Basic models to advanced systems: harnessing the power of organoids-based microphysiological models of the human brain

Understanding the complexities of the human brain’s function in health and disease is a formidable challenge in neuroscience. While traditional models like animals offer valuable insights, they often fall short in accurately mirroring human biology and drug responses. Moreover, recent legislation has underscored the need for more predictive models that more accurately represent human physiology. To address this requirement, human-derived cell cultures have emerged as a crucial alternative for biomedical research. However, traditional static cell culture models lack the dynamic tissue microenvironment that governs human tissue function. Advanced in vitro systems, such as organoids and microphysiological systems (MPSs), bridge this gap by offering more accurate representations of human biology. Organoids, which are three-dimensional miniaturized organ-like structures derived from stem cells, exhibit physiological responses akin to native tissues, but lack essential tissue-specific components such as functional vascular structures and immune cells. Recent endeavors have focused on incorporating endothelial cells and immune cells into organoids to enhance vascularization, maturation, and disease modeling. MPS, including organ-on-chip technologies, integrate diverse cell types and vascularization under dynamic culture conditions, revolutionizing brain research by bridging the gap between in vitro and in vivo models. In this review, we delve into the evolution of MPS, with a particular focus on highlighting the significance of vascularization in enhancing the viability, functionality, and disease modeling potential of organoids. By examining the interplay of vasculature and neuronal cells within organoids, we can uncover novel therapeutic targets and gain valuable insights into disease mechanisms, offering the promise of significant advancements in neuroscience and improved patient outcomes.


Introduction
The study of the human brain and its complex functioning in both health and disease states is a significant challenge in neuroscience research.Although traditional experimental models, including animal models, have provided valuable insights, their ability to replicate human biology and predict human responses to drugs is limited [1,2].Moreover, interspecies differences can significantly hinder the translation of findings to human patients.Recognizing the urgent need for more predictive and human-relevant models, recent legislations [3], such as the FDA Modernization Act 2.0 in the USA, the New Drugs and Clinical Trial Rules in India, the European Union Directive 2010/63/EU, and Canada's Bill C-47 amending the Food and Drugs Act, emphasize the demand for new alternative models that better represent human biology.
Human-derived cell culture models have increasingly become a common alternative method for biomedical research.Traditional cell culture approaches involve culturing monolayers of cells on flat plastic surfaces; however, they lack the ability to recreate the dynamic tissue microenvironment (TME) that governs human tissue and organ function.To address this limitation, advanced in vitro systems have emerged as promising avenues to bridge the gap and provide more accurate representations of human biology [4,5].
In the past 15 years, organoids have emerged as powerful tools in biomedical research, offering the potential to bridge the gap between traditional cell culture models and in vivo studies [6].These threedimensional (3D) miniaturized organ-like structures are derived from pluripotent stem cells (PSCs) or tissue-specific progenitor cells and aim to recapitulate the architecture and functionality of specific organs [7].The unique advantage of organoids lies in their ability to exhibit cellular diversity, selforganization, and physiological responses similar to native tissues [8].However, organoids also have limitations such as lack of biochemical and biomechanical stimuli provided by the surrounding tissuemicroenvironment and the absence of functional vascular structures [9].The absence of a functional vascular system within organoids restricts their size, limits nutrient diffusion, and may hinder their maturation [10].Recognizing the importance of vascularization in advancing organoid research, scientists have been actively exploring strategies to integrate vascular components into these structures.By introducing endothelial cells and promoting their interactions with other cell types, researchers aim to induce the formation of functional blood vessel networks within organoids.This approach facilitates cellular communication, ultimately enhancing organoid maturity, functionality, and their relevance in disease modeling [11,12].
Traditional static cell culture methods are often constrained in their capacity to maintain a consistent concentration of nutrients and oxygen [13], thereby limiting the possibility of growing organoids under physiologically relevant conditions.Recent advancements in microphysiological systems (MPSs) have opened the door to developing models that incorporate diverse cell types within dynamic microenvironments, faithfully emulating the biochemical and biomechanical characteristics of native TMEs.MPSs encompass various engineered in vitro models like Tissue Chips and Organs-on-Chips, integrating multiple cell types under dynamic cell culture conditions [14].These models are designed to more faithfully recapitulate human tissue and organ function compared to traditional static cell culture models.While the specifics of their design, materials and scope may vary significantly from one model to another, the presence of tissue/organ-level complexity stands out as a common characteristic of MPS.The convergence of organoids and MPS, such as organ-on-chip technologies, represents a powerful strategy for advancing our understanding of brain development, function, and disease [15].These integrated models serve as a versatile platform for exploring the intricate cellular interactions and dynamic processes underlying brain physiology and pathology.By bridging the gap between in vitro cell culture models and in vivo animal studies, this approach holds the potential to revolutionize brain research and contribute significantly to the development of innovative therapies for neurological disorders.
In this review, we explore the concept of vascularized cerebral organoids as a transformative approach to investigating human brain function.We track the evolution of MPS, examining their potential to replicate the complexities of the human brain and highlighting current limitations, as summarized in figure 1.Our emphasis is on bioengineering the TME to enhance viability, functionality, and the modeling of complex brain diseases linked to dysfunction, including multi-organ relationships.By shedding light on the interplay between the TME and the brain cells within organoids, we aim to provide valuable insights into disease mechanisms and novel therapeutic targets, ultimately advancing the field of neuroscience and improving patient outcomes.

Three-dimensional cell culture in neurobiology
The evolution of 3D cell culture methods (figure 2) for brain spheroids and organoids traces its origins to the classic hanging drop method, where cell clusters (spheroids) were cultured in droplets suspended from the lid of a petri dish.Initially employed for culturing brain cancer spheroids in 1970 [16,17], this method has since progressed to encompass non-cancerous cells and extracellular matrix (ECM) components, creating a more faithful representation of the brain microenvironment.One key advantage of 3D cell culture for neural cells, and other organ types, over traditional 2D approaches lies in its closer approximation of the in vivo brain environment.This approach facilitates a deeper understanding of development and disease mechanisms.Unlike 2D cultures, which adhere to flat surfaces, 3D cultures promote the formation of organotypic, intricate tissue structures.In the context of brain research, this is exemplified by the emergence of neural rosettes and distinct brain ventricle regions, enabling researchers to investigate critical cell-cell interactions pivotal for proper brain development and function.Diverse 3D neural cultures techniques have been developed, including cortical spheroids and cerebral organoids.

Neurospheres
One of the earliest techniques for culturing neural cells in a 3D format involves neurospheres, which comprise neural stem cells (NSCs) and neural progenitor cells (NPCs) grown in suspension cultures without ECM.These neural cell clusters proliferate within suspension culture, facilitating the study of primary cells or differentiated NSCs [18].Neurospheres offers a means to culture neural cells in 3D, an advancement over the 2D neural rosettes.However, both rely on the same NPC types derived from NSCs.While neurospheres were a significant milestone in modeling 3D brain in vitro, they lacked the structural organization and cellular diversity necessary for more faithful brain mimicry.To address this limitation, cortical spheroids were developed to better replicate the diverse cell population within the brain [19].In brief, cortical spheroids are derived from PSCs aggregated into a spheroid format in the absence of an ECM and then differentiated into the various neural cell types.Establishment of cortical spheroids through forced aggregation of PSCs relies on cellular self-organization, driven by chemical signaling mimicking corticogenesis, resulting in neural structures containing both deep and superficial neurons, interspersed with astrocytes to support proper neural development through synapse formation.

