Advances in Cortical Organoid Technology: From Microphysiological Systems to Biological Intelligence

Abstract

The capability to model the human brain in vitro has undergone a revolutionary transformation over the past decade. Driven by the convergence of stem cell biology, bioengineering, and computational neuroscience, the field has moved beyond simple monolayer cultures to complex, self-organizing three-dimensional tissues known as cortical organoids. These "mini-brains" offer a unique window into the cryptic processes of human neurodevelopment, allowing for the observation of cytoarchitectural formation, cellular migration, and network assembly that were previously inaccessible due to the limitations of animal models and the scarcity of fetal tissue. This comprehensive report explores the state-of-the-art advancements in cortical organoid technology as of 2024-2025. It provides a detailed examination of the engineering solutions developed to address critical physiological constraints, specifically vascularization and necrotic core formation, through the application of millifluidic and organ-on-a-chip systems. It further analyzes the emergence of "assembloids"—modular platforms that reconstruct long-range neural circuits to model inter-regional migration and complex disease phenotypes, including Timothy Syndrome and Parkinson's disease. Additionally, the report investigates the frontier of "Organoid Intelligence" (OI), scrutinizing the electrophysiological maturation of these tissues and their capacity for goal-directed behavior as demonstrated by the DishBrain system. Finally, it offers a rigorous evaluation of the ethical, legal, and social implications (ELSI) arising from the potential for consciousness and the moral status of bioengineered human neural systems.

1. Introduction: The Dimensional Shift in Neuroscience

For the vast majority of modern neuroscience history, our understanding of the human brain has been constrained by the tools available to study it. The intricate dance of neurodevelopment—where billions of neurons migrate, differentiate, and wire themselves into the most complex structure in the known universe—takes place shielded within the womb. Consequently, researchers have traditionally relied on two primary proxies: two-dimensional (2D) cell cultures and animal models. While foundational, both possess inherent limitations that have stifled the translation of basic science into clinical therapy.

Two-dimensional cultures, typically grown as monolayers of cells on rigid plastic substrates, fail to recapitulate the rich three-dimensional microenvironment of the native brain. In the developing cortex, cells do not exist in isolation on a flat plane; they are embedded within a soft, viscoelastic extracellular matrix (ECM), surrounded by a diverse milieu of other cell types, and subjected to complex mechanical forces and chemical gradients. The loss of this spatial context in 2D cultures alters cellular morphology, gene expression, and intercellular signaling, rendering them poor predictors of in vivo behavior.¹

Animal models, particularly rodents, have served as the workhorses of neuroscience, offering the systemic complexity that cell cultures lack. However, the evolutionary divergence between rodents and primates is profound, particularly in the cerebral cortex. The human brain possesses unique features absent or distinct in mice, such as the expanded outer subventricular zone (oSVZ)—a massive germinal layer responsible for the explosive expansion of cortical surface area—and the characteristic folding (gyrification) of the cortical sheet. Furthermore, the genetic background of human neurological disorders, which often involves subtle mutations in non-coding regulatory regions, is difficult to fully replicate in murine genomes.²

The paradigm shifted with the advent of human induced pluripotent stem cell (hiPSC) technology and the subsequent development of cerebral organoids. First described in seminal papers in the early 2010s, these 3D tissues are derived from pluripotent stem cells that, when provided with minimal cues in a suspension culture, self-organize into structures remarkably similar to the developing human brain. They generate defined cortical layers, distinct progenitor zones, and diverse neuronal subtypes.

Yet, as the field matures, the initial euphoria of "growing brains in a dish" has given way to a pragmatic recognition of the technology's limitations. Standard organoids often lack a vascular system, leading to the formation of a necrotic core that limits growth and health. They are frequently devoid of non-neuronal cells, such as microglia and endothelial cells, which are crucial for homeostasis and disease pathology. Moreover, single-region organoids cannot model the long-range connectivity between different brain areas that underlies complex behavior.

The current era of research, spanning 2024 and 2025, is defined by the integration of bioengineering to solve these biological deficits. We are witnessing the rise of "next-generation" organoids: vascularized, immune-competent, and computationally active. This report synthesizes the vast body of recent literature to provide a definitive account of these technological leaps, exploring how they are reshaping our understanding of the human brain in health and disease.

2. Bioengineering the Microenvironment: Vascularization and Necrotic Core Mitigation

One of the most significant biophysical barriers in organoid technology is the diffusion limit of oxygen and nutrients. In the developing human embryo, the vascular system develops concomitantly with the nervous system, ensuring that as the tissue grows, every cell remains within a few hundred micrometers of a capillary. Standard organoid protocols, however, rely on passive diffusion from the culture medium. As the organoid expands beyond approximately 400-500 micrometers in diameter, the diffusion distance becomes too great for oxygen to reach the center.

