A 3D Gut-Brain-Vascular Platform for Bidirectional Crosstalk in Gut-Neuropathogenesis

A 3D Gut-Brain-Vascular Platform for Bidirectional Crosstalk in Gut-Neuropathogenesis | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article A 3D Gut-Brain-Vascular Platform for Bidirectional Crosstalk in Gut-Neuropathogenesis Hansang Cho, Minh Tran, Hoe Jeong, Chaeyeon Been, Ariunzaya Jamsranjav, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6128513/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 07 Feb, 2026 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Abstract A ‘gut-brain axis’ is an intricate bidirectional connection between the gut and the central nervous system, serving as a key pathway for signal exchange. However, current in vitro models do not fully capture the dynamics of interactions between these organs, which limits mechanistic understanding and therapeutic exploration. Here, we present a 3D human gut-brain vascular (GBV) model that simulates bidirectional communication between these entities, allowing us to investigate disorders originating from both the gut and brain. We created a physiologically relevant gut-brain axis model, creating a 3D villus-like lumenized gut barrier, blood vascular-astrocyte interactions, and brain tissue that mimics neurovascular interactions. Next, we demonstrated gut-to-brain signaling by introducing bacterial-derived toxins into the gut side, allowing the toxins to penetrate the gut barrier and the neurovascular barrier, ultimately reaching the brain and leading to tauopathy, a key indicator of neurodegeneration. We observed brain-to-gut signaling by exposing the brain side to risk factors of Alzheimer’s (AD) and Parkinson’s (PD), which induces neuroinflammation, disrupts the vascular barrier, and subsequently affects gut epithelial integrity. Our gut-brain vascular model is a precisely engineered microphysiological system that emulates the human gut-brain-vascular axis and serves as a translational tool to identify therapeutic targets and evaluate pharmaceuticals for neurological and gastrointestinal disorders. Biological sciences/Biological techniques/Lab-on-a-chip Biological sciences/Cell biology/Cell signalling Health sciences/Diseases/Neurological disorders Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction The gut-brain axis (GBA), a bidirectional communication network between the gastrointestinal (GI) tract and the central nervous system (CNS), plays a crucial role in maintaining neurological homeostasis 1 – 4 . This interaction is mediated through neuronal (vagus nerve), immune system (immune cells/cytokine signaling), and circulatory (vasculature) pathways 1 , 3 , 5 , 6 . These pathways facilitate bidirectional communication, primarily mediated by neuroactive substances, microbial metabolites, and hormones that regulate neurological function and systemic homeostasis. Among these, the vasculature serves as a critical interface, regulating immune surveillance, molecular exchange, and inflammatory responses between the gut, the periphery, and the brain 3 , 4 . Mounting evidence underscores the microbiota–gut–brain axis as a key contributor to neurodegeneration, with early alterations in gut microbiota identified in Alzheimer’s disease (AD) and Parkinson’s disease (PD) patients, suggesting a potential causal link between gut dysbiosis and neurodegeneration 7 – 11 . Animal studies further support this hypothesis, demonstrating that microbiota-driven neuroinflammation and glial cell dysfunction contribute to disease progression through systemic immune activation, inflammatory cascades, and blood-brain barrier (BBB) disruption 12 – 14 . Microbial-derived toxins and inflammatory mediators have been implicated in BBB permeability loss, neuronal dysfunction, and protein aggregation, reinforcing the notion that gut dysbiosis is a key driver of neurodegenerative pathology. Despite these insights, a significant limitation in the field is the lack of physiologically relevant in vitro models that accurately replicate the complex, dynamic interactions among the gut, vasculature, and brain. Most studies depend on animal models, which, while informative, do not capture the human-specific mechanisms underlying gut-brain communication 15 – 17 . Existing in vitro systems typically concentrate on single-organ models, such as gut-on-chip, BBB-on-chip, or brain-on-chip platforms; however, these models fail to recapitulate the multi-organ interplay that underlies gut-brain dysfunction in neurodegeneration 18 – 20 . Additionally, traditional Transwell assays and organoid cultures do not integrate circulatory flow, barrier dynamics, and cellular interactions necessary to mimic the in vivo microenvironment 21 – 23 . Consequently, critical questions regarding the mechanistic underpinnings of gut-brain communication remain unresolved, hindering the development of targeted therapeutic interventions. To tackle these challenges, we developed a 3D microfluidic Gut-Brain-Vascular (GBV) platform that models bidirectional crosstalk in gut-neuropathogenesis. Our system features a 3D lumenized gut barrier with villus-like architecture to replicate the intestinal epithelium, a capillary-like structure compartment containing brain microvascular endothelial cells, and a 3D brain tissue compartment containing neurons and astrocytes that simulates endothelial-astrocyte interactions, facilitating dynamic molecular exchange. Our system enables real-time investigation of bidirectional inflammatory crosstalk by modeling gut-to-brain signaling and examining brain-to-gut interactions. With the gut-to-brain axis, bacterial-derived toxins (e.g., lipopolysaccharides, Helicobacter-derived toxins) affect the gut compartment, causing gut barrier dysfunction and increased vascular permeability, allowing toxins and pro-inflammatory factors from gut and vascular organs to infiltrate the brain. This ultimately leads to the accumulation of phosphorylated tau, which is a hallmark of neurodegeneration. In the brain-to-gut axis, AD-like (NPC-derived AD models) and PD-like (PFF α-synuclein aggregates) environments in the brain compartment induce vascular barrier breakdown and subsequent gut epithelial dysfunction. By integrating gut, vascular, and brain compartments within a physiologically relevant microfluidic system, this model serves as a sophisticated microphysiological tool for studying gut-brain interactions in neurodegeneration. Our findings offer critical insights into the role of gut-origin inflammation in neurodegeneration, identifying mechanisms of diseases such as Alzheimer’s and Parkinson’s and laying the groundwork for screening novel therapeutic interventions targeting the gut-brain-vascular axis. Results Engineering a physiologically relevant human gut-brain-vascular (hGBV) model The gut-brain axis (GBA) is a complex bidirectional communication network regulated by neuronal, immune, and vascular interactions 5 , 6 . However, current in vitro models lack the physiologically relevant microenvironment needed to study gut-brain-vascular crosstalk. To address this issue, we created a 3D microfluidic Human Gut-Brain-Vascular (hGBV) model that combines lumenized gut epithelium, perfusable vasculature, and a neurovascular brain compartment in a dynamic microfluidic system (Fig. 1 a). The hGBV system consists of three interconnected compartments: Gut Compartment (G.C.), Vascular Compartment (V.C.), and Brain Compartment (B.C.), designed to recapitulate key features of gut-brain interactions (Fig. 1 b). Each compartment was engineered using human-derived cell types to closely mimic native tissue architectures. The gut compartment was developed using Caco-2 cells, forming a 3D villus-like epithelium to replicate intestinal barrier functions. The vascular compartment consisted of induced brain microvascular endothelial cells (iBMECs), forming a capillary-like endothelium that enables selective molecular transport. The brain compartment incorporated neural progenitor cell (NPC)-derived neurons and astrocytes, which interact with vascular endothelial cells via astrocytic end-feet, representing the neurovascular unit (NVU). The stepwise assembly of the hGBV model (Fig. 1 c, Supplementary Fig. 2) involved the sequential differentiation of brain tissue, gut epithelium formation, and vascular compartment development, ensuring the formation of interconnected and physiologically relevant microenvironments. To confirm the physiological relevance of the hGBV model, we assessed the formation of 3D gut villus-like structures using immunofluorescence imaging. The gut compartment exhibited well-organized epithelial cell layers with tight junction integrity, closely resembling native intestinal architecture (Fig. 1 d-e). Immunostaining of the vascular and brain compartments further confirmed the presence of iBMECs in the vascular compartment and neurons with astrocyte support in the brain compartment, forming a multicellular microenvironment capable of bidirectional signaling (Fig. 1 d-e). Additionally, marker expression analysis confirmed ViL1 in the gut epithelium, Tuj1 in neurons, and GFAP in astrocytes, validating the structural integrity of the system. To evaluate barrier integrity and compartmental media separation, we performed FITC-Dextran permeability assays across all compartments. Under physiological conditions, the gut and vascular barriers demonstrated compartment-specific permeability, with approximately 15% FITC-Dextran leakage into the vascular compartment and < 5% leakage into the brain compartment (Fig. 1 f-g). This selective permeability indicates the presence of intact barrier functions, effectively preventing uncontrolled diffusion of across compartments. To validate the bidirectional crosstalk of the system, we performed two distinct experiments modeling gut-to-brain and brain-to-gut signaling. To investigate how bacterial-derived toxins from the gut contribute to neurodegeneration, we exposed the gut compartment to Helicobacter-conditioned media (HP CM ), mimicking gut dysbiosis-induced inflammation. Exposure to HP CM led to the disruption of barrier tight junctions and glial activation, as indicated by an approximate 50% reduction in gut epithelium OCLN expression, an approximate 50% reduction in vascular endothelium ZO-1 expression, and a 2-fold increase in GFAP expression in the brain compartment, suggesting neuroinflammatory responses triggered by gut-derived and vascular-derived signals (Fig. 1 h-i). To investigate brain-to-gut signaling, we modeled Alzheimer’s disease (AD)-like conditions in the brain compartment by differentiating NPCs that were transfected with the APP gene for six weeks. This resulted in pTau accumulation (3-fold increase), vascular barrier breakdown, and loss of gut epithelial integrity. Specifically, endothelium ZO-1 expression in the vascular compartment was reduced by 50%, while gut epithelial OCLN expression decreased by 30%, confirming that neuroinflammatory factors contribute to systemic vascular and gut dysfunction (Fig. 1 j-k). Together, these findings illustrate that the hGBV model effectively replicates bidirectional gut-brain-vascular interactions, facilitating the study of gut-to-brain and brain-to-gut signaling in both physiological and disease conditions. The capacity to model gut-derived inflammation, neurovascular dysfunction, and loss of systemic barrier integrity emphasizes the translational potential of this system for investigating neurodegenerative disorders, including Alzheimer's and Parkinson's disease. Physiological Properties of the hGBV Model To functionally validate the hGBV model, we assessed the structural, barrier, and transport properties of the gut, vascular, and brain compartments, comparing static and dynamic perfusion conditions. Dynamic flow enhances physiological relevance in microfluidic systems, supporting barrier integrity, tissue maturation, and functional molecular transport. Thus, we examined gut morphology, permeability, vascular endothelial properties, and neurovascular interactions to confirm that the hGBV model recapitulates key in vivo-like physiological properties. Immunofluorescence analysis confirmed that the hGBV system forms a compartmentalized gut-vascular-brain interface with distinct cellular identities in each region (Fig. 1 d, Fig. 2 a-b, Supplementary Fig. 4a). The gut compartment (G.C.) developed a villus-like epithelium, characterized by Gut-ViL-1 serving as villi markers, adjacent to the vascular compartment (C.C.), which contained induced brain microvascular endothelial cells (iBMECs; Fig. 2 a). The brain compartment (B.C.) exhibited an organized neurovascular interface, where iBMECs formed endothelial monolayers that interacted with astrocytic end-feet (ALDH1L1) and neurons (Tuj1), mimicking the neurovascular unit (NVU; Fig. 2 b, Supplementary Fig. 4l). To assess the functional gut barrier properties, we examined the expression of villi and tight junction proteins (ViL-1 and OCLN), and performed permeability assays. Immunostaining revealed that the gut epithelium formed tight junctions with well-distributed ViL-1 (left) and OCLN (right) expression (Fig. 2 c). The vascular compartment contained capillary-like endothelial structures expressing CD31 (left) and ZO-1 (right), indicating tightness of brain endothelial vascular junctions (Fig. 2 d). The brain compartment showed interactions between neurons and astrocytes, along with synaptic marker expression (SYN-1) and calcium signaling activity (OGB-1 Ca²⁺), suggesting functional neural connectivity (Fig. 2 e). A comparison of gut barrier function under static versus dynamic conditions revealed enhanced gut morphology and barrier integrity in dynamic conditions. Cross-sectional imaging of the gut compartment showed improved villus-like structures in the gut epithelium during dynamic conditions, with increased expression of ViL-1 and OCLN (Fig. 2 f). Morphometric analysis demonstrated that dynamic conditions significantly increased gut epithelium thickness (1.5-fold), surface area (1.7-fold), and uptake capacity (2.2-fold), highlighting improved epithelial structure and nutrient absorption (Fig. 2 g-k). Permeability measurements (P app ) showed significantly reduced FITC-dextran diffusion under dynamic perfusion, indicating a more restrictive and physiologically relevant gut barrier (Fig. 2 l). Time-lapse analysis of gut permeability over 24 hours further revealed that FITC-dextran penetration was significantly lower under dynamic conditions, confirming enhanced tight junction functionality and selective molecular transport (Fig. 2 m). These results demonstrate the successful formation of a gut epithelium-like villus structure that replicates key features of the in vivo system. Next, we evaluated vascular barrier properties, particularly endothelial tight junction integrity and dynamic-induced morphological adaptations in the vascular compartment. Immunofluorescence staining of ZO-1 confirmed that dynamic conditions promoted stronger endothelial junction formation (Fig. 2 n). Quantification of ZO-1 expression showed a significant upregulation (1.5-fold) in dynamic conditions, supporting improved BBB integrity (Fig. 2 o). In addition, endothelial cell (EC) morphology analysis revealed that dynamic perfusion induced an elongated, flow-adapted endothelial phenotype, as quantified by an increased Feret ratio (Fig. 2 p). This shift from static cuboidal morphology to an elongated form is characteristic of physiological endothelial responses to shear stress, confirming that the hGBV vascular compartment successfully recapitulates in vivo-like BBB characteristics. Together, these results demonstrate that a dynamic condition enhances the physiological properties of the hGBV model, improving gut barrier function, vascular endothelial integrity, and neurovascular interactions. The integration of flow-adapted villus-like gut epithelium, endothelial cells, and functional neurovascular interfaces establishes hGBV as a biomimetic platform for studying gut-brain-vascular dynamics. Modeling Gut-to-Brain Inflammation and Microbiota-Driven Neuroinflammation Dysbiosis and microbial-derived inflammation are increasingly recognized as key contributors to neurodegenerative disease progression 13 . To model gut-to-brain inflammation and microbiota-driven neuroinflammation, we exposed the gut compartment (G.C.) to either Helicobacter-conditioned media (HP CM , 10%) or pure lipopolysaccharide (LPS, 1 µg/mL) to mimic microbial-triggered inflammatory responses (Fig. 3 a). We evaluated the inflammatory cascade, barrier dysfunction, and neurodegeneration at sequential time points, including gut barrier disruption and local inflammatory response after 1 day of exposure to the gut epithelium; vascular inflammation and endothelial dysfunction after 4 days of treatment; neuroinflammation and neurodegeneration occur after approximately 10 days of treatment. To assess gut epithelial integrity in response to bacterial-derived toxins, we examined tight junction expression, epithelial permeability, and inflammatory factors release. HP CM and LPS treatment significantly increased gut permeability, as indicated by higher endotoxin levels in the vascular (V.C.) and brain (B.C.) compartments after 4-day treatment (Fig. 3 b). Increased epithelial permeability index (P app ) (Fig. 3 c) further confirmed barrier dysfunction, suggesting that bacterial-derived components translocated across compartments. Structural integrity of the gut barrier was assessed via immunofluorescence analysis, which revealed a significant reduction in villus marker ViL1 expression, decreasing from 1-fold (control) to ~ 0.8-fold upon HP CM and LPS treatment (Fig. 3 d, e), indicating epithelial damage. Concurrently, CD13 fluorescence intensity increased by approximately 3.2-fold ( p < 0.001) in HP CM - and LPS-treated conditions (Fig. 3 f), suggesting epithelial remodeling and gut inflammatory activation. Following exposure to HP CM /LPS, OCLN localization was reduced from 1.0-fold (control) to 0.7-fold, representing a 30% decrease relative to the control condition (Fig. 3 g), confirming tight junction breakdown and barrier disruption. A heatmap analysis of pro-inflammatory cytokines and chemokines secreted by the gut epithelium (Fig. 3 h, Supplementary Fig. 5) demonstrated distinct inflammatory signatures responding to bacterial stimuli. HP CM -treated condition induced C5a, IFN-γ, IL-5, IL-12, IL-16, and IL-21, whereas both HP CM - and LPS-treated conditions resulted in increased IL-1β, IL-6, IL-18, IL-32a, and TNF-α, indicating a robust inflammatory response. To determine whether gut-derived inflammation impacted vascular integrity, we assessed tight junction loss, endothelial nitric oxide synthase (iNOS) activation, and inflammatory cytokine release in the vascular compartment (C.C.). Immunofluorescence analysis revealed significant upregulation of iNOS expression (Fig. 3 i, j) and a corresponding increase in NO production (Fig. 3 i, k) under HP CM - and LPS-treated conditions, confirming the vascular inflammatory response. Vascular integrity was significantly disrupted, as evidenced by ZO-1 delocalization at endothelial junctions (Fig. 3 i, p). However, endothelial cell (EC) density remained unchanged (Fig. 3 o), indicating functional endothelial dysfunction rather than cell death. This was further supported by a significant increase in endothelial permeability (P app ) (Fig. 3 q), suggesting barrier dysfunction and enhanced vascular permeability. A heatmap analysis of pro-inflammatory cytokine secretion from endothelial cells (Fig. 3 l, Supplementary Fig. 6) confirmed an elevated inflammatory signature in response to bacterial-derived toxins. Direct exposure of Helicobacter-conditioned media (HP CM , 1%) to endothelial cells triggered increased IFN-γ and IL-12, while LPS (100 ng/mL) exposure induced CXCL1, IL-6, and IL-21 expression. Notably, both HP CM and LPS treatments led to increased levels of CXCL10, IL-1β, IL-16, and IL-18, suggesting that bacterial-derived toxins directly activate pro-inflammatory signaling in endothelial cells. To examine the impact of gut-origin inflammation on the brain compartment (B.C.), we evaluated complement activation (C3), synaptic integrity (Synapsin-1), and neurodegenerative markers (pTau). Immunofluorescence analysis revealed a 2.5-fold increase in C3 expression in both treatment conditions (Fig. 3 m, r), suggesting neuroinflammatory activation. Synaptic integrity was assessed by Synapsin-1 (SYN1) expression, which significantly decreased from 1 to approximately 0.7-fold under gut compartment-treated conditions, indicating a nearly 30% reduction in synaptic density and connectivity (Fig. 3 m, s) and suggesting synaptic loss. Furthermore, phosphorylated tau (pTau) expression was significantly upregulated, approximately 4.9-fold in HP CM -treated and 4-fold in LPS-treated gut compartment samples (Fig. 3 t), a hallmark of neurodegeneration-associated tau pathology. Notably, neuronal density was significantly reduced in the brain compartment (Fig. 3 u), confirming that gut-derived inflammation leads to neurodegenerative-like phenotypes. These findings confirm that gut-origin risk