Neurodegenerative diseases (NDDs) arise from the progressive loss of neurons, brain function, and cognition, affecting millions worldwide. Despite extensive research, no treatments currently halt or slow neurodegeneration. Advancing effective therapies requires a deeper understanding of brain complexity, NDD pathophysiology, and reliable in vitro drug-screening models. While animal models have enhanced our understanding of disease mechanisms, their translational success remains limited (Chang et al., 2020). Only about 5% of preclinical studies in animal models ultimately lead to regulatory approval for human use. For example, in diseases like Alzheimer’s (AD) and multiple sclerosis (MS), many therapies showing promise in preclinical stages fail to translate clinically (Ineichen et al., 2024). This gap stems from key differences in brain development (Quadrato et al., 2016), architecture, immune responses (Bjornson-Hooper et al., 2022), and metabolism between humans and animals, complicating cross-species extrapolation (Kuzawa et al., 2014).

Although human brain tissue is considered the gold standard for NDD research, its use is restricted by ethical and practical challenges, such as limited availability, sample sharing issues, and preservation difficulties (Quadrato et al., 2016). Post-mortem tissue also suffers from irreversible changes that may alter results, limiting its suitability for large-scale studies (Chang et al., 2020). These limitations highlight the urgent need for alternative models that accurately replicate human brain structure and function while being more ethically viable and accessible.

One promising solution is the use of three-dimensional (3D) organoids. Over the past few years, 3D organoid technology has gained significant relevance in NDD research. These organoids are small, in vitro 3D structures derived from human pluripotent stem cells (PSCs), such as embryonic stem cells (ESCs) and induced pluripotent stem cells (iPSCs) (Clevers, 2016; D’Antoni et al., 2023), or from adult stem cells (ASCs) (Clevers, 2016).

Since organoids are generated from human stem cells, which have the capacity to self-organize, they can recapitulate several aspects of the composition, organization and function of organs in vivo (Clevers, 2016), although they still lack full physiological complexity. Organoids retain key phenotypic traits of the original tissue, including a degree of cellular diversity and cell-to-cell interactions, and they have a long lifespan in culture (Quadrato et al., 2017; Trujillo et al., 2019). Furthermore, neurons within organoids have been shown to exhibit signs of polarity, migration (Clevers, 2016) and electrical activity (Quadrato et al., 2017; Trujillo et al., 2019). These characteristics make organoids a powerful platform for investigating selected cellular and molecular mechanisms of the organs they model.

Human brain organoids not only feature the development of neurons, but also glial cells, with transcriptional profiles and neurodevelopmental trajectories that closely resemble fetal brain development (Trujillo et al., 2019). This makes them a powerful tool for studying the patterning and specification of various neuronal and glial cell types. Therefore, while organoids can recapitulate important aspects of the structure and function of the human brain, they represent a simplified system that captures early developmental and disease-related processes, offering unique opportunities to explore features of human brain biology that cannot be effectively modeled in animals, and holding particular promise for studying neurological and neurodegenerative disorders (D’Antoni et al., 2023).

This review aims to provide a broad, up-to-date evaluation of the use of brain organoids in the study of NDDs and their potential applications in precision medicine. It is intended as a conceptual synthesis of different organoid models, addressing the main advances, opportunities, challenges and future perspectives in this rapidly evolving field. With this approach, we provide a comprehensive yet accessible resource for researchers and clinicians seeking to better understand the mechanisms of neurodegeneration and to advance the development of personalized therapeutic strategies.

In 1907, Henry Van Peters Wilson discovered that siliceous sponge cells reaggregated, even after being dissociated. These cells could self-organize and differentiate into fully formed sponges (Wilson, 1907). Since then, multiple research teams have shown that various types of dissociated cells can be reaggregated in a similar manner. For instance, Holtfreter (1948) demonstrated cellular reaggregation in early embryonic amphibian cells (Holtfreter, 1943). Similarly, Weiss and Taylor (1960) performed dissociation-reaggregation experiments with cells from multiple organ sources obtained from chicken embryos (Weiss and Taylor, 1960). In 1964, Malcolm Steinberg proposed the differential adhesion hypothesis, suggesting that cell sorting and rearrangement are governed by thermodynamic principles driven by variations in surface adhesion (Steinberg and Locke, 1964).

The 1980s represented a major milestone in organoid research, with studies focusing on cell-matrix interactions in organoid development (Li et al., 1987) and the isolation of PSCs from mouse embryos in 1981 (Evans, 1981; Martin, 1981). In 1987, scientists started to improve cell culture conditions by simulating the in vivo microenvironment. Breast epithelial cells were demonstrated to form 3D ducts and lumens when cultured on an extracellular matrix (ECM) extract derived from Engelbreth-Holm-Swarm mouse sarcoma, enabling them to synthesize and secrete milk proteins, a capability not observed in traditional two-dimensional (2D) cultures (Li et al., 1987).

It was in 1998 when scientists successfully isolated and cultured embryonic stem cells from human blastocysts for the first time (Thomson et al., 1998). In 2006 and 2007, a breakthrough occurred when Takahashi and Yamanaka were able to develop iPSCs through the reprogramming of mouse (Takahashi and Yamanaka, 2006) and human fibroblasts (Takahashi et al., 2007). The isolation of stem cells greatly propelled organoid research, showing superior effectiveness and providing deeper insights for disease modeling.

The shift from 2D to 3D organoid cultures happened in 2008 when polarized cerebral cortex tissue was generated for the first time by Eiraku et al. from ESCs using serum‐free embryoid bodies with quick reaggregation (Eiraku et al., 2008). In 2009, intestinal ASCs were demonstrated to self-organize and differentiate to crypt-villus structures in the absence of a mesenchymal niche and they were used for the first time to generate 3D intestinal organoids in Matrigel® (Sato et al., 2009). This marked the first successful establishment of 3D organoid cultures derived from a single ASC, paving the way for disease modeling and precision medicine.