Brain organoids
While adult stem cells have proven valuable in generating various tissue types like intestinal organoids [20], the possibility of studying human brain biology using adult stem cells has been hampered by the limited availability and accessibility of stem cells within the human brain.The discovery of induced PSCs (iPSCs) by Shinya Yamanaka and Kazutoshi Takahashi in the early 2000s addressed this obstacle [21].In contrast to primary cells, iPSCs can be derived from virtually any somatic cell in the human body, with skin cells and circulating immune cells being preferred sources.Furthermore, iPSCs can be obtained from any individual, enabling the modeling of inter-individual variability.This provides a powerful tool for understanding disease mechanisms and developing personalized therapies based on patientspecific genotypes and phenotypes.
PSC-derived cerebral organoids, commonly known as brain organoids, are 3D cell cultures derived from iPSCs or embryonic stem cells and differentiate into a variety of cell types found in distinct brain regions, including neurons, astrocytes, and oligodendrocytes [22].One of the pioneering models that successfully recapitulated the regional complexity of the brain and relied on Matrigel as the chosen ECM to facilitate the formation of desired structures was introduced by Lancaster et al in 2013 [23].In their groundbreaking study, the researchers cultivated 3D stem cell cultures over a span of 90 d, which exhibited tissue organization and cell types closely resembling the structures of cerebral cortex observed in vivo.Within the cerebral organoid model, these cultures formed layers of cortical neurons that closely resembled the cellular organization found in the brain.This intricate layering was achieved through the use of Matrigel as a 3D scaffold, promoting the development and organization of these brain-like structures.In contrast to 2D models, which often feature flat neural rosettes with limited layers and less physiologically relevant cellular organization, cerebral organoids provide a powerful method to mimic the complex structure of the brain in vitro.
When compared to cells derived from adult stem cells, iPSC-derived cells provide a solution to overcome the challenges associated with limited accessibility to primary or postmortem tissue sources.Moreover, the utilization of well-defined differentiation protocols enables the consistent generation of desired cell lineages or phenotypes.By employing cells from patients with specific genotypes, personalized cell lines featuring the desired genetic makeup can be generated.These personalized models are valuable tools for studying various aspects of development, genetic predisposition, disease states, drug therapies and identifying potential therapeutic targets [24].A notable example of the application of iPSCs to create personalized models of brain disorders was demonstrated by Penney et al in 2020.In their study, iPSCs were derived from skin fibroblasts obtained from individuals with Alzheimer's disease (A.D.), as well as healthy control subjects.These iPSCs were then differentiated into neurons enabling the establishment of a cellular model that specifically recapitulated the characteristics of A.D. in vitro [25].The researchers used these neurons and other cell types to study the mechanisms of A.D. by comparing the healthy against diseased cell types, including the effects of genetic mutations and the accumulation of amyloid-β peptides.Similarly, in 2021, Brennand et al developed a schizophrenia using iPSCs generated from fibroblasts of individuals with schizophrenia and control individual.These iPSCs were differentiated into neurons and astrocytes, which were subsequently used to study the mechanisms of schizophrenia.The study revealed diminished neuronal connectivity, decreased neuron counts, and neurotransmitter imbalances, marking one of the first instances where the pathogenesis of a late-stage neurodegenerative disorder was successfully modeled using iPSCs.

Assembloids
Brain organoids, originating from iPSc, are proving invaluable for studying the molecular dynamics of human brain development and neurological disorders.Despite this, the inherent complexity of the human brain, characterized by diverse multiregional structures, poses a challenge for current organoids in accurately recapitulating this intricate architecture, especially in establishing neural circuits across various brain regions.Assembloids represent 3D organoid models created by fusing together regional cell types to construct tissue structures with specific regional patterning [26].Initially, two or more regions of brain organoids are differentiated independently.Following the region-specific patterning, the organoids are selected and physically brought into close proximity, promoting spontaneous fusion and the formation of assembloids.
When compared to iPSc-derived cerebral organoids, assembloids offer better control over the inclusion of specific cell types in the model and their resulting organization [27].By integrating different brain regions within assembloids, researchers can gain valuable insights into the complex interactions and behaviors of diverse cell types, providing a powerful tool for studying neurodevelopmental processes and associated disorders.Studies published by Bagley et al [28] and Xiang et al [29] demonstrated interneuron migration within assembloids, showcasing their potential to replicate complex cellular interactions observed in the developing brain.Bagley et al achieved this by fusing dorsal and ventral forebrain cerebral organoids, resulting in the generation of a dorsal-ventral axis that facilitated interneuron migration from the ventral to the dorsal forebrain.Similarly, Xiang et al demonstrated the migration of interneurons from the medial ganglionic eminence organoids to cortical cerebral organoids within assembloids.These findings underscore assembloids' ability to replicate dynamic cellular processes, such as interneuron migration, critical for proper brain development and function.
Assembloids can be constructed in a modular fashion, allowing researchers to tailor models with specific cell types and structures.This flexibility facilitates a more targeted approach to study of disease mechanisms and drug discovery [30].In 2017, Birey et al pioneered a brain assembloid culture system using iPSCs derived from patients with Timothy syndrome (TS), a rare autosomal-dominant disorder linked to mutations in the CACNA1C gene.This breakthrough provided insights into inhibitory neuronal migration patterns during brain development [31].Real-time imaging of dorsal and ventral forebrain organoids from TS patients revealed defective tangential-like migration of GABAergic inhibitory neurons, a phenotype rescued by nimodipine through L-type calcium channel blockade.This marked the first use of brain assembloids to model neuronal cell interactions across different brain regions, enabling a deeper exploration of the molecular mechanisms underlying the disease [32].
Additionally, Andersen et al leveraged human iPSCs to reconstruct the cortico-spinal-muscle circuit, creating the first in vitro central nervous system (CNS) model of a three-unit assembloid.This innovative approach allowed the manipulation of motor control for muscle contraction in vitro [33].By fusing human cortical, spinal, and skeletal muscle spheroids, cortico-motor assembloids exhibited spontaneous muscle cell contractions, maintaining morphological and functional integrity for up to 10 weeks postfusion.Notably, these assembloids surpassed muscle cells alone or cortical-muscle assembloids in spontaneous contractions.Stimulation of cortical spheroids with glutamate induced a robust muscle contraction response, indicating the successful assembly of a functional cortico-spinal-muscle unit.This model holds promise for characterizing developmental abnormalities and dysfunctions in the cortico-motor circuit, offering new insights into the pathogenesis and treatment of diseases like amyotrophic lateral sclerosis.

Limitations of brain organoids and assembloids
While brain organoids and assembloids offer valuable insights into brain development and disease modeling, they come with various limitations that need consideration [9].One limitation is that they may not fully mature to resemble the complexity and functionality of the adult human brain, potentially limiting their accuracy in representing the natural progression of neurodegenerative processes primarily observed in elderly patients.Another drawback of typical iPSC-derived organoids is the absence of certain critical cell types.For example, brain organoids lack cerebrovascular endothelial cells, which are essential for forming the blood-brain barrier and regulating molecule exchange between the brain and blood.Additionally, brain organoids do not naturally host immune cells like microglia, originating from the yolk sac during development and contributing to immune responses within the brain.The absence of these crucial components restricts the ability of organoids to faithfully replicate the cellular diversity and complexity found in the human brain.Consequently, interactions between different cell types and the influence of the brain's microenvironment on disease progression may not be fully captured within organoid models.

