Modeling human brain development has been hindered due to the unprecedented complexity of neural epithelial tissue. Here, a method for the robust generation of brain organoids to delineate early events of human brain development and to model microcephaly in vitro is described.
The restricted availability of suitable in vitro models that can reliably represent complex human brain development is a significant bottleneck that limits the translation of basic brain research into clinical application. While induced pluripotent stem cells (iPSCs) have replaced the ethically questionable human embryonic stem cells, iPSC-based neuronal differentiation studies remain descriptive at the cellular level but fail to adequately provide the details that could be derived from a complex, 3D human brain tissue.
This gap is now filled through the application of iPSC-derived, 3D brain organoids, “Brains in a dish,” that model many features of complex human brain development. Here, a method for generating iPSC-derived, 3D brain organoids is described. The organoids can help with modeling autosomal recessive primary microcephaly (MCPH), a rare human neurodevelopmental disorder. A widely accepted explanation for the brain malformation in MCPH is a depletion of the neural stem cell pool during the early stages of human brain development, a developmental defect that is difficult to recreate or prove in vitro.
To study MCPH, we generated iPSCs from patient-derived fibroblasts carrying a mutation in the centrosomal protein CPAP. By analyzing the ventricular zone of microcephaly 3D brain organoids, we showed the premature differentiation of neural progenitors. These 3D brain organoids are a powerful in vitro system that will be instrumental in modeling congenital brain disorders induced by neurotoxic chemicals, neurotrophic viral infections, or inherited genetic mutations.
Human neurodevelopmental disorders, such as microcephaly, can only be poorly studied in animal models due to the fact that human brains have an extended cortical surface, a unique feature differing from non-human animals.
This aspect makes human brain development a complex process that cannot be sufficiently studied in a 2D, in vitro cell culture system. Emerging 3D culture techniques allow the generation of tissue-like organoids from induced pluripotent stem cells (iPSCs). The in vitro differentiation of pluripotent stem cells in a 3D suspension culture allows the formation of various cell types in a timely and region-specific manner, giving rise to an organized, stratified tissue1,2,3. Thanks to laboratories that pioneered 3D culture technologies and demystified the complexity of organ formation, starting from stem cells, we developed a robust method of generating brain organoids to delineate early events of human brain development and to model microcephaly in vitro1,2,3. It is noteworthy that we adapted the original method developed by Lancaster et al. to generate cerebral organoids1. This method was modified according to our experimental requirements.
The aim of a study from Gabriel et al. was to analyze the cellular and molecular mechanisms of neural stem cell maintenance during brain development. In order to do this, a mechanistic study was performed by analyzing neural progenitor cells (NPCs) in 3D brain organoids derived from a microcephaly patient4. This patient carried a mutation in CPAP, a conserved centrosomal protein required for centrosome biogenesis5. A widely accepted hypothesis is that microcephaly is the result of a depletion of the NPC pool, and this might be due either to cell death or to premature differentiation1,6,7,8,9.
By analyzing the ventricular zones (VZs) of microcephaly brain organoids, it was shown that a significant number of NPCs undergo asymmetrical cell division, unlike brain organoids derived from a healthy donor4. Extensive microscopic and biochemical analyses of microcephalic brain organoids revealed an unexpected role for CPAP in timely cilia disassembly4. Specifically, mutated CPAP is associated with retarded cilium disassembly and delayed cell cycle re-entry, leading to the premature differentiation of NPCs4. These results suggest a role for cilia in microcephaly and their involvement during neurogenesis and brain size control10.
The first part of this protocol is a description of a three-step method to generate homogenous brain organoids. As mentioned before, the original Lancaster protocol was adapted and modified to suit our purpose1. First, human iPSCs are cultured in a defined feeder-free condition on Engelbreth-Holm-Swarm (EHS) matrix. This step avoids the variations of feeder-dependent pluripotent stem cell cultures. In this protocol, the induction of neural differentiation to form neural epithelium starts directly from iPSCs. By skipping the embryoid body (EB) formation step, the neural differentiation proceeds in a more controlled and directed manner. This approach limits the spontaneous and undirected formation of other germ cell layers, such as mesoderm and endoderm. By applying this protocol, neurospheres containing neural rosettes can be harvested on day 5 for EHS matrix embedding and stationary suspension culture. The organoid medium used for the third step of our protocol is supplemented with dorsomorphin and SB431542. Dorsomorphin is a small-molecule inhibitor of bone morphogenic protein (BMP), and SB431542 inhibits the TGFβ/activin/nodal signaling pathway. The combination of these factors could promote neural differentiation more efficiently than retinoic acid alone11,12,13,14.