In 2013, Lancaster et al. generated the first brain organoids from human iPSCs (hiPSCs) upon growth in Matrigel® and with agitation, showing different brain regions (Lancaster et al., 2013). Since then, different studies have further refined organoid protocols to model specific brain areas, such as midbrain (Jo et al., 2016), hippocampus (Sakaguchi et al., 2015), and cerebellum (Muguruma et al., 2015), with innovations like the 3D-printing technology, which consists of a miniaturized spinning bioreactor, that has allowed the generation of forebrain organoids in a cost-effective manner (Qian et al., 2016).

The characteristics of an organoid, including its variability, heterogeneity, and functionality, depend significantly on the starting cell type, which can be either ASCs or PSCs, including both ESCs and iPSCs (Figure 1). PSC-derived organoids are generated by providing signaling cues that mimic in vivo development, and they are primarily used to study organogenesis and developmental events, as they resemble fetal-stage tissues, while ASCs-derived organoids reflect the self-renewal and differentiation capacity of somatic stem cells in tissue homeostasis (Watanabe et al., 2017).

ASC-based organoids are generated from biopsy samples obtained from either healthy or diseased tissues. The samples are dissociated into epithelia containing stem cells. These organoids closely simulate the original tissue’s dynamic stem cell behavior and are valuable models for studying monogenic diseases (Huch et al., 2015) and cancer (Matano et al., 2015) while also allowing molecular analysis and gene correction studies (e.g., using CRISPR/Cas9) (Schwank et al., 2013).

Robust protocols have been developed for the long-term cultivation, expansion, and cryopreservation of various ASC-derived organoid types. These protocols involve fewer steps, typically require less time overall and produce more mature structures that closely resemble the anatomy and function of adult tissue (Watanabe et al., 2017). This makes them highly suitable for regenerative medicine and disease modeling. ASCs are obtained either as isolated cells or from dissected tissue fragments and they are often used to generate organoids that maintain and repair their tissue of origin. For example, pancreas (Huch et al., 2013) and liver (Barker et al., 2010) organoids are used to model regeneration processes following injury. Besides, ASC-derived organoids can be generated from patient samples to assess responses to specific treatments, as seen in the colorectal cancer organoid biobank (Van de Wetering et al., 2015).

However, ASC-derived organoids have limitations. They have a restricted potency compared to PSCs (multipotency), as they are already predisposed to organ-specific differentiation and cannot generate multiple cell lineages, lacking the necessary tissue-tissue interactions to promote organ-level complexity. Moreover, as they require tissue samples containing stem cells, they have limited accessibility, and successful culture requires prior knowledge of tissue-specific conditions. While ASCs provide crucial insights into disease pathology, they are less suited for uncovering early disease mechanisms, as they primarily reflect tissue maintenance and regeneration in adulthood (Watanabe et al., 2017).

PSCs have an unlimited capacity to self-renew and differentiate into all cell types of the body, making them a powerful tool for organoid generation (Chang et al., 2020). PSCs can be obtained with less invasive procedures, unlike ASCs. This accessibility along with their ability to generate diverse cell types, makes PSC-derived organoids ideal for modeling complex diseases such as NDDs. Due to their pluripotency, PSCs have high proliferation potential and possess the ability to differentiate into the three primary germ layers -ectoderm, mesoderm, and endoderm-including mesenchymal (mesoderm), epithelial (endoderm/ectoderm), and endothelial (mesoderm) (Watanabe et al., 2017). This makes them particularly suited for studying organogenesis in terms of not only cell differentiation, but also spatial patterning and morphogenesis (Clevers, 2016).

ESCs are derived from the inner cell mass of the blastocyst and have the potential to differentiate into all cell types (De Wert and Mummery, 2003). ESCs-derived organoids are very effective models for investigating genetic disorders and infectious diseases, especially in organs with very limited regeneration power like the brain (Ahammed and Kalangi, 2024). Nevertheless, their use raises strong ethical concerns due to the requirement of using human embryos (De Wert and Mummery, 2003).

IPSCs can be derived from various somatic cells, including keratinocytes, dental pulp stem cells and mesenchymal stem cells (MSCs). However, fibroblasts are the most used source for iPSCs generation, as they are generally easy to obtain and handle. Therefore, iPSCs are more accessible and ethically favorable compared to ESCs, as they are derived from adult tissues and do not require the use of human embryos. Additionally, organoids derived from ESCs exhibit more advanced maturation compared to those derived from iPSCs (Yang et al., 2023).

Initially, reprogramming somatic cells into iPSCs involved four transcription factors, OCT4, SOX2, KLF4 and c-MYC, also known as the Yamanaka factors. Since then, different cocktails have been used to successfully reprogram iPSCs (Ali et al., 2023). These factors can be delivered using methods like viral vectors, liposomes, or transposons (Pazzin et al., 2024). Integrating methods like retroviral and lentiviral delivery insert foreign genetic material into the cell’s genome, leaving an undesirable footprint that can cause mutations and safety risks, making them less suitable for clinical applications. To address this, non-integrating reprogramming methods, such as episomal plasmids, have been developed. These plasmids introduce reprogramming factors without integrating into the genome and are naturally lost over time, ensuring no permanent genetic alterations (Ali et al., 2023).