Reconstituting the TME for achieving organoid maturation: lessons learned from animal studies
When cells and organoids are cultivated in vitro, they are typically grown under static conditions in artificial media.However, these conditions fail to accurately emulate the intricate and ever-changing nature of the in vivo microenvironment.In their natural niche, cells receive signals from neighboring cells, ECM components, and soluble factors that profoundly influence their behavior and function.These signals encompass physical cues like mechanical forces, topological features, and cell-cell interactions, as well as biochemical cues like growth factors, cytokines, and signaling molecules.Together, these cues collectively define the native TME where cells operate.Nevertheless, conventional in vitro cultures lack many of these dynamic elements, limiting their capacity to faithfully replicate the complexities of in vivo biology.

Animal models as hosts for enhancing organoid maturation and function
To address the constraints of in vitro culture systems, one strategy involves the transplantation of organoids into living animals, frequently employing mice as the host.A commonly chosen transplantation site is the kidney capsule, where the host environment supports organoid vascularization, thereby enhancing maturation and functionality [34][35][36].In the field of neurological diseases, researchers have made significant strides by transplanting brain organoids into the brains of mice, enabling the investigation of complex cellular interactions and disease mechanisms [37].Recent studies have demonstrated that this transplantation approach can lead to the simultaneous maturation of human cells within the organoids and their structural and functional integration into the animal, yielding genuine chimeric animals.

Chimera models for investigating brain development and regeneration in a native-like microenvironment
Pasca's research group has made notable advancements in leveraging transplantation experiments to explore the development and function of human neural organoids in vivo by successfully implanting human cortical organoids into the developing cerebral cortex of early-postnatal immunocompromised rats [38].The outcomes showcased the maturation of diverse cell types within the transplanted cerebral organoids, a phenomenon not observed in vitro.Importantly, the transplanted organoids anatomically and functionally integrated into the rodent brain, established connections with the host neural circuits and, the activity of human cells within the transplanted cerebral organoids was found to modulate the activity of rat neurons and drive behavioral responses.
Employing a similar strategy, researchers at the University of Pennsylvania successfully engrafted iPSC-derived forebrain organoids, which were cultured for 80 d, into the brains of adult rats with injuries to their visual cortex [39].Within three months, the transplanted organoids displayed integration within the host's brain.Notably, the grafted organoids underwent vascularization, increased in size and number, and extended neuronal projections, forming synapses with the host's neurons.These remarkable developments not only demonstrate the successful engraftment and integration of micro-tissues but also signify the enhancement of human cell maturation, resulting in the development of a functional component within the brain.
These transplantation methods hold immense potential in advancing organoid maturation and functionality, offering invaluable insights into neurodevelopmental disorders, neurodegenerative diseases, and other neurological conditions including brain tumors.Generating of chimera models is quite common practice in the field of oncology where immunosuppressed mice can serve as hosts for engrafting patient-derived xenografts (PDXs) [40].PDX models offer the advantage of retaining the original features of the tumor, recapitulating key characteristics such as the spatial structure and intratumor heterogeneity of cancer.Additionally, humanized PDX models can mimic the interactions between the tumor and the immune system, providing a platform for personalized patient medication management, especially in glioblastoma (GBM) and other types of brain tumors [41].

Immunocompetent organoids
The protocols commonly employed for generating iPSC-derived brain organoids aim to mimic the natural development of brain cells, with a focus on differentiating iPSCs into various neural cell types, such as neurons and astrocytes.However, they do not include microglia, the resident immune component cells of the brain.Microglial cells have a unique and distinct developmental origin compared to other immune cells in the body, primarily arising from bone marrow hematopoietic stem cells.Microglia originates from primitive myeloid progenitor cells in the yolk sac during early embryonic development [42].From there, they migrate to the brain and colonize the entire CNS before the blood brain barrier (BBB) is fully formed.Once in the brain, microglia undergo a maturation process and establish their unique phenotype and function.
In the mature brain, microglia cells play multifaceted roles beyond their immune response functions.They actively contribute to the clearance of cellular debris, surveillance of neuronal activity, and support of synaptic plasticity.Their dynamic interactions with neurons and astrocytes are crucial for maintaining a healthy brain environment.Consequently, the absence or dysfunction of microglia can contribute to chronic neuroinflammation, synaptic loss, and neuronal dysfunction observed in various neurodegenerative diseases.This highlights the importance of incorporating microglia into in vitro brain models to better understand and model these complex processes and diseases.
Recent studies have shown that the integration of microglia cells into human brain organoids results in increased neuronal maturation and functionality in vitro [43].This integration opens new avenues for investigating the contributions of microglia to the pathogenesis of neurodegenerative diseases.It allows researchers to explore the effects of microglial dysfunction on neuronal health, identify potential therapeutic targets for modulating microglial activity, and assess the efficacy of interventions aimed at restoring microglial function.
To obtain a more mature tissue structure and overcome some of the limitations associated with traditional in vitro models, Schafer et al transplanted brain organoids and microglia cells in living mice [44].Xenotransplantation of human brain organoids into animal hosts provides an opportunity to study the interactions between human microglia and the surrounding brain tissue in a more organotypic context.It allows for the exploration of how microglia develop, function, and respond to diseaseassociated factors within a living brain environment.While current iPSC-derived microglia-like cells have been used in vitro for disease modeling; their full characterization is still underway, making in vivo engraftment a more relevant model for studying disease [44].Schafer's research noted that when transplanting human microglia-like cells into mice, they developed transcriptomic identities similar to those found in humans, suggesting that a humanlike environment could be established with matured microglia.

Limitations of animal studies in neurobiology
Despite the advantages of xenotransplantation for enhancing stem cell maturation through physiological stimuli, this approach has limitations.In neurogenerative models, the inherent differences between brain structures in many common animal models and humans pose a significant challenge.These disparities in brain organization make it difficult to rely solely on animal models for neurodegeneration studies.Additionally, the analysis techniques available to assess neuron-microglia and other interactions are challenging to execute in vivo, further limiting the utility of animal models [45].In the case of PDX models, the under-representation of the human tumor stroma can impact therapy outcomes.To address this limitation, researchers have explored the use of 'humanized' mice expressing components of the human immune system and co-injection of human stromal/immune cells and cytokines with patientderived cancer cells [46].Furthermore, differences in pharmacokinetic properties of drugs across species pose challenges in predicting drug efficacy in patients using animal models.Additionally, PDX models may not fully replicate the metastatic process observed in patients due to host and species-specific factors, limiting their ability to capture all aspects of disease progression [47].
Ideally, a fully human-derived model could overcome these limitations.Recent advancements in reconstituting tumor heterogeneity have been achieved through the co-culturing of GBM organoids with iPSC-derived brain organoids and endothelial cells, known as the glioblastoma-in-a-cultureorganoid (GLICO) system [48].This innovative approach involves incorporating patient-derived GBM cells with normal brain cells, including astrocytes and endothelial cells, to recreate the tumor microenvironment within organoid cultures.The inclusion of astrocytes and endothelial cells in the GLICO system provides a more comprehensive representation of the cellular and molecular interactions within the tumor microenvironment.
Experiments that involve xenotransplantation are also limited in their throughput.While the number of organoids implanted can vary depending on experimental requirements, in the case of cerebral organoids, it is common to implant one organoid per mouse.However, it is important to note that not all xenotransplants are successful, and there may be instances where the organoid fails to engraft correctly in the host tissue.Previous research findings have highlighted the critical role of vascularization in supporting the successful engraftment and maturation of liver organoids, enabling their integration into the host environment [35].Similarly, in the context of brain organoids, the establishment of functional blood vessels is recognized as a pivotal event for enhancing their engraftment potential and promoting the development of mature and physiologically relevant structures.While achieving vascularization within brain organoids remains a formidable challenge, it represents a significant milestone in advancing their potential to mimic the complexity and functionality of the human brain.By incorporating a vascular network, organoids can better recapitulate the intricate interactions between neuronal cells and the surrounding vasculature, facilitating essential processes such as nutrient exchange, waste removal, and the modeling of diseases involving the BBB [49].