factors, triggered by bacterial-derived stimuli (HP CM and LPS), induce a cascade of gut-to-brain inflammatory signaling, leading to neurovascular dysfunction and neurodegeneration. The sequential progression from gut barrier breakdown to vascular, synaptic loss, and tau pathology highlights the critical role of microbiota-driven neuroinflammation in neurodegenerative diseases. This validates the hGBV platform as a translational tool for studying gut-brain interactions in neurodegenerative diseases. Modeling Brain-to-Gut Crosstalk in Neurodegenerative Disease Neurodegenerative diseases, including Alzheimer’s disease (AD) and Parkinson’s disease (PD), are increasingly linked to systemic inflammation and gut dysregulation 24 . To investigate brain-to-gut signaling and its effects on vascular and gut integrity, we modeled AD- and PD-like conditions in the hGBV system. AD-like conditions were induced by transfecting the APP gene into neural progenitor cells (NPCs) and differentiating them for six weeks, while PD was modeled by differentiating NPCs for three weeks, followed by treatment with preformed fibrillar (PFF) α-synuclein (1 µM) and assessment one-week post-treatment (Fig. 4 a). We evaluated brain inflammatory responses, vascular dysfunction, and gut structural alterations across compartments to determine the impact of brain-derived neuroinflammation on peripheral systems. To confirm the induction of AD- and PD-like pathology, we quantified the secretion of neurotoxic protein aggregates (Aβ 42 and α-Synuclein) across compartments. Both Aβ 42 (AD-like model) and α-Synuclein (PD-like model) were significantly elevated in the brain compartment (B.C.), with detectable translocation into the vascular (V.C.) and gut (G.C.) compartments in disease conditions (Fig. 4 b, c). This suggests that neurotoxic aggregates can permeate through the vasculature, potentially contributing to systemic neuroinflammation. Additionally, Immunofluorescence analysis of the brain compartment revealed pronounced complement activation (C3, Fig. 4 d), indicative of neuroinflammation in AD- and PD-like models. Synaptic integrity was significantly impaired, as evidenced by a reduction in Synapsin-1 (SYN1) expression (Fig. 4 f), confirming synaptic loss in both disease models. Additionally, phosphorylated tau (pTau) expression was significantly upregulated (3.8-fold in AD-like and 2.2-fold in PD-like conditions, Fig. 4 g), reinforcing neurodegenerative tau pathology. The cumulative impact of these changes resulted in a significant decline in neuronal density (Fig. 4 h), supporting neurodegenerative-associated neuronal loss. Additionally, a heatmap analysis of pro-inflammatory cytokines secreted from AD- and PD-like brain environments (Fig. 4 i, Supplementary Fig. 7) confirmed an elevated inflammatory signature in disease conditions. The AD-like brain environment triggered increased IL-16, MIF, and TNF-α, while both AD- and PD-like conditions led to elevated levels of CCL-1, CCL-2, MIP-1, CCL-5, CXCL1, CXCL10, ICAM-1, IFN-γ, IL-1β, IL-5, IL-6, IL-8, IL-12, IL-18, IL-21, and IL-32a, confirming widespread neuroinflammatory activation. Given the role of neuroinflammation in vascular dysfunction, we assessed endothelial inflammation, vascular barrier integrity, and permeability changes in the vascular compartment (V.C.). Immunofluorescence analysis of vascular endothelial cells revealed a significant approximate 2-fold increase in iNOS expression (Fig. 4 j, l) and a 2.5-fold increase in NO production (Fig. 4 j, m), suggesting endothelial dysfunction and oxidative stress response to inflammation. Tight junction integrity was broken, as evidenced by ZO-1 delocalization at endothelial junctions. In the AD-like model, ZO-1 localization was reduced from 1.0 to 0.6-fold, indicating a 40% loss of tight junction integrity, while in the PD-like model, it was reduced from 1.0 to 0.8-fold, indicating a 20% loss of junctional stability (Fig. 4 j, n), confirming vascular barrier breakdown. Consistent with these findings, endothelial cell (EC) density was significantly reduced in AD- and PD-like conditions (Fig. 4 o), indicating vascular damage and endothelial loss. Moreover, permeability assays demonstrated a significant increase in endothelial permeability (P app , Fig. 4 p), particularly in AD-like conditions, suggesting a severe breakdown of vascular integrity in response to brain-derived neuroinflammation. To determine whether brain-to-gut signaling disrupts gut homeostasis, we examined gut morphology, epithelial barrier integrity, and inflammatory activation in the gut compartment (G.C.). Unlike the pronounced vascular inflammation and endothelial barrier dysfunction observed in the V.C., the gut compartment exhibited only mild structural alterations. Immunofluorescence analysis revealed a moderate reduction in villus marker ViL-1 expression (Fig. 4 k, q), suggesting partial villus alteration in response to brain-derived inflammation. Additionally, CD13 expression was slightly increased with 1.7-fold in AD and 1.3-fold in PD-like conditions compared to WT (Fig. 4 k, r), indicating mild epithelial remodeling and inflammatory activation. However, tight junction integrity remained preserved, as evidenced by unchanged occludin (OCLN) localization (Fig. 4 k, s), and no significant reduction in epithelial cell density was observed (Fig. 4 t). These findings suggest that while brain-derived inflammatory signals propagate to the gut, they primarily induce structural remodeling rather than direct gut barrier disruption. These findings demonstrate that brain-origin inflammation in AD- and PD-like conditions leads to significant vascular dysfunction while only mildly affecting the gut barrier. The vascular compartment exhibited pronounced endothelial inflammation, tight junction disruption, and increased permeability, confirming brain-to-vascular inflammatory signaling. In contrast, the gut compartment showed only mild structural alterations, with reduced villus integrity and increased CD13 expression, but no significant tight junction disruption or epithelial loss. This study supports the hypothesis that brain-derived neuroinflammation propagates inflammatory signals to peripheral compartments, with strong effects on the vasculature but limited gut barrier breakdown. These results establish the hGBV platform as a valuable model for studying brain-to-gut crosstalk in neurodegenerative diseases. Discussion The gut-brain axis (GBA) is increasingly recognized as a key regulator of systemic homeostasis, with accumulating evidence implicating bidirectional communication between the gut, vasculature, and brain in the progression of neurodegenerative diseases 1 , 5 . Despite this recognition, current in vitro models fail to fully capture the complex, multi-compartmental interactions that underlie gut-brain inflammatory signaling. In this study, we developed and validated a 3D human Gut-Brain-Vascular (hGBV) model, successfully integrating a physiologically relevant, villus-like gut epithelium, perfusable vasculature, and a neurovascular brain compartment. Using this platform, we demonstrate that gut-origin inflammation (gut-to-brain axis) can drive neurovascular dysfunction and neurodegeneration, while brain-origin neuroinflammation (brain-to-gut axis) predominantly affects vascular integrity, resulting in only mild alterations to the gut barrier. These findings provide critical mechanistic insights into microbiota-driven neuroinflammation and brain-to-gut crosstalk, establishing the hGBV system as a valuable translational model for neurodegenerative disease research. Our results confirm that gut-derived inflammatory stimuli, including Helicobacter-conditioned media (HP CM ) and lipopolysaccharide (LPS), trigger systemic inflammation that spreads to the vasculature and brain. The observed gut barrier disruption, characterized by tight junction loss (OCLN, ZO-1), increased permeability, and the release of inflammatory factors, aligns with previous reports linking gut dysbiosis to intestinal barrier breakdown in neurodegenerative diseases 25 , 26 . Importantly, bacterial-derived endotoxins translocate into the vascular and brain compartments, leading to vascular dysfunction marked by endothelial nitric oxide synthase (iNOS) upregulation, increased NO production, and blood-brain barrier (BBB) disruption. These findings are consistent with preclinical and clinical studies reporting circulating endotoxins and inflammatory cytokines in Alzheimer’s disease (AD) and Parkinson’s disease (PD) patients 27 – 29 . At the neurovascular interface, prolonged exposure to gut-derived inflammatory stimuli results in synaptic dysfunction, complement C3 activation, and phosphorylated tau accumulation, all key hallmarks of neurodegeneration. This progression supports the hypothesis that chronic gut inflammation contributes to neurodegenerative pathology through systemic immune activation, changes in BBB permeability, and direct neuronal damage 11 , 13 , 30 , 31 . Notably, our model recapitulates the sequential progression observed in vivo, where gut inflammation precedes vascular impairment and neuronal loss, reinforcing the critical role of the gut-brain axis in neurodegeneration. In contrast to the gut-driven model, brain-origin neuroinflammation (brain-to-gut axis) in AD- and PD-like conditions primarily induced vascular dysfunction, with limited gut barrier disruption. APP-transfected NPCs (AD model) and α-synuclein-treated NPCs (PD model) exhibited neurotoxic protein accumulation (Aβ 42 , α-Syn), synaptic loss, complement activation, and pTau accumulation, consistent with postmortem brain analyses of AD and PD patients 32 – 35 . Brain inflammation triggered an endothelial response, including increased iNOS expression, NO production, ZO-1 delocalization, and elevated endothelial permeability, mirroring BBB dysfunction reported in AD and PD patients (ref). Surprisingly, despite inflammatory cytokine release into systemic circulation, gut epithelial tight junctions (OCLN) remained intact, with only mild structural alterations by showing slightly reduced ViL1 and increased CD13 expression. This suggests that brain-to-gut inflammatory signaling is primarily mediated through vascular dysfunction rather than direct gut barrier breakdown. These findings challenge previous hypotheses suggesting severe gut barrier dysfunction in neurodegenerative disease models, instead proposing that gut remodeling in response to brain inflammation is subtler 36 – 38 . This distinction underscores the need for physiologically relevant multicompartment models like hGBV, which allow for dynamic gut-brain-vascular interactions that are difficult to capture in traditional 2D cultures or in vivo models. By utilizing our hGBV platform, our findings have several translational implications. First, it provides a human-specific model to study gut-brain-vascular interactions in neurodegeneration, reducing reliance on animal models that often fail to replicate human pathophysiology. Second, our findings suggest that gut-to-brain signaling plays a more direct role in neurodegeneration, while brain-to-gut signaling primarily affects vascular integrity. This highlights the potential for gut-targeted therapeutics to prevent or slow neurodegenerative progression. In this case, several avenues for future research emerge from these findings. First, integrating patient-derived induced pluripotent stem cells (iPSCs) into the hGBV model could enable personalized disease modeling and drug screening. Second, investigating specific microbial metabolites (e.g., short-chain fatty acids) and their effects on neurovascular integrity could help elucidate mechanistic pathways linking gut dysbiosis to neurodegeneration. Finally, leveraging this platform for high-throughput drug screening could accelerate the identification of gut-targeted therapeutics for AD and PD. While our hGBV platform successfully recapitulates gut-brain-vascular interactions, however several limitations remained. First, the physical distance between the gut and brain compartments in our chip is shorter than in vivo , potentially accelerating inflammatory signaling and molecular transport. Second, the model lacks immune cells (e.g., monocytes, T-cells, microglia), which plays a crucial role in neuroinflammation and blood-brain barrier (BBB) regulation. Although the presence of endothelial cells and astrocytes provides a neurovascular interface, the addition of circulating peripheral immune cells could further enhance the physiological relevance of the platform. Third, gut microbiota interactions are simplified in this system, as only microbial-derived toxins (e.g., HP CM , LPS) were introduced. A system integrating live commensal and pathogenic microbes in a gut-microbiome-on-a-chip setup could provide deeper insights into host-microbe interactions in neurodegeneration. Lastly, while this study models AD and PD phenotypes, the system could be expanded to investigate other neurodegenerative conditions, such as ALS or multiple sclerosis, where gut-brain interactions are increasingly implicated. Incorporating genetic risk factors related AD and PD would enable disease modeling and therapeutic screening. Despite these limitations, our hGBV system represents a major advancement over traditional in vitro models, integrating gut, vascular, and brain compartments to create a more physiologically relevant platform. Addressing these limitations in future versions will further enhance its translational potential in neurodegenerative research and therapeutic development. Conclusion This study establishes the hGBV system as a physiologically relevant human model for investigating gut-brain-vascular interactions in neurodegeneration. Our findings highlight the differing effects of gut-origin and brain-origin inflammation, showing that gut-derived inflammation drives neurovascular dysfunction and neurodegeneration, while brain-derived neuroinflammation primarily leads to vascular dysfunction with minimal gut remodeling. These results provide critical insights into microbiota-driven neuroinflammation and brain-to-gut communication, supporting gut-targeted therapeutic strategies for AD and PD. By capturing the dynamic interplay between the gut, vasculature, and brain, the hGBV platform serves as a powerful translational tool for elucidating disease mechanisms, identifying biomarkers, and accelerating the development of targeted therapeutics for neurodegenerative diseases. Methods To systematically investigate the role of gut-brain-vascular interactions in neuroinflammation, we developed a 3D microfluidic platform that recapitulates the physiological architecture of the gut, vasculature, and brain. Below, we detail the fabrication of the microfluidic system, cellular components, and experimental conditions used to simulate gut-origin and brain-derived inflammatory processes. 1.1. Microfluidic Chip Fabrication The microfluidic chip was fabricated using soft lithography techniques, where a silicon substrate served as the base for device molding. The photolithography process sequentially patterned an SU-8 negative photoresist (MicroChem, Round Rock, TX) onto the silicon wafer, forming the microfluidic channel mold. To replicate the microfluidic structures, a polydimethylsiloxane (PDMS) mixture was prepared by combining PDMS elastomer with a curing agent (Sylgard 184 A/B, Dow Corning, Midland, MI) at a 9:1 ratio. The mixture was poured onto the SU-8 mold and degassed under vacuum to remove air bubbles. The PDMS was then thermally cured at 65°C for 2 hours to ensure complete polymerization. Once cured, the PDMS replica was carefully removed from the mold, and fluid reservoirs were created by punching inlet and outlet holes using a 2 mm biopsy punch. Additional plastic media reservoirs were fabricated using a computer-controlled laser cutter (Epilog Laser, Golden, CO) on a 5 mm thick acrylic plate to enhance experimental throughput and facilitate media handling. These acrylic reservoirs were bonded to the PDMS chip using an additional layer of liquid PDMS to ensure a stable, leak-free connection. For final assembly, the PDMS chip was bonded to a custom glass-bottomed uni-plate (MatTek, Ashland, MA) via oxygen plasma treatment (Plasma Etch, Carson City, NV), ensuring strong adhesion and optical transparency for high-resolution imaging (Supplementary Fig. 1). After assembly, the microfluidic chip was sterilized under UV light for 30 minutes before cell seeding and experimental use. 1.2. Functionalization of Microfluidic Devices Before Cell Seeding To promote cell adhesion and extracellular matrix (ECM) compatibility, the microfluidic compartments were functionalized with Poly-L-lysine (PDL) and collagen type I (Col I) following a well-established protocol [ref]. This functionalization step enhances cell attachment, barrier formation, and tissue-specific microenvironment stability within the microfluidic system. The entire microfluidic network, including all compartments and channels, was infused with 1 mg/mL Poly-L-lysine (Sigma Aldrich, St. Louis, MO) and incubated for 30 minutes at room temperature (RT) to facilitate surface charge modification and ECM adhesion. After incubation, the device was rinsed thoroughly with cold 1× PBS (phosphate-buffered saline) three times to remove excess PDL. Following PDL treatment, the gut compartment (G.C.) and vascular compartment (V.C.) were infused with 1 mg/mL collagen type I (Corning, Corning, NY), adjusted to pH 7.5. The functionalized device was incubated at 37°C for 1 hour to allow proper ECM polymerization and surface coating. After functionalization, all compartments were washed with fresh 1× PBS to remove unbound proteins. The microfluidic devices were then immediately seeded with cells to ensure optimal cell adhesion and tissue integration. 1.3. 3D Culture of Brain Tissue in the hGBV Microfluidic Device To establish a functional 3D neural microenvironment within the brain compartment (B.C.) of the hGBV model, human neural progenitor cells (ReN cell VM NPCs) were differentiated into neurons and astrocytes over a three-week period. This approach promoted cellular maturation, synaptic formation, and astrocyte-neuron interactions, ensuring a physiologically relevant brain tissue model. Human ReN cell VM neural progenitor cells (NPCs) were obtained from MilliporeSigma (Billerica, MA, USA) and maintained at passages < 20 to preserve their differentiation capacity. Cells were expanded in Dulbecco’s Modified Eagle Medium/Nutrient Mixture F-12 (DMEM/F-12, Life Technologies, Grand Island, NY, USA) supplemented with 2 mg/mL Heparin (StemCell Technologies, Vancouver, Canada), 2% (v/v) B27 neural supplement (Life Technologies, Grand Island, NY, USA), 20 ng/mL Epidermal Growth Factor (EGF) (Sigma-Aldrich, St. Louis, MO, USA), 20 ng/mL Basic Fibroblast Growth Factor (bFGF) (Stemgent, Cambridge, MA, USA), and 1% (v/v) Penicillin/Streptomycin (Sigma-Aldrich, St. Louis, MO, USA). NPCs were plated in T25 flasks coated with 1% Matrigel (BD Biosciences, San Jose, CA, USA) and cultured at 37°C with 5% CO₂ in a humidified incubator. The culture medium was refreshed every three days until confluence was reached. To generate a 3D mini-brain tissue model, ReN NPCs were suspended at 2 × 10 4 cells/device in a 10% (v/v) Matrigel-DMEM/F12 differentiation medium mixture (Supplementary Fig. 3). 10 µL of the NPC-Matrigel suspension was injected into brain compartment. The device was incubated at 37°C for 1 hour to facilitate gel polymerization and cell attachment. After polymerization, 100 µL of differentiation medium was added into both brain and vascular compartment. The devices were placed in a 37°C, 5% CO₂ incubator for long-term culture. Over 21 days, the NPCs progressively differentiated into neurons and astrocytes. Culture medium was replaced every two days to maintain optimal differentiation conditions. 1.4. 3D Gut Epithelium Barrier Generation To establish a functional gut epithelium within the gut compartment (G.C.) of the hGBV model, the Caco-2 epithelial cell line was cultured and differentiated under dynamic perfusion conditions to mimic the intestinal barrier. Human Caco-2 cells were obtained from the Cell Lines Service (CLS, Eppelheim, Germany) and maintained between passages 40–70 to ensure stable barrier formation properties. Cells were cultured in 75 cm² tissue culture flasks using high-glucsoe Dulbecco’s Modified Eagle’s Medium (high glucose DMEM, Gibco, Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 10% (v/v) Fetal Bovine Serum (Sigma-Aldrich, St. Louis, MO, USA), 1% (v/v) Penicillin/Streptomycin (Sigma-Aldrich, St. Louis, MO, USA). Cells were maintained at 37°C in a 5% CO₂/95% air atmosphere with constant humidity, and the culture medium was refreshed every two days until the cells reached 80–90% confluence. At day 7 of neuronal and astrocyte differentiation, Caco-2 cells were seeded into the gut compartment (G.C.) of the microfluidic chip to allow for co-culture with the neurovascular system. The cells were introduced at a density of 10 6 cells/mm 2 and allowed to attach for 4 hours before adding fresh media. The loaded cells adhered to the upper surface of the compartments when the platforms were inverted for 4 hours in a 37°C incubator with 5% CO 2 . After the cells adhered to the upper surface, newly cultured Caco-2 cells were gently reintroduced to the bottom surface in their normal orientation for an additional two hours to ensure proper adherence. Once both surfaces were covered with adherent cells, we added an extra 100 µL of fresh medium to the reservoirs. To enhance intestinal barrier function, dynamic perfusion conditions were introduced using a rocker integrated into the incubator. The culture medium was replaced daily to maintain optimal differentiation conditions. 