2.1 The Necrotic Core Phenomenon

The consequence of this diffusion limitation is the formation of a necrotic core—a central region of hypoxia and cell death. This is not merely a structural flaw; it is a metabolic disaster for the tissue. The dying cells in the core release toxic cellular debris, acidic byproducts, and pro-inflammatory signals that diffuse outward, compromising the health of the surviving cortical layers. This "bystander effect" can alter the transcriptional profile of the healthy neurons, inducing stress pathways that may confound disease modeling data.³

Furthermore, the necrotic core imposes a "fetal ceiling" on development. Because the tissue cannot grow larger or support high metabolic demands, organoids typically arrest at a developmental stage equivalent to the second trimester of fetal life. They fail to reach the size or maturity required to model late-stage neurodevelopment, postnatal maturation, or age-related neurodegeneration. Addressing this limitation has become a primary focus of bioengineering efforts.

2.2 Millifluidic Systems and Controlled Oxygenation

To mitigate the necrotic core without introducing complex vasculature, researchers have turned to fluidic technologies that mimic the dynamic environment of the body. While microfluidic "organ-on-a-chip" systems have been popular, they often utilize micron-scale channels that can be easily clogged by the cellular debris shed by growing organoids or restrict the physical expansion of the tissue.

A notable advancement in 2024-2025 is the adoption of millifluidic systems, such as the QuasiVivo900 platform. These devices operate on a slightly larger scale than microfluidics, utilizing millimeter-sized channels and chambers that accommodate the organoid's growth while maintaining precise control over fluid dynamics.³

Mechanism of Action:

The superiority of millifluidic systems over standard orbital shaking cultures lies in their ability to generate controlled laminar flow. In a standard dish, agitation is random and often turbulent, creating inconsistent shear forces. In a millifluidic chip, the culture medium flows continuously and unidirectionally over the organoid.

• Shear Stress: The controlled flow applies a physiological level of shear stress to the surface of the organoid. This mechanical cue is known to influence stem cell differentiation and tissue organization.

• Oxygen Availability: Computational Fluid Dynamics (CFD) simulations have revealed that the continuous replenishment of medium in a fluidic loop significantly increases the dissolved oxygen concentration available at the organoid surface compared to static or shaking conditions. This steeper oxygen gradient drives deeper diffusion into the tissue.³

Empirical Results:

The implementation of these systems has yielded measurable biological improvements. Studies comparing human midbrain organoids (hMOs) cultured in millifluidic chips versus standard plates demonstrated a statistically significant reduction in the size of the necrotic core in the fluidic group.3 Beyond survival, the fluidic environment appeared to enhance cellular identity; dopaminergic neuronal differentiation was more pronounced, and the metabolic maturation of the cells was accelerated. This suggests that the "hemodynamic-like" forces provided by the flow are not just supportive but instructive for neurodevelopment.7

2.3 Strategies for Vascularization

While fluidic chips improve external exchange, the ultimate solution to the diffusion limit is internal vascularization—the creation of a functional blood vessel network within the organoid. Several distinct bioengineering strategies are currently being pursued to achieve this hierarchy of perfusion.⁴

2.3.1 Co-culture and Cellular Mixing

The most straightforward approach involves introducing endothelial cells (ECs)—the building blocks of blood vessels—into the organoid system. This can be achieved by mixing hiPSCs with Human Umbilical Vein Endothelial Cells (HUVECs) or iPSC-derived endothelial progenitors during the initial aggregation phase.⁹

• Outcome: These protocols successfully generate organoids containing clusters of vascular cells. In some cases, these cells self-assemble into primitive tubular structures resembling capillaries.

• Limitations: A major hurdle is connectivity. The vascular networks formed via simple mixing often appear as isolated islands rather than a continuous, interconnected tree. They lack a functional inlet and outlet, meaning they cannot be perfused. Furthermore, without the supporting mural cells (pericytes and smooth muscle cells) and the appropriate basement membrane, these vessels are often leaky and immature.¹⁰

2.3.2 Fusion with Vascular Organoids

A more modular approach involves generating dedicated "blood vessel organoids"—structures composed entirely of vascular lineages—and fusing them with cerebral organoids.

• Mechanism: The vascular organoids are differentiated to a stage where they contain a rich network of ECs and pericytes. When placed in contact with a neural organoid, the vascular cells exploit their intrinsic invasive properties to migrate into the neural tissue.⁹

• Advantages: This method ensures that the vascular cells are properly specified before entering the brain environment, potentially reducing the risk of them differentiating into unwanted cell types.

2.3.3 Organ-on-a-Chip and Bio-printing

Advanced biofabrication techniques offer the highest level of control. 3D bioprinting allows researchers to deposit endothelial cells in precise patterns within a hydrogel scaffold, creating a pre-defined vascular geometry that can be perfused.⁹ Similarly, organ-on-a-chip platforms can be engineered with microchannels lined by endothelial cells. When a brain organoid is placed within this chip, the endothelial cells from the channels can sprout (angiogenesis) into the organoid, mimicking the way blood vessels invade the neural tube during embryonic development.¹¹

These engineered platforms allow for the study of the Blood-Brain Barrier (BBB). By co-culturing the organoids with endothelial cells under flow, researchers can recreate the tight junctions and selective permeability that characterize the human BBB, providing a critical tool for testing drug delivery into the central nervous system.¹¹