Since iPSCs can be derived from a patient’s own cells, they maintain the genetic and phenotypic traits of the donor tissue, making them an invaluable tool for personalized medicine, disease modeling, and patient-specific therapies (Halevy and Urbach, 2014). They better reflect disease-relevant phenotypic features, enabling more reliable disease modeling, allowing researchers to create patient and disease-specific cell lines, and avoiding immune rejection after transplantation. As iPSCs can differentiate into neural precursors, they subsequently generate diverse neuronal and glial cell types (Inoue et al., 2014). Numerous studies have already successfully reprogrammed iPSCs from fibroblasts obtained from individuals with various neurological disorders, including amyotrophic lateral sclerosis (ALS) (Dimos et al., 2008), Parkinson’s disease (PD) (Soldner et al., 2009) and AD (Choi et al., 2014). Furthermore, iPSC-derived organoids have been used for drug development and cell-based therapies by providing a renewable source of patient-matched cells (Dimos et al., 2008).

However, generating iPSC-based organoids remains complex, time-intensive, and expensive, requiring multiple reprogramming factors. Moreover, the resulting organoids are often immature and less functional, they have a limited lifespan and capacity to proliferate, frequently requiring additional culturing steps to achieve full maturation, and they lack interactions with other developing cell types (Yang et al., 2023). The use of oncogenes for reprogramming can also lead to de novo mutations (Halevy and Urbach, 2014). This instability increases tumorigenic risk, which presents a significant challenge for clinical applications, which may influence disease phenotype, posing a significant challenge for clinical application (Gutierrez-Aranda et al., 2010). To address these limitations, multiple biobanks have been established, supported by detailed patient clinical records. Additionally, utilizing established iPSC lines can help mitigate genetic variability (Evangelisti et al., 2024). These lines can be further modified using genome-editing technologies such as zinc finger nucleases, TALENs, and CRISPR/Cas to study specific genetic variants (Hotta and Yamanaka, 2015; Soldner and Jaenisch, 2018). Moreover, iPSCs fail to preserve the patient’s age, which hinders the study of aging-related diseases such as NDDs. This highlights the need to implement additional protocols to induce aging (Jothi and Kulka, 2024).

The wide variety of brain organoid cell sources can make it difficult to choose the most suitable one, as no single protocol fits all purposes. Researchers should select an organoid assay based on the complexity needed for their specific biological question and interpret results while considering the assay’s limitations.

Despite their advantages, like rapid growth, 2D culture systems, which are cultured on flat inert surfaces, face significant limitations (Saraswathibhatla et al., 2023), including the loss of normal polarity and shape and cell–cell and cell–extracellular interactions. This leads to cells that fail to accurately replicate the functions, development and behaviors observed in tissues or organs, compromising their biological relevance (Yang et al., 2023).

To overcome these limitations, more complex in vitro 3D culture systems were developed, replicating the in vivo physicochemical microenvironment (Drost and Clevers, 2018). These models maintain genetic stability and cellular structure while closely resembling the original cell types (Clevers, 2016; Drost and Clevers, 2018). They can be successfully cryopreserved and stored in living organoid biobanks and can be genetically characterized and used for drug screening and personalized medicine (Drost and Clevers, 2018). The ECM of the 3D culture enables inter-cellular signal transduction and physiological cues, simulating a more natural cell environment and providing a more reliable model system (Saraswathibhatla et al., 2023).

However, not all 3D neural culture systems qualify as brain organoids, as described by D’Antoni et al. in their review. Neurospheres, for instance, are 3D clusters of neural progenitor cells (NPCs) derived from central nervous system (CNS) primary tissue or iPSCs, but they lack the organization and complexity of true brain organoids. They grow in serum-free medium with FGF-2 and epidermal growth factor (EGF), without needing an adherent surface for expansion. Neurospheres are valuable for studying NPC behavior and neurogenic tissue in a simpler manner and are used in transplantation research (D’Antoni et al., 2023).

Thus, 3D brain organoids offer a more comprehensive model for investigating brain development and disease due to their structural complexity and capacity for long-term culture. They are generated from stem cells that self-organize through cell-sorting and form a heterogeneous population of cells that closely mimic the composition of the developing brain (“mini-brains”). As these structures evolve, they become more complex, leading to the formation of cortical progenitors. The most prominent structures emerging within the brain organoid tissue are ventricular-like zones and the organization of cortical layers (D’Antoni et al., 2023).

The creation of PSC-derived 3D brain organoids could be summarized in three different steps (Figure 2) (Kwak et al., 2024). First, iPSCs or ESCs are aggregated to form embryoid bodies (EBs) by seeding them onto low-attachment plates for approximately 1 week. This process promotes free-floating 3D cell clusters in medium supplemented with low concentrations of basic fibroblast growth factor (bFGF) and a ROCK inhibitor (Huch et al., 2013; Kwak et al., 2024).

In the second phase of brain organoid generation, the EBs are transferred into neural induction medium, such as Dulbecco’s modified Eagle’s medium/Ham’s F12 (Kwak et al., 2024; Gómez-Álvarez et al., 2023), which provides essential nutrients and maintains a stable environment (Gómez-Álvarez et al., 2023). Antibiotics like penicillin/streptomycin are added to prevent bacterial contamination (Soldner et al., 2009; Gómez-Álvarez et al., 2023), while serum components, such as those present in fetal bovine serum, supply biomolecules necessary for cell growth and survival. Additionally, a variety of soluble factors are incorporated to promote cell signaling, differentiation, growth, viability, and function (Gómez-Álvarez et al., 2023).