Altogether, these modifications enable the reproducible generation of brain organoids, with minimal variations across organoids. Importantly, this method was applied to robustly generate microcephalic brain organoids from patient iPSCs, which carry mutations in genes that affect centrosomes and cell-cycle dynamics.
The second part of this protocol gives instructions to prepare brain organoids for the analysis and interpretation of cellular defects in microcephaly. This includes fixation, cryosectioning, immunofluorescent staining, and confocal microscopic analysis. This protocol will provide the reader with a detailed description of expected results and with guidance for interpretation.
1. Generation of Brain Organoids (23 days)
2. Embedding Neurospheres in EHS Matrix (4 days)
3. Organoids in a Rotary Suspension Culture (14 Days)
4. Analysis of Brain Organoids
5. Immunofluorescent Staining of Organoid Sections
NOTE: For the general characterization of organoids, staining with nestin, a neural progenitor marker, and TUJ1, a pan-neuronal marker, is recommended. As additional examples, immunofluorescent staining with phospho-Vimentin (p-Vim), which labels mitotic apical radial glial cells, and Arl13b, for cilium, are described. To test apoptosis, use the Terminal deoxynucleotidyl transferase (TdT) dUTP Nick-End Labeling (TUNEL) assay. Place the slides in a plastic box during the incubations to protect them from dust, light, and drying out.
The generation of brain organoids requires at least three weeks of continuous culturing (Figure 1A). To accomplish reproducible results, we recommend that the researcher documents every step and, importantly, avoids any alterations regarding medium components, time points, and cell handling. Here, we give a summary of how to evaluate if critical milestones are reached in order to obtain organoids of sufficient quality at the end of the experiment. The formation of neurospheres in a 96-well plate should be clearly visible from day 4. Neurospheres can be recognized as well-defined spheres on the bottom of each well (Figure 1B, step 1).
On day 5, neurospheres should be ~500 µm in diameter and exhibit a smooth surface, with a bright rim surrounding a dark center. It might be possible to already observe neural rosette-like extended structures within this bright rim (Figure 1C, start of step 2). If the neurospheres fall apart or appear more like cell aggregates, differentiation should not be continued further. Finally, organoids should increase and exhibit a similar size during the differentiation process in spinner flasks (Figure 1D, step 3). See the troubleshooting table (Table 1) for further information.
The quality of day-14 organoids should be verified by light microscopy. VZs are composed of thick layers of nestin-positive NPCs/radial glial cells with a palisade nuclear shape at the apical side. On the other hand, the primitive cortical plate will contain abundant TUJ1-positive neurons on the basal side, spatially distinct from the apical side (Figure 1E). Importantly, TUJ1-positive neurons should not be seen at the apical side of the VZ.
To analyze the division plane of apical radial glial cells (aRGs), we recommend using phospho-vimentin (p-Vim) and Arl13b staining. Arl13b labels the primary cilia of the innermost aRGs lining the lumen of the VZ. Thus, the Arl13b-positive ciliated region serves as an orientation guide to locate and to analyze the division plane of p-Vim-positive aRGs in anaphase. Typically, the VZ of a 14-day-old control organoid displays a significant number of aRGs, whose division planes are horizontally oriented. A horizontally oriented division plane is essential for the early symmetric expansion of aRGs (Figure 1F and G)23,24. When sufficient aRG expansion is reached, neurogenesis begins, the dividing cells change their orientation towards the lumen, and the division plane switches from horizontal (symmetric) to vertical (asymmetric)4,12,24.