Vascularization as a cornerstone for generating physiological brain models in vitro
Vascularization is increasingly recognized as one of the key challenges or bottlenecks in fully harnessing the potential of organoids in vivo and in vitro.While organoids have made significant strides in replicating the structural and functional aspects of organs, the absence or limited presence of a functional vascular system within organoids poses a critical barrier.The lack of vascularization restricts the size and complexity of organoids that can be achieved.Without an adequate blood supply, nutrient and oxygen diffusion becomes limited, resulting in restricted growth and viability of cells within the organoid.As organoids grow larger and more complex, the distance between cells and the nearest nutrient supply increases, leading to compromised cell function and even cell death in the core regions of the organoid.This diffusion limitation prevents the development of fully mature and functional organoids that accurately recapitulate the intricate physiology and cellular interactions seen in native organs.Furthermore, the absence of vascularization poses a significant obstacle to accurately modeling diseases within organoids.Specifically, the lack of a functional vascular network significantly restricts the study of diseases that intricately involve vascular components, notably impacting areas like tumor angiogenesis and drug delivery across the BBB [50,51].Addressing these challenges is critical for developing more physiologically relevant disease models and advancing research in these critical fields.

Relevance of the vasculature in the brain
In the adult brain, the vascular endothelium of the BBB plays a critical role in maintaining homeostasis and functionality of the whole organ.This specialized endothelial layer forms a physical and biochemical barrier that separates the circulating blood from the brain [52] (figure 3).The BBB tightly regulates the exchange of molecules, ions, and cells between the blood and the brain, allowing for the maintenance of a stable and protected environment necessary for proper brain function.The endothelial cells of the BBB are unique in their structure and function.They possess tight junctions that restrict the paracellular movement of substances, preventing the entry of harmful molecules and pathogens into the brain.Additionally, the brain endothelial cells express various transporters and receptors that selectively facilitate the transport of essential nutrients, oxygen, and metabolic waste products across the barrier.
Beyond its role in regulating the transport of molecules, the vascular endothelium of the BBB also participates in important physiological processes in the brain.It contributes to the regulation of cerebral blood flow, ensuring an adequate supply of nutrients and oxygen to meet the high metabolic demand of neurons.Furthermore, the BBB actively participates in neurovascular coupling, the process by which changes in neuronal activity are coupled to adjustments in local blood flow [53].In addition to its vital role in normal brain physiology, dysfunction or disruption of the BBB has been implicated in various neurological disorders.Impaired BBB integrity can lead to increased permeability, allowing the entry of toxic substances, immune cells, and pathogens into the brain, which can trigger inflammation and contribute to the pathogenesis of conditions such as neurodegenerative diseases, stroke, and brain Figure 3.The blood-brain barrier in homeostasis and disease.Within the NVU, the healthy, intact BBB (left panel) limits paracellular permeability (via tight junctions) and maintains homeostasis in the brain by selectively allowing the uptake of glucose, nutrients, and oxygen, while limiting the uptake of noxious blood-borne agents like pathogens and xenobiotics.The BBB is also equipped with important efflux transporters like P-glycoprotein that limit the brain's accumulation of waste, such as proteinaceous aggregates like β-Amyloid (Aβ) peptide, and of foreign agents like drugs.On the other hand, the diseased or dysfunctional BBB (right panel) becomes 'leaky' due to weak or disrupted tight junctions, which allows peripheral immune cells (e.g.leukocytes) and blood proteins (e.g.fibrinogen) to enter the brain, eliciting an inflammatory response.Inflammation and oxidative stress in the BBB and brain are often associated with neurodegenerative diseases, including Alzheimer's, where the accumulation of insoluble Aβ peptide, due to ineffective or decreased extrusion via the BBB, forms neurotoxic Aβ plaque, an important hallmark of the disease.
tumors.Reconstituting the cell-cell interaction at the cerebrovascular interface is central to recapitulating the physiological function of the BBB in vitro.Several methods are currently employed to construct these models, including organoids obtained from stem cells genetically modified to overexpress endothelial factors as well as bioengineered systems as Transwell inserts and microfluidic-based platforms [54].
Endothelial cells do not typically form spontaneously into brain organoids.However, previous work by Cakir et al [55] has demonstrated that the ectopic expression of human ETS variant 2 (ETV2), a master regulator of vascular maturation, is sufficient to generate vascularized brain organoids in vitro.This occurs regardless of differentiation conditions and in the absence of growth factors, such as VEGF, which are essential for differentiating and maintaining mature endothelial cells in culture.While this approach represents an elegant solution to the challenging task of developing vascularized models of the brain, it is difficult to exclude potential negative impacts associated with the overexpression of the ETV2 gene throughout the entire organoid and how this might affect intercellular interactions.
Transwell inserts involve the use of permeable membranes to create separate compartments for different cell types, allowing for the study of direct or indirect cell-cell interactions on the BBB.On the other hand, microfluidic-based systems integrate multicellular cultures with controlled fluid flow, enabling the simulation of physiological conditions and the investigation of complex cell-cell interactions within the cerebrovascular tissue.These advanced techniques contribute to our understanding of BBB physiology and its role in various neurological processes and diseases.

Transwell inserts
Transwell models are considered as one of the most straightforward in vitro systems for recreating the natural cell-cell interactions commonly observed in organs with barrier functions [56,57], including the BBB [58].These systems utilize a permeable membrane in which endothelial cells are cultured alone or in juxtaposition to astrocytes and/or pericytes.Using this strategy, Shusta's research group has previously demonstrated the determinant role of astrocytes in regulating vascular barrier-function via modulating the expression of multiple markers and including tight junctions and molecular transporters using iPSC-derived brain endothelial cells (iHBMECs), astrocytes (iAstrocytes) and neurons (iNeurons) [45].While these models, such as the Transwell inserts, offer ease of use, they have limitations in mimicking the cell-cell contact and dynamic microenvironment observed in vivo, including fluid flow-induced shear stress.
Recent efforts have been made to address these limitations by integrating Transwell inserts into microfluidic setups using engineered microfluidic devices made of PDMS (Chips) (figure 4(A)) [59,60] or by incorporating pumps, tubing, and holders to enable continuous fluid flow over the Transwell inserts [61].Although not specifically designed to generate physiological relevant shear stress, the application of fluid flow in cell culture medium, often combined with media recirculation, has been demonstrated to support cell culture and promote the maturation of cell functions, often yielding better results compared to traditional static models (figure 4(B)).One advantage of fluid flow is that it provides a continuous and steady supply of nutrients to the cells, reducing the stress associated with sudden changes in growth factor concentrations and instantaneous depletion of autocrine factors that can occur during media changes in static systems.This constant nutrient supply helps maintain cell viability and supports optimal cellular function and maturation.These dynamic culture systems offer a more physiologically relevant environment for studying cell behavior and can provide valuable insights into the mechanisms underlying cell function and maturation.It should be noted, however, that the specific effects and optimal parameters of fluid flow in cell culture systems may vary depending on the cell type and desired outcomes.Further research is necessary to optimize these techniques and better understand the impact of fluid flow on brain cell culture and function.
In this context, microlfuidic platforms, especially Organ-on-Chips, emerge as microengineered systems tailored to replicate the physiological conditions of blood flow.These platforms maintain a laminar flow regime, characterized by streamlined and nonturbulent movement of fluids.This design choice is relevant for faithfully capturing the 'flow dynamics' of blood, including its velocity, patterns, and forces during circulation.

