Our understanding of brain development and physiology has primarily been studied using animal models because it is difficult to have a direct access to human brain (1, 2). However, these models do not fully represent molecular features, genetic characteristics, and functional complexity of the human brain (3). Particularly, research involving embryonic brains is essential for understanding human brain development but challenging to approach. Recently, the development of organoids derived from pluripotent stem cells has enhanced our understanding of the human brain (4-6). With the establishment of protocols inducing differentiation from human induced pluripotent stem cells through embryoid body to specific regions of the brain, research within the in vitro model now permits investigation into the complexity of the human brain, encompassing various cell types (7, 8). Single-cell RNA sequencing and immunohistochemistry have facilitated the identification of cell types such as astrocytes, neurons, and oligodendrocytes, as well as the observation of diverse inhibitory and excitatory neurons (9-11). Despite advancements in understanding differentiation processes of the human brain and structural characteristics of brain organoids, functional measurement of neuronal and glial cells in organoids remains the subject of ongoing research. Brain organoids form diverse and rhythmical cellular networks, mimicking distinctive and periodic brain signals observed in humans, known as neuronal oscillations (12-14). Therefore, electrophysiological investigations of brain organoids have the potential to reveal functionally unexplored measurements owing to the challenging accessibility of the human brain.

In this review, cellular compositions of brain organoids and resulting electrophysiological characteristics were first emphasized. Next, we discussed invasive or non-invasive methods for investigating electrical properties of brain organoids. Finally, we described various applications microelectrode arrays to offer non-invasive insights into the complexity and neuronal networks of brain organoids, ranging from developmental study to organoid intelligence.

Recently, numerous organoid protocols representing various brain regions have been developed, with cortical organoids (COs) effectively reflecting unique electrophysiological characteristics of the human brain. Cortical organoids generated from induced pluripotent stem cells (iPSCs) can differentiate from embryonic body under the influence of fibroblast growth factor 2 (FGF2). Epidermal growth factor (EGF) can be added to promote differentiation into neural cells. Subsequently, FGF2 and EGF are replaced with brain-derived neurotrophic factor (BDNF) and neurotrophic factor 3 (NT3) to further induce differentiation into mature neurons.

Cortical organoids were found to express PAX6, a marker for neural progenitor cells, in over 85% of cells by day 18. By day 25, organoids exhibited a lumen structure composed of cells expressing N-cadherin (NCAD), resembling the ventricular-like zone (VZ). Furthermore, it was observed that the VZ-like zone was surrounded by an intermediate zone similar to the subventricular zone (SVZ). The presence of glial cells such as astrocytes and oligodendrocyte progenitor cells, characteristic of the cerebral cortex, was also noted. This suggests that cortical organoids closely mimic developmental processes and structures of the human cerebral cortex, reproducing the complexity of neurons and thereby demonstrating the ability to recapitulate various physiological and functional aspects of the human brain in vitro (7).

Physiological complexity of the brain arises from integration of various organs such as blood vessels, the spinal cord, and the peripheral nervous system (9). Consequently, research has predominantly relied on animal models. However, human specific genes which is not expressed in animal models may influence brain physiology. Particularly, studies comparing human and chimpanzee brain organoids have found that genes are specifically expressed in humans, including those abundant in human radial glia, intermediate progenitors, and neurons. These differences imply that animal models may not fully represent the human brain. However, human brain organoids have been shown to faithfully reflect genes expressed in the actual human brain (15). We emphasize the significant role of organoids in human electrophysiological research.

Research on the human brain has mainly been conducted using postmortem brain tissues and animal models. While these methods have increased our understanding of the brain, it is difficult to access developmental stages and functional characteristics of the human brain using these methods (12, 15-17). Brain organoids have garnered considerable attention due to their resemblance to the human brain and potential for various applications (18). Brain organoids hold promise for diverse applications, notably in disease modeling. They can mimic congenital brain malformations and neurodegenerative disorders, reflecting genetic characteristics of patients’ cells. As such, brain organoids offer valuable tools for drug screening and therapy development (16, 19-24).

Research on the developmental process of the human brain is also feasible. Specifically, the outer subventricular zone (OSVZ), which is significantly underrepresented in the developmental process of mice, is abundantly found in brain organoids (25, 26). This enables studies on the complexity of the human cortical structure, reflecting key features of the human brain within laboratory settings. Additionally, inter-species comparisons allow for investigations into the evolution of the human brain. In such studies, electrophysiological investigations play a crucial role in confirming functional properties of brain organoids.

Calcium imaging

Calcium ions are utilized to generate intracellular signals regulating various functions in different cell types. Particularly in neurons, calcium concentration varies depending on neurotransmitters released at nerve terminals. In a resting state, most neurons maintain an intracellular calcium concentration of approximately 50-100 nM, which transiently increases to levels 10-100 times higher during action potentials. This can also be invoked in various fundamental events. It can form more complex intercellular signals such as brainwaves. Calcium imaging enables visualization of neuronal activity by labeling calcium with fluorescent calcium indicators such as Oregon Green BAPTA and Fluo-4 dyes, followed by observation using imaging equipment such as a two-photon microscope (27).