B27, 2-mercaptoethanol and insulin are supplemented to encourage the differentiation into neuroectodermal lineages (Kwak et al., 2024; Lancaster and Knoblich, 2014a). Differentiation may be either guided using patterning factors to maximize the features of a specific brain region, or unguided if stem cells are allowed to self-organize into brain organoids containing multiple brain areas. Dorsal forebrain (pallium, cerebral cortex) organoids are induced by inhibiting wingless-type MMTV integration site family (WNT) signaling pathways, while ventral forebrain organoids are fabricated by inhibiting WNT signaling and activating sonic hedgehog (SHH) signaling pathways, which regulate the specification of dorsal and ventral brain regions. Hypothalamus organoids are induced by activating WNT and SHH signaling, along with the addition of ciliary neurotrophic factors and FGF2, which is crucial for cell proliferation and survival, and ciliary neurotrophic factors. Midbrain organoids are formed by modulating WNT and SHH signaling while supplementing the culture with FGF8, and cerebellar organoids develop through the introduction of FGF8 and WNT1, with FGF19 promoting further maturation (Kwak et al., 2024). Therefore, the precise selection and combination of the culture components must be tailored according to the tissue type and developmental stage to ensure optimal organoid formation and functionality (Lancaster and Knoblich, 2014a).

After differentiation, neuroepithelial buds or neural rosettes are embedded in ECM droplets, an ECM that provides support for cell differentiation and tissue organization. Another alternative for the development of organoids are hydrogels, bioprinting, suspension cultures and flat organoids. Growing organoids can be placed in a spinning bioreactor, where continuous agitation promotes growth, by enhancing oxygen and nutrient exchange, which ultimately supports the development of diverse brain regions resembling cerebral cortex (Kwak et al., 2024; Lancaster and Knoblich, 2014a). Finally, brain organoids mature over time replicating neural functionality (Kwak et al., 2024).

On the other hand, the transition from ASCs to organoids begins with enzymatic dissociation and purification of stem cells from the source tissue, followed by their cultivation in a 3D ECM or a different culture environment. In the second phase, a specialized culture medium containing specific growth factors such as Noggin, EGF, Wnt3a, or R-spondin is added to stimulate cell proliferation and differentiation. Finally, under these controlled conditions, ASCs self-organize and form 3D structures that recapitulate the characteristics of the original organ (Günther et al., 2022).

As commented before, brain organoid generation has been generally categorized into two groups based on patterning approaches, the unguided differentiation and the guided differentiation.

Unguided differentiation, which does not employ region-specific patterning factors but exploits the spontaneous intrinsic signaling of human PSCs, generates “whole-brain organoids” containing cells of the three primary brain vesicles: midbrain, hindbrain and forebrain (D’Antoni et al., 2023; Qian et al., 2019). These organoids predominantly generate neuro-ectodermal tissue, but they also produce non-ectodermal cell types, such as microglia (D’Antoni et al., 2023). However, the stochastic nature of the unguided differentiation often leads to inconsistent cell proportions, disorganized spatial distribution, altered initial conditions, and non-physiological cell interactions, resulting in higher variability and more complex reproducibility, which complicate experimental outcomes (Qian et al., 2019).

In contrast, guided differentiation involves the use of brain-specific patterning factors that stimulate developmental signaling pathways at precise stages of differentiation, allowing for the generation of brain organoids with specific regional identities (D’Antoni et al., 2023; Qian et al., 2019). This approach reflects better the specific domains of the nervous system at the anatomical, cellular and molecular level with higher reproducibility and reliability. However, because interactions between different brain regions are crucial for understanding neurodegenerative and psychiatric disorders, guided protocols have also been adapted to generate multiple brain-region-specific organoids that can be fused into assembloids after differentiation (Qian et al., 2019). These fused structures facilitate the study of complex cellular interactions and neurodevelopmental processes across different brain areas.

Ultimately, the choice of protocol, whether whole-brain organoids, region-specific brain organoids, or assembloids—depends on the specific scientific questions being investigated and the intended application in developmental biology, disease modeling, or therapeutic screening (Jang et al., 2022).

Assembloids, considered the next-generation of organoids, are created by coculturing multiple cell types or combining organoids derived from different tissues (Yang et al., 2023). Small organoids with smooth, translucent edges, and strong integrity are selected for fusion. They are placed in close proximity within the same optimized culture media to facilitate adhesion and gradual integration. To prevent cell death and promote the formation of synaptic connections, physical and environmental conditions are regulated over time (Nityanandam et al., 2025).

Unlike single-cell-type organoids, assembloids reflect better interactions between subregions or cell lineages, facilitating the study of brain development processes such as long-distance projections and interneuron migration (Yang et al., 2023). Since interactions between brain regions and systems are critical in neurodegenerative and psychiatric diseases, establishing connections between brain organoids and other organ systems could significantly advance therapeutic target identification.

In the last couple of years, several studies have fused cerebral organoids from different brain regions, for example, Andersen et al. recently developed cortico-motor assembloids by combining cortical, hindbrain/cervical spinal cord, and skeletal muscle organoids, effectively modeling the cortico-motor pathway and the muscle contraction control (Andersen et al., 2020). Similarly, Ogawa et al. created a glioma model by integrating brain organoids with glioblastoma cells, offering insights into tumorigenesis and metastasis (Ogawa et al., 2018). Additionally, cortex-ganglionic cortex assembloids reproduced brain network formation and exhibited epileptiform-like activity, contributing to research on Rett syndrome (Samarasinghe et al., 2021).

Assembloids research has also been extended to the peripheral nervous system. iPSCs have been differentiated into spinal cord neurons and skeletal muscle cells that self-organize into neuromuscular organoids (Faustino Martins et al., 2020) and intestinal organoids cocultured with neural crest cells have modeled enteric nervous system development and intestinal motility disorders (Workman et al., 2017).

Understanding body-wide interactions is essential for studying neurological diseases, as key therapeutic targets may exist outside the brain (Yang et al., 2023). For example, liver disfunction has been linked to AD and PD (Bassendine et al., 2020; Ciulla et al., 2019). Organ-on-a-chip systems, which connect organoids through microfluidic channels and mimic microenvironment conditions (D’Antoni et al., 2023; Yang et al., 2023), enhance our understanding of neurological and psychiatric disease mechanisms by modeling inter-tissue and multi-organ interactions (Yang et al., 2023).

