane (PDMS) or any clear, cell- compatible plastic, are cast in a mold that can be micromachined from plastic or metal and fabricated by soft photolithography. A microfluidic system was developed that consists of two media channels and a center channel containing an island for 3D AD culture (FIG.1). This can be used to form a 3D BBB model consisting of a tri-culture of human primary brain ECs with pericytes and astrocytes. In a further embodiment, the BBB model includes a meningeal lymphatics component (FIG.2, FIG.5, FIG.6) Cells can be added to and cultured in the device as exemplified below, then used for screening for potential therapeutic and diagnostic agents by adding the agents to be tested to the seeded devices and monitoring effects on the cells and functions of the cells in the devices. Example 1: Fabrication of Microfluidic Device The mold used for the fabrication of the microfluidic devices was designed in AutoCAD (Autodesk, Inc.), and imported in Fusion 360 (Autodesk, Inc.) to generate corresponding tool paths, followed by milling a DELRIN®(polyoxymethylene homopolymer) block with a micro-computer numerical control (“CNC”) milling machine (Bantam Tools). Polydimethylsiloxane (PDMS) based microfluidic devices were then fabricated using the mold and bonded to a glass coverslip using the procedure established previously (https: / / doi.org / 10.1038 / s41596-021- 00635-w). Other methods of manufacture can also be used. The microfluidic device has four Ø 4 mm media channel holes, two Ø 1.2 mm holes for BBB-gel loading and one Ø 2 mm central (neuronal port) hole. The height of each chamber is 500 μm. If a lymphatic compartment is added, then the design has two extra Ø 4 mm media channel holes to provide access to the lymphatics and two extra Ø 1.2 mm holes for lymphatics-gel loading. See FIG. 4A-4C and 5A-5C for actual schematics of the microfluidic devices and connections to the pumps. 19 45630538.1 The pump 70 is shown in FIG.4A-4C. This includes an air supply that can be cycled and delivered to air channel inlet 72 to drive fluid into a chamber 74. The input chamber 74 together with deflection of a membrane (not shown) determines how much fluid is transported per pneumatic cycle. The air channel inlet 72 is the connection point to the pneumatic control system. Two one way valves 76a, 76b determine the direction of the flow and ensure there is no back flow through channels 86a, 86b and 90a, 90b. Capacitors 80a, 80b charge to a user-chosen pressure to ensure the pressure drop across the connected resistance is constant, with small fluctuations relative to the mean pressure, during operation. Device ports 82a, 82b are the connection points to the connected resistance, i.e., the microfluidic device(s). FIG.5A-5C are prospective schematics of the device 100 for forming a model with the meningeal lymphatics. FIG.5A shows wells 80a and 80b which the cell culture fluid is pumped in and out using the pump 70 shown in FIG.5B. FIG.5C is a prospective view of a mold 120 to make twelve of the devices 100. Example 2: Neurovascular unit model (Blood-brain barrier and neurons) Neural expansion medium was made using 500 mL of DMEM / F12 with L-glutamine and sodium bicarbonate (GIBCO, 11320-033), 500 μL of heparin (STEMCELL Technologies, 07980), 10 mL of B27 (GIBCO, 17504- 044), 400 μL of bFGF (Stemgent, 03-0002, reconstituted at 25 μg / mL and stored at -80°C), 500 μL of EGF (SIGMA, E9644, reconstituted at 20 μg / mL and stored at -80°C), and 5 mL of Penicillin-Streptomycin-Amphotericin B 100x (Lonza, 17-745E). Neural differentiation medium was made using 500 mL of DMEM / F12 with L-glutamine and sodium bicarbonate (GIBCO, 11320-033), 500 μL of heparin (STEMCELL Technologies, 07980), 10 mL of B27 (GIBCO, 17504-044), and 5 mL of Penicillin / Streptomycin / Amphotericin B 100x (Lonza, 17-745E). Microfluidic seeding medium was made using VascuLife (LifeLine Technologies, LL-0003) with ¼ of the supplemented heparin sulfate and 2% v / v FBS (complete). 20 45630538.1 One week before seeding (day -7) microfluidic devices, human astrocytes (SCIENCELL, 1800 at passage 4) were thawed at 8-10,000 cells / cm2into a Poly-L-Lysine (PLL) (SIGMA, P4707)-coated flask and maintained with astrocyte medium (SCIENCELL, 1801 at passage 4). The following day (day -6), human brain vascular pericytes (SCIENCELL, 1200) were thawed at 8-10,000 cells / cm2into a PLL-coated flask and maintained with pericyte medium (SCIENCELL, 1201). On day -6, neural progenitor cells (ReN cell) were also thawed at 20-25,000 cells / cm2and seeded into a MATRIGEL® (Corning, 354234)-coated (1:100 dilution of MATRIGEL®: DMEM / F12 medium) flask and maintained with neural expansion medium. Neural progenitor cells expressing familial AD mutations were engineered using ReNcell VM human neural progenitor cells (EMD Millipore, SCC008). Cells were all expanded and passaged once before seeding into microfluidic devices. Astrocytes and pericytes were disassociated with TrypLE express (ThermoFisher, 12605028), 1 mL per T25, and incubated at 37°C for 3-5 mins. Neural progenitor cells were disassociated with Accutase (Gibco, A1110501), 1 mL per T25, and incubated at 37°C for 3-5 mins. Four days before device seeding (day -4), human brain microvascular endothelial cells (HBMECs, Angioproteomie, cAP-002 at passage 8) were thawed at 3- 5,000 cells / cm2into fibronectin (Millipore Sigma, FC010)-coated flasks and maintained in complete VascuLife medium supplemented with total 10% v / v FBS and full heparin sulfate. To prepare the microfluidic devices for seeding, a triculture mixture of astrocytes, pericytes, and HBMECs was created. All BBB cells were dissociated with TrypLE express, 2 mL per T75, incubated at 37°C for 3-5 mins, and neutralized with microfluidic seeding medium. Cells were then resuspended in cold microfluidic seeding medium supplemented with 4 U / mL of thrombin (SIGMA, T4648-1KU, reconstituted at 100 U / mL in 0.1% w / v bovine serum albumin solution and stored at -80°C) at the following concentrations: 36 x 106 / mL (HBMECs), 3 x 106 / mL (pericytes), and 6 x 106 / mL (astrocytes). To form the BBB triculture, 5 μL from each cell type were mixed to create 15 μL of total cell mixture per microfluidic device.15 μL of fibrinogen (Sigma, F8630-1G, reconstituted at 6 mg / mL in PBS and stored at -80°C) was mixed with 15 μL of cells mixture and immediately 21 45630538.1 loaded into microfluidic devices through the gel-loading inlet. Before loading, an 8 mm wide PDMS disk was gently pressed on top of the neural media chamber to prevent BBB cells from overflowing through the neural inlet and to facilitate the smooth filling of the BBB compartment. The PDMS disk was then gently lifted, and the gel solution was slowly removed until the neural chamber