BiologyHealth and Medicine

3D Organoids as Models for Human Brain Research

By Rachel Hull, Biochemistry and Molecular Biology, ’19

Author’s Note
I first became interested in this topic when I read a news article about a team of scientists that had successfully integrated what the article called “mini human brains” into mice. Although the idea seemed novel to me, after a little digging, I found that the technology that generates such mini organs — or organoids — has existed for more than a decade. This technology is nevertheless constantly evolving, and I believe it will continue to do so for years to come.

Three-dimensional human brain organoids, cultures derived from stem cells to mimic organs, represent a promising avenue for future research. These organoid models allow for unparalleled investigation into human brains’ complex development and alterations in response to particular factors, and they overcome the problems associated with model organisms and two-dimensional models. There are currently several approaches for producing such organoids. Serum-free suspension culture (SFEBq) protocols rely upon the self-assembly of stem cells and may use either selective or non-selective differentiation methods. These protocols allow for the recapitulation of some features of cortical development. SpinΩ, a miniaturized spinning bioreactor, overcomes some of the challenges of SFEBq methods such as high cost, low throughput, and lack of consistency. By virtue of their spontaneous astrocyte dispersion, human cortical spheroids (hSCs) reflect robust synaptic activity in a way not seen in the other techniques. Neocortical organoids’ single self-organized structure and circumvention of an animal-derived gelatinous matrix act in their favor as a model. All of these techniques show potential for modeling the elaborateness of the human brain.

Human brain organoids, stem cells, brain development, brain disorders

The intricacies of the human brain make it a daunting organ for scientists looking to unravel its mysteries. Many turn to models to give them a glimpse into complex brain-related issues such as neurodevelopment, neurodegenerative disorders, and even virus exposure [1, 2]. However, using model organisms such as rodents presents clear drawbacks because of the stark differences between rodents and humans [1, 5]. Similarly, two-dimensional monolayer cultures fail to accurately mimic the three-dimensional structure of the brain [2, 5]. In response to such constraints, researchers have developed methods of producing human brain organoids, 3D culture systems made up of brain tissue. These organoids can be thought of as miniature organs derived from human stem cells, and their applications range from modeling the generation of cortical tissue to the effects of cocaine on the developing brain [1-5]. Ultimately, 3D organoids represent a promising field for uncovering the inner workings of the human brain.

In a serum-free suspension culture (SFEBq culture), dissociated human embryonic stem cells (hESCs) spontaneously reaggregate and differentiate into neural cells. To allow for cortex growth beyond what would be seen in the first trimester, Kadoshima et al. optimized the SFEBq method. To do so, they developed a culture of hESCs in a medium containing Rho kinase inhibitor, which mediates reaggregation. Additionally, supplementing the medium with chemically defined lipid concentrate, the carbohydrate heparin, and a low concentration of Matrigel — an animal-derived gelatinous protein mixture — helped maintain cells from which the ventricular zone would eventually develop. This zone is a tissue layer that contains neural stem cells belonging to the central nervous system. The cells’ telencephalic differentiation could be enhanced by transforming growth factor beta (TGF-β) inhibitor and Wnt protein inhibitor, both of which affect the pathways through which cell fate is determined [3].

In Kadoshima et al.’s optimized culture, the cortical neuroepithelium (NE) — neural cells that have yet to differentiate — spontaneously polarized along the dorsocaudal-ventrorostral axis. This result was demonstrated by the expression in the NE of brain region-specific genes, as visualized through immunostaining. The NE also displayed a curving morphology along the axis as seen through live imaging and Rho kinase inhibitor treatment, which blocks the pathway that normally constricts this morphology. Kadoshima et al. used layer markers to reveal that the NE separated into three cortical neuronal zones in the later days of the culture. At this stage, they also observed through immunostaining that the outer subventricular zone (oSVZ), a cortical layer in the central nervous system, contained a population of Tbr2− Sox2+ Pax6+ progenitors [3].

What is notable about all of these results is that they elucidate the SFEBq-derived cells’ ability to recapitulate human cortical generation, also known as corticogenesis. Developing embryos display both axial polarity and rounding morphogenesis. The distinct zones seen in the SFEBq cultures mimicked the fetal neocortex during development even up to the second trimester. Furthermore, the Tbr2− Sox2+ Pax6+ progenitors are specific to humans and, before the development of this method, were notoriously difficult to locate and distinguish from other progenitors in the oSVZ. The three-dimensional context that Kadoshima et al.’s SFEBq culture provides is thus instrumental in overcoming this difficulty [3].

Cerebral organoids
Lancaster et al. utilize a modified SFEBq protocol to yield what they term cerebral organoids, a three-dimensional culture system that could arise from either hESCs or human induced pluripotent stem cells (hiPSCs) [1]. This modified protocol uses a nonselective differentiation method, instead of the cortex-specific differentiation culture of Kadoshima et al. [3]. To generate the cerebral organoids, Lancaster et al. first induced the differentiation of stem cells into neuroepithelial cells through carefully controlled chemical conditions. They thereafter maintained these cells in 3D culture and embedded them in drops of Matrigel. The Matrigel droplets were then spun in a bioreactor in order to increase their nutrient absorption, a process that led to the rapid development of cerebral organoids [1].