1.5. 3D Vascular Endothelium Barrier Generation To establish a functional blood-brain barrier (BBB) within the vascular compartment (V.C.) of the hGBV model, human embryonic stem cell-derived endothelial cells (hECs) were differentiated into induced brain microvascular endothelial cells (iBMECs) and cultured under dynamic perfusion conditions to mimic physiological endothelial barrier function. Human embryonic stem cells (hEC, H9) were obtained from Wi-Cell (USA) and differentiated into iBMECs following a previously established protocol (ref). Cells were grown in mTeSR Plus medium (StemCell Technologies, Vancouver, Canada) supplemented with 10 µM Y-27632 dihydrochloride (Tocris #1254) to enhance survival and proliferation. Upon reaching ~ 70% confluency, differentiation into iBMECs was initiated. ells were switched to an unconditioned medium (Thermo Fisher Scientific #11330057) containing 20% KnockOut Serum Replacement (Thermo Fisher Scientific #10828010), 1% Non-Essential Amino Acids (NEAA, Thermo Fisher Scientific #11140050), 0.5% Glutamax (Thermo Fisher Scientific #35050079), 0.1 mM β-Mercaptoethanol (Sigma-Aldrich #M6250, St. Louis, MO, USA). This differentiation medium was maintained for 6 days to promote iBMEC-like characteristics. Cells were transitioned to EC + + medium, consisting of human Endothelial Serum-Free Medium (SFM, Thermo Fisher Scientific #11111044), 1% Platelet-Poor Plasma-Derived Serum, Bovine (Alfa Aesar #15406419, Haverhill, MA, USA), 20 ng/mL Basic Fibroblast Growth Factor (bFGF), 10 µM Retinoic Acid (Sigma-Aldrich #R2625). After 2 days, the medium was replaced with maturation medium (SFM without bFGF and retinoic acid) to stabilize endothelial barrier function. iBMECs were harvested using Accutase (Corning #25-058-CI, Corning, NY, USA) and prepared for seeding onto the vascular compartment (V.C.) of the hGBV microfluidic device. At day 18 of neuronal and astrocyte differentiation with Caco-2, a suspension of iBMECs (5 x 10 6 cell/mm²) was loaded into the vascular compartment (V.C.). The microfluidic chip was inverted for 4 hours at 37°C to facilitate cell attachment to the upper surface. After 4 hours, the chip was returned to its normal orientation, and a second suspension of iBMECs was gently reintroduced to the bottom surface for 2 additional hours to ensure bilayer coverage. Once both surfaces were fully covered with adherent iBMECs, 100 µL of fresh differentiation medium was added to one reservoir, allowing flow through the vascular compartment. To maintain a constant supply of fresh nutrients, the medium was replenished daily by adding 100 µL of fresh medium to one reservoir while removing the same volume from the opposite reservoir. To enhance endothelial barrier integrity, dynamic perfusion was introduced using a rocker integrated within the incubator. 1.6. Barrier Integrity Assessment To evaluate the permeability of the 3D barrier model, including the gut (G.C.) and vascular (V.C.) compartments, we performed a FITC-dextran permeability assay. This method assessed barrier integrity by quantifying the diffusion of fluorescently labeled dextran molecules across the epithelial and endothelial layers. FITC-conjugated dextran molecules of different molecular weights were used to evaluate paracellular permeability, including low molecular weight FITC-dextran (3–5 kDa) to assess small molecule permeability and high molecular weight FITC-dextran (70 kDa) to assess macromolecule leakage. Both dextran solutions were obtained from Sigma-Aldrich (St. Louis, MO, USA) and prepared in sterile 1× PBS at a final concentration of 10 µg/mL. FITC-dextran was introduced into the luminal side of either the gut compartment (G.C.) or vascular compartment (V.C.) of the microfluidic device. Time-lapse imaging was performed to monitor fluorescence intensity changes during the perfusion of FITC-dextran across the barrier layer. Imaging was conducted using a Nikon Eclipse Ti2-E microscope (Nikon Corporation, Tokyo, Japan) equipped with filter cube for FITC detection, 10× objective lens, and frames were captured at 1 second per frame to track dextran diffusion dynamics. Raw fluorescence intensity values were extracted using Nikon NIS-Elements software. The fluorescence intensity ratio between the perfused FITC-dextran and the initial luminal concentration was used to calculate the perfusion index. The apparent permeability coefficient (Papp, cm/s) was determined using the following equation: $$\:{P}_{app}=\:\frac{\left(\frac{dCi}{dt}\right)\times\:V}{{C}_{o}\:\times\:A}$$ 1 Where: P app is permeability coefficient (cm/s), dC i /dt is rate of change in FITC-dextran concentration in the receiving compartment over time (t), C o is initial FITC-dextran concentration in the luminal compartment (10 µg/mL), V is volume of the receiving compartment, A is surface area of microfluidic channel (0.013 cm²). Lower P app values indicate an intact and selectively permeable barrier. Higher P app values suggest barrier leakage. 1.7. Brain Calcium Signaling To assess neuronal activity within the brain compartment (B.C.), we monitored intracellular calcium dynamics using Oregon-Green-BAPTA-1 (OGB-1), a Ca²⁺ indicator (Life Technologies, Thermo Fisher Scientific). Calcium transients were analyzed using real-time fluorescence imaging, providing insights into neuronal and astrocytes interaction. Brain compartment were incubated in artificial cerebrospinal fluid (aCSF) at 37°C for 30 minutes to mimic physiological conditions. The aCSF composition included 25 mM NaHCO₃, 25 mM D-glucose, 125 mM NaCl, 2.5 mM KCl, 1.25 mM NaH₂PO₄, 1 mM MgCl₂, 2 mM CaCl₂. During incubation, Pluronic F-127 (20% w/v in DMSO, Sigma-Aldrich) was added to facilitate OGB-1 dye loading cytosol. After 30 minutes of dye loading, samples were rinsed twice with fresh aCSF to remove unbound dye. Time-lapse fluorescence imaging was performed using a Nikon Eclipse Ti2-E microscope (Nikon Corporation, Tokyo, Japan) equipped with Texas red filter cube (for Rhod-2 detection), 20× objective lens, and frames were captured at 1 frame per second to monitor dynamic changes in intracellular Ca²⁺ levels. Raw fluorescence intensity values were extracted using Nikon NIS-Elements software. 1.8. Helicobacter Bacteria Conditioned Media Collection H. pylori strain [NCTC 11637] was obtained from ATCC and cultured under microaerophilic conditions to maintain optimal bacterial growth. Bacteria were grown in Brain Heart Infusion (BHI) broth (BD Biosciences, San Jose, CA, USA) supplemented with 5% Fetal Bovine Serum (FBS, Gibco, Thermo Fisher Scientific), 10 µg/mL Vancomycin, 5 µg/mL Trimethoprim. Cultures were incubated at 37°C in a 5% CO₂, 10% O₂ atmosphere for 48–72 hours to reach the logarithmic growth phase. H. pylori-conditioned media (HPCM) was collected by centrifugation at 4,000 × g for 10 minutes at 4°C once cultures reached the log-phase. The collected supernatant was filter-sterilized using a 0.2 µm syringe filter to remove residual bacteria while retaining secreted bacterial toxins, virulence factors, and metabolic byproducts. Aliquots were stored at -80°C until use to prevent degradation of bioactive components. 1.9. Alzheimer’s Disease (AD) Model Construction in hGBV Platform To generate Alzheimer’s disease (AD) human neural progenitor cells (hNPCs), ReN hNPCs were transduced with a lentiviral vector encoding the APPSL mutation, resulting in the overexpression of a familial Alzheimer’s disease (FAD) variant of the amyloid precursor protein (APP). This model enabled the production of high levels of Aβ peptides, mimicking AD-like pathophysiology in the brain compartment (B.C.) of the hGBV system. ReN hNPCs were cultured in six-well plates and transduced using a commercially available APPSL-GFP Alzheimer’s lentivirus (EMD Millipore). The APPSL lentiviral vector encodes a mutated form of human APP containing two familial Alzheimer’s disease (FAD) mutations K670N/M671L (Swedish mutation) for increases β-secretase cleavage, leading to higher Aβ42 production and V717I (London mutation) for altering γ-secretase cleavage, promoting Aβ42 over Aβ40 production. ReN hNPCs were incubated with 5 µL of APPSL-GFP lentivirus (1 × 10⁹ IFU/mL) in the presence of 2 mg/mL polybrene (EMD Millipore) to enhance transduction efficiency. Cells were incubated for 24 hours at 37°C with 5% CO₂. GFP fluorescence was monitored to confirm initial viral expression. After three washes with PBS, a second round of viral infection was performed by adding fresh APPSL-GFP lentivirus to the culture medium. Cells were incubated for another 24 hours. After 48 hours, cells were washed three times with PBS and cultured in fresh medium to allow transgene expression stabilization. Fluorescence microscopy was used to confirm GFP-positive transgene expression. Transgene-positive cells were enriched using fluorescence-activated cell sorting (FACS, BD FACS Aria II, BD Biosciences) to ensure a homogeneous AD hNPC population. 1.10. Parkinson’s Disease (PD) Model Construction in the hGBV Platform To establish a Parkinson’s disease (PD) model within the hGBV platform, preformed fibril (PFF) α-synuclein aggregates were introduced into the brain compartment (B.C.), where they induced pathological α-synuclein aggregation, neuroinflammation, and neuronal dysfunction. This model enabled the study of brain-to-gut crosstalk in PD pathology. ReN-derived NPCs were differentiated into neurons and astrocytes for 3 weeks in the brain compartment, following the differentiation protocol described in Section 1.3. PFF α-Synuclein (PFF α-Syn, 1 µM, StressMarq Bioscience) added to the brain compartment (B.C.) for 6 hours to allow for fibril uptake by neurons and astrocytes. Cells were then washed and replace by fresh medium to remove excess extracellular fibrils. The cultures were maintained for 7 additional days to allow for α-synuclein aggregation and pathology development. 1.11. Chromogenic Endotoxin Measurement Assay To quantify endotoxin levels in conditioned media and microfluidic compartments, we performed a chromogenic endotoxin measurement assay using the Pierce™ Chromogenic Endotoxin Quant Kit (Thermo Fisher Scientific, Catalog #A39552S). This assay was conducted following the manufacturer’s user guide and is based on the Limulus Amebocyte Lysate (LAL) reaction. Samples were collected from the gut compartment (G.C.), vascular compartment (V.C.), and brain compartment (B.C.). Samples were diluted 1:10 in endotoxin-free water to ensure that measurements fell within the linear range of the standard curve. The optical density (OD) at 405 nm was measured using a microplate reader (Biotek, USA). Sample endotoxin levels were calculated using the standard curve and expressed in endotoxin units per milliliter (EU/mL). Endotoxin levels were compared between control (CT), Helicobacter-conditioned media (HP CM ), and LPS-treated conditions to assess gut-to-brain endotoxin translocation. Each condition was prepared in triplicate to ensure data reliability and reproducibility. 1.12. Amyloid beta 42 ELISA Assay Quantify amyloid beta 42 (Aβ 42 ) levels was performed an enzyme-linked immunosorbent assay (ELISA) using the Human Aβ 42 ELISA Kit (Abcam, Catalog #ab289832). The assay was conducted following the manufacturer’s user guide, which employs a sandwich ELISA method for the sensitive and specific detection of Aβ 42 in conditioned media. Conditioned media were collected from the brain compartment (B.C.), vascular compartment (V.C.), and gut compartment (G.C.) after 6 weeks of neuronal differentiation in wide type (WT) and AD-like conditions (APP-overexpressing NPCs, Section 1.9). Sample concentrations were interpolated from the standard curve and expressed as pg/mL Aβ 42 . Total Aβ 42 levels were compared between control (CT) and AD conditions to assess the release and potential translocation of Aβ 42 across the brain-gut axis. All samples were analyzed in triplicate to ensure data reproducibility. 1.13. Alpha-Synuclein ELISA Assay To quantify alpha-synuclein (α-Syn) levels in the brain compartment (B.C.) and assess its potential translocation into the vascular compartment (V.C.) and gut compartment (G.C.), we performed an enzyme-linked immunosorbent assay (ELISA) using the Human Alpha-Synuclein ELISA Kit (Abcam, Catalog #ab260052). The assay was conducted according to the manufacturer’s user guide and utilized a sandwich ELISA format for the specific and sensitive detection of α-Syn in conditioned media. Conditioned media were collected from the brain compartment (B.C.), vascular compartment (V.C.), and gut compartment (G.C.) following 7 days of PFF α-Synuclein treatment in the PD model (Section 1.10). Measured α-Syn levels were compared between wild type (WT) and PD-like conditions to assess neurotoxic α-Syn secretion and potential translocation across compartments. All samples were analyzed in triplicate to ensure data reproducibility. 1.14. Multicytokine Profiling Assay To investigate cellular responses to bacterial-derived toxins and neuroinflammatory stimuli, we analyzed the profiles of secreted cytokines and chemokines from conditioned media collected from the 3D gut epithelium (gut compartment), vascular endothelium (vascular compartment), and brain environment (brain compartment) using the Human Cytokine Array Kit (Catalog #ARY005B, R&D Systems, Minneapolis, MN, USA). Conditioned media were obtained from gut epithelial cells (Caco-2, gut compartment), endothelial cells (iBMECs, vascular compartment), and neurons and astrocytes (brain compartment) after exposure to bacterial toxins (Helicobacter-conditioned media, LPS) and neuroinflammatory stimuli (Aβ 42 , PFF α-Synuclein). Cytokine spot signals were detected using the iBright Imaging System (ThermoFisher, USA), and pixel density values were extracted using ImageJ (NIH, Wayne Rasband). Background subtraction was applied, and relative pixel intensity values were normalized against the positive control spots on the membrane. Cytokine expression levels were compared across control (CT), Helicobacter-conditioned media (HP CM )-treated, LPS-treated, and neuroinflammatory conditions (Aβ 42 , PFF α-Synuclein). A heatmap of log₂ fold-change cytokine expression was generated to visualize differential secretion patterns across experimental conditions. 1.15. Immunostaining Assay To visualize cellular markers and structural integrity within the 3D hGBV microfluidic platform, an immunostaining assay was performed following a standard fixation, permeabilization, blocking, and antibody incubation protocol. First, cells cultured within the microfluidic device were rinsed twice with 1× PBS and fixed at room temperature for 30 minutes using 4% paraformaldehyde (Electron Microscopy Sciences, Hatfield, PA, USA). After fixation, cells were washed three times with 1× PBS at 10-minute intervals. To enhance antibody penetration, cells were permeabilized for 30 minutes at room temperature with 1× PBS supplemented with 0.1% (v/v) Triton X-100 and 0.1% (v/v) Tween-20. Subsequently, cells underwent three additional 10-minute washes with 1× PBS before blocking to minimize non-specific antibody binding. For blocking, cells were incubated for 2 hours at 4°C in a solution of 1× PBS supplemented with 0.1% (v/v) Tween-20 and 4% (v/v) human serum albumin (HSA). Following blocking, cells were rinsed three times with 1× PBS containing 0.1% (v/v) Tween-20 at 10-minute intervals. Cells were then incubated with the primary antibody diluted in 1% (v/v) blocking solution for 6 hours at 4°C. The dilution ratios of all primary and secondary antibodies are provided in Table S1 (Supplementary Information). After the primary antibody incubation, cells were washed three times with 1× PBS containing 0.1% (v/v) Tween-20 at 10-minute intervals before the addition of the secondary antibody. Secondary antibodies were diluted in 1× PBS supplemented with 0.1% (v/v) Tween-20 and 4% (v/v) HSA and incubated for 2 hours at 4°C. Following secondary antibody binding, cells were washed three times with 1× PBS supplemented with 0.1% (v/v) Tween-20 at 10-minute intervals to remove excess antibodies. Immunostained samples were visualized using a Nikon Eclipse Ti2-E fluorescence microscope (Nikon Corporation, Tokyo, Japan) and a K1-Fluo Confocal Laser Microscope (Nanoscope, Korea). Regions of interest (ROIs) were manually selected for each compartment (gut, vascular, and brain) to quantify the mean fluorescence intensity (MFI). Image acquisition and MFI quantification were performed using NIS-Elements software (Nikon, Japan) and Fiji/ImageJ (NIH, Wayne Rasband, USA). To account for variations in cell density, MFI values were normalized to the counts of DAPI-stained nuclei. Cell morphology and tight junction integrity were assessed by measuring ZO-1 and OCLN fluorescence intensity at cell junctions, which were processed in Fiji/ImageJ using segmentation-based analysis to determine tight junction localization and disruption. 1.16. Quantification and Statistical Analysis All statistical analyses were conducted using SPSS (IBM SPSS Statistics, Armonk, NY, USA) and GraphPad Prism (GraphPad Software, San Diego, CA, USA). We employed independent Student’s t-tests (for comparisons between two groups/conditions) or one-way ANOVA followed by post hoc tests (for comparisons of three or more groups/conditions), or two-way ANOVA followed by post hoc tests (for comparisons involving three or more groups/two or more conditions) to assess the statistical significance among experimental groups. Data are presented as means ± standard deviation (SD), with statistical significance established at a 95% confidence interval (CI) (p < 0.05, p < 0.01, p < 0.001, p < 0.0001 for two-group or three-group analyses). Details of the statistical analyses are provided in the figure legends. No statistical methods were utilized to pre-determine sample sizes; however, experiments were repeated to improve reproducibility and minimize confidence intervals. The data collection and analysis were not blinded to the experimental conditions, and randomization and exclusions were not implemented. Abbreviations AD Alzheimer's disease CNS Central Nervous System GS Glutamine Synthetase CSF Cerebrospinal fluid NVU neurovascular unit ROS reactive oxygen species 3D Three-dimensional ALDH1L1 aldehyde dehydrogenase 1 family member L1 Tuj1 neuron-specific class III beta-tubulin NeuN neuronal nuclear protein BBB blood-brain barrier EC Endothelial cell AC Astrocytes Neu Neuron NH 4 CH 3 CO 2 ammonia acetate NH 4 Cl ammonium chloride H 2 O 2 hydrogen peroxide GLN glutamine pTau phosphor-tau ZO-1 Zonula occludens-1 PECAM Platelet endothelial cell adhesion molecule-1 IFN-ɣ Interferon gamma IL-2 Interleukin 2 MIF Macrophage migration inhibitory factor TNF-α Tumor necrosis factor alpha IL-1ra Interleukin 1 receptor antagonist IL-4 Interleukin 4 IL-5 Interleukin 5 IL-13 Interleukin 13 IL-1β Interleukin-1 beta IL-10 Interleukin 10 IL-6 Interleukin 6 IL-8 Interleukin 8 NPC Neural progenitor cells GFAP Glial fibrillary acidic protein CCL2 chemokine (C-C motif) ligand 2 CT Control WT Wide-type iNOS Inducible nitric oxide synthase LPS Lipopolysaccharide Ca 2+ calcium OCLN Occludin Declarations Data Availability All data generated or analyzed during this study are included in this published article (and its supplementary information files) or are available from the corresponding author on reasonable request. Acknowledgments We acknowledge MicroTas2024 for supporting this project through the CBMS Student/Young Researcher Grant and the Biomicrofluidics Best Paper Award (M.T.). Author Contribution M.T. developed the microfluidic device, designed and performed the experiments, analyzed the data, and wrote the manuscript. H.W.J. designed and developed the microfluidic device. C.B. conducted experiments using the fabricated microfluidic devices, while V.T.A.T. was responsible for collecting Helicobacter-conditioned media. A.J. performed experiments related to Alzheimer’s disease (AD) modeling, and M.J.A. conducted experiments on Parkinson’s disease (PD) modeling. L.P.L. provided insightful feedback and contributed to the overall supervision of the project. H.C. conceived, initiated, provided insightful feedback, and supervised the entire study. Ethics Declarations All authors were fully involved in the study, and the material within has not been and will not be submitted for publication elsewhere. All authors declare no competing interests. Funding This study was funded by the National Research Foundation of Korea (NRF) grant, funded by the Korea government (MSIT) (RS-2024-00336758), the Korea Dementia Research Project through the Korea Dementia Research Center (KDRC), funded by the Ministry of Health & Welfare and Ministry of Science and ICT, Republic of Korea (RS-2022-KH126808), the Technology development Program, funded by the Ministry of SMEs and Startups (MSS, Korea) (RS-2023-00224842), and the Fourth Stage of Brain Korea 21 Project in Department of Intelligent Precision Healthcare, Sungkyunkwan University (SKKU) to H.C. References Loh JS et al (2024) Microbiota–gut–brain axis and its therapeutic applications in neurodegenerative diseases. Signal Transduct Target Ther 9 Parker A, Fonseca S, Carding SR (2020) Gut microbes and metabolites as modulators of blood-brain barrier integrity and brain health. Gut Microbes 11:135–157 Yadav H et al (2023) Unveiling the role of gut-brain axis in regulating neurodegenerative diseases: A comprehensive review. Life Sci 330 Knox EG, Aburto MR, Clarke G, Cryan JF, O’Driscoll CM (2022) The blood-brain barrier in aging and neurodegeneration. Mol Psychiatry 27:2659–2673 Agirman G, Yu KB, Hsiao EY (2021) Signaling inflammation across the gut-brain axis. Sci (80-) 374:1087–1092 Yang D, Almanzar N, Chiu IM (2023) The role of cellular and molecular neuroimmune crosstalk in gut immunity. Cell Mol Immunol 20:1259–1269 H., A. et al. Microbiota-derived metabolites as drivers of gut-brain communication. Gut Microbes 14, 2102878 (2022) Vogt NM et al (2017) Gut microbiome alterations in Alzheimer’s disease. Sci Rep 7 Jung JH et al (2022) Gut microbiome alterations in preclinical Alzheimer’s disease. PLoS ONE 17 Zhang F et al (2020) Altered gut microbiota in Parkinson’s disease patients/healthy spouses and its association with clinical features. Park Relat Disord 81:84–88 Romano S et al (2021) Meta-analysis of the Parkinson’s disease gut microbiome suggests alterations linked to intestinal inflammation. npj Park Dis 7 Mou Y et al (2022) Gut Microbiota Interact With the Brain Through Systemic Chronic Inflammation: Implications on Neuroinflammation, Neurodegeneration, and Aging. Front Immunol 13 Bairamian D et al (2022) Microbiota in neuroinflammation and synaptic dysfunction: a focus on Alzheimer’s disease. Mol Neurodegener 17 Warren A, Nyavor Y, Zarabian N, Mahoney A, Frame LA (2024) The microbiota-gut-brain-immune interface in the pathogenesis of neuroinflammatory diseases: a narrative review of the emerging literature. Front Immunol 15 T.L.A., N., S., V.