Unlike neurons, which originate from the neuroectoderm, microglia derive from embryonic yolk sac progenitors and migrate to the brain, where they self-renew with the help of cytokines (IL-34, CSF-1) and transcription factors (PU.1, IRF8). As the brain’s primary immune cells, microglia play a crucial role in neuroinflammation, synaptic regulation, and neural circuit formation. However, since brain organoids lack native microglia, researchers have introduced microglial precursors to create microglia-containing assembloids, useful for modeling NDDs (D’Antoni et al., 2023).

Multiple strategies have integrated microglia into brain organoids, advancing research on brain-immune interactions. These studies range from co-culture systems to genetic engineering techniques. In 2017, Abud et al. introduced microglia into iPSC-derived organoids via a co-culture system, showing mature microglial behavior validated by cytokine secretion and synaptic pruning assays (Abud et al., 2017). In 2018, Ormel et al. generated microglia-containing organoids by reducing heparin levels and delaying ECM embedment, resulting in microglia with realistic morphology and immune responses (Ormel et al., 2018). In 2019, Song et al. introduced iPSC-derived microglia into cerebral organoids, demonstrating immune responses and their ability to stimulate cell proliferation and reduce oxidative stress (Song et al., 2019). In 2021, Ao et al. developed a tubular organoid-on-a-chip system with isogenic microglia, reducing hypoxia and modelling neuro-immune interactions by supporting microglial activation after exposure to an opioid receptor agonist (Ao et al., 2021). In 2022, Cakir et al. improved microglia integration by inducing PU.1 expression, enhancing reproducibility and consistency in organoid models (Cakir et al., 2022). In the future, advanced protocols are expected to incorporate region-specific glia, aiming to enhance the molecular and functional characteristics of neurons in brain organoids. These findings collectively demonstrate the feasibility of integrating microglia into brain organoids, although challenges in consistency and functionality remain (D’Antoni et al., 2023).

Another challenge that arises is cell death in the internal regions due to lack of vascularization that alters oxygen and nutrient delivery, metabolite elimination, and cell signaling. This restricts the organoids lifespan, making it difficult to use organoids for long-term studies (Lancaster and Knoblich, 2014b). To overcome this, new strategies have been developed.

Coculturing brain organoids in a ECM with endothelial cells has promoted vascularization by using FGF2, CHIR99012, BMP4, and VEGF to promote proper endothelial differentiation (Lancaster and Knoblich, 2012; Pham et al., 2018). Other studies have been based on the co-culture of human umbilical vein endothelial cells (HUVECs) with PSCs before neural induction, producing vascularized brain organoids (HUVEC-BOs) with enhanced neurogenesis and maturation (Shi et al., 2020). An alternative approach is generating brain-specific vascular organoids. In a guided protocol, human ESCs were differentiated into human blood vessel organoids, which were then fused with cerebral organoids stimulating vascularization (Sun et al., 2022). Different studies have developed BBB organoids by combining glial, vascular, and neuronal cells in co-culture systems, or by producing BBB assembloids through the fusion of brain and blood vessel organoids. This type of model could be valuable for studying vascular dysfunction and BBB disruption associated with some NDDs, such as AD, PD and MS, as it closely mimics the interactions between neural and vascular components (Summers et al., 2024). Bergmann et al. created BBB organoids by coculturing brain endothelial cells, astrocytes and pericytes under low conditions, enabling studies on drug permeability and neurodegenerative therapies (Bergmann et al., 2018). Microfluidic devices have also been proven to recapitulate the BBB complexity, allowing the study of the interaction of compounds with endothelial cells, pericytes and astrocytes, and the transition across the BBB(101).

Several additional approaches have been proposed to enhance vascularization in organoids. As commented before, one strategy involves utilizing microfluidics devices or organ-on-a-chip devices with periodic flow that improve nutrient and oxygen exchange in brain organoids while reducing cell death (Cho et al., 2021). Pericytes and endothelial cells derived from human PSCs spontaneously form vascular networks that physically integrate with cerebral organoids, creating fully connected neurovascular organoids on a chip (Castiglione et al., 2022). Another approach involves genetically engineered cortical organoids to express human ETS variant transcription factor 2 (hETV2) to spontaneously form a vascular-like network in vitro which dramatically reduces markers of cell death and hypoxia without transplantation (Cakir et al., 2022).

Current brain organoids do not have cerebrospinal fluid (CSF). Co-culturing vascularized brain organoids and choroid plexus organoids has allowed the production of a more complete vasculature in cerebral organoids (Pellegrini et al., 2020).

Other progresses that have been made in this area are spinning bioreactors that enhance nutrient exchange and enable growth up to a few millimeters in size (Lancaster and Knoblich, 2014a), ALI cultures (Cho et al., 2021), and slicing or cutting methods, to culture organoids as small pieces, reducing hypoxia and necrosis (Choe et al., 2021). Silk scaffolds (Sozzi et al., 2022), PCL scaffolds (Rothenbücher et al., 2021) and carbon fibers (Tejchman et al., 2020) have also been used as innovative frameworks to reduce necrosis and promote neuronal survival in organoids by improving oxygen flow and waste removal. Lastly, the transplantation of organoids in animals has also resulted in vascularization (Mansour et al., 2018; Revah et al., 2022).

The integration of microglia and vascularization in brain organoids represents a major step toward more physiologically relevant models. Advances in co-culture techniques, microfluidics, and genetic engineering are progressively overcoming current limitations, paving the way for improved disease modeling and therapeutic testing.