Table 1: Comparative Analysis of Organoid Vascularization and Perfusion Techniques

Technique	Method Description	Key Outcome	Physiological Relevance	Reference
Millifluidic Chip	Culturing organoids in millimeter-scale chambers with controlled laminar fluid flow (e.g., QuasiVivo900).	Increased interstitial O_2 concentration; Significant reduction in necrotic core size; Enhanced dopaminergic differentiation.	Moderate; mimics interstitial fluid dynamics but lacks vessels.	3
In Vivo Grafting	Transplanting the organoid into the cortex of an immunodeficient host (e.g., mouse).	Rapid invasion by host vasculature; Full systemic perfusion; Elimination of necrotic core; Enhanced neuronal size and complexity.	High; provides true blood supply and systemic factors.	12
Co-culture/Mixing	Mixing iPSCs with HUVECs or iPSC-derived ECs during the aggregation phase.	Formation of vascular clusters and primitive tubes; often lack functional lumen or connectivity.	Low to Moderate; establishes cell-cell contact but rarely perfusion.	9
Organ-on-a-Chip	Placing organoids in microfluidic chips lined with endothelial cells.	Improved viability; Facilitates angiogenesis from the chip into the tissue; Mimics the BBB interface.	Moderate to High; allows for controlled perfusion and drug testing.	11
Bio-printing	Using 3D printers to pattern vascular channels within the organoid or scaffold.	High control over vascular geometry; Enables immediate perfusion through printed channels.	High (Structural); technically demanding.	9

3. Completing the Cellular Ecosystem: Non-Neuronal Integration

The brain is not merely a collection of neurons. It is a complex ecosystem where glial cells, immune cells, and vascular cells outnumber neurons in many regions and are indispensable for function. However, the classic "Lancaster protocol" and its derivatives for generating brain organoids rely on neuroectodermal induction. This signaling pushes stem cells to become neural, effectively suppressing the mesodermal and endodermal lineages.

The consequence is that standard brain organoids are "immune-privileged" in an unnatural way: they lack microglia. Microglia, the resident macrophages of the brain, do not arise from the neural tube. Instead, they originate from the yolk sac mesoderm during primitive hematopoiesis and migrate into the developing brain at a very early embryonic stage.¹ Because induced pluripotent stem cells (iPSCs) in neuroectoderm protocols do not pass through a yolk sac-like stage, microglia are absent unless specifically introduced.

3.1 The Microglia Imperative

The absence of microglia is a critical deficiency for modeling both development and disease. Microglia are not just passive defenders; they are active sculptors of the neural landscape.

1. Synaptic Pruning: During development, neurons form an excess of synapses. Microglia identify and "eat" (phagocytose) weak or inappropriate connections, a process called pruning. This refinement is essential for the formation of precise neural circuits. Without microglia, organoids may exhibit hyper-connectivity or immature synaptic profiles.¹⁵

2. Debris Clearance: In the context of the necrotic core or general cellular turnover, microglia are the cleanup crew. They engulf apoptotic cells and cellular debris. In their absence, this waste accumulates, creating a toxic environment.¹⁶

3. Trophic Support: Microglia secrete cytokines and growth factors (e.g., IGF-1, BDNF) that support neuronal survival, axonal growth, and oligodendrocyte differentiation.¹⁶

3.2 Methodologies for Integration

To bridge this gap, researchers have developed three primary strategies to introduce immune competence into organoid models.¹⁶

3.2.1 The Addition Approach

This method involves generating microglia (or microglial progenitors) separately from iPSCs using a dedicated hematopoietic differentiation protocol. Once these cells are available, they are added to an already established brain organoid.

• Advantages: This mimics the developmental timing of microglia invading the brain. It also allows for "chimeric" experiments—for example, adding microglia carrying an Alzheimer's risk gene (like TREM2 or APOE4) to a "healthy" brain organoid to isolate the specific contribution of the immune cells to the disease pathology.

• Challenges: The process of handling and transferring microglia can activate them, potentially causing them to adopt an inflammatory state before they even enter the tissue.

3.2.2 The Co-culture Approach

In this strategy, neural progenitor cells and macrophage/microglia progenitors are mixed together at the very beginning of organoid generation (Day 0).

• Advantages: This ensures that the two cell types interact throughout the entire developmental trajectory.

• Challenges: Finding a single culture medium that supports the optimal growth and differentiation of both neuroectodermal and mesodermal lineages can be difficult.

3.2.3 The Spontaneous (Innate) Approach

Some recent protocols have modified the initial induction steps to be less restrictive. By omitting specific inhibitors (like dual-SMAD inhibition) or altering the extracellular matrix (e.g., specific Matrigel formulations), researchers can allow a small population of mesodermal cells to co-differentiate alongside the neural tissue. These cells naturally acquire a microglial identity.¹⁶

• Advantages: This is the least labor-intensive method and eliminates the need for separate differentiation protocols.

• Challenges: It suffers from high variability. The number and distribution of microglia can vary significantly between batches, making standardization difficult.