-S., A., L. & J., R. How informative is the mouse for human gut microbiota research? DMM Dis Model Mech 8, 1–16 (2015) Oriá RB et al (2018) Revisiting inbred mouse models to study the developing brain: The potential role of intestinal microbiota. Front Hum Neurosci. 12 Kim MH, van Noort D, Sung JH, Park S (2021) Organ-on-a-chip for studying gut-brain interaction mediated by extracellular vesicles in the gut microenvironment. Int J Mol Sci 22 Ashammakhi N et al (2020) Gut-on-a-chip: Current progress and future opportunities. Biomaterials 255 Taavitsainen S et al (2024) Gut-on-chip devices as intestinal inflammation models and their future for studying multifactorial diseases. Front Lab Chip Technol 2 Berry D et al (2018) Organ-On-A-Chip in vitro Models of the Brain and the Blood-Brain Barrier and Their Value to Study the Microbiota-Gut-Brain Axis in Neurodegeneration. Curr Opin Biotechnol 13:247–257 A guide to (2022) the organ-on-a-chip. Nat Rev Methods Prim 2 Porter RJ, Murray GI, McLean MH (2020) Current concepts in tumour-derived organoids. Br J Cancer 123:1209–1218 Urzì O et al (2023) Three-Dimensional Cell Cultures: The Bridge between In Vitro and In Vivo Models. Int J Mol Sci 24 Denman CR, Park SM, Jo J (2023) Gut-brain axis: gut dysbiosis and psychiatric disorders in Alzheimer’s and Parkinson’s disease. Front Neurosci 17 Park AM, Tsunoda I (2022) Helicobacter pylori infection in the stomach induces neuroinflammation: the potential roles of bacterial outer membrane vesicles in an animal model of Alzheimer’s disease. Inflamm Regen 42 Guo S, Al-Sadi R, Said HM, Ma TY (2013) Lipopolysaccharide causes an increase in intestinal tight junction permeability in vitro and in vivo by inducing enterocyte membrane expression and localization of TLR-4 and CD14. Am J Pathol 182:375–387 Jaworowicz DJ, Korytko PJ, Lakhman SS, Boje KM (1998) K. Nitric oxide and prostaglandin E2 formation parallels blood-brain barrier disruption in an experimental rat model of bacterial meningitis. Brain Res Bull 46:541–546 Brown GC, Camacho M, Williams-Gray CH (2023) The Endotoxin Hypothesis of Parkinson’s Disease. Mov Disord 38:1143–1155 Brown GC, Heneka MT (2024) The endotoxin hypothesis of Alzheimer’s disease. Mol Neurodegener 19 Mulak A, Bonaz B (2015) Brain-gut-microbiota axis in Parkinson’s disease. World J Gastroenterol 21:10609–10620 Zhang X, Tang B, Guo J (2023) Parkinson’s disease and gut microbiota: from clinical to mechanistic and therapeutic studies. Transl Neurodegener 12 Zhang W, Xiao D, Mao Q, Xia H (2023) Role of neuroinflammation in neurodegeneration development. Signal Transduct Target Ther 8 Janelidze S et al (2018) CSF biomarkers of neuroinflammation and cerebrovascular dysfunction in early Alzheimer disease. Neurology 91:e867–e877 Gilbert MAG et al (2024) CryoET of β-amyloid and tau within postmortem Alzheimer’s disease brain. Nature 631:913–919 Irwin DJ, Lee VMY, Trojanowski JQ (2013) Parkinson’s disease dementia: Convergence of α-synuclein, tau and amyloid-β pathologies. Nat Rev Neurosci 14:626–636 Pei Y, Lu Y, Li HZ, Jiang CY, Wang L (2023) Gut microbiota and intestinal barrier function in subjects with cognitive impairments: a cross-sectional study. Front Aging Neurosci 15 Kowalski K, Mulak A (2019) Brain-gut-microbiota axis in Alzheimer’s disease. J Neurogastroenterol Motil 25:48–60 Guan J et al (2013) Vascular degeneration in parkinsons disease. Brain Pathol 23:154–164 Additional Declarations There is NO Competing Interest. Supplementary Files GraphicalAbstract.png SupplementaryInformation.docx Cite Share Download PDF Status: Published Journal Publication published 07 Feb, 2026 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6128513","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":427405979,"identity":"0d97ef96-a1e1-46c1-8187-cc3ae2f90eb0","order_by":0,"name":"Hansang Cho","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAuUlEQVRIiWNgGAWjYBACAwYGNgaGCiCLhw3EZyZWyxmStTC2kaLFnIH52YOP82oZ5HuOpUkwVFgnNhDSYtnAZm44c9txBoOzbcckGM6kE9ZicICHTZp32zEGA372NgnGtsPEaplzjEG+H6TlH9FaGmoYGEAOY2wgRsthNjPJGccO8BicOZZskXAs3ZiwluPNzyQ+1NTJyfekGd74UGMtS1ALNCIO84CpBILKEaCOBLWjYBSMglEw4gAA8ss3Pjl94iMAAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0003-1829-2462","institution":"Institute of Quantum Biophysics, Sungkyunkwan University, Suwon","correspondingAuthor":true,"prefix":"","firstName":"Hansang","middleName":"","lastName":"Cho","suffix":""},{"id":427405980,"identity":"ecc2e855-0559-4ca3-9adf-1e389401f910","order_by":1,"name":"Minh Tran","email":"","orcid":"https://orcid.org/0000-0002-7657-8541","institution":"Sungkyunkwan University","correspondingAuthor":false,"prefix":"","firstName":"Minh","middleName":"","lastName":"Tran","suffix":""},{"id":427405981,"identity":"ee653b3e-5119-40f8-b99d-dd58cad2bb3c","order_by":2,"name":"Hoe Jeong","email":"","orcid":"","institution":"Sungkyunkwan University","correspondingAuthor":false,"prefix":"","firstName":"Hoe","middleName":"","lastName":"Jeong","suffix":""},{"id":427405982,"identity":"b8ecc216-c18e-4ca0-b778-f0e09b4a9949","order_by":3,"name":"Chaeyeon Been","email":"","orcid":"","institution":"Sungkyunkwan University","correspondingAuthor":false,"prefix":"","firstName":"Chaeyeon","middleName":"","lastName":"Been","suffix":""},{"id":427405983,"identity":"33d788c2-4dfa-499d-a932-d1d25a0200e0","order_by":4,"name":"Ariunzaya Jamsranjav","email":"","orcid":"","institution":"Sungkyunkwan University","correspondingAuthor":false,"prefix":"","firstName":"Ariunzaya","middleName":"","lastName":"Jamsranjav","suffix":""},{"id":427405984,"identity":"7a3584c0-228a-4949-b036-dce4e2b7c8c8","order_by":5,"name":"Minjoon An","email":"","orcid":"","institution":"Sungkyunkwan University","correspondingAuthor":false,"prefix":"","firstName":"Minjoon","middleName":"","lastName":"An","suffix":""},{"id":427405985,"identity":"d7756a4e-ac99-4002-9256-1dda51b0c0d9","order_by":6,"name":"Van Tran","email":"","orcid":"https://orcid.org/0000-0003-3521-6701","institution":"Sungkyunkwan University","correspondingAuthor":false,"prefix":"","firstName":"Van","middleName":"","lastName":"Tran","suffix":""},{"id":427405986,"identity":"2e225971-39b4-4791-92b8-fe8084ce2804","order_by":7,"name":"Luke Lee","email":"","orcid":"https://orcid.org/0000-0002-1436-4054","institution":"Harvard Medical School, Harvard University","correspondingAuthor":false,"prefix":"","firstName":"Luke","middleName":"","lastName":"Lee","suffix":""}],"badges":[],"createdAt":"2025-02-28 12:15:27","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6128513/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6128513/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41467-026-69318-y","type":"published","date":"2026-02-07T05:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":78368898,"identity":"64fdd03c-4554-4ec2-b878-a94716a78f8f","added_by":"auto","created_at":"2025-03-12 13:39:34","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1795064,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eHuman gut-brain-vascular (hGBV) physiomimetic system for modeling gut-brain axis interactions\u003c/strong\u003e. (a) Schematic representation of the Gut-Brain Axis via circulation vessel system. (b) Gut-Brain-Vascular-on-a-chip platform. (c) Stepwise assembly of the hGBV model. (d) Immunofluorescence images of the tetra-culture model, showing gut epithelium (left: gut-red and right: Gut ViL1-green), capillary-like vascular structures (iBMEC-blue), and brain tissue microenvironment (left: neuron and astrocyte-cyan, right: neurons/Tuj1-red and astrocytes/GFAP-yellow). (e)\u0026nbsp; Zoom-in 3D gut villus-like structures (Gut-ViL1-green, DAPI). (f) FITC-Dextran (4 kDa) leakage assays. (g) Quantification of leakage over 24 hours. (h) Immunofluorescence of gut tight junction (OCLN-red), vascular tight junction (ZO-1-green) and astrocyte (GFAP-yellow) and neuron (Tuj1-red) in gut-to-brain axis. (e) Fold change of gut OCLN localization, vascular ZO-1 localization, and GFAP expression in gut-to-brain axis. (j) Immunofluorescence of phosphorylated tau (pTau, red) expression, vascular tight junction (ZO-1-green), and gut tight junction (OCLN-red) in brain-to-gut axis. (k) Fold change of pTau expression, vascular ZO-1 localization, gut OCLN localization, in brain-to-gut axis. Scale bars, 500 μm (d), 100 μm (d, f), 20 μm (e, h, j). Sample sizes, n=3 (g), 10 (i, k).\u0026nbsp; Data represent Mean ± SD. ns, non-significant difference; p-value calculated by independent Student’s t-tests. CT represents control; HP\u003csub\u003eCM\u003c/sub\u003e, Helicobacter-conditioned media; WT, Wide-type; AD, Alzheimer’s disease; G.C., Gut compartment; V.C., Vascular compartment; B.C., Brain compartment; NPC, Neural progenitor cell.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6128513/v1/50a5f7225077bcda92ed9d1c.png"},{"id":78369456,"identity":"19ea22f1-4771-4cd9-addf-82df2ddaf590","added_by":"auto","created_at":"2025-03-12 13:47:34","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":2021209,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePhysiological validation of gut barrier and vascular barrier integrity in the hGBV model.\u003c/strong\u003e (a) Immunofluorescence images of the gut-vascular connection, showing compartmentalized gut epithelium (Gut-Vil1-green, G.C.) and vascular endothelium (iBMECs-blue, V.C.) connected by multi-microchannel. (b) The neurovascular interface shows endothelial cells (iBMECs-blue, V.C.) interacting with astrocytic end-feet (ALDH1L1-yellow) in multi-microchannel and neurons (Tuj1-red) in the brain compartment (B.C.), mimicking the neurovascular unit (NVU). Characterization of individual compartmental features: (c) The gut epithelium villus-like markers (left: ViL1, right: OCLN); (d) The vascular capillary-like endothelium marker (left: CD31, right: ZO-1); (e) The brain environment communication through synaptic marker expression (left: SYN-1) and calcium signaling activity (right: OGB-1 Ca²⁺). (f) Cross-sectional images of the gut villus-like structures (ViL1-green, OCLN-red, DAPI-blue). Quantification of (g) gut thickness, (h) gut surface area, and (i) gut uptake capacity, (j-k) tight junction of ViL1 and OCLN expression, (l) gut permeability (P\u003cem\u003eapp\u003c/em\u003e), (m) FITC-dextran diffusion over 24h hrs. Brain vascular endothelium integrity of (n) ZO-1 expression, (o) quantification of ZO-1 expression, and (p) endothelial cell morphology analysis. Scale bars, 100 μm (a, b), 20 μm (c-f, n). Sample sizes, n=10 (g-k, o), 3 (l-m), 30 (p). Data represent Mean ± SD. ns, non-significant difference; p-value calculated by independent Student’s t-tests. G.C., Gut compartment; V.C., Vascular compartment; B.C., Brain compartment.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6128513/v1/31bb30fc1b5314d6dec2f2eb.png"},{"id":78368896,"identity":"bc47c29a-17e6-40ba-90e9-c680e9e7f563","added_by":"auto","created_at":"2025-03-12 13:39:34","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1790342,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eGut-originated inflammation induces vascular dysfunction and neurodegeneration via gut-to-brain pathways signaling. \u003c/strong\u003e(a) Schematic representation of gut-to-brain signaling, illustrating barrier disruption and leakage into the brain following HP\u003csub\u003eCM\u003c/sub\u003e and LPS treatment in the gut. Quantification of (b) endotoxin levels and (c) gut permeability index (Papp). (d) Immunofluorescence images of gut Epi villus marker ViL1 (green) expression, Epi CD13 (yellow) expression, Epi tight junction OCLN (green) expression and nuclear (DAPI, blue). Quantification of (e) ViL-1 expression, (f) CD13 expression, and (g) OCLN localization analysis. (h) Heatmap analysis of pro-inflammatory factors release (log₂ fold-change) from the gut epithelium. (i) Immunofluorescence images of vascular EC iNOS (yellow) expression, NO (green) expression, EC tight junction ZO-1 (red) expression and nuclear (DAPI, blue). Quantification of (j) iNOS expression and (k) NO release. (l) Heatmap analysis of pro-inflammatory factors release from endothelial cells. (m) Immunofluorescence images of complement activation marker C3 (red) expression, synaptic marker Synapsin-1 (red) expression, phosphorylated tau (pTau, red) expression, and nuclear (DAPI, blue). Quantification of vascular barrier breakdown upon gut stimuli treatment, (n) EC density, (o) EC ZO-1 localization, and (p) EC permeability (P\u003cem\u003eapp\u003c/em\u003e). Quantification of (q) C3 expression; (r) Syn-1 expression; (s) pTau expression; and (t) neuronal density. Scale bars, 20 μm (d, I, m). Sample sizes, n=5 (b-c, e-g, j-k, o-u), 2 (h, l). Data represent Mean ± SD. ns, non-significant difference; p-value calculated by one‐way ANOVA with Tukey post‐hoc correction for multiple comparisons. CT represents control; HP\u003csub\u003eCM\u003c/sub\u003e, Helicobacter-conditioned media; and LPS, Lipopolysaccharide; G.C., Gut Compartment; V.C., Vascular Compartment; B.C., Brain Compartment.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6128513/v1/4637934e4217b98d07396bc3.png"},{"id":78368900,"identity":"fe3fa689-2b3c-40d2-993b-af2b17479873","added_by":"auto","created_at":"2025-03-12 13:39:34","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":1793200,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eModeling brain-to-gut crosstalk in neurodegenerative disease. \u003c/strong\u003e(a) Schematic representation of brain-to-gut crosstalk in neurodegenerative disease. Brain-derived protein aggregates contribute to vascular and gut dysfunction, leading to inflammatory cytokine leakage across compartments.\u0026nbsp; Quantification of (b) Aβ\u003csub\u003e42\u003c/sub\u003e and (c) α-Syn levels across compartments. (d) Immunofluorescence images of complement activation marker C3 (red) expression, synaptic marker Synapsin-1 (red) expression, phosphorylated tau (pTau, red) expression, and nuclear (DAPI, blue). Quantification of (e) C3 expression, (f) SYN1 expression, (g) pTau expression, (h) neuronal density (i) Heatmap analysis of pro-inflammatory factors release in the brain compartment. Immunofluorescence images of (j) vascular EC iNOS (yellow) expression and NO (green) expression, EC tight junction ZO-1 (red) expression; and (k) gut Epi CD13 (yellow) expression and villus marker ViL1 (green) expression, Epi tight junction OCLN (green) expression, and nuclear (DAPI, blue). Quantification of (l) iNOS expression; (m) NO expression, (n) EC tight junction ZO-1 localization analysis; (o) EC density; (p) EC permeability (P\u003cem\u003eapp\u003c/em\u003e).\u0026nbsp; Quantification of gut q) Villus marker ViL1 expression; (r) CD13 expression, (s) gut Epi tight junction localization and (t) Epi cell density. Scale bars, 20 μm (d, j, k). Sample sizes, n=3 (b-c), 5 (e-h, l-m, n-t), 2 (i). Data represent Mean ± SD. ns, non-significant difference; p-value calculated by one‐way ANOVA with Tukey post‐hoc correction for multiple comparisons. WT, Wild-type; AD, Alzheimer’s disease; PD, Parkinson’s disease; G.C., Gut Compartment; V.C., Vascular Compartment; B.C., Brain Compartment.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-6128513/v1/88c2c046e6e71f987107e460.png"},{"id":104866249,"identity":"dadb3f33-c33e-494c-bef7-d3c1546ec2ae","added_by":"auto","created_at":"2026-03-18 07:07:08","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":7948952,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6128513/v1/d5737251-1252-4d5d-8c45-0d1cca23e9e6.pdf"},{"id":78368902,"identity":"f0c5e85b-6958-4fe0-9c89-afe877323c15","added_by":"auto","created_at":"2025-03-12 13:39:34","extension":"png","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1341102,"visible":true,"origin":"","legend":"","description":"","filename":"GraphicalAbstract.png","url":"https://assets-eu.researchsquare.com/files/rs-6128513/v1/a4082ad57e11bff3d3bb17bd.png"},{"id":78368903,"identity":"33304b4f-851c-4910-8a9d-c2b311569d98","added_by":"auto","created_at":"2025-03-12 13:39:34","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":4597279,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryInformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-6128513/v1/a6fd97cf1defe7bcecc561f4.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"A 3D Gut-Brain-Vascular Platform for Bidirectional Crosstalk in Gut-Neuropathogenesis","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe gut-brain axis (GBA), a bidirectional communication network between the gastrointestinal (GI) tract and the central nervous system (CNS), plays a crucial role in maintaining neurological homeostasis\u003csup\u003e\u003cspan additionalcitationids=\"CR2 CR3\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. This interaction is mediated through neuronal (vagus nerve), immune system (immune cells/cytokine signaling), and circulatory (vasculature) pathways\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. These pathways facilitate bidirectional communication, primarily mediated by neuroactive substances, microbial metabolites, and hormones that regulate neurological function and systemic homeostasis. Among these, the vasculature serves as a critical interface, regulating immune surveillance, molecular exchange, and inflammatory responses between the gut, the periphery, and the brain\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. Mounting evidence underscores the microbiota\u0026ndash;gut\u0026ndash;brain axis as a key contributor to neurodegeneration, with early alterations in gut microbiota identified in Alzheimer\u0026rsquo;s disease (AD) and Parkinson\u0026rsquo;s disease (PD) patients, suggesting a potential causal link between gut dysbiosis and neurodegeneration\u003csup\u003e\u003cspan additionalcitationids=\"CR8 CR9 CR10\" citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. Animal studies further support this hypothesis, demonstrating that microbiota-driven neuroinflammation and glial cell dysfunction contribute to disease progression through systemic immune activation, inflammatory cascades, and blood-brain barrier (BBB) disruption\u003csup\u003e\u003cspan additionalcitationids=\"CR13\" citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. Microbial-derived toxins and inflammatory mediators have been implicated in BBB permeability loss, neuronal dysfunction, and protein aggregation, reinforcing the notion that gut dysbiosis is a key driver of neurodegenerative pathology.