Organs-on-chips
A key limitation in traditional static culture systems is the restricted diffusion of nutrients and signaling molecules beyond a tissue thickness of around 200 µm, which can impair cellular function and viability [62].While neural tissue models and brain organoids can be co-cultured with human endothelial cells to mimic cell-cell interactions found in living organs [12,63,64], the functionality of any developed vasculature remains a subject of debate.In drug testing applications, drugs are administered to the entire microtissue rather than being perfused through the vasculature.Consequently, traditional static organoids and spheroids models are inadequate for assessing drug penetration and delivery through the BBB, a critical step that impacts drug availability and effectiveness in humans and is not always effectively captured by animal models.
In the last 12 years, the Organ-on-Chip technology has revolutionized tissue engineering and regenerative medicine.These devices replicate human tissue structure and function by integrating microfluidics, multi-cell culture, and biomaterials.Organon-Chip devices contain microfluidic channels lined with diverse cell types representing specific organs.These channels facilitate nutrient exchange and waste removal, mimicking the organ's vasculature and interstitial flow.These systems aim to replicate tissue and organ-level functions.The microfluidic framework, often created using techniques like soft lithography from the electronic chip industry, has led to their nickname 'Organ-on-Chips' [14].In brain research, Organ-on-Chip models offer unique advantages for studying interactions between cerebrovascular tissue, neurons, astrocytes, and other supporting cells.By incorporating these cell types in a controlled microenvironment, Organ-on-Chip models provide a platform to investigate their dynamic interplay and responses to biomechanical cues [45,65].
Vatine et al [65] pioneered the first isogenic model of BBB-on-Chip.The group combined iNeurons, iAstrocytes obtained from dissociated brain organoids (EZ-Spheres), with iHBMECs generated from the same patient (isogenic), paving the way for developing patient-derived models (also named Avatars [14]) of the brain.The Chip design featured a microfluidic device with two parallel chambers separated by a porous membrane.Both chambers were coated with ECM proteins Collagen IV and Fibronectin and then seeded with cells.Astrocytes and neurons were seeded on the top chamber, while endothelial cells were seeded on the lower chamber and allowed to grow to cover the entire surface of the microfluidic compartment.By progressively increasing the flow rate (0.1-6 dyne cm −2 ), the team demonstrated that physiological shear stress enhances the maturation of iHBMECs.The system was also used to show, for the first time, that human blood and plasma can be perfused through the vascular compartment to study mechanisms of molecular transport in a physiologically relevant manner.In a similar approach, the group of Donald Ingber, who pioneered the 'Sandwich-Chip' design used by Vatine et al and widely adopted in the field [66,67], demonstrated the possibility of mimicking vascular shear stress on the chip by increasing the viscosity of the cell culture medium, introducing dextran into the vascular medium [68].This strategy provides a more viscous solution that can generate high levels of shear with a lower flow rate facilitating the use of physiological shear stress in these models.
Using a different Chip design incorporating multiple parallel chambers, Campisi et al in the research group of Roger Kamm, generated an Organ-on-Chip model of the BBB incorporating capillary-like structures by leveraging the self-assembling ability of human cells when cultured in a fibrin hydrogel [69].The Chip design incorporates three parallel microfluidic chambers, with the central one used for casting a cell-laden hydrogel, including primary astrocytes, pericytes, and non-tissue-specific iPSCderived endothelial cell (iPSC-EC).The lateral chambers are seeded only with iPSC-EC.The formation of a perfusive capillary network in the central channel is achieved via vasculogenesis.The Chip design is optimized for promoting the physical interconnection of capillaries with the lateral chambers via vascular anastomosis.The result is a perfusable selfassembled network of capillaries that can be perfused with cell culture medium under controlled flow regimes.The group also achieved the recapitulation of interstitial fluid flow by applying pressure to the media inlets into the lateral chambers.The differential pressure between the microfluidic chambers was used to generate interstitial fluid flow across the 3D hydrogel located in the central compartment of the device [70].
While these devices differ significantly in their design, both systems are well-suited for modeling physiologically relevant parameters of vascular permeability of the BBB, which is crucial for gaining insights into how molecules and therapeutics can be effectively delivered into the brain.When compared, the Sandwich-Chip design developed in Ingber's lab offers a better solution for sampling media effluent and conducting analytical measurements of barrier function and molecular transport.Moreover, the presence of a stretchable membrane, a key feature of this design, could be used in the future for modeling specific aspects of brain trauma injury currently difficult to capture in vitro.One limitation of this Chip design is that the vascular compartment, although lined with iHBMECs, measures a diameter of nearly 1 mm what is significantly larger than a human capillary, typically measuring 5-10 µm.On the other hand, Kamm's group design, incorporating complex architectures and image-based measurements, provides an opportunity to study the dynamics of molecular transport at a single capillary resolution.The presence of capillary-like structures in the second design, moreover, offers a more physiological relevant method for studying the cellular dynamics guiding migration of circulatory cells in capillaries [71].Other considerations regarding these and other models for brain research have been thoroughly discussed in previous reviews [72][73][74].
Here, we will emphasize the recent trend of integrating organoids into Organ-on-Chip technologies.This integration of dynamic flow with organoids holds great promise for creating more physiologically relevant and functional models of the human brain, facilitating a deeper understanding of brain development, function, and disease.Key features of vascularized models discussed in the present review are captured in figure 5.