Figure 1: Generation of Brain Organoids from Human iPS Cells. (A) Workflow of the differentiation protocol. Step 1: Start of differentiation. During this phase, the formation of neurospheres from human iPSCs in a 96-well plate occurs. The time duration is 5 days. Step 2: Embedding the neurospheres in EHS matrix droplets. A stationary suspension culture of EHS matrix-embedded neurospheres has a time duration of 4 days. Step 3: Organoids in spinner flasks. The transfer of EHS matrix-embedded neurospheres to spinner flasks takes place. The time duration is 14 days. (B) A neurosphere in a multi-well plate. Representative image of a neurosphere on day 4 in a 96-well plate during step 1 is shown. The sphere should be clearly visible on the bottom of the well, with a smooth, round surface (red arrow). (C) Morphology of neurospheres prior to EHS matrix embedding. Neurospheres collected on day 5 of step 1 should be uniform in size and display a smooth surface with a bright rim (bracket and arrow). (D) Organoids in spinner flasks. A spinner flask with brain organoids during step 3 (red arrows) is shown here. (E) Immunofluorescent imaging of cryosectioned organoids displaying a typical VZ and primitive cortical plate. The left panel shows the nuclear staining of cells at the VZ. The VZ spans from the apical side (lumen L) to the basal side (yellow line). Note that cells within the VZ display palisade-like nuclei, suggesting that they are radial glial cells. The right panel shows the immunofluorescent staining of nestin-positive NPCs (green) within the VZ and TUJ1-positive neurons (magenta) in the primitive cortical plate. Scale bar = 50 µm. (F) Immunofluorescent imaging of cryosectioned organoids. Two examples of VZs stained with p-Vim and Arl13b (i and iii). The regions marked by white squares are magnified at the right for each image inset (ii and iv). The division plane of anaphase apical radial glial cells (aRGs). Dividing apical radial glial cells at the apical side of the VZ are p-Vim-positive (magenta). Examples for symmetrically dividing aRGs are shown (white squares and insets). The division plane is given as a white line. The division plane of an anaphase cell is horizontal to the lumen surface line (yellow dotted line) and is marked by Arl13b staining (green). Scale bar = 50 µm. (G) This schematic provides the vertical or horizontal orientation of aRGs in anaphase relative to the lumen. The dividing cells of control organoids on day 14 are mostly horizontally oriented (0-30°) relative to the lumen surface of the VZ. The apical side of the lumen is lined by the primary cilia of aRGs, which specify the lumen surface of the VZ (green line). In contrast, most of the radial glial cells of microcephaly organoids display a vertically oriented (60-90°) division plane. The white line shows the axis of the division plane. Please click here to view a larger version of this figure.
Problem: | Possible reasons: | Suggestions: |
Step 1.1.1.2 | Poor efficiency of reprogramming | Check hiPSCs for pluripotency markers: if quality is low, manually pick undifferentiated colonies for few passages to enrich pluripotent colonies |
Human iPSCs differentiate before start of differentiation | ||
Stress due to early passaging or passaging as single cells | Do not passage before 80% confluency is reached. Passage as aggregates, not single cells | |
Contamination with mycoplasma | Test for mycoplasma contamination | |
Step 2 | Poor quality of hiPSCs | Improve quality of hiPSCs (see above) |
Neurospheres do not form at all or fall apart at day 5 after start of differentiation | Starting number of cells per well is too low or too high | Accurately count the number of cells and distribute them equally in each well |
hiPSC medium was used instead of neural differentiation medium | Use neural differentiation medium | |
Centrifugation step was too harsh or to mild | Spin 96-well plate 500 x g for 3 min and check if cells accumulated centrally at the bottom of each well | |
No Y-27632 at start of differentiation | Use Y-27632 to enhance the cell survival | |
Cells attached and grew at the bottom of the plate | Make sure that a non-adherent 96-well v-bottom plate is used. | |
Step 2 | No daily medium change | Change half amount of medium daily |
Neurospheres are not homogenous in size | Starting number of cells per well was not equal | Mix the tube with single cell suspension each time before taking out 100 µL cell suspension |
Medium change was not done for each well daily | Make sure every well gets medium change daily | |
Step 2 | hiPSC were not dissociated into single cells before the start of differentiation | Check under the microscope if hiPSC are dissociated to single cells after accutase treatment. If there are still aggregates, pipette cells 10 times up and down with 100 µL micropipette and check again or repeat accutase treatment |
Neurospheres do not display a bright rim or round surface; they form large cysts at the surface | Starting number of cells per well was too high | Accurately count the number of cells and distribute them equally in each well |
Medium change was not done daily for each well | Change half the medium daily | |
Step 2.7 | Poor quality of hiPSCs | Improve quality of hiPSCs (see above) |
EHS matrix embedded neurospheres stick and clump together | Not enough medium (volume) | Provide a minimum of 10 mL medium in a 100 mm petridish |
Shelf of incubator not even | Place dish on an even surface | |
After placing into incubator, move the dish so that neurospheres are distributed evenly |
Table 1: Troubleshooting Table.
MCPH is a complex human neurodevelopmental disorder that cannot be recapitulated in animal models in vivo or in simple human cell culture approaches in vitro. The clinical manifestation of MCPH begins to appear during the first trimester, when early neurogenesis begins. Thus, 3D brain organoids represent a reliable experimental system to model MCPH development. In addition, 3D human brain organoids are an ideal approach since i) they allow for the adaptation of a spectrum of patient samples with various genetic backgrounds, ii) they display organized tissue containing different neural cell types, and, importantly, iii) various differentiation stages of neurodevelopment in this in vitro approach are linked to their in vivo counterparts25,26,27. For example, the neural rosette is an equivalent structure to the developing neural tube.