3.3 Functional Impact of Integration

The successful integration of microglia transforms the organoid. Studies utilizing human midbrain organoids demonstrated that the incorporation of iPSC-derived microglia (resulting in ~6% IBA1+ cell population) had profound effects.¹⁵ These cells adopted a ramified morphology characteristic of resting microglia and actively surveyed the tissue. Most importantly, the presence of microglia significantly reduced the number of apoptotic cells and the amount of cellular debris within the organoid, effectively "cleaning" the tissue.¹⁶ Furthermore, these immune-competent organoids exhibited enhanced neuronal maturation and more synchronized electrical activity, confirming that microglia provide essential cues for functional circuit formation.¹⁷

4. Assembloids: Reconstructing Neural Circuits and Migration

The human brain is characterized by long-range connectivity. Neurons in the cortex project axons to the spinal cord to control movement; neurons in the thalamus project to the cortex to relay sensory information. A single, spherical organoid, which typically represents a single brain region (e.g., dorsal forebrain), cannot model these inter-regional interactions.

To overcome this, scientists have developed "Assembloids." The concept is modular: generate organoids representing different brain regions separately, then fuse them together. The cells at the interface fuse, and axons from one organoid grow into the other, establishing functional circuits.¹⁸

4.1 The Forebrain Assembloid and Interneuron Migration

One of the most powerful applications of assembloids is modeling the migration of inhibitory interneurons. In the developing brain, excitatory glutamatergic neurons are born in the dorsal forebrain (pallium), while inhibitory GABAergic interneurons are born in the ventral forebrain (subpallium). To form a balanced circuit, the GABAergic neurons must migrate a long distance tangentially to populate the cortex.

By fusing a dorsal forebrain organoid (resembling the cortex) with a ventral forebrain organoid (resembling the ganglionic eminence), researchers can recreate this journey in a dish.

• Saltatory Migration: Live imaging of these assembloids reveals that human interneurons migrate in a "saltatory" (jumping) fashion. They extend a leading process, the nucleus swells and translocates forward, and then they pause before repeating the cycle.¹⁸

• Human Specificity: Crucially, assembloid studies have shown that human interneurons migrate more slowly and with more complex branching patterns than those in rodents, highlighting the importance of human-specific models.¹⁸

• Circuit Integration: Once the interneurons reach the dorsal side, they functionally integrate. Calcium imaging shows that they form inhibitory synapses onto the excitatory neurons, establishing the Excitation/Inhibition (E/I) balance that is critical for preventing seizures and processing information.²⁰

4.2 Cortico-Motor Assembloids: The Output Circuit

To model the brain's control over the body, researchers have engineered "cortico-spinal-muscular" assembloids—a triple-fusion system.

1. Cortical Organoid: Contains upper motor neurons.

2. Spinal Cord Organoid: Contains lower motor neurons.

3. Skeletal Muscle Spheroid: Contains myotubes (muscle fibers).

When these three components are fused, they self-assemble into a functional motor circuit. Axons from the cortical neurons grow into the spinal organoid and synapse onto spinal motor neurons. The spinal motor neurons, in turn, project axons out to the muscle spheroid.²¹

Validation via Optogenetics:

The functionality of this circuit has been elegantly demonstrated using optogenetics, a technique where neurons are genetically modified to express light-sensitive ion channels.

• Experiment: Researchers shined blue light on the cortical organoid to activate the upper motor neurons.

• Result: The stimulation traveled down the circuit—from cortex to spinal cord to muscle—causing the muscle tissue to visibly twitch (contract).²²

• Significance: This provides a complete in vitro model of the human motor unit, allowing for the study of diseases like Amyotrophic Lateral Sclerosis (ALS) where this specific pathway degenerates.

5. Modeling Neurological Disorders: From Virology to Neurodegeneration

The enhanced physiological relevance of vascularized, immune-competent, and multi-regional organoids has made them indispensable tools for understanding human disease. They bridge the gap between genetic association and pathological mechanism.

5.1 Zika Virus and Microcephaly: A Case Study in Virology

The 2015-2016 Zika virus (ZIKV) outbreak in the Americas presented a medical mystery: a surge in babies born with microcephaly (abnormally small heads) in regions with high ZIKV transmission. While epidemiological data suggested a link, proving causality and understanding the mechanism was urgent. Brain organoids were pivotal in this effort.

Researchers exposed human brain organoids to the Zika virus and observed a striking phenotype: the infected organoids were significantly smaller than controls, mimicking microcephaly in a dish.²⁴

• Mechanism of Action: Through single-cell RNA sequencing (scRNA-seq), scientists identified that the virus specifically targets neural progenitor cells (NPCs), particularly the radial glia that act as the stem cells for the cortex.

• The Receptor: The study identified AXL, a receptor tyrosine kinase highly expressed on these progenitors, as the primary entry gate for the virus.²⁵

• Pathology: Once inside, the virus hijacks the cell's machinery, causing cell cycle arrest and inducing massive apoptosis (cell death) via the activation of Caspase-3.²⁶ This depletion of the progenitor pool means there are fewer cells available to generate neurons, leading to a smaller brain.²⁷ This rapid deployment of organoid technology to solve a global health crisis demonstrated the platform's agility and relevance.