was empty, but the BBB chamber remained full. Devices were stored in sterilized and autoclaved boxes with a small petri dish filled with PBS to avoid them drying out. Microfluidic devices were kept in the incubator at 37°C for at least 15 mins to allow gel crosslinking, and then the vascular media channels were filled with 150 μL of complete VascuLife medium. Next, the neural progenitor cells (Ctrl and AD) were dissociated using ACCUTASE®and neutralized with neural expansion medium. Cells were then resuspended at 1 x 106 / 7 μL in neural differentiation medium and were mixed with 7 μL of cold MATRIGEL®(1:1) and loaded by pipet into the neural chamber of the microfluidic devices. Devices were then incubated at room temperature for 5-10 mins, followed by 30 mins at 37°C to allow the gel to form. The neural media chambers were then filled with 75 μL of neural differentiation medium. One day after seeding, the vascular medium was replaced with 150 μL of complete VASCULIFE®medium and 50 μL of neural differentiation medium was added to the neural cells. Two days after seeding, flow was introduced to the BBB culture by placing syringes (BH Supplies 1 mL Luer Slip Tip Syringes-BH1LS) into the media ports. Half of the tip of sterile 1 mL syringes was cut at 45 degrees to avoid blocking the vascular chambers. Similarly, they were cut to hold a maximum of 300 μL media and fit within storage boxes. Four syringes were inserted per device, and the vascular medium was added to one side up to 250 μL on each syringe (flow entry chamber) while keeping the other side ~50 μL medium (flow exit chamber). Vascular medium was removed daily from the flow exit chamber and added to the flow entry chamber to produce a transient flow of medium through the microfluidic devices. Additional differentiation medium was also added to the neural compartment as needed. Twice a week, all vascular medium was removed and replaced with fresh medium, while only half of the neural 22 45630538.1 differentiation medium was refreshed twice a week. If studies are focused only on the BBB, the microfluidic device can be fabricated without the central hole for the neuronal compartment. Example 3: Formation of neurospheres and transfer to the microfluidic device One day before initiating the co-culture in the microfluidic device, 10–12 neurospheres were collected in a microcentrifuge tube. The number of neurospheres was adjusted to match the number of cells employed in the previous microphysiological AD study (Shin, et al. Adv. Sci.6(20) August 2019). Equal volumes of the following three stock solutions were mixed to generate a homogenous cell suspension: 36 × 106primary human brain microvascular endothelial cells / mL, 6 × 106primary human astrocytes / mL, and 3 × 106primary human brain vascular pericytes / mL. All stock solutions were prepared using 4 U / mL of thrombin (Sigma Aldrich, T4648) in VASCULIFE®medium (LifeLine Technologies, LL-0003) supplemented with 0.19 U / mL of heparin.15 μL of the cell suspension were mixed with 15 µL of a 6 mg / mLfibrinogen stock solution (Sigma Aldrich, F8630) and loaded into the central BBB chamber via the BBB loading port. To prevent the hydrogel from escaping through the microfluidic device’s neurosphere chamber, a PDMS stamp was placed on top of the PDMS reservoir prior to loading. After the channel wasfilled, the remaining hydrogel in the neurosphere compartment was removed by aspirating 6 µL of the hydrogel mixture from the inlet of the microfluidic device. After polymerization for 15 min at 37°C and the addition of 140 µL of VASCULIFE® medium (LifeLine Technologies, LL-0003) supplemented with 0.19 U / mL of heparin to each side of the microfluidic device, the neurosphere chamber of the microfluidic device was carefully wetted using neurosphere media. After wetting, the neurospheres were transferred into the central neurosphere chamber by the use of a cut pipette tip. After the neurospheres settled to the bottom of the neurosphere chamber the remaining medium was carefully removed employing a gel loading tip. Thereafter, 10 µL of MATRIGEL®(Corning, 356230) were carefully pipetted into the neurosphere chamber and polymerized for 45 min at 37°C, 23 45630538.1 before adding 50 µL of conditioned neurosphere medium. After 2 days of static culture, a pressure gradient was established in the microfluidic devices by inserting two cut syringes into the media reservoirs on one side of the microfluidic device and adding 600 µL media. Note that as the vascular network developed, the mediaflow increasingly passed through the low resistance vascular network rather than the gel, causing the pressure to equalize more rapidly, in ~30 min once fully perfusable. The syringes were refilled every 24 h using fresh VASCULIFE®medium (LifeLine Technologies, LL-0003) supplemented with 0.19 U / mL of heparin. Neurosphere media was replaced every second day with conditioned media collected from the 96-well plates. Example 4: Development of a Microfluidic Co-Culture System for Modeling BBB in Alzheimer’s disease Incorporating established methodologies and the design that allows co-culturing using two different hydrogels, microfluidic chips were fabricated from polydimethylsiloxane (PDMS) elastomer, which facilitate the co-culture of BBB and neurons / astrocytes to model AD in vitro. Methods The mold used for the fabrication of the microfluidic devices was designed in AutoCAD (Autodesk, Inc.), and imported in Fusion 360 (Autodesk, Inc.) to generate corresponding tool paths, followed by milling a DELRIN®block with a micro-CNC milling machine (Bantam Tools). Polydimethylsiloxane (PDMS) based microfluidic devices were then fabricated and bonded to a glass coverslip using the procedure established previously (https: / / doi.org / 10.1038 / s41596-021-00635-w). The microfluidic device has four Ø 4 mm media channel holes, two Ø 1.2 mm holes for BBB- gel loading and one Ø 2 mm central (neuronal port) hole. The height of each chamber is 500 μm. Neural expansion medium was made using 500 mL of DMEM / F12 with L-glutamine and sodium bicarbonate (Gibco, 11320-033), 500 μL of heparin (STEMCELL Technologies, 07980), 10 mL of B27 (Gibco, 17504- 044), 400 μL of bFGF (Stemgent, 03-0002, reconstituted at 25 μg / mL and stored at -80°C), 500 μL of EGF (Sigma, E9644, reconstituted at 20 μg / mL and stored at -80°C), and 5 mL of Penicillin-Streptomycin-Amphotericin B 24 45630538.1 100x (Lonza, 17-745E). Neural differentiation medium was made using 500 mL of DMEM / F12 with L-glutamine and sodium bicarbonate (Gibco, 11320-033), 500 μL of heparin (STEMCELL Technologies, 07980), 10 mL of B27 (Gibco, 17504-044), and 5 mL of Penicillin / Streptomycin / Amphotericin B 100x (Lonza, 17-745E). Microfluidic seeding medium was made using VASCULIFE®(LifeLine