Integrating microfluidics for cerebrovascular and neural development studies in brain models
Cerebrovascular tissue in mammalian brains has evolved to meet the growing metabolic demands and unique functions of the brain throughout millions of years of evolution.During CNS development, vasculature is primarily formed through angiogenesis, which involves the growth of new blood vessels from pre-existing ones, rather than de novo vasculogenesis observed in other tissues [76].The angiogenic process is tightly controlled by a complex intercellular communication between the developing brain and the vascular endothelium, although the precise mechanisms are not fully understood.Indeed, this intercellular cross talk is believed to play a pivotal role in coordinating and synchronizing vascularization with brain maturation.The increasing body of evidence supports the notion that cerebrovascular endothelial cells play a key role in the maturation of NSCs during embryogenesis, further emphasizing their critical involvement in crucial developmental processes [77].
One Chip design that captures the intricate relationship between neural development and In this design, an induced NSC spheroid was cultured within a fibrin gel, alongside fibroblasts and endothelial cells, which were fed from reservoirs via microfluidic channels.These cells coexisted in the gels for a duration of two weeks, during which the endothelial cells organized themselves into a network of blood vessels that surrounded the spheroids.When compared to non-vascularized controls, the NSCs within the vasculature-rich environment displayed heightened and sustained expression of markers associated with neural and supporting cell differentiation.Additionally, the presence of these penetrating blood vessels seemed to enhance cell viability while reducing cell death within the spheroid.Although promising, the use of spheroids, which are smaller and less structurally organized than organoids, may limit the applicability of this approach to organoid research [78].
These interactions are also evident in the Chip designed by Salmon et al, which combined cerebral organoids with human iPSC-derived vascular cells.Notably, iPSC-derived endothelial cells, when compared to standard human umbilical vein endothelial cells, were not terminally differentiated and could potentially be guided down tissue-specific pathways.In this innovative setup, endothelial cells and pericytes, both derived from the same iPSC cell line, were incorporated into the channels surrounding the central organoid chamber.Matrigel served as the sole hydrogel for 3D cell culture, facilitating the sprouting and invasion of these cells into the core and, notably, the cortices of the organoids.Intriguingly, this integration led to the presence of more mature NPCs after only 30 d of culture.Furthermore, these NPCs generated neurites that associated and aligned with the vascular sprouts.While it was confirmed that the sprouts were perusable, it is important to note that long-term perfusion and flow studies were not conducted on this chip [75].
Future development and validation of such a human-relevant models can offer insights into the intricate interactions between the vasculature and brain cells, shedding light on the underlying mechanisms of brain function and the pathogenesis of neurological disorders.Organs-on-Chips and other microfluidic platforms offer unique advantages that address current in vitro limitations, such as size and complexity, making them more applicable for developmental or disease models compared to traditional alternatives.These platforms provide precise control over physical and chemical variables, including flow rates, shear stress, hydrogel stiffness, and nutrient and signaling gradients [79,80].This level of control allows researchers to tease out developmental signals and model diseases with a high degree of repeatability.Moreover, the incorporation of hydrogels enables more controlled designs for co-culturing different cell lineages [81,82], like incorporating vasculature, or modeling specific structures, such as the BBB, which is challenging with conventional cultures.These advanced interactions would be difficult to implement or test with traditional methods or would have limited applicability with animal models.[83] with permission from the Royal Society of Chemistry.Adapted from [84].CC BY 4.0.Adapted from [85].CC BY 4.0.

Microfluidic devices for improved brain organoid maturation and functionality
Incorporating biomechanical cues, such as substrate stiffness, vascular shear stress, and interstitial flow, allows researchers to mimic the physiological conditions experienced by the brain and other organs [80].This enhanced mimicry enables a deeper understanding of the underlying mechanisms of brain development and provides valuable insights for the development of novel treatments for brain disorders.
A microfluidic device developed by Wang et al in 2018 enabled precise control over the differentiation and cultivation of iPSC-derived brain organoids under continuous flow conditions (figure 6(A)) [83].This device utilized a syringe pump to facilitate the even diffusion of media over the organoids embedded in Matrigel, ensuring a consistent supply of nutrients and growth factors.Comparative analysis with static Chip cultures and conventional petri dish cultures revealed that the perfused organoids exhibited notable advantages.These included larger size, enhanced maturity in terms of neural expression and structural organization, and reduced incidence of cellular death.Using a similar approach, in 2021, Cho et al designed a microfluidic device for the cultivation of cerebral organoids, utilizing human brain tissuederived ECM and harnessing gravity-driven fluid flow (figure 6(B)) [84].The findings from this research shed light on the advantages of dynamic organoid culture when compared with conventional static techniques.The authors observed a reduction in apoptotic cell death and necrotic region formation, attributed to improved distribution of oxygen and nutrients throughout the tissue leading to superior structural and functional maturation.Notably, neural and astrocyte markers were expressed as early as 45 d into the culture, a significant advancement compared to the 6 months required by standard methods.In another study conducted by Seiler et al in 2022, a multiplexed continuous flow design was employed for the differentiation and maintenance of cerebral organoids (figure 6(C)).These organoids were cultured for six days post-induction in independently perfused wells, with precise control over the flow rate.When compared to control organoids grown using conventional orbital shakers, both systems exhibited robust neural progenitor differentiation.However, results obtained from the dynamic culture conditions demonstrated significantly reduced expression of gene markers associated with cellular stress that are crucial for enhancing the viability and applicability of organoid models [85].Cumulatively, these findings highlight the critical role of fluid flow in promoting maturation within cerebral organoids.

Recent advancements in automated organoid-based drug testing
Organoids, with their ability to faithfully replicate the intricate 3D architecture and microenvironment of living organs, hold immense promise for unraveling complex biological phenomena.However, this transition towards greater complexity demands not only advanced laboratory techniques for their generation and manipulation but also a robust infrastructure for handling the wealth of data they produce.The techniques employed in the generation and manipulation of organoids surpass those of standard cell culture, and the resulting wealth of data, including microscopic readouts and transcriptomic analyses, coupled with the heightened cellular complexity of 3D organoid structures, often presents a daunting challenge in terms of data management and processing.In this dynamic landscape of organoid-based research, automation stands as a transformative force, poised to unlock the full potential of this cutting-edge field.In this section, we will review some of the recent advancements at the intersection of automation and brain organoid research.
In another study, Darrigues et al demonstrated the feasibility of conducting high-throughput invasion screening [86].They employed both a well-established cell line and 3D model systems derived from patients to screen 22 anti-invasive compounds on GBM U-251MG cell spheroids.Their findings unveiled that among these compounds, tubulin inhibitors exhibited the highest efficacy with U-251MG cells.