For instance, spontaneous calcium dynamics have been detected on the 76th day using a two-photon microscope to investigate neuronal functions of cerebral organoids. Furthermore, neuronal networks isolated from organoids exhibit synchronized activity in some small cell groups on post-harvest day 16, followed by synchronized network activity on day 30. Additionally, the addition of glutamate resulted in an explosively observed synchronized response, whereas its absence in the same field led to a decrease. Conversely, the addition of bicuculline methochloride, a GABA_A receptor antagonist (−), significantly increased the neuronal activity (28). Calcium imaging allows for the observation of neuronal activity and networks using spatiotemporal calcium dynamics. However, calcium imaging is limited by temporal resolution and the measurable depth of field (29).

Patch clamp

Patch-clamp techniques allow us to record electrophysiological characteristics of organoids. Unlike calcium imaging, patch-clamp allows for the recording of potentials occurring in neurons at the millisecond level (30, 31). Sivitilli et al. have employed patch clamp to assess neuronal function (32). Whole-cell patch clamping was performed to measure electrophysiological outputs in fresh slices prepared from 12- and 24-week cerebral organoids for comparing mature and developing neurons in organoids. Developing neurons exhibited fired action potential with slow kinetics and low amplitude. On the other hand, mature neurons fired action potentials with fast kinetics and high amplitude. Moreover, they generated high amplitudes and stable spontaneous action potentials.

Patch clamping provides neuronal activity and amplitude voltage data induced from organoid’s neuronal cell. However, to measure electrophysiological properties of organoids by patch clamp, it is necessary to slice organoids or isolate single neurons. Additionally, patch-clamp recordings are restricted to capturing signals from single cells in organoids, thereby lacking the ability to provide insight into neuronal networks in living organoids (32).

Microelectrode array (MEA)

Microelectrode array (MEA) is a method of collecting neuronal electrophysiological data through microelectrodes arranged on the bottom of a culture plate. When in contact with organoids, electrodes can record neuronal activity and network patterns. Electrophysiological signals collected through MEA undergo amplification and filtering processes, enabling various types of analyses. Primitive MEA data can extract relatively high-frequency data ranging from 300 to 3,000 Hz, allowing for the generation of spike raster plots. These plots enable the analysis of neuronal bursts and network activity. Additionally, analyzing the local field potential of organoids is achievable low-frequency ranging from 1 Hz to 1,000 Hz. Therefore, it is feasible to measure both the overall oscillation of brain organoids and the spike activity of individual neurons over long-term culture periods (Table 1) (33).

Recording complicate electrophysiological properties of brain organoids is a challenging task. It is also indispensable for understanding the brain. While calcium imaging allows visualization of overall neuronal activity, it suffers from limitations such as temporal resolution, shallow penetration depth of fluorescent dyes, and an inability to observe over prolonged culture periods. Similarly, techniques such as patch clamp struggle to reflect complex neuronal networks and oscillatory patterns within organoids. Recent research has turned to the use of multi-electrode array (MEA) to investigate spontaneous neural activity and oscillations in long-term and sustained culture environments. This approach holds promise for playing a crucial role in human brain differentiation and disease modeling. The current advancement of MEA involves parallel development of devices to reflect three-dimensional characteristics of organoids and the utilization of MEA for functional measurement and analysis of various organoids (Fig. 1).

Electrophysiological changes along brain development

The difficulty in accessing developmental changes in fetal and neonatal brains has led to the substitution of animal models. However, as animal models do not fully represent human brain development, measuring neural oscillations in organoids generated from embryonic bodies can enhance our understanding of human brain development. For instance, Trujillo et al. measured characteristics of oscillations at different developmental stages of brain organoids using MEA (33). Cortical organoids exhibited high phase-amplitude coupling between oscillatory delta (1-4 Hz) and broadband gamma activity (100-400 Hz). Additionally, 10-month-old organoids showed robust spikes and increased burst frequency, a pattern not observed in spheroids composed of differentiated neurons, thus demonstrating continuously evolving neural networks in organoids. Furthermore, local field potential of organoids has been reported to closely resemble preterm neonatal EEG patterns (1).

Drug screening

MEA offers a non-invasive means to measure electrophysiological properties, making it an ideal platform for recording drug responses while culturing organoids. Yokoi et al. have induced epileptic seizures in cerebral organoids by treating them with pentylenetetrazol (PTZ) and screened effects of antiepileptic drugs (AEDs) (34). Electrophysiological signals of organoids in response to drugs were recorded via MEA and analyzed within a frequency range of < 500 Hz. After PTZ treatment, raw waveforms were recorded 10 minutes later. The total spike rate of neurons showed a concentration-dependent increase in response to PTZ. Furthermore, when analyzing waveforms between 4 and 500 Hz before and after PTZ administration, the power of frequencies at 4-10 Hz increased in the PTZ-treated group. Additionally, as the concentration of PTZ increased, the frequency power exhibited a significant increase. Furthermore, when administering perampanel, an antiepileptic drug, oscillation data obtained through MEA showed an increase in frequencies around 0-30 Hz. This demonstrates that perampanel inhibits epileptic seizures in organoids (34). This study demonstrates that brain organoids exhibit unique characteristics in frequency components below 500 Hz and that these characteristics can be altered by drugs. If responses of organoids to compounds can be identified through frequency component analysis obtained via MEA, it will be possible to predict how new drugs for brain disorders will react in the human brain.