Technologies, LL-0003) with ¼ of the supplemented heparin sulfate and 2% v / v FBS (complete). One week before seeding (day -7) microfluidic devices, human astrocytes (ScienCell, 1800 at passage 4) were thawed at 8-10,000 cells / cm2into a Poly-L-Lysine (PLL) (Sigma, P4707)-coated flask and maintained with astrocyte medium (ScienCell, 1801 at passage 4). The following day (day -6), human brain vascular pericytes (ScienCell, 1200) were thawed at 8- 10,000 cells / cm2into a PLL-coated flask and maintained with pericyte medium (ScienCell, 1201). On day -6, neural progenitor cells (ReN cell) were also thawed at 20-25,000 cells / cm2and seeded into a MATRIGEL®(Corning, 354234)-coated (1:100 dilution of MATRIGEL®: DMEM / F12 medium) flask and maintained with neural expansion medium. Neural progenitor cells expressing familial AD mutations were engineered using ReNcell VM human neural progenitor cells (EMD Millipore, SCC008). Cells were all expanded and passaged once before seeding into microfluidic devices. Astrocytes and pericytes were disassociated with TrypLE express (ThermoFisher, 12605028), 1 mL per T25, and incubated at 37°C for 3-5 mins. Neural progenitor cells were disassociated with ACCUTASE®(Gibco, A1110501), 1 mL per T25, and incubated at 37°C for 3-5 mins. Four days before device seeding (day -4), human brain microvascular endothelial cells (HBMECs, Angioproteomie, cAP-002 at passage 8) were thawed at 3-5,000 cells / cm2into fibronectin (Millipore Sigma, FC010)-coated flasks and maintained in complete VascuLife medium supplemented with total 10% v / v FBS and full heparin sulfate. To prepare the microfluidic devices for seeding, a triculture mixture of astrocytes, pericytes, and HBMECs was created. All BBB cells were disassociated with TrypLE express, 2 mL per T75, incubated at 37°C for 3-5 mins, and neutralized with microfluidic seeding medium. Cells were then 25 45630538.1 resuspended in cold microfluidic seeding medium supplemented with 4 U / mL of thrombin (SIGMA®, T4648-1KU, reconstituted at 100 U / mL in 0.1% w / v bovine serum albumin solution and stored at -80°C) at the following concentrations: 36 x 106 / mL (HBMECs), 3 x 106 / mL (pericytes), and 6 x 106 / mL (astrocytes). To form the BBB triculture, 5 μL from each cell type were mixed to create 15 μL of total cell mixture per microfluidic device.15 μL of fibrinogen (Sigma, F8630-1G, reconstituted at 6 mg / mL in PBS and stored at -80°C) was mixed with 15 μL of cells mixture and immediately loaded into microfluidic devices through the gel-loading inlet. Before loading, an 8 mm wide PDMS disk was gently pressed on top of the neural media chamber to prevent BBB cells from overflowing through the neural inlet and to facilitate the smooth filling of the BBB compartment. The PDMS disk was then gently lifted, and the gel solution was slowly removed until the neural chamber was empty, but the BBB chamber remained full. Devices were stored in sterilized and autoclaved boxes with a small petri dish filled with PBS to avoid them drying out. Microfluidic devices were kept in the incubator at 37°C for at least 15 mins to allow gel crosslinking, and then the vascular media channels were filled with 150 μL of complete VascuLife medium. Next, the neural progenitor cells (Control and AD) were dissociated using ACCUTASE®and neutralized with neural expansion medium. Cells were then resuspended at 1 x 106 / 7 μL in neural differentiation medium and were mixed with 7 μL of cold MATRIGEL®(1:1) and loaded by pipet into the neural chamber of the microfluidic devices. Devices were then incubated at room temperature for 5-10 mins, followed by 30 mins at 37°C to allow the gel to form. The neural media chambers were then filled with 75 μL of neural differentiation medium. One day after seeding, the vascular medium was replaced with 150 μL of complete VASCULIFE®medium and 50 μL of neural differentiation medium was added to the neural cells. Two days after seeding, flow was introduced to the BBB culture by placing syringes (BH Supplies 1 mL Luer Slip Tip Syringes-BH1LS) into the media ports. Half of the tip of sterile 1 mL syringes was cut at 45 degrees to avoid blocking the vascular chambers. Similarly, they were cut to hold a maximum of 300 μL media and fit within 26 45630538.1 storage boxes. Four syringes were inserted per device, and the vascular medium was added to one side up to 250 μL on each syringe (flow entry chamber) while keeping the other side ~50 μL medium (flow exit chamber). Vascular medium was removed daily from the flow exit chamber and added to the flow entry chamber to produce a transient flow of medium through the microfluidic devices. Additional differentiation medium was also added to the neural compartment as needed. Twice a week, all vascular medium was removed and replaced with fresh medium, while only half of the neural differentiation medium was refreshed twice a week. If studies are focused only on the BBB, the microfluidic device can be fabricated without the central hole for the neuronal compartment. Results Incorporating established methodologies and the design that allows co-culturing using two different hydrogels, microfluidic chips fabricated from polydimethylsiloxane (PDMS) elastomer which facilitate the co-culture of BBB and neurons / astrocytes to model AD in vitro. The microfluidic chip offers versatility through the integration of three separate but communicating compartments, each serving a specific purpose within the co-culture system: i) BBB vasculature compartment: a central circular chamber containing the BBB vasculature, with the intention to accurately reproduce both the structural and functional characteristics of the BBB, ii) neuronal cell culture compartment: within the BBB vasculature compartment lies another circular chamber in which neurons and astrocytes are embedded. This chamber is open and accessible from the top, which allows for separate media introduction and extraction, facilitating precise sampling of the neuronal culture, and iii) two media channels: two additional channels flank the central compartment, specifically designed for controlled media administration to the co-culture system. FIG.6A-6D are confocal fluorescence micrographs of the cells in the microfluidic device. FIG.6A shows the meningeal lymphatics over the blood brain barrier; FIG.6B shows the neurons over the blood brain barrier over the meningeal lymphatics. FIG.6C shows the meningeal lymphatic cells. FIG.6D shows the endothelial cells forming the blood brain barrier, using dextran as vascular perfusability proof.. 