Modeling gut-brain axis using a multi-organ on chip approach
Given the significant role the gut plays in CNS health, in vitro multi-organ systems (MOSs) to model the gut-brain axis (GBA) have recently gained traction in the fields of neuropharmacology and brain disease modeling.MOSs, a type of MPS, are novel tools used to study the interaction between various human organ systems, like the GI tract and the brain, in vitro.The GBA is defined as the bidirectional communication pathway between the GI tract, including its commensal microbiome, and the CNS; and it is further recognized as playing a significant role in regulating homeostasis between the two systems, as well as in modulating the immune and endocrine systems [87] (figures 7(A) and (B)).Importantly, the three main communication pathways within the GBA are (1) the immune system, which primarily signals via cytokines; (2) the Vagus nerve, which directly carries neuronal signals; and (3) the neuroendocrine system, which signals via neurotransmitters and GI hormones [88].Importantly, conventional in vivo GBA models (e.g.gnotobiotic mice) have been instrumental in unveiling the GI/microbiological impact on CNS health and disease progression; however, given the known limitations of animal models and the multifactorial nature of CNS diseases, there is a widespread need for highly controllable and complex humanbased pre-clinical in vitro models, such as MOSs, to dissect the convoluted signals within the GBA.While an in-depth review of novel biofabricated gut models for the study of neurological diseases is published elsewhere [87,88].In this section of this review we focus on reviewing a few of the most recent GBA MPS models.It is important to note that there are no organoid based GBA MPSs published to date.
For instance, the MINVERVA Project was funded by the European Research Council in 2020 with the goal of elucidating the effect of the microbiota secretome on brain function, specifically in the context of A.D. Briefly, this MOS encompasses five microfluidic organs-on-a-chip (figure 7(C)), including the commensal microbiome (from A.D. patients or healthy donors) grown in a 3D collagen hydrogel, the gut, the immune system, the BBB, and the brain.Each compartment allows for both sampling and cell perfusion, and a microporous membrane separates the cells from the culture medium that flows into the next compartment.Therefore, the culture media that reaches the last compartment, the brain and glial cells, is enriched with the secretome from all the preceding compartments.This platform is the first of its kind and may be useful for detangling the complex biochemical and signaling pathways in the GBA and for identifying new therapeutic targets and interventions based on microbiota and/or dietary management.For a more in depth review of the MINERVA Project, see Raimondi et al [87].
In 2021, Kim et al developed an in vitro GBA MPS containing gut (human Caco-2 or murine bEnd.3) and human brain (primary) endothelial cells connected via microfluidic channels to recapitulate the gut-epithelial barrier and BBB [91].To stimulate an inflammatory response in this GBA, the MPS was treated with LPS, a widely used bacterial endotoxin, and the interaction between the two organ systems was observed by a decrease in gut barrier integrity and an increase of IL-8 production, thought to be generated by BBB [91] (figure 7(D)).Additionally, this prototypical GBA chip demonstrated that gut and BBB cell responses were partially correlated with other in vitro and animal models (previously reported), and that the exosome transport across the barriers was influenced by the fluidic environment.This simple in vitro GBA model offered a useful demonstration of the soluble-factor communication pathway that exists between the gut and the brain.
Finally, given the evidence for the role of the GBA specifically in neurodegenerative disease like Alzheimer's and Parkinson's [90,92].Trapecar et al generated a 3X human gut-liver-brain axis MPS that included innate and adaptive immune cells to study the multi-organ-interactions that could potentially result in phenotypic changes associated with Parkinson's disease [90] (figure 7(E)).By connecting a primary gut and liver cell MPS to an iPSCderived cerebral MPS using 'systemically circulated' common culture media containing CD4+ regulatory T cells and TH17 cells, they demonstrated that this multi-organ-system interaction enhanced invivo like features of cerebral MPSs (e.g.increased expression of neuronal, astrocyte and microglial homeostatic genes and metabolic pathways), and that microbiome-associated short-chain fatty acids increase the expression of disease-associated pathways in PD.Additionally, using hiPSCs from a donor carrying the familiar PD A53T mutation, they recreated various clinical markers of familial PD, including aggregation of α-Synuclein, changes in lipid metabolism and neuronal pathology, etc.These clinically relevant models showcase the utility of a multiorgan-on-a-chip approach to the field and feasibility for future disease modeling studies.Although only a few MPS GBA models currently exist, they are advantageous in that they offer a wider range of cell type interactions and greater anatomical complexity, including dynamic flow and sheer stress.Overall, this novel MOS approach to study the GBA is a promising avenue for the improved modeling of human physiology both in health and disease and can help advance translational research.

Recent advancements in automated organoid-based drug testing
Organoids, with their ability to faithfully replicate the intricate 3D architecture and microenvironment of living organs, hold immense promise for unraveling complex biological phenomena.However, this transition towards greater complexity demands not only advanced laboratory techniques for their generation and manipulation but also a robust infrastructure for handling the wealth of data they produce.The techniques employed in the generation and manipulation of organoids surpass those of standard cell culture, and the resulting wealth of data, including microscopic readouts and transcriptomic analyses, coupled with the heightened cellular complexity of 3D organoid structures, often presents a daunting challenge in terms of data management and processing.In this dynamic landscape of organoid-based research, automation stands as a transformative force, poised to unlock the full potential of this cutting-edge field.In this section, we will review some of the recent advancements at the intersection of automation and brain organoid research.
In another study, Darrigues et al demonstrated the feasibility of conducting high-throughput invasion screening [86].They employed both a wellestablished cell line and 3D model systems derived from patients to screen 22 anti-invasive compounds on GBM U-251MG cell spheroids.Their findings unveiled that among these compounds, tubulin inhibitors exhibited the highest efficacy with U-251MG cells.Interestingly, their results also shed light on the considerable variability in compound effectiveness among GBM organoids sourced from different patients, indicating a potential use for this technology in the field of personalized medicine.Durens et al presented a thorough workflow that leverages high-throughput screening (HTS) approaches with 3D human stem cell platforms, focusing on serum-free embryoid bodies (SFEBs) derived from human iPSCs [93].The study utilizes high-content imaging to analyze SFEBs, assessing parameters such as neurite outgrowth and cellular composition.Additionally, electrophysiological measurements employing multi-electrode arrays capture spontaneous and evoked spiking, effectively emulating cortical network aberrations.The integration of these assays enables an efficient investigation into diverse properties of 3D organoids at both network and single-cell levels, providing a solid foundation for modeling neurodevelopmental disorders.The research emphasizes the scalability of HTS, providing extensive morphological and electrophysiological data mirroring in vivo physiological processes, facilitating thorough analysis and the identification of potential drug targets.However, the authors highlight the importance of further validation, standardization, and automation to reduce experimental variability and enhance cost-effectiveness in 3D culture analysis.
While these studies collectively represent significant strides, they also emphasize the ongoing need for further validation, standardization, and automation to reduce experimental variability and enhance costeffectiveness in 3D culture analysis.Together, these advancements signify a paradigm shift in drug testing methodologies, heralding a new era of precision medicine and personalized therapeutics.