3D MEA for brain organoid electrophysiology

To overcome limitations of planner MEA, it is essential to develop MEA that reflects the three-dimensional characteristic of organoids. For example, inspired by human EEG measurements, one study has produced MEA wrapped in the form of a shell with self-folding polymer bilayer (35). This device was designed to adjust curvature depending on the diameter of the organoid. In that study, compared to existing 2D MEA, the signal-to-noise ratio (SNR) was significantly increased. Increased electrophysiological responses to stimulation such as glutamate were also captured. In conclusion, recording three-dimensional signals rather than signals in one plane will be advantageous for electrophysiological understanding of brain organoids.

Organoids can increase in thickness from 500 μm to 4 mm, unlike 2D cell cultures (36). There is ongoing research utilizing stretchable mesh nanoelectronics to obtain stable 3D biological-electrical interfaces for developing organoids (37, 38). This technology is, softer than brain tissue, accommodates compression, folding, and expansion during organoid formation. Stretchable mesh nanoelectronics can seamlessly integrate with organoid tissues, uniformly distributing across the entire area to record electrophysiological signals. Consequently, rhythmical oscillations in spontaneous local field potentials (LFPs) were observed at one month after differentiation and from tissue-wide embedded mesh electrodes in the range of 0-1 kHz three months later.

Reflecting the 3D nature of organoids necessitates material technologies that are sensitive and flexible without compromising cell viability (39). Continued research in this direction is expected to contribute to a more comprehensive understanding of neuron networks distributed within organoids.

Organoid intelligence (OI)

Human brain consisting of 86 to 100 billion neurons with over 1,015 connections can surpass machines in processing sparse or uncertain data despite being slower than supercomputers in simple data computation (40, 41). A report has suggested that a supercomputer takes 40 minutes to model 1% of human brain activity for 1 second in complex decision-making scenarios (42). Thus, modeling the human brain has high efficiency for inference using sparse and uncertain data. Organoids mimic neuronal networks of the human brain, exhibiting high neuron density and synaptic generation crucial for learning. They can be cultured for over a year. Through electrophysiological recordings obtained via MEA, unsupervised learning via input-output patterns becomes feasible.

In the case that organoids are placed on MEA electrodes for unsupervised learning, changes in neuronal connectivity in response to different patterns of stimulations have been reported (43, 44). Additionally, nonlinear regression analysis demonstrated precise prediction results as organoids underwent learning. This demonstrates that organoids can replicate the brain’s nonlinear properties for AI machine learning and computing (43).

The Baltimore Declaration toward the exploration of organoid intelligence, established in February 2022, is gaining recognition as an emerging scientific field. Expectations lie in the electrophysiological understanding of brain organoids fostering the direction and advancement of organoid computing (44).

Other applicable bio-convergence technologies in electrophysiology

Although research utilizing organoids for electrophysiological studies offers many advantages for understanding the functionality of the human brain, significant advancements are still needed to faithfully mimic the activity of the adult brain (33, 45). To overcome these limitations, recent efforts have been actively conducted to enhance the complexity and connectivity of organoids using technologies such as bioprinting and microfluidic chips (46, 47). These studies could bridge the gap between human brains and organoids, opening possibilities to investigate electrophysiological features reminiscent of those observed in the human brain because of its complexity.

Moreover, brain waves in living organisms can vary in response to different stimuli (48, 49). Recently, there has been a growing interest in utilizing invasive/noninvasive stimuli to treat neurodegenerative diseases (47). Techniques to overcome pathological oscillations through appropriate stimulation of brain organoids are likely necessary. Optogenetics and sonogenetics, which regulate neuronal activity through light and ultrasound stimulation, respectively, have gained attention in research aiming to suppress pathological activity in the brain. Their application in organoids is being explored (50).

In this review, we briefly introduced brain organoid’s electrophysiological features and measuring methods. Utilizing electrophysiological characteristics of organoids for functional measurements is a crucial perspective in understanding them. The process by which unique neuronal networks and waves in the human brain form is not yet fully understood. Therefore, it is necessary to develop methods that reflect three-dimensional features of organoids and the connectivity of neuronal and glial cell populations.

Recently, there has been a surge in research utilizing and developing microelectrode arrays (MEAs) as a method to investigate such topics. This approach allows for the measurement of spikes, confirming neuronal activity and LFP, which can enable the observation of neuronal networks from organoids. These recent advancements will not only contribute to our understanding of the human brain, but also provide valuable insights for drug screening and disease modeling using organoids.

This work was supported by a grant of Ministry of Education of the Republic of Korea and the National Research Foundation of Korea (RS-2023-00266110). The graphical arts are generated by Biorender.com under Jong-Chan Park, Ph.D.

The authors have no conflicting interests.