27 45630538.1 Example 5: Development of a Microfluidic Co-Culture System for Modeling BBB in Alzheimer’s disease using neurospheres The same methodology and materials were used as described above but instead of using suspended neurons and differentiating them inside the microfluidic device, neurospheres that had already started to differentiate were produced. Methods One day before initiating the co-culture in the microfluidic device, 10–12 neurospheres were collected in a microcentrifuge tube. The number of neurospheres was adjusted to match the number of cells employed in the previous microphysiological AD study described above, based on Shin et a. 2019. Equal volumes of the following three stock solutions were mixed to generate a homogenous cell suspension: 36 × 106primary human brain microvascular endothelial cells / mL, 6 × 106primary human astrocytes / mL, and 3 × 106primary human brain vascular pericytes / mL. All stock solutions were prepared using 4 U / mL of thrombin (Sigma Aldrich, T4648) in VASCULIFE®medium (LifeLine Technologies, LL-0003) supplemented with 0.19 U / mL of heparin.15 μL of the cell suspension were mixed with 15 µL of a 6 mg / mLfibrinogen stock solution (Sigma Aldrich, F8630) and loaded into the central BBB chamber via the BBB loading port. To prevent the hydrogel from escaping through the microfluidic device’s neurosphere chamber, a PDMS stamp was placed on top of the PDMS reservoir prior to loading. After the channel wasfilled, the remaining hydrogel in the neurosphere compartment was removed by aspirating 6 µL of the hydrogel mixture from the inlet of the microfluidic device. After polymerization for 15 min at 37°C and the addition of 140 µL of VASCULIFE®medium (LifeLine Technologies, LL-0003) supplemented with 0.19 U / mL of heparin to each side of the microfluidic device, the neurosphere chamber of the microfluidic device was carefully wetted using neurosphere media. After wetting, the neurospheres were transferred into the central neurosphere chamber by the use of a cut pipette tip. After the neurospheres settled to the bottom of the neurosphere chamber, the remaining medium was carefully removed employing a gel loading tip. Thereafter, 10 µL of MATRIGEL®28 45630538.1 (CORNING®, 356230) were carefully pipetted into the neurosphere chamber and polymerized for 45 min at 37°C, before adding 50 µL of conditioned neurosphere medium. After 2 days of static culture, a pressure gradient was established in the microfluidic devices by inserting two cut syringes into the media reservoirs on one side of the microfluidic device and adding 600 µL media. Note that as the vascular network developed, the mediaflow increasingly passed through the low resistance vascular network rather than the gel, causing the pressure to equalize more rapidly, in ~30 min once fully perfusable. The syringes were refilled every 24 h using fresh VASCULIFE®medium (LifeLine Technologies, LL-0003) supplemented with 0.19 U / mL of heparin. Neurosphere media was replaced every second day with conditioned media collected from the 96-well plates. Example 6: Development of a Microfluidic Co-Culture System for Modeling BBB and meningeal lymphatics (ML) in Alzheimer’s disease The microfluidic design consists of a vascular component with three channels (a central channel for 3D BBB and neuron co-culture and two side channels for BBB medium that run parallel to it) and a lymphatic component, composed of one channel that forms a T-junction with the central channel where media can come into contact with the gel in the BBB compartment. The BBB compartment incorporates a neuronal culture islet with an open-top well for seeding brain spheroids to come into contact with three media channels. Conditions were as described by Serrano, et al. (2023) DOI: 10.1002 / advs.202302903. To form the 3D AD ML, the device is seeded with a mixture of bEC 6: PC 0.5: AC 1 encapsulated in a fibrin gel (3 mg / mL) in the central channel, fill the side channels with media, and allow the formation of perfusable vessels open to the media channels over 5 days, recapitulating the BBB. Neurons (suspended neurons or neurospheres) are seeded separately into the open- top well after BBB formation, allowing the neuronal cells to be in contact with a fully mature, brain-like microvasculature. Finally, the third media channel is filled with cell culture medium, and a monolayer of primary lymphatic endothelial cells is seeded onto the side of the gel. Once the lymphatic cells are seeded, a syringe-pressure height is applied to the side ports that connect with the vasculature, which leads to 29 45630538.1 leakage from the vessels into the surrounding matrix and exposes the lymphatic monolayer to pathological levels of interstitial flow (4-10 μm / s), thus potentiating lymphangiogenesis into the gel region, generating the lymphatic microvasculature. The lymphatic vessels will sprout in the gel region. The system allows simultaneous measurement of molecular and fluid transport across the vascular and lymphatic endothelia under physiological and pathological conditions. After 4-6 days of continuous flow, this lymphatic vascularized gel region provides the basis for modeling the role of meningeal lymphatics during AD progression. Lymphangiogenic growth factors (50 ng / ml) coupled with or replaced by low interstitial flow (0.1 to 11 µm / sec) stimulate the lymphatic cells to sprout into the gel.. Example 7: Neurovascular unit with meningeal lymphatics model An on-chip lymphatic microvasculature in a neurovascular model that mimics the morphological and functional properties of in vivo lymphatic capillaries was also developed. The formation of lymphatic vasculature in this microfluidic platform is achieved by first seeding a confluent monolayer of human lymphatic endothelial cells at the media-gel interface and, subsequently, introducing a pro-angiogenic stimulus (growth factors or interstitial flow) at the opposite media channel to promote lymphangiogenesis into the central 3D gel channel. By continuously exposing the lymphatic endothelial cells to pro-angiogenic factors over 4-6 days, the resultant lymphatic network that invades the 3D gel region recapitulates key physiological characteristics, including vessel density, diameter, and dead-ended vessels. To form the 3D AD BBB-lymphatics model, the device is seeded with a mixture of bEC 6: PC 0.5: AC 1 encapsulated in a fibrin gel (3 mg / mL) in the central channel, the side channels filled with media, and the formation of perfusable vessels open to the media channels allowed over a period of 5 days, recapitulating the BBB. Neural spheroids are seeded separately into the open- top well after BBB formation, allowing the neuronal cells to be in contact with a fully mature, brain-like microvasculature. Finally, the third media channel is filled with cell culture medium, and a monolayer of primary lymphatic endothelial cells is seeded 30 45630538.1 onto the side of the gel. Once the lymphatic cells are