\u003c/p\u003e \u003cp\u003eDespite these insights, a significant limitation in the field is the lack of physiologically \u003cem\u003erelevant in vitro\u003c/em\u003e models that accurately replicate the complex, dynamic interactions among the gut, vasculature, and brain. Most studies depend on animal models, which, while informative, do not capture the human-specific mechanisms underlying gut-brain communication\u003csup\u003e\u003cspan additionalcitationids=\"CR16\" citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. Existing \u003cem\u003ein vitro\u003c/em\u003e systems typically concentrate on single-organ models, such as gut-on-chip, BBB-on-chip, or brain-on-chip platforms; however, these models fail to recapitulate the multi-organ interplay that underlies gut-brain dysfunction in neurodegeneration\u003csup\u003e\u003cspan additionalcitationids=\"CR19\" citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. Additionally, traditional Transwell assays and organoid cultures do not integrate circulatory flow, barrier dynamics, and cellular interactions necessary to mimic the \u003cem\u003ein vivo\u003c/em\u003e microenvironment\u003csup\u003e\u003cspan additionalcitationids=\"CR22\" citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e. Consequently, critical questions regarding the mechanistic underpinnings of gut-brain communication remain unresolved, hindering the development of targeted therapeutic interventions. To tackle these challenges, we developed a 3D microfluidic Gut-Brain-Vascular (GBV) platform that models bidirectional crosstalk in gut-neuropathogenesis. Our system features a 3D lumenized gut barrier with villus-like architecture to replicate the intestinal epithelium, a capillary-like structure compartment containing brain microvascular endothelial cells, and a 3D brain tissue compartment containing neurons and astrocytes that simulates endothelial-astrocyte interactions, facilitating dynamic molecular exchange.\u003c/p\u003e \u003cp\u003eOur system enables real-time investigation of bidirectional inflammatory crosstalk by modeling gut-to-brain signaling and examining brain-to-gut interactions. With the gut-to-brain axis, bacterial-derived toxins (e.g., lipopolysaccharides, Helicobacter-derived toxins) affect the gut compartment, causing gut barrier dysfunction and increased vascular permeability, allowing toxins and pro-inflammatory factors from gut and vascular organs to infiltrate the brain. This ultimately leads to the accumulation of phosphorylated tau, which is a hallmark of neurodegeneration. In the brain-to-gut axis, AD-like (NPC-derived AD models) and PD-like (PFF α-synuclein aggregates) environments in the brain compartment induce vascular barrier breakdown and subsequent gut epithelial dysfunction. By integrating gut, vascular, and brain compartments within a physiologically relevant microfluidic system, this model serves as a sophisticated microphysiological tool for studying gut-brain interactions in neurodegeneration. Our findings offer critical insights into the role of gut-origin inflammation in neurodegeneration, identifying mechanisms of diseases such as Alzheimer\u0026rsquo;s and Parkinson\u0026rsquo;s and laying the groundwork for screening novel therapeutic interventions targeting the gut-brain-vascular axis.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eEngineering a physiologically relevant human gut-brain-vascular (hGBV) model\u003c/h2\u003e \u003cp\u003eThe gut-brain axis (GBA) is a complex bidirectional communication network regulated by neuronal, immune, and vascular interactions\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. However, current \u003cem\u003ein vitro\u003c/em\u003e models lack the physiologically relevant microenvironment needed to study gut-brain-vascular crosstalk. To address this issue, we created a 3D microfluidic Human Gut-Brain-Vascular (hGBV) model that combines lumenized gut epithelium, perfusable vasculature, and a neurovascular brain compartment in a dynamic microfluidic system (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). The hGBV system consists of three interconnected compartments: Gut Compartment (G.C.), Vascular Compartment (V.C.), and Brain Compartment (B.C.), designed to recapitulate key features of gut-brain interactions (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb). Each compartment was engineered using human-derived cell types to closely mimic native tissue architectures. The gut compartment was developed using Caco-2 cells, forming a 3D villus-like epithelium to replicate intestinal barrier functions. The vascular compartment consisted of induced brain microvascular endothelial cells (iBMECs), forming a capillary-like endothelium that enables selective molecular transport. The brain compartment incorporated neural progenitor cell (NPC)-derived neurons and astrocytes, which interact with vascular endothelial cells via astrocytic end-feet, representing the neurovascular unit (NVU). The stepwise assembly of the hGBV model (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec, Supplementary Fig.\u0026nbsp;2) involved the sequential differentiation of brain tissue, gut epithelium formation, and vascular compartment development, ensuring the formation of interconnected and physiologically relevant microenvironments.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo confirm the physiological relevance of the hGBV model, we assessed the formation of 3D gut villus-like structures using immunofluorescence imaging. The gut compartment exhibited well-organized epithelial cell layers with tight junction integrity, closely resembling native intestinal architecture (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed-e). Immunostaining of the vascular and brain compartments further confirmed the presence of iBMECs in the vascular compartment and neurons with astrocyte support in the brain compartment, forming a multicellular microenvironment capable of bidirectional signaling (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed-e). Additionally, marker expression analysis confirmed ViL1 in the gut epithelium, Tuj1 in neurons, and GFAP in astrocytes, validating the structural integrity of the system. To evaluate barrier integrity and compartmental media separation, we performed FITC-Dextran permeability assays across all compartments. Under physiological conditions, the gut and vascular barriers demonstrated compartment-specific permeability, with approximately 15% FITC-Dextran leakage into the vascular compartment and \u0026lt;\u0026thinsp;5% leakage into the brain compartment (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ef-g). This selective permeability indicates the presence of intact barrier functions, effectively preventing uncontrolled diffusion of across compartments.\u003c/p\u003e \u003cp\u003eTo validate the bidirectional crosstalk of the system, we performed two distinct experiments modeling gut-to-brain and brain-to-gut signaling. To investigate how bacterial-derived toxins from the gut contribute to neurodegeneration, we exposed the gut compartment to Helicobacter-conditioned media (HP\u003csub\u003eCM\u003c/sub\u003e), mimicking gut dysbiosis-induced inflammation. Exposure to HP\u003csub\u003eCM\u003c/sub\u003e led to the disruption of barrier tight junctions and glial activation, as indicated by an approximate 50% reduction in gut epithelium OCLN expression, an approximate 50% reduction in vascular endothelium ZO-1 expression, and a 2-fold increase in GFAP expression in the brain compartment, suggesting neuroinflammatory responses triggered by gut-derived and vascular-derived signals (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eh-i). To investigate brain-to-gut signaling, we modeled Alzheimer\u0026rsquo;s disease (AD)-like conditions in the brain compartment by differentiating NPCs that were transfected with the APP gene for six weeks. This resulted in pTau accumulation (3-fold increase), vascular barrier breakdown, and loss of gut epithelial integrity. Specifically, endothelium ZO-1 expression in the vascular compartment was reduced by 50%, while gut epithelial OCLN expression decreased by 30%, confirming that neuroinflammatory factors contribute to systemic vascular and gut dysfunction (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ej-k).\u003c/p\u003e \u003cp\u003eTogether, these findings illustrate that the hGBV model effectively replicates bidirectional gut-brain-vascular interactions, facilitating the study of gut-to-brain and brain-to-gut signaling in both physiological and disease conditions. The capacity to model gut-derived inflammation, neurovascular dysfunction, and loss of systemic barrier integrity emphasizes the translational potential of this system for investigating neurodegenerative disorders, including Alzheimer's and Parkinson's disease.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003ePhysiological Properties of the hGBV Model\u003c/h3\u003e\n\u003cp\u003eTo functionally validate the hGBV model, we assessed the structural, barrier, and transport properties of the gut, vascular, and brain compartments, comparing static and dynamic perfusion conditions. Dynamic flow enhances physiological relevance in microfluidic systems, supporting barrier integrity, tissue maturation, and functional molecular transport. Thus, we examined gut morphology, permeability, vascular endothelial properties, and neurovascular interactions to confirm that the hGBV model recapitulates key in vivo-like physiological properties. Immunofluorescence analysis confirmed that the hGBV system forms a compartmentalized gut-vascular-brain interface with distinct cellular identities in each region (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed, Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea-b, Supplementary Fig.\u0026nbsp;4a). The gut compartment (G.C.) developed a villus-like epithelium, characterized by Gut-ViL-1 serving as villi markers, adjacent to the vascular compartment (C.C.), which contained induced brain microvascular endothelial cells (iBMECs; Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). The brain compartment (B.C.) exhibited an organized neurovascular interface, where iBMECs formed endothelial monolayers that interacted with astrocytic end-feet (ALDH1L1) and neurons (Tuj1), mimicking the neurovascular unit (NVU; Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb, Supplementary Fig.\u0026nbsp;4l).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo assess the functional gut barrier properties, we examined the expression of villi and tight junction proteins (ViL-1 and OCLN), and performed permeability assays. Immunostaining revealed that the gut epithelium formed tight junctions with well-distributed ViL-1 (left) and OCLN (right) expression (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec). The vascular compartment contained capillary-like endothelial structures expressing CD31 (left) and ZO-1 (right), indicating tightness of brain endothelial vascular junctions (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed). The brain compartment showed interactions between neurons and astrocytes, along with synaptic marker expression (SYN-1) and calcium signaling activity (OGB-1 Ca\u0026sup2;⁺), suggesting functional neural connectivity (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee). A comparison of gut barrier function under static versus dynamic conditions revealed enhanced gut morphology and barrier integrity in dynamic conditions. Cross-sectional imaging of the gut compartment showed improved villus-like structures in the gut epithelium during dynamic conditions, with increased expression of ViL-1 and OCLN (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef). Morphometric analysis demonstrated that dynamic conditions significantly increased gut epithelium thickness (1.5-fold), surface area (1.7-fold), and uptake capacity (2.2-fold), highlighting improved epithelial structure and nutrient absorption (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eg-k). Permeability measurements (P\u003cem\u003eapp\u003c/em\u003e) showed significantly reduced FITC-dextran diffusion under dynamic perfusion, indicating a more restrictive and physiologically relevant gut barrier (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003el). Time-lapse analysis of gut permeability over 24 hours further revealed that FITC-dextran penetration was significantly lower under dynamic conditions, confirming enhanced tight junction functionality and selective molecular transport (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003em). These results demonstrate the successful formation of a gut epithelium-like villus structure that replicates key features of the in vivo system.\u003c/p\u003e \u003cp\u003eNext, we evaluated vascular barrier properties, particularly endothelial tight junction integrity and dynamic-induced morphological adaptations in the vascular compartment. Immunofluorescence staining of ZO-1 confirmed that dynamic conditions promoted stronger endothelial junction formation (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003en). Quantification of ZO-1 expression showed a significant upregulation (1.5-fold) in dynamic conditions, supporting improved BBB integrity (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eo). In addition, endothelial cell (EC) morphology analysis revealed that dynamic perfusion induced an elongated, flow-adapted endothelial phenotype, as quantified by an increased Feret ratio (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ep). This shift from static cuboidal morphology to an elongated form is characteristic of physiological endothelial responses to shear stress, confirming that the hGBV vascular compartment successfully recapitulates in vivo-like BBB characteristics.\u003c/p\u003e \u003cp\u003eTogether, these results demonstrate that a dynamic condition enhances the physiological properties of the hGBV model, improving gut barrier function, vascular endothelial integrity, and neurovascular interactions. The integration of flow-adapted villus-like gut epithelium, endothelial cells, and functional neurovascular interfaces establishes hGBV as a biomimetic platform for studying gut-brain-vascular dynamics.\u003c/p\u003e\n\u003ch3\u003eModeling Gut-to-Brain Inflammation and Microbiota-Driven Neuroinflammation\u003c/h3\u003e\n\u003cp\u003eDysbiosis and microbial-derived inflammation are increasingly recognized as key contributors to neurodegenerative disease progression\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. To model gut-to-brain inflammation and microbiota-driven neuroinflammation, we exposed the gut compartment (G.C.) to either Helicobacter-conditioned media (HP\u003csub\u003eCM\u003c/sub\u003e, 10%) or pure lipopolysaccharide (LPS, 1 \u0026micro;g/mL) to mimic microbial-triggered inflammatory responses (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea). We evaluated the inflammatory cascade, barrier dysfunction, and neurodegeneration at sequential time points, including gut barrier disruption and local inflammatory response after 1 day of exposure to the gut epithelium; vascular inflammation and endothelial dysfunction after 4 days of treatment; neuroinflammation and neurodegeneration occur after approximately 10 days of treatment.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo assess gut epithelial integrity in response to bacterial-derived toxins, we examined tight junction expression, epithelial permeability, and inflammatory factors release. HP\u003csub\u003eCM\u003c/sub\u003e and LPS treatment significantly increased gut permeability, as indicated by higher endotoxin levels in the vascular (V.C.) and brain (B.C.) compartments after 4-day treatment (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb). Increased epithelial permeability index (P\u003cem\u003eapp\u003c/em\u003e) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec) further confirmed barrier dysfunction, suggesting that bacterial-derived components translocated across compartments. Structural integrity of the gut barrier was assessed via immunofluorescence analysis, which revealed a significant reduction in villus marker ViL1 expression, decreasing from 1-fold (control) to ~\u0026thinsp;0.8-fold upon HP\u003csub\u003eCM\u003c/sub\u003e and LPS treatment (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed, e), indicating epithelial damage. Concurrently, CD13 fluorescence intensity increased by approximately 3.2-fold (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001) in HP\u003csub\u003eCM\u003c/sub\u003e- and LPS-treated conditions (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ef), suggesting epithelial remodeling and gut inflammatory activation. Following exposure to HP\u003csub\u003eCM\u003c/sub\u003e/LPS, OCLN localization was reduced from 1.0-fold (control) to 0.7-fold, representing a 30% decrease relative to the control condition (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eg), confirming tight junction breakdown and barrier disruption. A heatmap analysis of pro-inflammatory cytokines and chemokines secreted by the gut epithelium (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eh, Supplementary Fig.\u0026nbsp;5) demonstrated distinct inflammatory signatures responding to bacterial stimuli. HP\u003csub\u003eCM\u003c/sub\u003e-treated condition induced C5a, IFN-γ, IL-5, IL-12, IL-16, and IL-21, whereas both HP\u003csub\u003eCM\u003c/sub\u003e- and LPS-treated conditions resulted in increased IL-1β, IL-6, IL-18, IL-32a, and TNF-α, indicating a robust inflammatory response.\u003c/p\u003e \u003cp\u003eTo determine whether gut-derived inflammation impacted vascular integrity, we assessed tight junction loss, endothelial nitric oxide synthase (iNOS) activation, and inflammatory cytokine release in the vascular compartment (C.C.). Immunofluorescence analysis revealed significant upregulation of iNOS expression (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ei, j) and a corresponding increase in NO production (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ei, k) under HP\u003csub\u003eCM\u003c/sub\u003e- and LPS-treated conditions, confirming the vascular inflammatory response. Vascular integrity was significantly disrupted, as evidenced by ZO-1 delocalization at endothelial junctions (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ei, p). However, endothelial cell (EC) density remained unchanged (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eo), indicating functional endothelial dysfunction rather than cell death. This was further supported by a significant increase in endothelial permeability (P\u003csub\u003e\u003cem\u003eapp\u003c/em\u003e\u003c/sub\u003e) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eq), suggesting barrier dysfunction and enhanced vascular permeability. A heatmap analysis of pro-inflammatory cytokine secretion from endothelial cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003el, Supplementary Fig.\u0026nbsp;6) confirmed an elevated inflammatory signature in response to bacterial-derived toxins. Direct exposure of Helicobacter-conditioned media (HP\u003csub\u003eCM\u003c/sub\u003e, 1%) to endothelial cells triggered increased IFN-γ and IL-12, while LPS (100 ng/mL) exposure induced CXCL1, IL-6, and IL-21 expression. Notably, both HP\u003csub\u003eCM\u003c/sub\u003e and LPS treatments led to increased levels of CXCL10, IL-1β, IL-16, and IL-18, suggesting that bacterial-derived toxins directly activate pro-inflammatory signaling in endothelial cells.\u003c/p\u003e \u003cp\u003eTo examine the impact of gut-origin inflammation on the brain compartment (B.C.), we evaluated complement activation (C3), synaptic integrity (Synapsin-1), and neurodegenerative markers (pTau). Immunofluorescence analysis revealed a 2.5-fold increase in C3 expression in both treatment conditions (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003em, r), suggesting neuroinflammatory activation. Synaptic integrity was assessed by Synapsin-1 (SYN1) expression, which significantly decreased from 1 to approximately 0.7-fold under gut compartment-treated conditions, indicating a nearly 30% reduction in synaptic density and connectivity (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003em, s) and suggesting synaptic loss. Furthermore, phosphorylated tau (pTau) expression was significantly upregulated, approximately 4.9-fold in HP\u003csub\u003eCM\u003c/sub\u003e-treated and 4-fold in LPS-treated gut compartment samples (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003et), a hallmark of neurodegeneration-associated tau pathology. Notably, neuronal density was significantly reduced in the brain compartment (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eu), confirming that gut-derived inflammation leads to neurodegenerative-like phenotypes.\u003c/p\u003e \u003cp\u003eThese findings confirm that gut-origin risk factors, triggered by bacterial-derived stimuli (HP\u003csub\u003eCM\u003c/sub\u003e and LPS), induce a cascade of gut-to-brain inflammatory signaling, leading to neurovascular dysfunction and neurodegeneration. The sequential progression from gut barrier breakdown to vascular, synaptic loss, and tau pathology highlights the critical role of microbiota-driven neuroinflammation in neurodegenerative diseases. This validates the hGBV platform as a translational tool for studying gut-brain interactions in neurodegenerative diseases.