5.2 Timothy Syndrome: Unlocking the Mechanism of Autism

Timothy Syndrome (TS) is a rare, severe neurodevelopmental disorder caused by a specific mutation in the CACNA1C gene, which encodes the CaV1.2 calcium channel. Patients often present with autism spectrum disorder (ASD) and cardiac arrhythmias.

Using forebrain assembloids derived from TS patient iPSCs, researchers uncovered a specific defect in neuronal migration.¹⁸

• The Defect: In TS assembloids, the interneurons migrated inefficiently. Their saltatory movements were more frequent but shorter in distance, resulting in delayed cortical population.

• Molecular Mechanism: The mutation causes the calcium channel to remain open too easily or for too long, leading to chronically elevated intracellular calcium levels. This calcium overload disrupts the cytoskeletal machinery (actin and myosin) required for the smooth translocation of the nucleus during migration.²⁹

• Therapeutic Proof-of-Concept: Remarkably, the researchers treated the assembloids with an antisense oligonucleotide (ASO) designed to modulate the splicing of the calcium channel gene. This treatment corrected the calcium defect and restored normal migration patterns, serving as a powerful proof-of-concept for gene therapies in neurodevelopmental disorders.¹⁸

5.3 Alzheimer’s Disease: Modeling the Sporadic Majority

Alzheimer’s Disease (AD) has been notoriously difficult to model in mice. Rodents do not naturally develop plaques and tangles; they must be genetically engineered with aggressive mutations found in familial AD (fAD). However, 95% of human cases are sporadic (sAD), with no single causative mutation.

Human organoids offer a platform to model sAD.

• Inducing Pathology: Since sAD organoids do not always develop pathology spontaneously, researchers have developed methods to "age" them or expose them to pathological cues. One effective method is exposing the organoids to serum or brain extracts from AD patients.

• Results: These treated organoids develop the classic hallmarks of AD: Amyloid-beta (Aβ) plaques, Tau neurofibrillary tangles, and synaptic loss.³⁰

• The Role of Microglia: Crucially, recent studies have shown that the presence of microglia is essential to recapitulate the full disease phenotype. In AD organoids containing microglia, the immune cells engulf Aβ but also aggressively prune synapses, driving the cognitive decline associated with the disease. This confirms that neuroinflammation is a driver, not just a bystander, of AD pathology.

• Drug Screening: These models are now being used to test drugs. For example, treatment with Lecanemab, an FDA-approved antibody targeting amyloid, significantly reduced the amyloid burden in these organoid models, validating their predictive power for clinical outcomes.³⁰

5.4 Parkinson’s Disease: The Alpha-Synuclein Loop

In Parkinson’s Disease (PD), the focus is on the dopaminergic neurons of the midbrain. Organoid models have helped elucidate the relationship between alpha-synuclein (α-Syn) aggregation and dopamine metabolism.

• The Toxic Loop: Research suggests a vicious cycle: α-Syn accumulation inhibits the enzymes responsible for dopamine synthesis and transport (VMAT2 and DAT). Conversely, dysregulated cytosolic dopamine promotes the formation of toxic α-Syn oligomers.³¹

• Microglial Involvement: Assembloid studies integrating microglia have shown that immune cells from PD patients (carrying specific mutations or risk factors) can induce α-Syn pathology in healthy neurons. This suggests that in some cases, the immune system may be the primary instigator of the neurodegenerative process.¹⁷

6. Functional Maturation and Organoid Intelligence (OI)

Perhaps the most provocative frontier in organoid research is the investigation of their functional capabilities. Can a piece of tissue in a dish process information? Can it learn?

6.1 Electrophysiological Maturation

As organoids mature in culture, their electrical activity evolves. Early-stage organoids show sparse, random spiking. However, over months of culture (up to 10 months), they develop synchronized "network bursts."

• Oscillations: mature organoids exhibit complex oscillatory activity, including Delta waves (1-4 Hz) and Gamma waves (100-400 Hz). These oscillations are critical for information processing in the intact brain.³³

• Preterm EEG: Analyses of Local Field Potentials (LFPs) from these organoids have revealed spectral features that are strikingly similar to the EEG patterns of preterm infants (25-35 weeks gestation). They exhibit "trace discontinu"—periods of silence punctuated by high-amplitude bursts—suggesting that the organoid is following an intrinsic genetic program of functional maturation.³⁴

6.2 DishBrain and the Game of Pong

In a landmark study that captured global attention, researchers at Cortical Labs developed a system called "DishBrain." They cultured human neural cells on a high-density Multi-Electrode Array (MEA)—a chip capable of both recording neural activity and delivering electrical stimulation.

They integrated this biological hardware with a computer simulation of the arcade game Pong.

• The Interface: The position of the ball was encoded as electrical pulses delivered to specific electrodes (sensory input). The firing rate of neurons in a different region of the chip controlled the position of the paddle (motor output).