seeded, flow is added to the microfluidic system which exposes the lymphatic monolayer to pathological levels of interstitial flow (4-10 μm / s), thus potentiating lymphangiogenesis into the gel region, generating the lymphatic microvasculature. The lymphatic vessels will sprout in between the gel region next to the BBB compartment. The system allows simultaneous measurement of molecular and fluid transport across the vascular and lymphatic endothelia under physiological and pathological conditions. After 4-6 days of continuous flow, this lymphatic vascularized gel region provides the basis for modeling the role of meningeal lymphatics during AD progression. The microfluidic device facilitates the compartmentalization of culture media, extracellular matrix (ECM), and cells within a 3D environment. Additionally, the PDMS-based device allowed for high- resolution imaging of microscale, cellular events via confocal microscopy. Diffusive transport of a 70 kDa FITC-dextran through the ECM was determined by measuring the fluorescence intensity profile across the gel channel 120 min after introducing the fluorophore to the media channel. The resulting profile was fitted to the 1D unsteady solution of Fick's Second Law from which the effective diffusion coefficient (D) of ≈45 µm2s−1was obtained that is within the range of values measured ex vivo for the same fluorescent tracer in tissues. Using this computed diffusion coefficient, the timescale required for diffusion of growth factors across the gel region as ≈w2 / D (with w the width of the gel channel) was estimated to be about two hrs, which is a small fraction of the several days required for lymphatic vascular growth, sufficiently short to ensure the adequate delivery of growth factors. FIG.6A-6D are confocal fluorescent micrographs of the cells in the microfluidic device. FIG.6A shows the meningeal lymphatics adjacent to the blood brain barrier; FIG.6B shows the neurons adjacent to the blood brain barrier and the meningeal lymphatics. FIG.6C shows the meningeal lymphatic cells. FIG.6D shows the endothelial cells forming the blood brain barrier, after perfusion with fluorescent dextran to demonstrate vascular perfusability. 31 45630538.1 FIG.6C shows a projected confocal image of a lymphatic monolayer seeded at the media channel with podoplanin staining the cell membrane demonstrating lymphatic sprouting for experimental conditions corresponding to stimulus by growth factors and / or interstitial flow and corresponding to in vivo values. Representative growth factors include VEGF-C, ANG-1, and HGF. High interstitial fluid flow is approximately 5 µm s−1. The solute drainage through the engineered lymphatic vasculature is characterized by adapting a standard in vivo clearance assay. Fluorescent tracers of varying molecular weights and composition (10 kD and 70 kDa Dextran, and BSA) were introduced into the BBB vessels, which entered into the surrounding matrix and were convected by interstitial flow toward and into the lymphatic vasculature by controlling the pressure difference between the inside of the BBB and the outlet of the lymphatic channel. The increase in fluorescent intensity was measured over time at the opposite lymphatic-media channel under physiological levels of interstitial flow. This showed that the engineered 3D lymphatic vasculature exhibited higher solute drainage rates than that of microfluidic models with a lymphatic monolayer or a bare gel (devoid of lymphatic cells) with values approaching in vivo measurements. In addition to the diffusive transport of growth factors, vascular fluid flow drives the mass transfer of biomolecules and proteins throughout tissues. This convective transport of fluid through tissue ECM (interstitial flow) and into the lymphatics facilitates the percolation and drainage of extracellular fluid and solutes via the lymphatics to maintain hydrostatic and oncotic pressure homeostasis in tissues. The microfluidic platform mimics this convective transport by introducing a hydrostatic pressure imbalance between media channels, which drives flow across the compartmentalized ECM region. The pressure differences imposed drive interstitial fluid flow at about 1 µm s-1representing physiological / homeostatic conditions or, at higher velocities at about 4 µm s-1relevant to an inflamed tissue and tumor microenvironments. Starting from a confluent monolayer of human lymphatic endothelial cells at the media-gel interface, growth factors 32 45630538.1 previously identified as key mediators of developmental lymphangiogenesis (VEGF-C, ANG-1, HGF) were introduced into the opposite media channel to generate a localized source of factors, that would steadily diffuse toward the lymphatic monolayer, resulting in lymphatic sprouting into the central ECM compartment. The addition of lymphangiogenic growth factors, individually or in combination, consistently resulted in lymphatic sprouts with diameters well-within in vivo values. However, only the simultaneous addition of all three growth factors (100 ng mL−1each) led to the full range of values corresponding to in vivo lymphatic vascular coverage area. Interstitial flow stimulates the growth of lymphatic capillaries. Sprouting is induced against the direction of interstitial flow as it passes from the extracellular space toward the vascular compartment. The high flow (HF) regime of pathological nature (3–6 µm s−1) produced maximal vascular invasion was significantly greater under HF conditions with HF matching. This finding is in line with in vivo pathological microenvironments (i.e., a developing tumor or an inflamed wound site) where a buildup of interstitial fluid pressure, from leaky blood vessels, leads to higher interstitial fluid flow toward the lymphatics, thus evoking lymphangiogenesis. Blind / blunt-ended, small-scale (≈25 µm), 3D, and lumenized structures were also observed, along with lymphatic-specific markers such as lymphatic vascular endothelial receptor-1 (LYVE-1), and the upregulation of transcriptional factor PROX-1 each of which is indicative of in vivo lymphatic features. In summary, the lymphatic system serves an integral role in maintaining tissue homeostasis by clearing excess fluid, plasma proteins, pathogenic agents (antigens), and endo- / exogenous carriers (vesicles / exosomes, therapeutics) from the peripheral tissues into the systemic circulation. Physiologically, these factors are exchanged between the blood vasculature and the tissue interstitium due to differences in oncotic and hydrostatic pressure. This biological mass transfer is proceeded by their transit from the tissue interstitial space toward the lymphatics due to lower intraluminal concentration and pressure, thus resulting in lymphatic drainage. 