Myelination is a process where oligodendrocytes wrap axons in the CNS with myelin, a fatty substance that speeds up the transmission of nerve impulses (López-Muguruza and Matute, 2023). This process is crucial for the proper functioning of the nervous system, and its disruption is associated with various NDDs, such as MS (Nasrabady et al., 2018). Despite significant advancements, many aspects of myelin formation and function remain poorly understood, partly because models fail to replicate accurately human physiology. Consequently, there has been a long-standing need for robust human cell-based models of myelination to study myelin formation and the pathological processes underlying human diseases.

Organoids are particularly well-suited for research as they mimic aspects of spatial complexity and cellular diversity found in living organisms. In 2018, Madhavan et al. developed “oligocortical spheroids” by differentiating human PSCs and adding compounds, such as T3, clemastine or ketoconazole. These spheroids matured into functional oligodendrocytes, demonstrating myelin compaction and response to promyelinating drugs. Additionally, patient-derived spheroids mimicked disease-related abnormalities, making them a valuable tool for studying disease mechanisms (Madhavan et al., 2018). In 2019, Kim et al. created forebrain organoids using human PSCs engineered with an OLIG2-GFP knockin. These organoids successfully modeled human oligodendrogenesis, demonstrating that oligodendrocytes can originate from both ventral and dorsal regions. This model provided a valuable platform for investigating cortical myelin defects and regional oligodendroglial differences (Kim et al., 2019a). In the same year, human iPSCs were used by Marton et al. to generate brain organoids containing oligodendrocytes, neurons and astrocytes. Oligodendrocytes matured and began myelinating nearby neurons (Marton et al., 2019).

In 2021, James et al. developed myelinating organoids (“myelinoids”) derived from human iPSCs with genetic mutations to study oligodendrogenesis, compact myelin formation and myelinated axon organization. Pharmacologic perturbations altered myelin formation by reducing neuronal synaptic vesicles (James et al., 2021).

In 2023, Feng et al. generated human iPSC-derived myelin spheroids to model Canavan disease, a demyelinating disorder. By treating them with N-acetyl-aspartate, they were able to mimic key pathological features of the disease, establishing this model as a tool to study myelin diseases (Feng et al., 2023).

Myelinoids have been demonstrated to provide a robust platform for studying oligodendrocyte development, mechanisms of myelination, and cell-cell interactions in the CNS, holding potential for advancing research into white matter disorders, such as MS, and for developing therapeutic interventions.

Organoids are rapidly becoming a key tool in cell culture for a wide range of biomedical research. Their ability to represent diverse tissue types, sustain long-term growth, and recapitulate physiological 3D structures makes them an innovative technology with numerous biological and clinical applications. Importantly, organoids have found extensive use in areas such as disease modeling, precision medicine, toxicology testing, and regenerative medicine (Corrò et al., 2020). Next, we will describe their applications in disease modeling, biobanking, personalised medicine, neurotoxicity, identification and validation of biomarkers, drug screening, gene editing and biocompatibility (Figure 3).

As brain organoids are a relatively new technology, several challenges must be overcome to improve their effectiveness and reliability as models of the human brain. The primary issues include limited reproducibility and maturation, the formation of a necrotic core, and the absence of key brain characteristics such as gyrification, as well as essential non-neuronal cell types like microglia and oligodendrocytes.

A major challenge in organoid research is the lack of standardized protocols for reproducibility. Current methods often rely on poorly documented steps, require skilled personnel, and involve high costs, leading to variability across laboratories (Soldner and Jaenisch, 2018). The incorporation of diverse growth factors or nutritional components further complicates standardization, as each laboratory uses its own components (Kim et al., 2020). Commercial certified kits could minimize batch-to-batch variability. Simplifying protocols has shown promise, as evidenced by improvements in dorsal forebrain organoid generation (Velasco et al., 2019). Novel techniques like using micropillar arrays that enable the generation of contamination-free 3D cultures have improved neuronal differentiation and regionalization (Zhu et al., 2017). Hydrogels (Lancaster and Knoblich, 2014a) and 3D bioprinting (D’Antoni et al., 2023) are also emerging as alternatives to improve reproducibility and structural complexity. Moreover, the use of oncogenes for reprogramming iPSCs introduces risks, such as gene mutations, that may distort study outcomes, as they lead to genotype heterogeneity, resulting in diverse phenotypes (Halevy and Urbach, 2014). Culturing hPSCs in defined and xenogen-free conditions before detaching and seeding them to form EBs can enhance reproducibility and stem cell quality (Rothenbücher et al., 2021). Establishing accessible biobanks with clinical data for researchers could be a valuable approach, applicable to a variety of conditions, and form the foundation for future modeling strategies (Evangelisti et al., 2024). Variability can also be reduced by using genome-editing strategies for isogenic controls, considering polygenic risk scores (Soldner et al., 2009), and guiding regional fate with cytokines (Kim et al., 2020). These improvements could enhance reproducibility and translational value for disease modeling and drug testing. However, despite its transformative impact, CRISPR-Cas9 also faces challenges, including off-target effects, dependency on protospacer adjacent motif sequences, and immune responses to Cas9 components. Therefore, advances in engineering and delivery methods are essential to enhance its precision and applicability (Moon et al., 2019).