\u003c/p\u003e\n\u003ch3\u003eModeling Brain-to-Gut Crosstalk in Neurodegenerative Disease\u003c/h3\u003e\n\u003cp\u003eNeurodegenerative diseases, including Alzheimer\u0026rsquo;s disease (AD) and Parkinson\u0026rsquo;s disease (PD), are increasingly linked to systemic inflammation and gut dysregulation\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. To investigate brain-to-gut signaling and its effects on vascular and gut integrity, we modeled AD- and PD-like conditions in the hGBV system. AD-like conditions were induced by transfecting the APP gene into neural progenitor cells (NPCs) and differentiating them for six weeks, while PD was modeled by differentiating NPCs for three weeks, followed by treatment with preformed fibrillar (PFF) α-synuclein (1 \u0026micro;M) and assessment one-week post-treatment (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea). We evaluated brain inflammatory responses, vascular dysfunction, and gut structural alterations across compartments to determine the impact of brain-derived neuroinflammation on peripheral systems.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo confirm the induction of AD- and PD-like pathology, we quantified the secretion of neurotoxic protein aggregates (Aβ\u003csub\u003e42\u003c/sub\u003e and α-Synuclein) across compartments. Both Aβ\u003csub\u003e42\u003c/sub\u003e (AD-like model) and α-Synuclein (PD-like model) were significantly elevated in the brain compartment (B.C.), with detectable translocation into the vascular (V.C.) and gut (G.C.) compartments in disease conditions (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb, c). This suggests that neurotoxic aggregates can permeate through the vasculature, potentially contributing to systemic neuroinflammation. Additionally, Immunofluorescence analysis of the brain compartment revealed pronounced complement activation (C3, Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ed), indicative of neuroinflammation in AD- and PD-like models. Synaptic integrity was significantly impaired, as evidenced by a reduction in Synapsin-1 (SYN1) expression (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ef), confirming synaptic loss in both disease models. Additionally, phosphorylated tau (pTau) expression was significantly upregulated (3.8-fold in AD-like and 2.2-fold in PD-like conditions, Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eg), reinforcing neurodegenerative tau pathology. The cumulative impact of these changes resulted in a significant decline in neuronal density (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eh), supporting neurodegenerative-associated neuronal loss. Additionally, a heatmap analysis of pro-inflammatory cytokines secreted from AD- and PD-like brain environments (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ei, Supplementary Fig.\u0026nbsp;7) confirmed an elevated inflammatory signature in disease conditions. The AD-like brain environment triggered increased IL-16, MIF, and TNF-α, while both AD- and PD-like conditions led to elevated levels of CCL-1, CCL-2, MIP-1, CCL-5, CXCL1, CXCL10, ICAM-1, IFN-γ, IL-1β, IL-5, IL-6, IL-8, IL-12, IL-18, IL-21, and IL-32a, confirming widespread neuroinflammatory activation.\u003c/p\u003e \u003cp\u003eGiven the role of neuroinflammation in vascular dysfunction, we assessed endothelial inflammation, vascular barrier integrity, and permeability changes in the vascular compartment (V.C.). Immunofluorescence analysis of vascular endothelial cells revealed a significant approximate 2-fold increase in iNOS expression (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ej, l) and a 2.5-fold increase in NO production (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ej, m), suggesting endothelial dysfunction and oxidative stress response to inflammation. Tight junction integrity was broken, as evidenced by ZO-1 delocalization at endothelial junctions. In the AD-like model, ZO-1 localization was reduced from 1.0 to 0.6-fold, indicating a 40% loss of tight junction integrity, while in the PD-like model, it was reduced from 1.0 to 0.8-fold, indicating a 20% loss of junctional stability (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ej, n), confirming vascular barrier breakdown. Consistent with these findings, endothelial cell (EC) density was significantly reduced in AD- and PD-like conditions (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eo), indicating vascular damage and endothelial loss. Moreover, permeability assays demonstrated a significant increase in endothelial permeability (P\u003cem\u003eapp\u003c/em\u003e, Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ep), particularly in AD-like conditions, suggesting a severe breakdown of vascular integrity in response to brain-derived neuroinflammation.\u003c/p\u003e \u003cp\u003eTo determine whether brain-to-gut signaling disrupts gut homeostasis, we examined gut morphology, epithelial barrier integrity, and inflammatory activation in the gut compartment (G.C.). Unlike the pronounced vascular inflammation and endothelial barrier dysfunction observed in the V.C., the gut compartment exhibited only mild structural alterations. Immunofluorescence analysis revealed a moderate reduction in villus marker ViL-1 expression (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ek, q), suggesting partial villus alteration in response to brain-derived inflammation. Additionally, CD13 expression was slightly increased with 1.7-fold in AD and 1.3-fold in PD-like conditions compared to WT (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ek, r), indicating mild epithelial remodeling and inflammatory activation. However, tight junction integrity remained preserved, as evidenced by unchanged occludin (OCLN) localization (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ek, s), and no significant reduction in epithelial cell density was observed (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003et). These findings suggest that while brain-derived inflammatory signals propagate to the gut, they primarily induce structural remodeling rather than direct gut barrier disruption.\u003c/p\u003e \u003cp\u003eThese findings demonstrate that brain-origin inflammation in AD- and PD-like conditions leads to significant vascular dysfunction while only mildly affecting the gut barrier. The vascular compartment exhibited pronounced endothelial inflammation, tight junction disruption, and increased permeability, confirming brain-to-vascular inflammatory signaling. In contrast, the gut compartment showed only mild structural alterations, with reduced villus integrity and increased CD13 expression, but no significant tight junction disruption or epithelial loss. This study supports the hypothesis that brain-derived neuroinflammation propagates inflammatory signals to peripheral compartments, with strong effects on the vasculature but limited gut barrier breakdown. These results establish the hGBV platform as a valuable model for studying brain-to-gut crosstalk in neurodegenerative diseases.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThe gut-brain axis (GBA) is increasingly recognized as a key regulator of systemic homeostasis, with accumulating evidence implicating bidirectional communication between the gut, vasculature, and brain in the progression of neurodegenerative diseases\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. Despite this recognition, current in vitro models fail to fully capture the complex, multi-compartmental interactions that underlie gut-brain inflammatory signaling. In this study, we developed and validated a 3D human Gut-Brain-Vascular (hGBV) model, successfully integrating a physiologically relevant, villus-like gut epithelium, perfusable vasculature, and a neurovascular brain compartment. Using this platform, we demonstrate that gut-origin inflammation (gut-to-brain axis) can drive neurovascular dysfunction and neurodegeneration, while brain-origin neuroinflammation (brain-to-gut axis) predominantly affects vascular integrity, resulting in only mild alterations to the gut barrier. These findings provide critical mechanistic insights into microbiota-driven neuroinflammation and brain-to-gut crosstalk, establishing the hGBV system as a valuable translational model for neurodegenerative disease research.\u003c/p\u003e \u003cp\u003eOur results confirm that gut-derived inflammatory stimuli, including Helicobacter-conditioned media (HP\u003csub\u003eCM\u003c/sub\u003e) and lipopolysaccharide (LPS), trigger systemic inflammation that spreads to the vasculature and brain. The observed gut barrier disruption, characterized by tight junction loss (OCLN, ZO-1), increased permeability, and the release of inflammatory factors, aligns with previous reports linking gut dysbiosis to intestinal barrier breakdown in neurodegenerative diseases\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e,\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. Importantly, bacterial-derived endotoxins translocate into the vascular and brain compartments, leading to vascular dysfunction marked by endothelial nitric oxide synthase (iNOS) upregulation, increased NO production, and blood-brain barrier (BBB) disruption. These findings are consistent with preclinical and clinical studies reporting circulating endotoxins and inflammatory cytokines in Alzheimer\u0026rsquo;s disease (AD) and Parkinson\u0026rsquo;s disease (PD) patients\u003csup\u003e\u003cspan additionalcitationids=\"CR28\" citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. At the neurovascular interface, prolonged exposure to gut-derived inflammatory stimuli results in synaptic dysfunction, complement C3 activation, and phosphorylated tau accumulation, all key hallmarks of neurodegeneration. This progression supports the hypothesis that chronic gut inflammation contributes to neurodegenerative pathology through systemic immune activation, changes in BBB permeability, and direct neuronal damage\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e,\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. Notably, our model recapitulates the sequential progression observed in vivo, where gut inflammation precedes vascular impairment and neuronal loss, reinforcing the critical role of the gut-brain axis in neurodegeneration.\u003c/p\u003e \u003cp\u003eIn contrast to the gut-driven model, brain-origin neuroinflammation (brain-to-gut axis) in AD- and PD-like conditions primarily induced vascular dysfunction, with limited gut barrier disruption. APP-transfected NPCs (AD model) and α-synuclein-treated NPCs (PD model) exhibited neurotoxic protein accumulation (Aβ\u003csub\u003e42\u003c/sub\u003e, α-Syn), synaptic loss, complement activation, and pTau accumulation, consistent with postmortem brain analyses of AD and PD patients\u003csup\u003e\u003cspan additionalcitationids=\"CR33 CR34\" citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. Brain inflammation triggered an endothelial response, including increased iNOS expression, NO production, ZO-1 delocalization, and elevated endothelial permeability, mirroring BBB dysfunction reported in AD and PD patients (ref). Surprisingly, despite inflammatory cytokine release into systemic circulation, gut epithelial tight junctions (OCLN) remained intact, with only mild structural alterations by showing slightly reduced ViL1 and increased CD13 expression. This suggests that brain-to-gut inflammatory signaling is primarily mediated through vascular dysfunction rather than direct gut barrier breakdown. These findings challenge previous hypotheses suggesting severe gut barrier dysfunction in neurodegenerative disease models, instead proposing that gut remodeling in response to brain inflammation is subtler\u003csup\u003e\u003cspan additionalcitationids=\"CR37\" citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. This distinction underscores the need for physiologically relevant multicompartment models like hGBV, which allow for dynamic gut-brain-vascular interactions that are difficult to capture in traditional 2D cultures or in vivo models.\u003c/p\u003e \u003cp\u003eBy utilizing our hGBV platform, our findings have several translational implications. First, it provides a human-specific model to study gut-brain-vascular interactions in neurodegeneration, reducing reliance on animal models that often fail to replicate human pathophysiology. Second, our findings suggest that gut-to-brain signaling plays a more direct role in neurodegeneration, while brain-to-gut signaling primarily affects vascular integrity. This highlights the potential for gut-targeted therapeutics to prevent or slow neurodegenerative progression. In this case, several avenues for future research emerge from these findings. First, integrating patient-derived induced pluripotent stem cells (iPSCs) into the hGBV model could enable personalized disease modeling and drug screening. Second, investigating specific microbial metabolites (e.g., short-chain fatty acids) and their effects on neurovascular integrity could help elucidate mechanistic pathways linking gut dysbiosis to neurodegeneration. Finally, leveraging this platform for high-throughput drug screening could accelerate the identification of gut-targeted therapeutics for AD and PD.\u003c/p\u003e \u003cp\u003eWhile our hGBV platform successfully recapitulates gut-brain-vascular interactions, however several limitations remained. First, the physical distance between the gut and brain compartments in our chip is shorter than \u003cem\u003ein vivo\u003c/em\u003e, potentially accelerating inflammatory signaling and molecular transport. Second, the model lacks immune cells (e.g., monocytes, T-cells, microglia), which plays a crucial role in neuroinflammation and blood-brain barrier (BBB) regulation. Although the presence of endothelial cells and astrocytes provides a neurovascular interface, the addition of circulating peripheral immune cells could further enhance the physiological relevance of the platform. Third, gut microbiota interactions are simplified in this system, as only microbial-derived toxins (e.g., HP\u003csub\u003eCM\u003c/sub\u003e, LPS) were introduced. A system integrating live commensal and pathogenic microbes in a gut-microbiome-on-a-chip setup could provide deeper insights into host-microbe interactions in neurodegeneration. Lastly, while this study models AD and PD phenotypes, the system could be expanded to investigate other neurodegenerative conditions, such as ALS or multiple sclerosis, where gut-brain interactions are increasingly implicated. Incorporating genetic risk factors related AD and PD would enable disease modeling and therapeutic screening. Despite these limitations, our hGBV system represents a major advancement over traditional in vitro models, integrating gut, vascular, and brain compartments to create a more physiologically relevant platform. Addressing these limitations in future versions will further enhance its translational potential in neurodegenerative research and therapeutic development.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis study establishes the hGBV system as a physiologically relevant human model for investigating gut-brain-vascular interactions in neurodegeneration. Our findings highlight the differing effects of gut-origin and brain-origin inflammation, showing that gut-derived inflammation drives neurovascular dysfunction and neurodegeneration, while brain-derived neuroinflammation primarily leads to vascular dysfunction with minimal gut remodeling. These results provide critical insights into microbiota-driven neuroinflammation and brain-to-gut communication, supporting gut-targeted therapeutic strategies for AD and PD. By capturing the dynamic interplay between the gut, vasculature, and brain, the hGBV platform serves as a powerful translational tool for elucidating disease mechanisms, identifying biomarkers, and accelerating the development of targeted therapeutics for neurodegenerative diseases.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eTo systematically investigate the role of gut-brain-vascular interactions in neuroinflammation, we developed a 3D microfluidic platform that recapitulates the physiological architecture of the gut, vasculature, and brain. Below, we detail the fabrication of the microfluidic system, cellular components, and experimental conditions used to simulate gut-origin and brain-derived inflammatory processes.\u003c/p\u003e\n\u003ch3\u003e1.1. Microfluidic Chip Fabrication\u003c/h3\u003e\n\u003cp\u003eThe microfluidic chip was fabricated using soft lithography techniques, where a silicon substrate served as the base for device molding. The photolithography process sequentially patterned an SU-8 negative photoresist (MicroChem, Round Rock, TX) onto the silicon wafer, forming the microfluidic channel mold. To replicate the microfluidic structures, a polydimethylsiloxane (PDMS) mixture was prepared by combining PDMS elastomer with a curing agent (Sylgard 184 A/B, Dow Corning, Midland, MI) at a 9:1 ratio. The mixture was poured onto the SU-8 mold and degassed under vacuum to remove air bubbles. The PDMS was then thermally cured at 65\u0026deg;C for 2 hours to ensure complete polymerization. Once cured, the PDMS replica was carefully removed from the mold, and fluid reservoirs were created by punching inlet and outlet holes using a 2 mm biopsy punch. Additional plastic media reservoirs were fabricated using a computer-controlled laser cutter (Epilog Laser, Golden, CO) on a 5 mm thick acrylic plate to enhance experimental throughput and facilitate media handling. These acrylic reservoirs were bonded to the PDMS chip using an additional layer of liquid PDMS to ensure a stable, leak-free connection. For final assembly, the PDMS chip was bonded to a custom glass-bottomed uni-plate (MatTek, Ashland, MA) via oxygen plasma treatment (Plasma Etch, Carson City, NV), ensuring strong adhesion and optical transparency for high-resolution imaging (Supplementary Fig.\u0026nbsp;1). After assembly, the microfluidic chip was sterilized under UV light for 30 minutes before cell seeding and experimental use.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e1.2. Functionalization of Microfluidic Devices Before Cell Seeding\u003c/h2\u003e \u003cp\u003eTo promote cell adhesion and extracellular matrix (ECM) compatibility, the microfluidic compartments were functionalized with Poly-L-lysine (PDL) and collagen type I (Col I) following a well-established protocol [ref]. This functionalization step enhances cell attachment, barrier formation, and tissue-specific microenvironment stability within the microfluidic system. The entire microfluidic network, including all compartments and channels, was infused with 1 mg/mL Poly-L-lysine (Sigma Aldrich, St. Louis, MO) and incubated for 30 minutes at room temperature (RT) to facilitate surface charge modification and ECM adhesion. After incubation, the device was rinsed thoroughly with cold 1\u0026times; PBS (phosphate-buffered saline) three times to remove excess PDL. Following PDL treatment, the gut compartment (G.C.) and vascular compartment (V.C.) were infused with 1 mg/mL collagen type I (Corning, Corning, NY), adjusted to pH 7.5. The functionalized device was incubated at 37\u0026deg;C for 1 hour to allow proper ECM polymerization and surface coating. After functionalization, all compartments were washed with fresh 1\u0026times; PBS to remove unbound proteins. The microfluidic devices were then immediately seeded with cells to ensure optimal cell adhesion and tissue integration.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e1.3. 3D Culture of Brain Tissue in the hGBV Microfluidic Device\u003c/h2\u003e \u003cp\u003eTo establish a functional 3D neural microenvironment within the brain compartment (B.C.) of the hGBV model, human neural progenitor cells (ReN cell VM NPCs) were differentiated into neurons and astrocytes over a three-week period. This approach promoted cellular maturation, synaptic formation, and astrocyte-neuron interactions, ensuring a physiologically relevant brain tissue model. Human ReN cell VM neural progenitor cells (NPCs) were obtained from MilliporeSigma (Billerica, MA, USA) and maintained at passages\u0026thinsp;\u0026lt;\u0026thinsp;20 to preserve their differentiation capacity. Cells were expanded in Dulbecco\u0026rsquo;s Modified Eagle Medium/Nutrient Mixture F-12 (DMEM/F-12, Life Technologies, Grand Island, NY, USA) supplemented with 2 mg/mL Heparin (StemCell Technologies, Vancouver, Canada), 2% (v/v) B27 neural supplement (Life Technologies, Grand Island, NY, USA), 20 ng/mL Epidermal Growth Factor (EGF) (Sigma-Aldrich, St. Louis, MO, USA), 20 ng/mL Basic Fibroblast Growth Factor (bFGF) (Stemgent, Cambridge, MA, USA), and 1% (v/v) Penicillin/Streptomycin (Sigma-Aldrich, St. Louis, MO, USA). NPCs were plated in T25 flasks coated with 1% Matrigel (BD Biosciences, San Jose, CA, USA) and cultured at 37\u0026deg;C with 5% CO₂ in a humidified incubator. The culture medium was refreshed every three days until confluence was reached.\u003c/p\u003e \u003cp\u003eTo generate a 3D mini-brain tissue model, ReN NPCs were suspended at 2 \u0026times; 10\u003csup\u003e4\u003c/sup\u003e cells/device in a 10% (v/v) Matrigel-DMEM/F12 differentiation medium mixture (Supplementary Fig.