33 45630538.1 The lymphatic vasculature is essential for immune cell trafficking during host immune responses. In the event of tissue infection and inflammation, local stromal cells are activated by pathogenic signals, such as bacterial lipopolysaccharide, and respond with the release of inflammatory cytokines including tumor necrosis factor-alpha (TNF-α), transforming growth factor-β (TGF-β), interleukins, amongst others. Once the tissue site is primed by pro-inflammatory signals, innate and adaptive immune cells are recruited by blood and lymphatic endothelial-secreted chemokines, which then initiates immune activation and response. Thus, in order to adequately study the signaling and migratory events that lead to the recruitment of immune cells to the lymphatics, the full range of fluid, protein, and cellular transport phenomena need to be recapitulated in in vitro models. This platform can recreate the pathological microenvironment that facilitates the intravascular recruitment of immune cells by the lymphatics. Recombinant-human cytokines can be introduced into the system to mimic stromal-secreted inflammatory factors. Additionally, immune cells can be introduced into the system at the adjacent media channel while establishing the appropriate hydraulic pressure difference across the gel compartment to impart a pathological level of interstitial flow toward the lymphatic channel. As a model for immune cell response, isolated human peripheral blood mononuclear cells (PBMCs) which include a broad population of immune cells that are naturally found in the systemic circulation, and are known to extravasate and migrate into the lymphatic periphery. The magnitude of lymphatic-elicited immune response is assessed by counting the number of PBMCs recruited to the lymphatic channel after 24 hours of induction. In initial experiments, PBMCs consistently migrated into the opposite channel at higher numbers for devices incorporating lymphatics (high-flow stimulated, growth factor-grown, and monolayer), in comparison to the bare gel system. To validate the increased lymphatic recruitment of immune cells under inflammatory conditions, devices with high flow-engineered lymphatics were treated with TNF-α prior to PBMC perfusion, in parallel with a set of devices without TNF-α exposure serving as the control. It was confirmed that, under inflammatory stimulus (TNF-α pre- treatment), lymphatics significantly increase the recruitment of PBMCs. 34 45630538.1 Additionally, decellularized samples consistently exhibited lower infiltration numbers regardless of whether the devices were preconditioned or not with TNF-α. To further examine the preferential migration and infiltration of PBMCs, confocal imaging of the gel region was done to visualize the spatial distribution of PBMCs immediately after performing the infiltration assay. Both images and quantitative data indicate that PBMCs were preferentially localized closer to the media region where they are introduced, for all conditions except the TNF-α-stimulated lymphatics. For the latter, the highest population of PBMCs corresponded to the region co-localized with lymphatic vessels. These data indicate that, under inflammatory stimulus, the on-chip lymphatics provide directional cues that elicit the preferential migration and infiltration of PBMCs. Immune cell recruitment to specific sites is guided by chemokines. These chemotactic cytokines provide coordinated delivery of cell-secreted ligands to specific immune cell surface receptors. The lymphatic secreted ligands CCL21 and CCL19 attract immune cells via their CCR7 receptor in various pathological conditions including immunogenic response and cancer metastasis, and stromal / endothelial secreted CXCL12 coordinates the homing of immune cells for immune maintenance and development by their CXCR4 surface receptor. To determine if these receptors are elicited during inflammatory immune recruitment in the tissue-engineered system, receptor- blocking experiments were performed by incubating PBMCs in a buffer solution containing CCR7 and / or CXCR4 antibodies, or control isotypes, prior to performing the infiltration assay with HF-grown lymphatics stimulated with TNF-α. A consistent and significant decrease in the number of infiltrated PBMCs was observed when their cell surface receptors, CCR7 and / or CXCR4, were functionally blocked, compared to the IgG isotype control, despite being introduced into devices with inflammatory-stimulated lymphatics. Furthermore, analysis of the spatial distribution of PBMCs within the gel region revealed that solely for the IgG isotype control samples, the PBMCs preferentially migrated and co-localized at the inflamed- lymphatics region. All other samples with neutralizing antibody treatment displayed higher PBMC numbers closer to the media channel. Thus, the CCR7 and CXCR4 immune cell surface receptors, responsible for immune 35 45630538.1 homing, are elicited in the engineered pathological model for immune cell recruitment by the lymphatics in response to local inflammatory stimulus. Example 8: Model System Including immune Cells Immune cells can be added to the lymphatics, such as dendritic cells, macrophages, T cells and Natural Killer (NK) cells. The systems can be used to investigate immune system reaction in several cases such as AD by adding cells involved in inflammation in the brain, specifically microglia. Microglia cells can be introduced in the ReN cell compartments of the model on top of a layer of MATRIGEL® to determine if they are activated and recruited to the ReN cells in the 3D AD BBB-ML. Human immortalized microglial cell lines (SV-40) and human iPSC-derived microglia-like cells (iMGL) can be used. To quantify the recruitment of microglial cells, immunofluorescence staining can be used to visualize microglia-specific markers, such as Iba1, CD11b, and CD33, using confocal microscopy. To characterize migration, individual microglial cells will be monitored for 3 days using time-lapse confocal z-stack imaging. CD33 is particularly crucial because, in microglia, CD33 may play a role in the phagocytosis of Aβ. One can also examine the interaction of microglia cells with the neurovascular and meningeal lymphatic endothelial cells and any changes in this interaction with disruption to the neurovascular or meningeal lymphatic endothelial cells. The secretion of cytokines (V-PLEX Neuroinflammation Panel 1 Human Kit, MSD) can be screened to determine if it is the language used to communicate between these cell types. Furthermore, the study after the addition of immune cells in the meningeal compartment can be assessed to further dissect the interaction between different cell types during pathological conditions. An anti-CCL2 antibody can be used to modulate microglia recruitment and the crosstalk with BBB