Lancaster et al. next carried out reverse transcription polymerase chain reaction (RT-PCR) for several markers to test the pluripotency and neural identity of the organoids. As the organoids became more differentiated, unsurprisingly, the pluripotency markers were downregulated and the neural identity markers were upregulated. Subsequent RT-PCR for forebrain and hindbrain markers revealed that both were present in the tissue. Additionally, the hindbrain markers diminished over time, just as they would during normal human brain development. Lancaster et al. utilized immunohistochemical staining to visualize an abundance of brain regions from the forebrain, midbrain, and hindbrain. They found that not only had these broad regions developed discretely, but that more refined regions and structures such as the choroid plexus and immature retina were also displayed [1].

Through staining and fluorescent live imaging, Lancaster et al. also observed that the dorsal cortical tissue organization and the radial glia’s morphology and mitotic behavior were reminiscent of what would be seen during human cortical development. Additionally, fluorescent live imaging revealed that the cortical neurons were able to exhibit neural activity as seen through spontaneous calcium spikes in individual cells [1]. Importantly, Lancaster et al. observed the unique distribution of human-specific progenitors within the oSZV just as Kadoshima et al. did using their own methodology [1, 3]. Thus they affirmed that their cerebral organoid model could capture cerebral cortical organization as would be seen in a human brain [1].

Lancaster et al. pioneered the method of using a spinning bioreactor to support the growth of cerebral organoids that represent key features of the developing human brain [1]. However, the large size, maintenance cost, and lack of experimental suitability act as severe limitations to this technology [2]. Furthermore, because SFEBq methodology involves the intrinsic self-assembly of human stem cells, the resulting organoids are typically made up of an abundance of cell types found in the forebrain, hindbrain, and retina [1]. What this diversity of cell types means is that cell samples vary greatly, making quantitative analyses more difficult and reducing the applicability of such organoids. Additionally, though these methods have shown promise toward shedding light on the elusive oSVZ, their progenitor populations have been sparse and the oSVZ layer underdeveloped [2].

To address this range of problems, Qian et al. designed a miniaturized spinning bioreactor called SpinΩ. Unlike the researchers using SFEBq methods, Qian et al. treated hiPSCs with SMAD protein inhibitors in order to pre-pattern the cells to the fate of particular brain regions. They next embedded the cells in Matrigel for seven days, removed them from the Matrigel, and spun them in SpinΩ. Immunohistological analysis showed that the generated organoids were forebrain-specific and homogeneous. Using layer neuron markers, they also found that the forebrain organoids effectively represented both the organization of multiple progenitor zones — including the oSVZ layer — and all six neuronal subtypes that would be observed during human cortical development; transcriptome comparisons validated these findings [2].

Further assessment revealed that the cortical neurons were functional and mature, as measured through their postsynaptic current activity and hyperpolarizing responses to gamma-aminobutyric acid. Additionally, Qian et al. used electrophysiological recordings and immunohistological analysis to uncover that forebrain organoids contain a variety of neuronal and other subtypes just as would be found in developing human brains. They even applied the SpinΩ technique toward producing midbrain and hypothalamic organoids. Overall, SpinΩ’s lower cost, higher throughput, and better reproducibility make it an advantageous method for the generation of 3D organoids [2].

Human cortical spheroids
Human cortical spheroids (hSCs), derived from hiPSCs, are three-dimensional structures that resemble the cerebral cortex. Unlike the previous methods, this one does not involve cell plating, embedding in extracellular matrices, or culturing under complex conditions. To generate the hSCs, Paşca et al. grew hiPSC colonies on inactivated mouse embryonic fibroblasts and then transferred them to low attachment plates. The floating colonies folded into spherical structures within hours and could thereafter be induced to differentiate using a bone morphogenetic protein inhibitor and TGF-β inhibitor. These structures were then moved to a serum-free medium containing fibroblast growth factor 2 (FGF2) and epidermal growth factor (EGF) [4].

Paşca et al. used probes, markers, and flow cytometry to confirm that the hSCs contained neurons from both the deep and superficial cortical layers; their corticogenesis also resembled that of in vivo human development. Imaging and current-clamp recordings revealed that the neurons displayed spontaneous calcium spikes and action potentials, pointing to their functional maturation. Subsequent investigations demonstrated that these neurons could also exhibit spontaneous synaptic activity that is linked to postsynaptic neuronal spike firing. Additionally, by increasing the hSCs’ exposure time to FGF2 and EGF, Paşca et al. could bring about the generation of glial cells called astrocytes, which play a critical role in synapse formation and function [4].

Significantly, the other aforementioned organoid generation methodologies do not allow for the spontaneous dispersion of astrocytes. The generation and spatial closeness of the astrocytes in the hSCs likely contributed to the robust synaptic activity that Paşca et al. observed. This method’s simplicity and reproducibility work in its favor as a model. Its amenability to slice physiology techniques adds to its favorability, as these techniques can be used to study complex synaptic events [4].