Cerebral organoids capture early stages of brain development, but they fail to form later structures, such as cortical plate layers. This limits their ability to study late-onset neurodegenerative stages such as synaptic loss or chronic gliosis. However, they provide a powerful platform to study early disease onset, including amyloid aggregation, mitochondrial stress, and neuroinflammatory responses (Raja et al., 2016). Efforts to induce maturation include using physiological glucose levels (Rocktäschel et al., 2019), telomerase inhibition (Vera et al., 2016), progerin overexpression (Miller et al., 2013) or removing antioxidant components from the culture (Kim et al., 2019b). Organoid fusion (Ahammed and Kalangi, 2024) and in vivo transplantation (Mansour et al., 2018; Revah et al., 2022) have also been shown to enhance maturity by exposing cultures to natural microenvironments. Genome-editing technologies could also be used to introduce mitochondrial dysfunctions associated with aging (Soldner et al., 2009). In addition to these approaches, direct reprogramming techniques, such as cellular transdifferentiation, offer the potential for age-preserved models by generating post-mitotic cells like neurons and glial cells, which could enhance aging studies in organoids. These advancements could enable the creation of multicellular, age-equivalent organoid models for studying aging and age-related diseases, overcoming some of the current limitations such as incomplete aging profiles and limited vascularization. Furthermore, by combining age-equivalent tissue models with heterochronic paradigms, future research could explore how aging processes interact with proliferative cells in 3D environments, opening new avenues for therapeutic interventions (Pitrez et al., 2024).

Beyond maturation issues, another critical limitation is the lack of vascularization, which leads to cell death in the core of the tissue, by restricting oxygen and nutrient delivery, hindering metabolite elimination, and disrupting cell signaling (Lancaster and Knoblich, 2014b). Therefore, many teams are working on different solutions. Coculturing organoids with endothelial cells, embedding them in ECMs or using HUVECs (Shi et al., 2020) fosters vascularization and enhances neurogenesis and maturation. Transplantation of organoids into animals (Mansour et al., 2018; Revah et al., 2022) and genetic engineering, such as using hETV2- a key transcription factor in vascular development-to induce vascular-like networks (Cakir et al., 2022), have also been successful in reducing hypoxia and necrosis. Other teams simply slice organoids to mitigate hypoxia (Choe et al., 2021), coculture vascularized brain organoids with choroid plexus organoids to produce cerebrospinal fluid and more complete vasculature (Pellegrini et al., 2020) Other solutions include spinning bioreactors, microfluidic devices (Lancaster and Knoblich, 2014a), ALI cultures (Cho et al., 2021), silk scaffolds (Sozzi et al., 2022), PCL scaffolds (Rothenbücher et al., 2021) and carbon fibers (Tejchman et al., 2020). All these advancements improve vascularization and nutrient delivery, creating more functional organoids for precise and cost-effective neurodegenerative disease research.

The lack of microglia, oligodendrocytes, and other essential cell types limits organoid utility for studying neuroinflammation and myelination. Introducing microglia or oligodendrocyte precursors into organoids has enabled the modeling of neuroinflammatory interactions and myelination processes (D’Antoni et al., 2023; López-Muguruza and Matute, 2023). Moreover, BBB organoids that include glial, vascular and neural cells may be a novel approach to model dysregulated neuroinflammatory processes implicated in NDDs such as MS(99,100).

Organoids lack cortical gyrification, a central developmental feature of the human brain, limiting their ability to successfully model human brain structure. Organ-on-a-chip systems (Karzbrun et al., 2018) and engineered flat brain organoids (Rothenbücher et al., 2021) may simulate mechanical and biological conditions driving the brain folding, offering insights into neocortical development and related disorders, and increasing the brain organoid’s relevance as a human brain model.

Lastly, the moral status of brain organoids requires consideration from multiple ethical perspectives. While current organoids are not sentient, ethical discussions aim to anticipate future challenges related to the possibility that brain organoids could 1 day exhibit consciousness-like features. The debate around brain organoids and consciousness arises because they might develop structures or functions similar to those of a developing human brain. If they reach a level comparable to a 20-week-old fetus, they could potentially have subjective experiences or suffer—even without feeling pain. This means they might have moral status, and research on them should be limited accordingly. Given the uncertainty, many argue that we should adopt a precautionary approach and treat potentially conscious organoids with ethical consideration. Categorizing brain organoids based on their maturation and interaction capabilities may help ensure ethical research (Koplin and Savulescu, 2019). Additionally, brain organoids offer a promising alternative to animal testing, aligning with ethical principles like the 3Rs (Replacement, Reduction, Refinement) to minimize animal use in neurobiological studies.

The development of brain organoids has revolutionized neuroscience by providing 3D models that reproduce some aspects of human brain complexity. Unlike traditional animal models, which often fail to fully reproduce human-specific neurophysiological processes due to species differences (Quadrato et al., 2016; Bjornson-Hooper et al., 2022), organoids offer a more physiologically relevant—though still simplified—representation of human brain development and disease mechanisms. Additionally, compared to 2D cultures, which lack the 3D cellular interactions and structural complexity required to capture key aspects of in vivo organization (Yang et al., 2023; Saraswathibhatla et al., 2023), organoids enable better cell-cell communication and microenvironment interactions. Compared to neurospheres, another 3D model, organoids offer superior complexity, replicating the organization, regional specification and cortical layering of the CNS, making them more suitable for studying complex neurodevelopmental processes (D’Antoni et al., 2023). Their ability to self-organize, generate diverse neuronal and glial cell populations (Clevers, 2016), and sustain long-term cultures (Quadrato et al., 2017; Trujillo et al., 2019) has significantly enhanced our understanding of neurodevelopmental and pathological processes, by capturing key cellular and molecular characteristics, facilitating the study of interactions between different cell types and assessing the impact of genetic and environmental factors on disease on-set (Choi et al., 2014; Raja et al., 2016; Kwart et al., 2019). Additionally, their use reduces reliance on animal experimentation, aligning with ethical principles.