and meningeal lymphatics. Studies show that neutralization of CCL2 chemokine signaling using an anti- CCL2 antibody reduced microglia migration in response to neuron and astrocyte co-culture. IV. Methods of Use of the Model The brain model can be used to investigate the effect of disease related products and / or compounds such as therapeutic, prophylactic or 36 45630538.1 diagnostic agents, for their effects on normal, diseased, defective, injured or aging brains. In vitro model for Alzheimer’s Disease The blood–brain barrier (BBB) serves as a selective filter that prevents harmful substances from entering the healthy brain. Dysfunction of this barrier is implicated in several neurological diseases. In the context of Alzheimer’s disease (AD), BBB breakdown plays a significant role in both the initiation and progression of AD. This study introduces a three- dimensional (3D) self-assembled in vitro model of the human neurovascular unit to recapitulate some of the complex interactions between the BBB and AD pathologies. It incorporates primary human brain endothelial cells, pericytes and astrocytes, and stem cell-derived neurons and astrocytes harboring Familial AD (FAD) mutations. Over an extended co-culture period, the model demonstrates an increase in BBB permeability, dysregulation of key endothelial and pericyte markers, and morphological alterations mirroring AD pathologies. The model enables visualization of amyloid-beta (Aβ) accumulation in both neuronal and vascular compartments. This model may serve as a versatile tool for neuroscience research and drug development to provide insights into the dynamic relationship between vascular dysfunction and AD pathogenesis. This model allows one to determine the mechanism(s) by which adult-born neurons and brain-derived neurotrophic factor (“BDNF”) collaboratively play a critical role in resilience to AD-related pathology and cognitive deficits; screen for pharmacological and natural compounds that increase neurogenesis and BDNF or that reduce Aβ and tau pathologies; study the roles of genetics and other factors such as hormones, in AD pathogenesis; explore the roles of astrocytes in tauopathy; and develop 3D neurovascular, meningeal lymphatics-on-a chip to study the functions of blood-brain barrier (BBB) and lymphatics during AD progression. Further developments incorporate advances and introduces a 3D self-assembled human BBB in vitro model within microfluidic devices, utilizing primary human brain ECs, ACs, PCs, and stem cell-derived neurons and astrocytes harboring Familial AD (FAD) mutations. 37 45630538.1 Modeling of Impact on BBB Permeability, neuronal activity and lymphatic drainage; Impact of Immune cells By introducing proteins, particles and other agents in the model, the effect of these on the BBB permeability can be tested. The detailed permeability protocol and the image analysis code can be found in Hajal et al., 2022. For BBB permeability, several sizes of fluorescent tracers can be used with time intervals of 6 min and 12 min (depending on the molecular size), between the two imaging timepoints. Neuronal and / or lymphatic compartment can be added using different microfluidic design and test BBB permeability. Sensing neuronal activity from neuronal compartment (suspended neurons or neurospheres), such as calcium imaging, can provide a more comprehensive understanding of the neuronal response in AD environments, and it can also reveal neurovascular interactions depending on the model. The solute drainage through the engineered lymphatic vasculature is characterized by adapting a standard in vivo clearance assay. Fluorescent tracers of varying molecular weights and composition (10 kD and 70 kDa Dextran, and BSA) were introduced into the non-endothelial media channel and measured the increase in fluorescent intensity over time at the opposite lymphatic-media channel under physiological levels of interstitial flow. This technique was validated and showed the engineered 3D lymphatic vasculature exhibited higher solute drainage rates than that of microfluidic models with a lymphatic monolayer or a bare gel (devoid of lymphatic cells) with values approaching in vivo measurements. This demonstrates the high degree of physiological mimicry achieved by the current platform. Media can be collected from the microfluidic devices every week for MSD analysis.50 μL of media are collected from both the neural and vascular chambers of each microfluidic device and stored at -20°C until ready for analysis. Levels of different soluble factors in the media of both the neural and vascular compartments can be measured using kit and associated protocols for MSD analysis. Media from microfluidic devices are carefully removed, and devices are rinsed with PBS. The devices are then fixed with 4% PFA overnight and 38 45630538.1 are washed three times with PBS. Microfluidic devices are blocked overnight and washed twice with PBS. After permeabilization with 1X TBST (Boston BioProducts - IBB-580X) containing 0.5% Triton X-100 (Millipore Sigma, X-100) and 4% donkey serum (ABCAM, ab7475), primary antibodies, including 3D6 antibody (a gift from Lilly, 1:500), VE-Cadherin (R&D, AF938, 1:500), PDGFRα + PDGFRβ (ABCAM, ab32570, 1:100) and S-100 (SIGMA-ALDRICH, S2532-.2ML, 1:50), are added and incubated overnight. The next day, microfluidic devices are washed twice with PBS. Then secondary antibodies, including Cy5 anti-rabbit (Jackson, 711-175-152, 1:400), 405 anti-mouse (Jackson, 715-475-150, 1:400), anti-mouse (THERMOFISHER®, A32728, 1-100) and anti-rabbit (THERMOFISHER®, A11011, 1-100), and DAPI (4',6-Diamidino-2-Phenylindole, Dilactate) are added to the microfluidic devices and incubated overnight. The next day they are washed twice with PBS and then once overnight. Microfluidic devices are then imaged using a NIKON®A1R HD25 and an OLYMPUS®FV1200 confocal microscope, and images are processed using ImageJ and ICY. The morphological analysis of the vascular network can be conducted through a multi-step process to extract structural parameters. Maximal projection of ~85 slices with 5 μm z-step is acquired using ImageJ. Initially, vessel diameters are measured using the AutoTube®tool (MATLAB). Subsequently, the mask generated by AutoTube®is employed in ImageJ®to further analyze the vasculature. The Skeleton®tool in ImageJ®is utilized to skeletonize the vascular network, producing a simplified representation of the vessel structure. This skeletonized image is then subjected to quantitative analysis to determine the number of branches, junctions, triple points, quadruple points, and branch lengths. Drugs can be added to the system to look at the effect on BBB vascularity, permeability, inflammation, levels of amyloid beta protein, and other factors. 39 45630538.1