Neocortical organoids
Matrigel provides a platform for extracellular scaffolding and plays an instrumental role in the 3D culture systems developed by both Lancaster et al. and Qian et al [1, 2]. The use of this animal-derived gelatinous matrix, however, may introduce undefined animal factors and disrupt the delivery of drugs and other compounds into the generated organoids. While Paşca et al. avoid using Matrigel, their hSCs contain ventricular zone regions that vary in size and number, thus limiting their reproducibility and cell architecture quantification. In response to these challenges, Lee et al. developed a technique to produce what they call neocortical organoids from either hESCs or hiPSCs. These organoids, as their name implies, emulate the developing human neocortex [5].

To generate the neocortical organoids, Lee et al. cultured human stem cell colonies using irradiated mouse embryonic fibroblast feeder cells. After a period of growth in adherent culture, the researchers manually dissected rosettes — round groupings of cells — of a certain size and primed them in suspended culture. From this culture arose self-organized neocortical organoids. Treating the stem cells with combined dual SMAD and FGF inhibitors allowed for the generation of NE rosettes of an appropriate size for organoid development. The differentiation of these NE rosettes mimicked that of the human fetal brain by the middle of the second trimester, and astrocyte generation could be observed in the organoids. Additionally, whole-cell electrophysiological recording demonstrated the functional activity of the neocortical neurons [5].

What makes this model advantageous compared to the others is its avoidance of Matrigel and its production of a single self-organized neocortex-like structure. The former makes neocortical organoids an attractive option for quantitative pharmacological purposes, and the latter increases the organoids’ reproducibility by allowing for the generation of a predictable number and size of NE rosettes. Since the organoids effectively capture several important features of human cortical development, they could potentially act as a platform for further investigation of such development [5].

With the techniques behind each method outlined thusly, their various applications may now be explored. Kadoshima et al.’s SFEBq system lends itself well to investigation into early human corticogenesis. Future applications may include studying the formation of human cortical interneurons, the involved separation of cortical zones, and the mechanism of cortical NE growth [3]. Lancaster et al. focused on microcephaly as a worthwhile avenue to pursue using their cerebral organoids; experiments using organoids derived from a patient with microcephaly revealed that these organoids could provide some insight into the pathogenesis of this neurodevelopmental disorder [1].

Qian et al. utilized their forebrain organoid generated from SpinΩ as a model for Zika virus (ZIKV) exposure. Their results brought to light that ZIKV targets neural precursor cells and leads to deficits in cortical development [2]. The hSCs of Paşca et al. act as a promising platform for research into the patterning and specification of different neuronal and glial cell types, the screening of drugs in vitro, and the study of neuropsychiatric disorders [4]. Finally, Lee et al. applied their neocortical organoids toward elucidating the pathways through which cocaine affects human brain development. Other applications of neocortical organoids include uncovering the negative effects of psychostimulants, as well as modeling neurodevelopmental disorders that are found only in humans, such as autism spectrum disorder and schizophrenia [5].

Three-dimensional human brain organoids allow for modeling of the complexities of the human brain in a way that model organisms and two-dimensional models do not. Several methodologies for producing these organoids are currently available, termed SFEBq, cerebral organoids, SpinΩ, hSCs, and neocortical organoids. Each methodology has distinct assets and drawbacks, but all of them provide crucial insight into the human brain. Already some of their applications have been explored, ranging from studying human corticogenesis to delving into neurodevelopmental disorders to examining cocaine’s effects on the brain. Future avenues for research using these methods are abundant and include illuminating how NE thickness is controlled, how glial cells become patterned, and how pharmaceutical agents adversely affect humans. Through the use of 3D organoid models, researchers may find themselves one step closer to unlocking the secrets of the human brain.


  1. Lancaster MA, Renner M, Martin C-A et al. Cerebral organoids model human brain development and microcephaly. Nature 2013;501(7467):353-379. doi:10.1038/nature12517.
  2. Qian X, Nguyen HN, Song MM et al. Brain region-specific organoids using mini-bioreactors for modeling ZIKV exposure. Cell 2016;165(5):1238-1254. doi:10.1016/cell.2016.04.032.
  3. Kadoshima T, Sakaguchi H, Nakano T et al. Self-organization of axial polarity, inside-out layer pattern, and species-specific progenitor dynamics in human ES cell-derived neocortex. Proc Natl Acad Sci USA 2013;110(50):20284-20289. doi:10.1073/pnas.1315710110.
  4. Paşca AM, Sloan SA, Clarke LE et al. Functional cortical neurons and astrocytes from human pluripotent stem cells in 3D culture. Nat Methds 2015;12(7):671-678. doi:10.1038/nmeth.3415.
  5. Lee C-T, Chen J, Kindberg AA et al. CYP3A5 mediates effects of cocaine on human neocorticogenesis: Studies using an in vitro 3D self-organized hPSC model with a single cortex-like unit. Neuropsychopharmacology 2017;42(3):774-784. doi:10.1038/npp.2016.156.