• The Learning: Remarkable, within five minutes of real-time gameplay, the neuronal culture adapted its firing patterns to move the paddle and hit the ball. It exhibited goal-directed behavior, improving its performance over time.³⁵

6.3 Mechanism: The Free Energy Principle

How does a culture of cells "know" how to play a video game? The learning mechanism is hypothesized to be distinct from the reinforcement learning typically used in AI (which relies on reward functions). Instead, it is based on the Free Energy Principle (FEP) and Active Inference, theories proposed by neuroscientist Karl Friston.

• Minimizing Surprise: The theory posits that biological systems strive to minimize "free energy," which can be equated to "surprise" or unpredictability.

• The Experiment: In the Pong simulation, the system was programmed such that missing the ball resulted in a chaotic, unpredictable electrical stimulus (high surprise). Hitting the ball resulted in a predictable, consistent stimulus (low surprise).

• Conclusion: The neurons rewired their connections not to get a "reward," but to stabilize their world. They learned to hit the ball to avoid the chaos of the "miss" signal. This suggests that intelligence may be an emergent property of any self-organizing neural system coupled to an environment.³⁷

This field, now termed Organoid Intelligence (OI), aims to harness this biological computing power. Biological neurons are vastly more energy-efficient than silicon processors and excel at parallel processing and learning from sparse data. The vision is to create "biocomputers" that can solve problems intractable for traditional AI.⁴⁰

7. In Vivo Transplantation: The Ultimate Bioreactor

Despite the advances in fluidics and bioreactors, no in vitro system can yet match the complexity of the living brain. To push organoid maturation to the limit, researchers have turned to transplantation.

7.1 Integration and Vascularization

When human cortical organoids are transplanted into the cortex of immunodeficient mice, the results are transformative.

• Vascularization: The host (mouse) blood vessels rapidly invade the human graft. Within weeks, the organoid is perfused by a functional capillary network connected to the host circulation. This effectively eliminates the necrotic core and supports the survival of the graft for months.¹²

• Maturation: The vascularized graft grows significantly larger and develops more complex neuronal morphologies than its in vitro counterparts. Glial cells, which are often stunted in culture, mature fully in the transplant environment.¹²

7.2 Functional Connectivity and Sensory Response

The integration is not just structural; it is functional.

• Synaptic Integration: Viral tracing experiments show that host mouse neurons project axons into the human graft and form synapses. Conversely, human axons project out into the mouse brain, reaching distant targets like the thalamus and spinal cord.⁴³

• Sensory Processing: In a definitive demonstration of integration, researchers placed transparent graphene microelectrodes over the transplanted organoid. When they presented visual stimuli (flashing lights) to the mouse, they recorded evoked potentials within the human organoid.¹³ This proves that the human tissue had integrated into the mouse's visual processing circuit and was actively processing sensory information from the host's eyes.

These transplantation studies serve two purposes:

1. Modeling: They provide the most physiologically relevant model of human neural tissue currently possible.

2. Therapy: They act as a proof-of-concept for cell replacement therapies. If an organoid can integrate and function in a mouse brain, it could theoretically be used to repair damaged cortex in human patients following a stroke or traumatic brain injury.⁴²

8. Ethical, Legal, and Social Implications (ELSI)

The rapid progression from "blobs of tissue" to "learning brain models" capable of processing sensory input has precipitated a robust ethical debate. As the biological fidelity of these models increases, so does the moral weight of their use.

8.1 Consciousness and Moral Status

The DishBrain experiment and the observation of complex brain wave oscillations raise an uncomfortable question: Are these organoids conscious?

Current Consensus:

Most neuroscientists and ethicists agree that current organoids are not conscious. They lack the necessary sensory organs, the complex internal architecture (e.g., a fully developed thalamocortical loop), and the subjective experience (qualia) that define sentience.44 The "intelligence" displayed in Pong is viewed as local adaptive computation—a physical property of the network—rather than conscious awareness.

The Precautionary Principle:

However, the line is becoming blurred. As we engineer organoids with vascularization (supporting larger size), microglia (supporting complex circuitry), and sensory input (via OI or transplantation), we are systematically removing the barriers that prevent higher cognition. Some ethicists argue for a precautionary approach. If there is even a remote possibility that a mature, vascularized organoid could experience something akin to pain or distress (e.g., via "nociceptive" signaling during necrosis), it may deserve a moral status distinct from a skin biopsy. This might not be full "personhood," but a status similar to that of laboratory animals, requiring oversight by committees similar to IACUC (Institutional Animal Care and Use Committee).45

8.2 Chimeras and Humanization

Transplantation studies create human-animal chimeras. While the goal is to study the graft, there is a theoretical concern about "humanizing" the host animal. Could a mouse with a large chunk of human cortex possess enhanced cognition? While current evidence suggests the graft adopts the host's neural rhythms rather than imposing human intelligence, this remains a sensitive area. Guidelines currently restrict the breeding of such chimeras to prevent the germline transmission of human traits.⁴⁰

8.3 Consent and Data Privacy

With the advent of Organoid Intelligence, new privacy concerns emerge. If an organoid is derived from a donor's iPSCs and is used for computation or memory storage, does that organoid contain proprietary or sensitive information? There is a movement towards "embedded ethics," where ethicists are integrated directly into research labs to navigate these issues in real-time. Informed consent forms for donors are being revised to explicitly mention the potential creation of "neural systems" or "biological computing entities" to ensure donors understand the unique trajectory of their cells.⁴⁰

9. Conclusion and Future Outlook

The field of cortical organoids has matured from a novelty of developmental biology into a robust platform for engineering, disease modeling, and computing. The years 2024 and 2025 have been defined by the successful integration of non-neuronal cells (microglia, vasculature) and the application of fluidic systems to overcome physical constraints. The rise of assembloids has unlocked the study of complex circuits and migration-dependent disorders like Timothy Syndrome, while the DishBrain system has opened the door to biological computing.