\u0026nbsp;3). 10 \u0026micro;L of the NPC-Matrigel suspension was injected into brain compartment. The device was incubated at 37\u0026deg;C for 1 hour to facilitate gel polymerization and cell attachment. After polymerization, 100 \u0026micro;L of differentiation medium was added into both brain and vascular compartment. The devices were placed in a 37\u0026deg;C, 5% CO₂ incubator for long-term culture. Over 21 days, the NPCs progressively differentiated into neurons and astrocytes. Culture medium was replaced every two days to maintain optimal differentiation conditions.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e1.4. 3D Gut Epithelium Barrier Generation\u003c/h2\u003e \u003cp\u003eTo establish a functional gut epithelium within the gut compartment (G.C.) of the hGBV model, the Caco-2 epithelial cell line was cultured and differentiated under dynamic perfusion conditions to mimic the intestinal barrier. Human Caco-2 cells were obtained from the Cell Lines Service (CLS, Eppelheim, Germany) and maintained between passages 40\u0026ndash;70 to ensure stable barrier formation properties. Cells were cultured in 75 cm\u0026sup2; tissue culture flasks using high-glucsoe Dulbecco\u0026rsquo;s Modified Eagle\u0026rsquo;s Medium (high glucose DMEM, Gibco, Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 10% (v/v) Fetal Bovine Serum (Sigma-Aldrich, St. Louis, MO, USA), 1% (v/v) Penicillin/Streptomycin (Sigma-Aldrich, St. Louis, MO, USA). Cells were maintained at 37\u0026deg;C in a 5% CO₂/95% air atmosphere with constant humidity, and the culture medium was refreshed every two days until the cells reached 80\u0026ndash;90% confluence.\u003c/p\u003e \u003cp\u003eAt day 7 of neuronal and astrocyte differentiation, Caco-2 cells were seeded into the gut compartment (G.C.) of the microfluidic chip to allow for co-culture with the neurovascular system. The cells were introduced at a density of 10\u003csup\u003e6\u003c/sup\u003e cells/mm\u003csup\u003e2\u003c/sup\u003e and allowed to attach for 4 hours before adding fresh media. The loaded cells adhered to the upper surface of the compartments when the platforms were inverted for 4 hours in a 37\u0026deg;C incubator with 5% CO\u003csub\u003e2\u003c/sub\u003e. After the cells adhered to the upper surface, newly cultured Caco-2 cells were gently reintroduced to the bottom surface in their normal orientation for an additional two hours to ensure proper adherence. Once both surfaces were covered with adherent cells, we added an extra 100 \u0026micro;L of fresh medium to the reservoirs. To enhance intestinal barrier function, dynamic perfusion conditions were introduced using a rocker integrated into the incubator. The culture medium was replaced daily to maintain optimal differentiation conditions.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e1.5. 3D Vascular Endothelium Barrier Generation\u003c/h2\u003e \u003cp\u003eTo establish a functional blood-brain barrier (BBB) within the vascular compartment (V.C.) of the hGBV model, human embryonic stem cell-derived endothelial cells (hECs) were differentiated into induced brain microvascular endothelial cells (iBMECs) and cultured under dynamic perfusion conditions to mimic physiological endothelial barrier function.\u003c/p\u003e \u003cp\u003eHuman embryonic stem cells (hEC, H9) were obtained from Wi-Cell (USA) and differentiated into iBMECs following a previously established protocol (ref). Cells were grown in mTeSR Plus medium (StemCell Technologies, Vancouver, Canada) supplemented with 10 \u0026micro;M Y-27632 dihydrochloride (Tocris #1254) to enhance survival and proliferation. Upon reaching\u0026thinsp;~\u0026thinsp;70% confluency, differentiation into iBMECs was initiated. ells were switched to an unconditioned medium (Thermo Fisher Scientific #11330057) containing 20% KnockOut Serum Replacement (Thermo Fisher Scientific #10828010), 1% Non-Essential Amino Acids (NEAA, Thermo Fisher Scientific #11140050), 0.5% Glutamax (Thermo Fisher Scientific #35050079), 0.1 mM β-Mercaptoethanol (Sigma-Aldrich #M6250, St. Louis, MO, USA). This differentiation medium was maintained for 6 days to promote iBMEC-like characteristics. Cells were transitioned to EC\u0026thinsp;+\u0026thinsp;+\u0026thinsp;medium, consisting of human Endothelial Serum-Free Medium (SFM, Thermo Fisher Scientific #11111044), 1% Platelet-Poor Plasma-Derived Serum, Bovine (Alfa Aesar #15406419, Haverhill, MA, USA), 20 ng/mL Basic Fibroblast Growth Factor (bFGF), 10 \u0026micro;M Retinoic Acid (Sigma-Aldrich #R2625). After 2 days, the medium was replaced with maturation medium (SFM without bFGF and retinoic acid) to stabilize endothelial barrier function. iBMECs were harvested using Accutase (Corning #25-058-CI, Corning, NY, USA) and prepared for seeding onto the vascular compartment (V.C.) of the hGBV microfluidic device.\u003c/p\u003e \u003cp\u003eAt day 18 of neuronal and astrocyte differentiation with Caco-2, a suspension of iBMECs (5 x 10\u003csup\u003e6\u003c/sup\u003e cell/mm\u0026sup2;) was loaded into the vascular compartment (V.C.). The microfluidic chip was inverted for 4 hours at 37\u0026deg;C to facilitate cell attachment to the upper surface. After 4 hours, the chip was returned to its normal orientation, and a second suspension of iBMECs was gently reintroduced to the bottom surface for 2 additional hours to ensure bilayer coverage. Once both surfaces were fully covered with adherent iBMECs, 100 \u0026micro;L of fresh differentiation medium was added to one reservoir, allowing flow through the vascular compartment. To maintain a constant supply of fresh nutrients, the medium was replenished daily by adding 100 \u0026micro;L of fresh medium to one reservoir while removing the same volume from the opposite reservoir. To enhance endothelial barrier integrity, dynamic perfusion was introduced using a rocker integrated within the incubator.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e1.6. Barrier Integrity Assessment\u003c/h2\u003e \u003cp\u003eTo evaluate the permeability of the 3D barrier model, including the gut (G.C.) and vascular (V.C.) compartments, we performed a FITC-dextran permeability assay. This method assessed barrier integrity by quantifying the diffusion of fluorescently labeled dextran molecules across the epithelial and endothelial layers. FITC-conjugated dextran molecules of different molecular weights were used to evaluate paracellular permeability, including low molecular weight FITC-dextran (3\u0026ndash;5 kDa) to assess small molecule permeability and high molecular weight FITC-dextran (70 kDa) to assess macromolecule leakage. Both dextran solutions were obtained from Sigma-Aldrich (St. Louis, MO, USA) and prepared in sterile 1\u0026times; PBS at a final concentration of 10 \u0026micro;g/mL. FITC-dextran was introduced into the luminal side of either the gut compartment (G.C.) or vascular compartment (V.C.) of the microfluidic device. Time-lapse imaging was performed to monitor fluorescence intensity changes during the perfusion of FITC-dextran across the barrier layer. Imaging was conducted using a Nikon Eclipse Ti2-E microscope (Nikon Corporation, Tokyo, Japan) equipped with filter cube for FITC detection, 10\u0026times; objective lens, and frames were captured at 1 second per frame to track dextran diffusion dynamics. Raw fluorescence intensity values were extracted using Nikon NIS-Elements software. The fluorescence intensity ratio between the perfused FITC-dextran and the initial luminal concentration was used to calculate the perfusion index. The apparent permeability coefficient (Papp, cm/s) was determined using the following equation:\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$$\\:{P}_{app}=\\:\\frac{\\left(\\frac{dCi}{dt}\\right)\\times\\:V}{{C}_{o}\\:\\times\\:A}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eWhere: P\u003cem\u003eapp\u003c/em\u003e is permeability coefficient (cm/s), dC\u003cem\u003ei\u003c/em\u003e/dt is rate of change in FITC-dextran concentration in the receiving compartment over time (t), C\u003csub\u003eo\u003c/sub\u003e is initial FITC-dextran concentration in the luminal compartment (10 \u0026micro;g/mL), V is volume of the receiving compartment, A is surface area of microfluidic channel (0.013 cm\u0026sup2;). Lower P\u003cem\u003eapp\u003c/em\u003e values indicate an intact and selectively permeable barrier. Higher P\u003cem\u003eapp\u003c/em\u003e values suggest barrier leakage.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e1.7. Brain Calcium Signaling\u003c/h2\u003e \u003cp\u003eTo assess neuronal activity within the brain compartment (B.C.), we monitored intracellular calcium dynamics using Oregon-Green-BAPTA-1 (OGB-1), a Ca\u0026sup2;⁺ indicator (Life Technologies, Thermo Fisher Scientific). Calcium transients were analyzed using real-time fluorescence imaging, providing insights into neuronal and astrocytes interaction. Brain compartment were incubated in artificial cerebrospinal fluid (aCSF) at 37\u0026deg;C for 30 minutes to mimic physiological conditions. The aCSF composition included 25 mM NaHCO₃, 25 mM D-glucose, 125 mM NaCl, 2.5 mM KCl, 1.25 mM NaH₂PO₄, 1 mM MgCl₂, 2 mM CaCl₂. During incubation, Pluronic F-127 (20% w/v in DMSO, Sigma-Aldrich) was added to facilitate OGB-1 dye loading cytosol. After 30 minutes of dye loading, samples were rinsed twice with fresh aCSF to remove unbound dye. Time-lapse fluorescence imaging was performed using a Nikon Eclipse Ti2-E microscope (Nikon Corporation, Tokyo, Japan) equipped with Texas red filter cube (for Rhod-2 detection), 20\u0026times; objective lens, and frames were captured at 1 frame per second to monitor dynamic changes in intracellular Ca\u0026sup2;⁺ levels. Raw fluorescence intensity values were extracted using Nikon NIS-Elements software.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e1.8. Helicobacter Bacteria Conditioned Media Collection\u003c/h2\u003e \u003cp\u003e \u003cem\u003eH. pylori\u003c/em\u003e strain [NCTC 11637] was obtained from ATCC and cultured under microaerophilic conditions to maintain optimal bacterial growth. Bacteria were grown in Brain Heart Infusion (BHI) broth (BD Biosciences, San Jose, CA, USA) supplemented with 5% Fetal Bovine Serum (FBS, Gibco, Thermo Fisher Scientific), 10 \u0026micro;g/mL Vancomycin, 5 \u0026micro;g/mL Trimethoprim. Cultures were incubated at 37\u0026deg;C in a 5% CO₂, 10% O₂ atmosphere for 48\u0026ndash;72 hours to reach the logarithmic growth phase. H. pylori-conditioned media (HPCM) was collected by centrifugation at 4,000 \u0026times; g for 10 minutes at 4\u0026deg;C once cultures reached the log-phase. The collected supernatant was filter-sterilized using a 0.2 \u0026micro;m syringe filter to remove residual bacteria while retaining secreted bacterial toxins, virulence factors, and metabolic byproducts. Aliquots were stored at -80\u0026deg;C until use to prevent degradation of bioactive components.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e1.9. Alzheimer\u0026rsquo;s Disease (AD) Model Construction in hGBV Platform\u003c/h2\u003e \u003cp\u003eTo generate Alzheimer\u0026rsquo;s disease (AD) human neural progenitor cells (hNPCs), ReN hNPCs were transduced with a lentiviral vector encoding the APPSL mutation, resulting in the overexpression of a familial Alzheimer\u0026rsquo;s disease (FAD) variant of the amyloid precursor protein (APP). This model enabled the production of high levels of Aβ peptides, mimicking AD-like pathophysiology in the brain compartment (B.C.) of the hGBV system.\u003c/p\u003e \u003cp\u003eReN hNPCs were cultured in six-well plates and transduced using a commercially available APPSL-GFP Alzheimer\u0026rsquo;s lentivirus (EMD Millipore). The APPSL lentiviral vector encodes a mutated form of human APP containing two familial Alzheimer\u0026rsquo;s disease (FAD) mutations K670N/M671L (Swedish mutation) for increases β-secretase cleavage, leading to higher Aβ42 production and V717I (London mutation) for altering γ-secretase cleavage, promoting Aβ42 over Aβ40 production. ReN hNPCs were incubated with 5 \u0026micro;L of APPSL-GFP lentivirus (1 \u0026times; 10⁹ IFU/mL) in the presence of 2 mg/mL polybrene (EMD Millipore) to enhance transduction efficiency. Cells were incubated for 24 hours at 37\u0026deg;C with 5% CO₂. GFP fluorescence was monitored to confirm initial viral expression. After three washes with PBS, a second round of viral infection was performed by adding fresh APPSL-GFP lentivirus to the culture medium. Cells were incubated for another 24 hours. After 48 hours, cells were washed three times with PBS and cultured in fresh medium to allow transgene expression stabilization. Fluorescence microscopy was used to confirm GFP-positive transgene expression. Transgene-positive cells were enriched using fluorescence-activated cell sorting (FACS, BD FACS Aria II, BD Biosciences) to ensure a homogeneous AD hNPC population.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003e1.10. Parkinson\u0026rsquo;s Disease (PD) Model Construction in the hGBV Platform\u003c/h2\u003e \u003cp\u003eTo establish a Parkinson\u0026rsquo;s disease (PD) model within the hGBV platform, preformed fibril (PFF) α-synuclein aggregates were introduced into the brain compartment (B.C.), where they induced pathological α-synuclein aggregation, neuroinflammation, and neuronal dysfunction. This model enabled the study of brain-to-gut crosstalk in PD pathology. ReN-derived NPCs were differentiated into neurons and astrocytes for 3 weeks in the brain compartment, following the differentiation protocol described in Section 1.3. PFF α-Synuclein (PFF α-Syn, 1 \u0026micro;M, StressMarq Bioscience) added to the brain compartment (B.C.) for 6 hours to allow for fibril uptake by neurons and astrocytes. Cells were then washed and replace by fresh medium to remove excess extracellular fibrils. The cultures were maintained for 7 additional days to allow for α-synuclein aggregation and pathology development.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003e1.11. Chromogenic Endotoxin Measurement Assay\u003c/h2\u003e \u003cp\u003eTo quantify endotoxin levels in conditioned media and microfluidic compartments, we performed a chromogenic endotoxin measurement assay using the Pierce\u0026trade; Chromogenic Endotoxin Quant Kit (Thermo Fisher Scientific, Catalog #A39552S). This assay was conducted following the manufacturer\u0026rsquo;s user guide and is based on the Limulus Amebocyte Lysate (LAL) reaction. Samples were collected from the gut compartment (G.C.), vascular compartment (V.C.), and brain compartment (B.C.). Samples were diluted 1:10 in endotoxin-free water to ensure that measurements fell within the linear range of the standard curve. The optical density (OD) at 405 nm was measured using a microplate reader (Biotek, USA). Sample endotoxin levels were calculated using the standard curve and expressed in endotoxin units per milliliter (EU/mL). Endotoxin levels were compared between control (CT), Helicobacter-conditioned media (HP\u003csub\u003eCM\u003c/sub\u003e), and LPS-treated conditions to assess gut-to-brain endotoxin translocation. Each condition was prepared in triplicate to ensure data reliability and reproducibility.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003e1.12. Amyloid beta 42 ELISA Assay\u003c/h2\u003e \u003cp\u003eQuantify amyloid beta 42 (Aβ\u003csub\u003e42\u003c/sub\u003e) levels was performed an enzyme-linked immunosorbent assay (ELISA) using the Human Aβ\u003csub\u003e42\u003c/sub\u003e ELISA Kit (Abcam, Catalog #ab289832). The assay was conducted following the manufacturer\u0026rsquo;s user guide, which employs a sandwich ELISA method for the sensitive and specific detection of Aβ\u003csub\u003e42\u003c/sub\u003e in conditioned media. Conditioned media were collected from the brain compartment (B.C.), vascular compartment (V.C.), and gut compartment (G.C.) after 6 weeks of neuronal differentiation in wide type (WT) and AD-like conditions (APP-overexpressing NPCs, Section 1.9). Sample concentrations were interpolated from the standard curve and expressed as pg/mL Aβ\u003csub\u003e42\u003c/sub\u003e. Total Aβ\u003csub\u003e42\u003c/sub\u003e levels were compared between control (CT) and AD conditions to assess the release and potential translocation of Aβ\u003csub\u003e42\u003c/sub\u003e across the brain-gut axis. All samples were analyzed in triplicate to ensure data reproducibility.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003ch2\u003e1.13. Alpha-Synuclein ELISA Assay\u003c/h2\u003e \u003cp\u003eTo quantify alpha-synuclein (α-Syn) levels in the brain compartment (B.C.) and assess its potential translocation into the vascular compartment (V.C.) and gut compartment (G.C.), we performed an enzyme-linked immunosorbent assay (ELISA) using the Human Alpha-Synuclein ELISA Kit (Abcam, Catalog #ab260052). The assay was conducted according to the manufacturer\u0026rsquo;s user guide and utilized a sandwich ELISA format for the specific and sensitive detection of α-Syn in conditioned media. Conditioned media were collected from the brain compartment (B.C.), vascular compartment (V.C.), and gut compartment (G.C.) following 7 days of PFF α-Synuclein treatment in the PD model (Section 1.10). Measured α-Syn levels were compared between wild type (WT) and PD-like conditions to assess neurotoxic α-Syn secretion and potential translocation across compartments. All samples were analyzed in triplicate to ensure data reproducibility.\u003c/p\u003e \u003cdiv id=\"Sec23\" class=\"Section3\"\u003e \u003ch2\u003e1.14. Multicytokine Profiling Assay\u003c/h2\u003e \u003cp\u003eTo investigate cellular responses to bacterial-derived toxins and neuroinflammatory stimuli, we analyzed the profiles of secreted cytokines and chemokines from conditioned media collected from the 3D gut epithelium (gut compartment), vascular endothelium (vascular compartment), and brain environment (brain compartment) using the Human Cytokine Array Kit (Catalog #ARY005B, R\u0026amp;D Systems, Minneapolis, MN, USA). Conditioned media were obtained from gut epithelial cells (Caco-2, gut compartment), endothelial cells (iBMECs, vascular compartment), and neurons and astrocytes (brain compartment) after exposure to bacterial toxins (Helicobacter-conditioned media, LPS) and neuroinflammatory stimuli (Aβ\u003csub\u003e42\u003c/sub\u003e, PFF α-Synuclein). Cytokine spot signals were detected using the iBright Imaging System (ThermoFisher, USA), and pixel density values were extracted using ImageJ (NIH, Wayne Rasband). Background subtraction was applied, and relative pixel intensity values were normalized against the positive control spots on the membrane. Cytokine expression levels were compared across control (CT), Helicobacter-conditioned media (HP\u003csub\u003eCM\u003c/sub\u003e)-treated, LPS-treated, and neuroinflammatory conditions (Aβ\u003csub\u003e42\u003c/sub\u003e, PFF α-Synuclein). A heatmap of log₂ fold-change cytokine expression was generated to visualize differential secretion patterns across experimental conditions.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec24\" class=\"Section2\"\u003e \u003ch2\u003e1.15. Immunostaining Assay\u003c/h2\u003e \u003cp\u003eTo visualize cellular markers and structural integrity within the 3D hGBV microfluidic platform, an immunostaining assay was performed following a standard fixation, permeabilization, blocking, and antibody incubation protocol.\u003c/p\u003e \u003cp\u003eFirst, cells cultured within the microfluidic device were rinsed twice with 1\u0026times; PBS and fixed at room temperature for 30 minutes using 4% paraformaldehyde (Electron Microscopy Sciences, Hatfield, PA, USA). After fixation, cells were washed three times with 1\u0026times; PBS at 10-minute intervals. To enhance antibody penetration, cells were permeabilized for 30 minutes at room temperature with 1\u0026times; PBS supplemented with 0.1% (v/v) Triton X-100 and 0.1% (v/v) Tween-20. Subsequently, cells underwent three additional 10-minute washes with 1\u0026times; PBS before blocking to minimize non-specific antibody binding. For blocking, cells were incubated for 2 hours at 4\u0026deg;C in a solution of 1\u0026times; PBS supplemented with 0.1% (v/v) Tween-20 and 4% (v/v) human serum albumin (HSA). Following blocking, cells were rinsed three times with 1\u0026times; PBS containing 0.1% (v/v) Tween-20 at 10-minute intervals. Cells were then incubated with the primary antibody diluted in 1% (v/v) blocking solution for 6 hours at 4\u0026deg;C. The dilution ratios of all primary and secondary antibodies are provided in Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e (Supplementary Information). After the primary antibody incubation, cells were washed three times with 1\u0026times; PBS containing 0.1% (v/v) Tween-20 at 10-minute intervals before the addition of the secondary antibody. Secondary antibodies were diluted in 1\u0026times; PBS supplemented with 0.1% (v/v) Tween-20 and 4% (v/v) HSA and incubated for 2 hours at 4\u0026deg;C. Following secondary antibody binding, cells were washed three times with 1\u0026times; PBS supplemented with 0.1% (v/v) Tween-20 at 10-minute intervals to remove excess antibodies.