Claims

We claim:

1. A microfluidic device for modeling the brain comprising A polymeric substrate having A first chamber for growth of microvascular cells having connected thereto a first and a second channel for media, the first and second channels each having one or two ports for media into and out of the first and second channels; A second chamber for growth of neuronal tissue, the second chamber being surrounded by and fluidically accessible from the first chamber; A third chamber for meningeal lymphatic culture, the third chamber being fluidically accessible to the second chamber through the first chamber, wherein the third chamber has connected thereto a third channel for media, the third channel each having two ports for media into and out of the first and second channels; fluidic connections to each end of the first, second and third channels for media having a one-way valve to control fluid flow through the channel into or out of the media channels.

2. The microfluidic device of claim 1 further comprising one or more pumps for controlling flow through the media channels.

3. The microfluidic device of claim 1 further comprising cells, wherein the cells are selected from the group consisting of endothelial cells, perivascular cells, neural cells, lymphatic cells and immune cells.

4. The microfluidic device of claim 3 comprising microvascular endothelial cells, neural cells, meningeal lymphatic cells and immune cells.

5. The microfluidic device of claim 3 or 4 wherein the neural cells are in the form of suspended cells or neurospheres.

6. The microfluidic device of claim 3 or 4 wherein the immune cells are selected from the group consisting of T cells, macrophages or microglial cells, lymphocytes, and dendritic cells.

7. The microfluidic device of claim 1 comprising microvascular endothelial cells forming a blood brain barrier. 40 45630538.

18. The microfluidic device of claim 7 further comprising within the blood brain barrier neurospheres or suspended neurons in the first and second chambers, respectively.

9. The microfluidic device of claim 8 further comprising meningeal lymphatic structures within the third chamber.

10. The microfluidic device of claim 9 further comprising immune cells selected from the group consisting of T cells, macrophages or microglial cells, lymphocytes, and dendritic cells.

11. The microfluidic device of any of claims 1-10 wherein the cells are from an individual or cell line from an individual, the individual having a neurodegenerative disease or disorder, having had a stroke or other interruption in flow of blood to the brain, a human that is over the age of 40, 50, 60 or 70, or a combination thereof.

12. A method of making the device of any of claims 1-11 comprising providing a microfluidic device for modeling the brain comprising A polymeric substrate having A first chamber for growth of microvascular cells having connected thereto a first and a second channel for media, the first and second channels each having two ports for media into and out of the first and second channels; A second chamber for growth of neuronal tissue, the second chamber being surrounded by and fluidically accessible from the first chamber; A third chamber for meningeal lymphatic culture, the third chamber being fluidically accessible to the second chamber through the first chamber, the third chamber having connected thereto a third channel for media, the third channel each having two ports for media into and out of the first and second channels; fluidic connections to each end of the first, second and third channels for media having a one-way valve to control fluid flow through the channel into or out of the media channels, one or more pumps for controlling flow through the media channels, and cells selected from the group consisting of endothelial cells, neural cells, meningeal lymphatic cells and immune cells, 41 45630538.1wherein the endothelial cells are added to the first chamber, wherein the neural cells are added to the second chamber, wherein the meningeal lymphatic cells are added to the third chamber, and optionally wherein immune cells are added to any of the chambers or to the media.

13. A method for screening for the effect of an agent on the brain or components thereof comprising adding the agent to a chamber and / or media of a microfluidic device for modeling the brain comprising A polymeric substrate having A first chamber for growth of microvascular cells having connected thereto a first and a second channel for media, the first and second channels each having two ports for media into and out of the first and second channels; A second chamber for growth of neuronal tissue, the second chamber being surrounded by and fluidically accessible from the first chamber; A third chamber for meningeal lymphatic culture, the third chamber being fluidically accessible to the second chamber through the first chamber, the third chamber having connected thereto a third channel for media, the third channel each having two ports for media into and out of the first and second channels; fluidic connections to each end of the first, second and third channels for media having a one-way valve to control fluid flow through the channel into or out of the media channels, one or more pumps for controlling flow through the media channels, and cells selected from the group consisting of endothelial cells, neural cells, meningeal lymphatic cells and immune cells, wherein the endothelial cells are added to the first chamber, wherein the neural cells are added to the second chamber, wherein the meningeal lymphatic cells are added to the third chamber, and optionally wherein immune cells are added to any of the chambers or to the media. 42 45630538.1

Images