Although some brain organoids are derived from ESCs, which raises additional ethical considerations, the use of iPSCs offers a valuable alternative, avoiding the moral implications associated with embryo use (De Wert and Mummery, 2003; Yang et al., 2023). Furthermore, as discussions about the potential for consciousness-like properties in brain organoids continue, it is expected that the scientific community will establish clear ethical and regulatory frameworks to guide future research in this area (Koplin and Savulescu, 2019), improving reproductivity and protocol harmonization across laboratories.

Organoids have demonstrated significant applications in biomedical research, including disease modeling, precision medicine, neurotoxicity evaluations, biomarker discovery, drug screening, gene editing, and biocompatibility testing (Corrò et al., 2020; Tang et al., 2022; Zhou et al., 2021; Romero et al., 2019). These models have enhanced our ability to study human-specific pathologies, test therapeutic compounds in a physiologically relevant environment, and develop personalized treatments based on patient-derived samples (Evangelisti et al., 2024; Tang et al., 2022).

One of the main advantages of organoids is their ability to model human diseases with greater fidelity than previous models. In the context of NDDs, the use of iPSCs is particularly advantageous, as they retain the phenotypic, genetic and epigenetic traits of the patient, allowing for personalized disease modeling by creating patient-specific and disease-specific cell lines (Halevy and Urbach, 2014; Dimos et al., 2008) and avoiding immune rejection after transplantation (Inoue et al., 2014). For conditions like PD and ALS, region-specific organoids provide the most effective approach, as they allow the generation of midbrain or spinal cord structures, which are the most affected regions in these diseases (Soldner et al., 2009; Andersen et al., 2020). By differentiating iPSCs into midbrain organoids for PD or spinal cord organoids for ALS, researchers can study region-specific neurodegeneration and screen targeted therapies more effectively (Jo et al., 2016; Faustino Martins et al., 2020).

For AD and MS, assembloids—fused organoids that interregional interactions—are the most effective models (D’Antoni et al., 2023; Qian et al., 2019). This is because AD involves widespread neurodegeneration across multiple brain regions, including the cortex and hippocampus, requiring interconnected organoids to simulate the complexity of disease mechanisms (Samarasinghe et al., 2021; Raja et al., 2016). Similarly, MS is characterized by immune system dysregulation and demyelination, which can be better studied in assembloids incorporating microglia and oligodendrocytes to model neuroinflammation and myelination processes (López-Muguruza and Matute, 2023; Madhavan et al., 2018).

Despite these benefits, organoid technology still faces multiple challenges. Before addressing these limitations, it is important to highlight additional benefits. Organoids provide a renewable and scalable source of human-relevant tissue, allowing long-term culture and high-throughput screening (Drost and Clevers, 2018). They also enable the study of patient-specific genetic mutations, the testing of gene-editing therapies like CRISPR-Cas9, and the development of patient-derived biobanks for future research (Evangelisti et al., 2024; Zhou et al., 2021; Li G. et al., 2022). In addition, generating assembloids (D’Antoni et al., 2023; Yang et al., 2023; Qian et al., 2019) and incorporating microglia (Cakir et al., 2022) and myelin-producing cells (López-Muguruza and Matute, 2023) have significantly improved the biological relevance of these models.

Current limitations of organoids include the lack of vascularization, leading to cell death in deeper layers due to insufficient nutrient and oxygen supply (Lancaster and Knoblich, 2014b), and limited neuronal maturation, which affects the modeling of late-onset diseases like AD and PD (Ahammed and Kalangi, 2024). For this last reason, the primary strength of organoids lies in their capacity to reproduce early-stage neurodegeneration, making them highly relevant for studying early on-set mechanisms, as it has been evidenced in PD (Morrone et al., 2024) and ALS brain organoids (Wang et al., 2022).

Strategies to address these issues involve organoid-on-a-chip systems with microfluidics for better vascularization and nutrient exchange (Cho et al., 2021; Castiglione et al., 2022), as well as maturation techniques such as adjusting glucose levels (Rocktäschel et al., 2019), telomerase inhibition (Vera et al., 2016), progerin overexpression (Miller et al., 2013), and reducing antioxidants (Kim et al., 2019b). Additionally, organoid fusion (Ahammed and Kalangi, 2024), in vivo transplantation (Mansour et al., 2018; Revah et al., 2022), and genome-editing for mitochondrial dysfunctions (Soldner et al., 2009) have been used to mimic aging processes. Co-culturing with endothelial cells (Lancaster and Knoblich, 2012; Pham et al., 2018; Shi et al., 2020), spinning bioreactors (Lancaster and Knoblich, 2014a), ALI cultures (Cho et al., 2021), scaffolds (Sozzi et al., 2022; Rothenbücher et al., 2021), and hydrogels (Rossi et al., 2018; Ranga et al., 2014) improve oxygen diffusion and support the growth of larger, more mature organoids. Standardization remains a challenge due to variability between cultures, but efforts to address this include simplified differentiation protocols (Velasco et al., 2019), automation (Evangelisti et al., 2024; Zhu et al., 2017), and bioengineering approaches (Soldner et al., 2009) to improve reproducibility and scalability.

In conclusion, brain organoids represent a powerful and continuously evolving tool for studying neuronal biology and nervous system diseases. While challenges remain, ongoing technological innovations—including organoid-on-a-chip systems, bioreactors, gene editing techniques and standardized protocols—suggest that these models could play a key role in translational medicine, bridging the gap between preclinical research and effective clinical applications (Yang et al., 2023; Qian et al., 2019; Corrò et al., 2020). Furthermore, integrating organoids with artificial intelligence and omics technologies could facilitate biomarkers identification and the development of personalized therapies for NDDs (Tang et al., 2022; Romero et al., 2019). The establishment of organoid biobanks would also expand access to these models for research and preclinical testing, fostering advancements in precision medicine (Evangelisti et al., 2024; Zhou et al., 2021).