\u003c/p\u003e \u003cp\u003eImmunostained samples were visualized using a Nikon Eclipse Ti2-E fluorescence microscope (Nikon Corporation, Tokyo, Japan) and a K1-Fluo Confocal Laser Microscope (Nanoscope, Korea). Regions of interest (ROIs) were manually selected for each compartment (gut, vascular, and brain) to quantify the mean fluorescence intensity (MFI). Image acquisition and MFI quantification were performed using NIS-Elements software (Nikon, Japan) and Fiji/ImageJ (NIH, Wayne Rasband, USA). To account for variations in cell density, MFI values were normalized to the counts of DAPI-stained nuclei. Cell morphology and tight junction integrity were assessed by measuring ZO-1 and OCLN fluorescence intensity at cell junctions, which were processed in Fiji/ImageJ using segmentation-based analysis to determine tight junction localization and disruption.\u003c/p\u003e \u003cdiv id=\"Sec25\" class=\"Section3\"\u003e \u003ch2\u003e1.16. Quantification and Statistical Analysis\u003c/h2\u003e \u003cp\u003eAll statistical analyses were conducted using SPSS (IBM SPSS Statistics, Armonk, NY, USA) and GraphPad Prism (GraphPad Software, San Diego, CA, USA). We employed independent Student\u0026rsquo;s t-tests (for comparisons between two groups/conditions) or one-way ANOVA followed by post hoc tests (for comparisons of three or more groups/conditions), or two-way ANOVA followed by post hoc tests (for comparisons involving three or more groups/two or more conditions) to assess the statistical significance among experimental groups. Data are presented as means\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviation (SD), with statistical significance established at a 95% confidence interval (CI) (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05, p\u0026thinsp;\u0026lt;\u0026thinsp;0.01, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001 for two-group or three-group analyses). Details of the statistical analyses are provided in the figure legends. No statistical methods were utilized to pre-determine sample sizes; however, experiments were repeated to improve reproducibility and minimize confidence intervals. The data collection and analysis were not blinded to the experimental conditions, and randomization and exclusions were not implemented.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Abbreviations","content":"\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"624\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eAD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eAlzheimer\u0026apos;s disease\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eCNS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eCentral Nervous System\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eGS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eGlutamine Synthetase\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eCSF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eCerebrospinal fluid\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eNVU\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eneurovascular unit\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eROS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003ereactive oxygen species\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e3D\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eThree-dimensional\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eALDH1L1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003ealdehyde dehydrogenase 1 family member L1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eTuj1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eneuron-specific class III beta-tubulin\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eNeuN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eneuronal nuclear protein\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eBBB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eblood-brain barrier\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eEC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eEndothelial cell\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eAC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eAstrocytes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eNeu\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eNeuron\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eNH\u003csub\u003e4\u003c/sub\u003eCH\u003csub\u003e3\u003c/sub\u003eCO\u003csub\u003e2\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eammonia acetate\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eNH\u003csub\u003e4\u003c/sub\u003eCl\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eammonium chloride\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eH\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003ehydrogen peroxide\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eGLN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eglutamine\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003epTau\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003ephosphor-tau\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eZO-1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eZonula occludens-1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003ePECAM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003ePlatelet endothelial cell adhesion molecule-1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eIFN-ɣ\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eInterferon gamma\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eIL-2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eInterleukin 2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eMIF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eMacrophage migration inhibitory factor\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eTNF-\u0026alpha;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eTumor necrosis factor alpha\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eIL-1ra\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eInterleukin 1 receptor antagonist\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eIL-4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eInterleukin 4\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eIL-5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eInterleukin 5\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eIL-13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eInterleukin 13\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eIL-1\u0026beta;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eInterleukin-1 beta\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eIL-10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eInterleukin 10\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eIL-6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eInterleukin 6\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eIL-8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eInterleukin 8\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eNPC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eNeural progenitor cells\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eGFAP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eGlial fibrillary acidic protein\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eCCL2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003echemokine (C-C motif) ligand 2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eCT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eControl\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eWT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eWide-type\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eiNOS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eInducible nitric oxide synthase\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eLPS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eLipopolysaccharide\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eCa\u003csup\u003e2+\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003ecalcium\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eOCLN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 480px;\"\u003e\n \u003cp\u003eOccludin\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData Availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll data generated or analyzed during this study are included in this published article (and its supplementary information files) or are available from the corresponding author on reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe acknowledge MicroTas2024 for supporting this project through the CBMS Student/Young Researcher Grant and the Biomicrofluidics Best Paper Award (M.T.).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contribution\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eM.T. developed the microfluidic device, designed and performed the experiments, analyzed the data, and wrote the manuscript. H.W.J. designed and developed the microfluidic device. C.B. conducted experiments using the fabricated microfluidic devices, while V.T.A.T. was responsible for collecting Helicobacter-conditioned media. A.J. performed experiments related to Alzheimer\u0026rsquo;s disease (AD) modeling, and M.J.A. conducted experiments on Parkinson\u0026rsquo;s disease (PD) modeling. L.P.L. provided insightful feedback and contributed to the overall supervision of the project. H.C. conceived, initiated, provided insightful feedback, and supervised the entire study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics Declarations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll authors were fully involved in the study, and the material within has not been and will not be submitted for publication elsewhere. All authors declare no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was funded by the National Research Foundation of Korea (NRF) grant, funded by the Korea government (MSIT) (RS-2024-00336758), the Korea Dementia Research Project through the Korea Dementia Research Center (KDRC), funded by the Ministry of Health \u0026amp; Welfare and Ministry of Science and ICT, Republic of Korea (RS-2022-KH126808), the Technology development Program, funded by the Ministry of SMEs and Startups (MSS, Korea) (RS-2023-00224842), and the Fourth Stage of Brain Korea 21 Project in Department of Intelligent Precision Healthcare, Sungkyunkwan University (SKKU) to H.C.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eLoh JS et al (2024) Microbiota\u0026ndash;gut\u0026ndash;brain axis and its therapeutic applications in neurodegenerative diseases. Signal Transduct Target Ther 9\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eParker A, Fonseca S, Carding SR (2020) Gut microbes and metabolites as modulators of blood-brain barrier integrity and brain health. Gut Microbes 11:135\u0026ndash;157\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYadav H et al (2023) Unveiling the role of gut-brain axis in regulating neurodegenerative diseases: A comprehensive review. Life Sci 330\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKnox EG, Aburto MR, Clarke G, Cryan JF, O\u0026rsquo;Driscoll CM (2022) The blood-brain barrier in aging and neurodegeneration. Mol Psychiatry 27:2659\u0026ndash;2673\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAgirman G, Yu KB, Hsiao EY (2021) Signaling inflammation across the gut-brain axis. Sci (80-) 374:1087\u0026ndash;1092\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang D, Almanzar N, Chiu IM (2023) The role of cellular and molecular neuroimmune crosstalk in gut immunity. Cell Mol Immunol 20:1259\u0026ndash;1269\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eH., A. et al. Microbiota-derived metabolites as drivers of gut-brain communication. Gut Microbes 14, 2102878 (2022)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVogt NM et al (2017) Gut microbiome alterations in Alzheimer\u0026rsquo;s disease. Sci Rep 7\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJung JH et al (2022) Gut microbiome alterations in preclinical Alzheimer\u0026rsquo;s disease. PLoS ONE 17\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang F et al (2020) Altered gut microbiota in Parkinson\u0026rsquo;s disease patients/healthy spouses and its association with clinical features. Park Relat Disord 81:84\u0026ndash;88\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRomano S et al (2021) Meta-analysis of the Parkinson\u0026rsquo;s disease gut microbiome suggests alterations linked to intestinal inflammation. npj Park Dis 7\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMou Y et al (2022) Gut Microbiota Interact With the Brain Through Systemic Chronic Inflammation: Implications on Neuroinflammation, Neurodegeneration, and Aging. Front Immunol 13\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBairamian D et al (2022) Microbiota in neuroinflammation and synaptic dysfunction: a focus on Alzheimer\u0026rsquo;s disease. Mol Neurodegener 17\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWarren A, Nyavor Y, Zarabian N, Mahoney A, Frame LA (2024) The microbiota-gut-brain-immune interface in the pathogenesis of neuroinflammatory diseases: a narrative review of the emerging literature. Front Immunol 15\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eT.L.A., N., S., V.-S., A., L. \u0026amp; J., R. How informative is the mouse for human gut microbiota research? DMM Dis Model Mech 8, 1\u0026ndash;16 (2015)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOri\u0026aacute; RB et al (2018) Revisiting inbred mouse models to study the developing brain: The potential role of intestinal microbiota. Front Hum Neurosci. 12\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim MH, van Noort D, Sung JH, Park S (2021) Organ-on-a-chip for studying gut-brain interaction mediated by extracellular vesicles in the gut microenvironment. Int J Mol Sci 22\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAshammakhi N et al (2020) Gut-on-a-chip: Current progress and future opportunities. Biomaterials 255\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTaavitsainen S et al (2024) Gut-on-chip devices as intestinal inflammation models and their future for studying multifactorial diseases. Front Lab Chip Technol 2\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBerry D et al (2018) Organ-On-A-Chip in vitro Models of the Brain and the Blood-Brain Barrier and Their Value to Study the Microbiota-Gut-Brain Axis in Neurodegeneration. Curr Opin Biotechnol 13:247\u0026ndash;257\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eA guide to (2022) the organ-on-a-chip. Nat Rev Methods Prim 2\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePorter RJ, Murray GI, McLean MH (2020) Current concepts in tumour-derived organoids. Br J Cancer 123:1209\u0026ndash;1218\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUrz\u0026igrave; O et al (2023) Three-Dimensional Cell Cultures: The Bridge between In Vitro and In Vivo Models. Int J Mol Sci 24\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDenman CR, Park SM, Jo J (2023) Gut-brain axis: gut dysbiosis and psychiatric disorders in Alzheimer\u0026rsquo;s and Parkinson\u0026rsquo;s disease. Front Neurosci 17\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePark AM, Tsunoda I (2022) Helicobacter pylori infection in the stomach induces neuroinflammation: the potential roles of bacterial outer membrane vesicles in an animal model of Alzheimer\u0026rsquo;s disease. Inflamm Regen 42\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuo S, Al-Sadi R, Said HM, Ma TY (2013) Lipopolysaccharide causes an increase in intestinal tight junction permeability in vitro and in vivo by inducing enterocyte membrane expression and localization of TLR-4 and CD14. Am J Pathol 182:375\u0026ndash;387\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJaworowicz DJ, Korytko PJ, Lakhman SS, Boje KM (1998) K. Nitric oxide and prostaglandin E2 formation parallels blood-brain barrier disruption in an experimental rat model of bacterial meningitis. Brain Res Bull 46:541\u0026ndash;546\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrown GC, Camacho M, Williams-Gray CH (2023) The Endotoxin Hypothesis of Parkinson\u0026rsquo;s Disease. Mov Disord 38:1143\u0026ndash;1155\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrown GC, Heneka MT (2024) The endotoxin hypothesis of Alzheimer\u0026rsquo;s disease. Mol Neurodegener 19\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMulak A, Bonaz B (2015) Brain-gut-microbiota axis in Parkinson\u0026rsquo;s disease. World J Gastroenterol 21:10609\u0026ndash;10620\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang X, Tang B, Guo J (2023) Parkinson\u0026rsquo;s disease and gut microbiota: from clinical to mechanistic and therapeutic studies. Transl Neurodegener 12\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang W, Xiao D, Mao Q, Xia H (2023) Role of neuroinflammation in neurodegeneration development. Signal Transduct Target Ther 8\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJanelidze S et al (2018) CSF biomarkers of neuroinflammation and cerebrovascular dysfunction in early Alzheimer disease. Neurology 91:e867\u0026ndash;e877\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGilbert MAG et al (2024) CryoET of β-amyloid and tau within postmortem Alzheimer\u0026rsquo;s disease brain. Nature 631:913\u0026ndash;919\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIrwin DJ, Lee VMY, Trojanowski JQ (2013) Parkinson\u0026rsquo;s disease dementia: Convergence of α-synuclein, tau and amyloid-β pathologies. Nat Rev Neurosci 14:626\u0026ndash;636\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePei Y, Lu Y, Li HZ, Jiang CY, Wang L (2023) Gut microbiota and intestinal barrier function in subjects with cognitive impairments: a cross-sectional study. Front Aging Neurosci 15\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKowalski K, Mulak A (2019) Brain-gut-microbiota axis in Alzheimer\u0026rsquo;s disease. J Neurogastroenterol Motil 25:48\u0026ndash;60\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuan J et al (2013) Vascular degeneration in parkinsons disease. Brain Pathol 23:154\u0026ndash;164\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6128513/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6128513/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eA \u0026lsquo;gut-brain axis\u0026rsquo; is an intricate bidirectional connection between the gut and the central nervous system, serving as a key pathway for signal exchange. However, current \u003cem\u003ein vitro\u003c/em\u003e models do not fully capture the dynamics of interactions between these organs, which limits mechanistic understanding and therapeutic exploration. Here, we present a 3D human gut-brain vascular (GBV) model that simulates bidirectional communication between these entities, allowing us to investigate disorders originating from both the gut and brain. We created a physiologically relevant gut-brain axis model, creating a 3D villus-like lumenized gut barrier, blood vascular-astrocyte interactions, and brain tissue that mimics neurovascular interactions. Next, we demonstrated gut-to-brain signaling by introducing bacterial-derived toxins into the gut side, allowing the toxins to penetrate the gut barrier and the neurovascular barrier, ultimately reaching the brain and leading to tauopathy, a key indicator of neurodegeneration. We observed brain-to-gut signaling by exposing the brain side to risk factors of Alzheimer\u0026rsquo;s (AD) and Parkinson\u0026rsquo;s (PD), which induces neuroinflammation, disrupts the vascular barrier, and subsequently affects gut epithelial integrity. Our gut-brain vascular model is a precisely engineered microphysiological system that emulates the human gut-brain-vascular axis and serves as a translational tool to identify therapeutic targets and evaluate pharmaceuticals for neurological and gastrointestinal disorders.\u003c/p\u003e","manuscriptTitle":"A 3D Gut-Brain-Vascular Platform for Bidirectional Crosstalk in Gut-Neuropathogenesis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-03-12 13:39:29","doi":"10.21203/rs.3.rs-6128513/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"44d0380a-11c7-4c11-b55d-19d0eac41343","owner":[],"postedDate":"March 12th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":45542596,"name":"Biological sciences/Biological techniques/Lab-on-a-chip"},{"id":45542597,"name":"Biological sciences/Cell biology/Cell signalling"},{"id":45542598,"name":"Health sciences/Diseases/Neurological disorders"}],"tags":[],"updatedAt":"2026-03-18T07:06:58+00:00","versionOfRecord":{"articleIdentity":"rs-6128513","link":"https://doi.org/10.1038/s41467-026-69318-y","journal":{"identity":"nature-communications","isVorOnly":false,"title":"Nature Communications"},"publishedOn":"2026-02-07 05:00:00","publishedOnDateReadable":"February 7th, 2026"},"versionCreatedAt":"2025-03-12 13:39:29","video":"","vorDoi":"10.1038/s41467-026-69318-y","vorDoiUrl":"https://doi.org/10.1038/s41467-026-69318-y","workflowStages":[]},"version":"v1","identity":"rs-6128513","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6128513","identity":"rs-6128513","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}