Moving forward, the convergence of these technologies—creating vascularized, immune-competent, multi-regional assembloids trained on active inference tasks—promises to bridge the gap between the petri dish and the patient. While the creation of a fully sentient "brain in a jar" remains in the realm of science fiction, the creation of high-fidelity human neural systems is now a scientific reality. This technology holds the promise of decoding the deepest mysteries of the human mind and curing its most devastating afflictions, provided we navigate the ethical landscape with the same rigor we apply to the biology.

Works cited

1. All Together Now: Modeling the Interaction of Neural With Non-neural Systems Using Organoid Models - PubMed Central, accessed January 5, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC6598414/

2. Cerebral Organoids in Neurodevelopment, Neurodegeneration and Virology Research, accessed January 5, 2026, https://www.tocris.com/resources/tocris-blog/cerebral-organoids-neurodevelopment-neurodegeneration-virology-research

3. Organoids-on-a-chip: microfluidic technology enables culture of organoids with enhanced tissue function and potential for disease modeling - Frontiers, accessed January 5, 2026, https://www.frontiersin.org/journals/bioengineering-and-biotechnology/articles/10.3389/fbioe.2025.1515340/full

4. (PDF) Recent advancements and future requirements in vascularization of cortical organoids - ResearchGate, accessed January 5, 2026, https://www.researchgate.net/publication/365090701_Recent_advancements_and_future_requirements_in_vascularization_of_cortical_organoids

5. Human Brain Organoids-on-Chip: Advances, Challenges, and Perspectives for Preclinical Applications - PubMed Central, accessed January 5, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC9699341/

6. Organoids-on-a-chip: microfluidic technology enables culture of organoids with enhanced tissue function and potential for disease modeling - PubMed Central, accessed January 5, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC11933005/

7. Millifluidic culture improves human midbrain organoid vitality and differentiation - Lab on a Chip (RSC Publishing) DOI:10.1039/C8LC00206A, accessed January 5, 2026, https://pubs.rsc.org/en/content/articlehtml/2018/lc/c8lc00206a

8. Reduced "dead core" size in hMO cultured in the millifluidic device.... - ResearchGate, accessed January 5, 2026, https://www.researchgate.net/figure/Reduced-dead-core-size-in-hMO-cultured-in-the-millifluidic-device-Ai-Experimental_fig1_327587480

9. Cerebral Organoids with Integrated Endothelial Networks Emulate the Neurovascular Unit and Mitigate Core Necrosis | bioRxiv, accessed January 5, 2026, https://www.biorxiv.org/content/10.1101/2025.04.23.650161v1.full-text

10. Recent advancements and future requirements in vascularization of ..., accessed January 5, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC9669755/

11. Brain Organoid Transplantation: A Comprehensive Guide to the Latest Advances and Practical Applications—A Systematic Review - MDPI, accessed January 5, 2026, https://www.mdpi.com/2073-4409/14/14/1074

12. Vascularization and Engraftment of Transplanted Human Cerebral Organoids in Mouse Cortex | eNeuro, accessed January 5, 2026, https://www.eneuro.org/content/5/6/ENEURO.0219-18.2018

13. Multimodal monitoring of human cortical organoids implanted in mice reveal functional connection with visual cortex - PubMed Central, accessed January 5, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC9792589/

14. Microglia-containing neural organoids as brain microphysiological systems for long-term culture - Frontiers, accessed January 5, 2026, https://www.frontiersin.org/journals/cellular-neuroscience/articles/10.3389/fncel.2025.1616470/full

15. Microglia integration into human midbrain organoids leads to increased neuronal maturation and functionality | bioRxiv, accessed January 5, 2026, https://www.biorxiv.org/content/10.1101/2022.01.21.477192v1.full-text

16. Generation, interrogation, and future applications of microglia ..., accessed January 5, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC11974650/

17. Parkinson's disease microglia induce endogenous α-synuclein pathology in patient-specific midbrain organoids | bioRxiv, accessed January 5, 2026, https://www.biorxiv.org/content/10.1101/2025.07.01.662685v1.full-text

18. Assembloid models of cell-cell interaction to study tissue and ..., accessed January 5, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC12143640/

19. Cortical assembloids support the development of fast-spiking human PVALB+ cortical interneurons and uncover schizophrenia-associated defects | bioRxiv, accessed January 5, 2026, https://www.biorxiv.org/content/10.1101/2024.11.26.624368v1.full-text