Major limitations of current human brain MPS models for biomedical research and drug testing: challenges and opportunities
Despite the significant advancements made over the last decade, current human brain MPSs face several challenges that hinder their applicability and in vivo fidelity in biomedical research and drug testing.First, brain MPSs have not yet achieved complete replication of all components of the CNS.Recent efforts to combine organoids with Organ-on-Chip technology have yielded promising results, suggesting that the integration of organoids with Organ-on-Chip platforms can mitigate some of the limitations associated with traditional static cell culture methods and accelerate or even enhance the maturation of iPSC-derived cells.This integration facilitates more accurate and physiologically relevant studies of drug penetration and delivery, ultimately enhancing our understanding of brain diseases and drug development.
However, certain limitations still remain unaddressed such as: neuronal tissue, the interstitial system, and cerebrovascular tissue.While many brain MPSs focus on the neuronal and interstitial components, they lack key NVU cells (e.g.endothelial cells and pericytes) and/or the immune cells (microglia), all of which are extremely important in physiological/pathophysiological modeling and drug toxicity screening [51,94].Importantly, the BBB within the NVU is a highly dynamic, tightly controlled barrier with physical, transport, and enzymatic properties that become dysfunctional under CNS disease and in neurotoxicity.Therefore, the inclusion of all relevant cell types and their adequate association and function alongside the neural and interstitial systems in the brain MPS is critical for the development of an organotypic brain MPS.
As evidenced by a substantial body of literature, biomechanical forces acting on endothelium and other cells can significantly impact cell maturation and function [95][96][97].Although Organon-Chips and other dynamic systems have the potential to mimic vascular shear stress and other biomechanical components of the TME, only a few groups have precisely tuned parameters such as flow rates and medium viscosity to adequately recapitulate physiologically relevant shear stress on chips [65,68,69].These strategies should be consistently adopted to better simulate the physiological shear rate that brain cells experience in our body and to determine whether mechanical forces can enhance stem cell maturation and function.
Another limitation of MPSs is the undefined or inadequate interstitium that lacks the relevant ECM proteins that support brain parenchymal and cerebrovascular basement membranes (BMs).While many MPSs attempt to incorporate the brain's interstitium through the use of ECM proteins, such as Collagen IV or Matrigel (laminin-rich gel) [98], the brain and cerebrovascular BMs in MPSs may not accurately recapitulate the in vivo brain and cerebrovascular BMs, and often lack one or more of the four major glycoprotein families (collagen IV isoforms, laminins, nidogen, and heparan sulfate proteoglycans) [99,100].These are important in ECM remodeling and tissue patterning in both health and disease, and their presence in the correct isoforms and quantities are essential for modeling disease states and studying drug-toxicity because many cellular responses, such as inflammation and oxidative stress, depend on the ECM's composition and ability to metabolize and degrade waste or injurious agents.
Within the fields of neurobiology and neuropharmacology, preclinical drug development studies aim to apply predictive, physiologically relevant brain model systems to evaluate the permeability, efficacy and safety of CNS-targeting drugs.Given its ability to prevent xenobiotics from entering the brain, the BBB is considered a major bottleneck in pharmacotherapy development, whose goal is to identify potential therapeutics or compounds that are capable of translocating across the barrier into the brain [101].In his foundational work, Pardridge noted that 100% of large molecule drugs and 98% of small molecule drugs cannot cross the BBB; for which the development of accurate modeling systems to evaluate drug transport into the brain is necessary [102].While current in vitro BBB models cannot fully recapitulate the impact of a drug's effect in vivo if they retain enough properties of in vivo physiology they can serve as good predictive models [101].One such example from 2008 used a combination of pericytes, endothelial cells and astrocytes from rats to create a triple co-culture monolayer system capable of accurately evaluating the permeability of various known drug compounds [103].High throughput drug discovery and development systems have been generated and have become commonly used in studies as a first line of screening.Recently, several research groups have showcased the feasibility of culturing either primary or iPSC-derived cells to bioengineer specific components of the BBB-on-Chip [50,65,72].These models represent innovative tools for evaluating BBB function in complex human pathologies, including patient-specific approaches.However, many of these models rely on silicon-based scaffolds, often made from materials like PDMS, which are known for their poor scalability and propensity to absorb small molecules [104].This limitation has hindered the widespread adoption of these technologies.
Finally, a significant limitation of brain MPSs (and single-organ MPSs in general), particularly in drug toxicity studies, is the lacking or inadequate incorporation of drug absorption, distribution, metabolism, and excretion (ADME) pathways, which significantly alter the delivered dose (availability), efficacy, and potency of drugs, most of which are delivered through the circulatory system and are metabolized (from pro-drugs) in the intestine and/or liver prior to reaching the BBB/brain.Some multiorgan MPSs that mimic systemic circulation have attempted to overcome this challenge (i.e.human-on-a-chip), but they have not yet been able to capture the complex ADME process [105].In the context of brain disease and toxicity studies, dynamic organ crosstalk can only be obtained through the use of interconnected MPSs that closely resemble human physiology and enable researchers to investigate pharmacokinetics and efficacy/toxicity [106].Nevertheless, these models do not consistently replicate the physiological shear stress experienced by the human brain.The primary obstacle to integrating MPS technologies into the existing industry pipeline is the lack of validation against a standardized library of compounds and established biomarkers [107].
In this direction, the integration of automation and 3D brain organoids marks a transformative leap in clinical drug testing.Automation, with its capability for real-time live imaging and simultaneous fluorescent marker detection, not only offers invaluable insights into organoids and spheroids but also enables comprehensive toxicity screening.This empowers researchers to conduct thorough evaluations of potential risks associated with substances, a crucial step in drug development.By harnessing the power of automation, the field is poised to establish a new generation of cellular 3D in vitro assays and personalized medicine approaches.These models, providing unbiased, quantitative, and highthroughput access to human tissue surrogates, hold immense potential for advancing research and applications in this domain.Moreover, when combined with artificial intelligence, automated organoid culture processes have the capacity to revolutionize disease modeling and significantly streamline drug discovery for clinical trials.This integration promises to yield valuable insights and accelerate progress in the quest for more effective therapies [108][109][110][111].This integration promises to yield valuable insights and accelerate progress in the quest for more effective therapies [109][110][111].Another branch of bioengineering research is currently and rapidly moving toward the integration of sensors that are capable of real-time monitoring of essential parameters, such as trans-endothelial electrical resistance, within a scalable microfluidic platform [112,113].These platforms have the potential of accelerating the use of MPS across a variety of applications while expediting and standardizing the readouts.
In conclusion, the collective efforts of bioengineers in automating organoid culture, standardizing analytical endpoints, and integrating real-time monitoring technologies represent a significant leap forward in brain organoid research.These advancements not only enhance the reliability and reproducibility of organoid models but also hold the promise of revolutionizing drug testing and personalized medicine approaches for neurological disorders.This dynamic intersection of engineering and neuroscience is poised to shape the future of biomedical research and therapeutic development.

Figure 1 .
Figure 1.Overview of key topics in this review paper.

Figure 2 .
Figure 2. Comparison of experimental strategies for brain modeling.

Figure 4 .
Figure 4. Examples of common Transwell insert in vitro methods by (A) modeling fluid shear stress with a channel insert; (B) mimicking the flow dynamics of circulating natural killer (NK) cells.Adapted from [59].CC BY 4.0.Adapted from [61].CC BY 4.0.

Figure 5 .
Figure 5. Vascularized physiological brain models in vitro.Main structural and functional characteristics of vascaulrized models of brain described in this review.Adapted from[58].CC BY 4.0.Adapted from[69].CC BY 4.0.Reproduced from[75] with permission from the Royal Society of Chemistry.

Figure 6 .
Figure 6.Examples of microfluidic devices for brain organoid maturation and functionality by (A) showcasing how organoids subjected to continuous fluid flow improves neural regional identities (TBR1, CTIP2); (B) microfluidic devices can increase deep layer cortical development in organoids compared to other matrix types; (C) The fluid flow shear stress caused an increase in neural differentiation and decreased ER stress in cerebral organoids compared to cultures on an orbital shaker.Reproduced from[83] with permission from the Royal Society of Chemistry.Adapted from[84].CC BY 4.0.Adapted from[85].CC BY 4.0.

Figure 7 .
Figure 7. Modeling gut-brain axis.(A) Diagram of the anatomical structures of the gut-brain axis (GBA), including the gut microbiome.The brain and gut communicate bi-directionally via the Vagus nerve, the immune system, and the neuroendocrine system.(B) Diagram of the micro-anatomical structures of the GBA: (bottom) the microbiota in the intestinal lumen and mucosal layer interacts with epithelial cells (e.g.enterocytes) of the intestinal brush border, vascularized by capillaries that carry red blood cells and lymphocytes traveling via systemic circulation to reach the blood-brain barrier, which protects and communicates with the neurons, pericytes, astrocytes, microglia, and oligodendrocytes within the brain.(C) Diagram of the MINERVA Multi-Organ Chip, which connects multiple MPSs including (bottom) the microbiota-on-a-chip, the gut-on-a-chip (i.e.enterocytes), the immune system-on-a-chip (i.e.macrophages and lymphocytes), followed by the blood-brain barrier-on-a-chip (i.e.cerebrovascular endothelial cells), and finally (top) the brain-on-a-chip (with neurons, microglia, and astrocytes seeded in a hydrogel).Adapted from[87].CC BY 4.0.(D) Example of a GBA MPS using Caco-2 and hBMECs co-culture showing modulation of gut barrier permeability (IL-8 expression) and exosome transport across the BBB when subjected to fluid flow.Reprinted from [89], Copyright (2021), with permission from © 2021 The Korean Society of Industrial and Engineering Chemistry.Published by Elsevier B.V. All rights reserved.(E) Example of gut-liver-brain axis MPS for PD that can model the interactions between the three organs and maintenance of disease phenotypes (IBA2, S100B, Tuj1).From [90].Reprinted with permission from AAAS.