20. Imaging neuronal migration and network activity in human forebrain assembloids - PMC, accessed January 5, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC9234084/

21. Human neuromuscular junction three-dimensional organoid models and the insight in motor disorders - PubMed Central, accessed January 5, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC8782584/

22. Generation of Functional Human 3D Cortico-Motor Assembloids | Request PDF - ResearchGate, accessed January 5, 2026, https://www.researchgate.net/publication/347663145_Generation_of_Functional_Human_3D_Cortico-Motor_Assembloids

23. Generation of Functional Human 3D Cortico-Motor Assembloids - PubMed - NIH, accessed January 5, 2026, https://pubmed.ncbi.nlm.nih.gov/33333020/

24. Using brain organoids to understand Zika virus-induced microcephaly | Development, accessed January 5, 2026, https://journals.biologists.com/dev/article/144/6/952/48342/Using-brain-organoids-to-understand-Zika-virus

25. Expression Analysis Highlights AXL as a Candidate Zika Virus Entry Receptor in Human Neural Stem Cells - PMC, accessed January 5, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC4860115/

26. Zika Virus Neuropathogenesis: The Different Brain Cells, Host Factors and Mechanisms Involved - Frontiers, accessed January 5, 2026, https://www.frontiersin.org/journals/immunology/articles/10.3389/fimmu.2022.773191/full

27. Potential mechanisms of Zika‐linked microcephaly - PMC - PubMed Central, accessed January 5, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC5516183/

28. Brain assembloid: a human model for neural circuits research - PMC - PubMed Central, accessed January 5, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC11749076/

29. Aberrant calcium channel splicing drives defects in cortical differentiation in Timothy syndrome - PMC - NIH, accessed January 5, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC6964969/

30. Alzheimer's Disease Patient Brain Extracts Induce Multiple Pathologies in Vascularized Neuroimmune Organoids for Disease Model - bioRxiv, accessed January 5, 2026, https://www.biorxiv.org/content/10.1101/2024.10.28.620680v1.full.pdf

31. Mechanisms of α-Synuclein Induced Synaptopathy in Parkinson's Disease - Frontiers, accessed January 5, 2026, https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2018.00080/full

32. Mechanisms of α-Synuclein Induced Synaptopathy in Parkinson's Disease - PMC, accessed January 5, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC5825910/

33. Electrophysiological insights with brain organoid models: a brief review - BMB Reports, accessed January 5, 2026, https://www.bmbreports.org/view.html?uid=2026&vmd=Full

34. Complex oscillatory waves emerging from cortical organoids model ..., accessed January 5, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC6778040/

35. Brain Cells in a Lab Dish can Play Pong - UTMB, accessed January 5, 2026, https://www.utmb.edu/mdnews/podcast/episode/brain-cells-in-a-lab-dish-can-play-pong

36. Brain cells in a dish learn to play Pong - Medicine, Nursing and ..., accessed January 5, 2026, https://www.monash.edu/medicine/news/latest/2022-articles/brain-cells-in-a-dish-learn-to-play-pong

37. What DishBrain Gets Right (and Wrong): How Living Neurons Are Rewiring the Future of AI Efficiency | by Octavian Boji | Medium, accessed January 5, 2026, https://medium.com/@octavian.boji/what-dishbrain-gets-right-and-wrong-how-living-neurons-are-rewiring-the-future-of-ai-efficiency-dd54ec026473

38. Free energy principle - Wikipedia, accessed January 5, 2026, https://en.wikipedia.org/wiki/Free_energy_principle

39. Why I'm not into the Free Energy Principle — LessWrong, accessed January 5, 2026, https://www.lesswrong.com/posts/MArdnet7pwgALaeKs/why-i-m-not-into-the-free-energy-principle

40. Brain organoids and organoid intelligence from ethical ... - Frontiers, accessed January 5, 2026, https://www.frontiersin.org/journals/artificial-intelligence/articles/10.3389/frai.2023.1307613/full

41. Organoid intelligence (OI): the new frontier in biocomputing and intelligence-in-a-dish - Frontiers, accessed January 5, 2026, https://www.frontiersin.org/journals/science/articles/10.3389/fsci.2023.1017235/full

42. Vascularized human cortical organoids model cortical development in vivo - bioRxiv, accessed January 5, 2026, https://www.biorxiv.org/content/10.1101/682104v1.full

43. An in vivo model of functional and vascularized human brain organoids - PMC, accessed January 5, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC6331203/

44. Human Brain Organoids and Consciousness: Moral Claims and Epistemic Uncertainty, accessed January 5, 2026, https://www.mdpi.com/2674-1172/2/1/4

45. A Legal and Ethical Perspective on Human Consciousness and Human Brain Organoids, accessed January 5, 2026, https://law.stanford.edu/2023/07/27/a-legal-and-ethical-perspective-on-human-consciousness-and-human-brain-organoids/

46. Human brain organoids and their ethical issues: Navigating the moral and social challenges between hype and underestimation - PubMed Central, accessed January 5, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC10897434/