Lancaster et al. used twice the number of patient iPSCs compared to control iPSCs in order to succeed in generating patient organoids. Of note, with the protocol described here, by modifying the existing methods, we could generate control and microcephaly brain organoids starting from the same cell numbers4. This was possible due to the defined culture conditions of iPSCs and due to a more directed differentiation. In this protocol, organoid generation was improved by replacing the commonly performed EB formation step by initiating neural differentiation directly from iPSCs. Secondly, in step 3, dorsomorphin and SB431542 were supplemented into the culture medium instead of retinoic acid alone. Retinoic acid is readily isomerized when applied to cell culture medium28. Therefore, its biological activity is less defined than that of more stable compounds such as dorsomorphin. Dorsomorphin inhibits the BMP4 signaling pathway, and SB431542 is an inhibitor of the activin/nodal signaling pathway (TGF-β1 ALK inhibitor). A combination of both compounds leads to the inhibition of BMP and the activin/nodal signaling pathway, induces neural differentiation, and reduces the differentiation into other lineages in various cell lines from human ES or human iPS cells with a similar efficiency29. Compound stability and universal cell response to a compound are important points for establishing a protocol that can be applied to different patient cell lines to model a disease in vitro. Thus, a robust and efficient differentiation results in homogenous brain organoid cultures and, eventually, in more reproducible data.
With these modifications, critical steps within the protocol are minimized to the initiation of differentiation in step 1. At this point, it is important to gently dissociate the iPSCs into a single-cell suspension and to distribute them accurately and evenly to each well of the 96-well plate. Rho-associated protein kinase inhibitor (Y-27632) must be added to medium B during the first 24 h, as it supports the survival of iPSCs upon their dissociation to single cells. It is important to bring cells in close contact with each other by spinning them down to the bottom of the wells. Once neurospheres are formed at the end of step 1, the successful completion of further experimental procedures can be expected. In case neurospheres do not form, the pluripotency of the human iPSCs, the non-adherence of the cells in the 96-well plate, and the centrifuging conditions must be checked. The whole procedure on day 0 of step 1 should take no longer than 1 h due to the sensitivity of human iPSCs. Medium changes during the following days must be done carefully, avoiding sheer forces, in order not to destroy freshly formed cell contacts between the cells.
It is noteworthy that centrosomal mutants have defective cell proliferation and altered cell-cycle dynamics. Thus, modeling MCPH requires a powerful protocol that can withstand the compromised cellular functions. Hence, the current protocol allows for the generation of homogenous organoids of high quality and serves as a unique tool to study stem cell homeostasis at the VZ. In contrast to other protocols, this protocol enables the generation of organoids from control and patient iPSCs, starting with equal cell numbers. For studies based on in vitro differentiation, identical culture conditions for different iPSCs are a basic requirement to identify disease-related alterations and to avoid culture-condition artifacts. Besides modeling microcephaly of genetic origin, this protocol can also be applied to microcephaly of non-genetic origin, including from neurotrophic viral infections, chemicals, or radiation.30 In addition, modern molecular biology tools, such as CRISPR/Cas9 genome-editing technologies, can be applied to 3D organoids to dissect specific aspects of the human brain development paradigm in vitro31,32,33.
On the other hand, the generation of brain organoids to date has still not gone beyond the first and early second trimester of human development34. Surpassing this limitation and enabling the generation of mature brain organoids in vitro would open new avenues to model neurodegenerative disorders that manifest at later stages, such as Parkinson's or Alzheimer's disease. For future applications, protocols directing differentiation to more specific regions, like the forebrain or midbrain, might be of interest to study complex neurological disorders like autism and schizophrenia35,36,37,38.
Importantly, certain aspects should be taken into consideration when drawing conclusions from organoid studies. The first limitation is that there is no obvious vascularization in organoids. Thus, gaseous exchange and nutrient supply in vitro only approximate in vivo conditions. Second, since brain organoids are not connected to complex organ systems, the organoid model lacks a complete immune, metabolic, and hormonal system. Nevertheless, the absence of these aspects sometimes provides an advantage. For example, the organoids described here could replace immunosuppressed in vivo models when addressing the impact of the immune system in the pathogenesis of brain disorders.
With the use of spinner flasks, a large number of organoids can be generated for biochemical experiments, whole transcriptome sequencing analysis, and high-throughput drug screenings. Taken together, by using this current protocol, one can generate brain organoids from human iPSCs cells, which can then be utilized in a broad range of applications, from disease modeling to drug testing platforms, in order to reliably replace animal trials in the future.
The authors have nothing to disclose.
This work was supported by the Fritz Thyssen Foundation (Az.10.14.2.152). We are grateful to the tissue embedding facility and the microscope core facility of CMMC. We are grateful for the discussions and technical support provided by the members of the Laboratory for Centrosome and Cytoskeleton Biology. We thank Li Ming Gooi for proofreading the manuscript.
Anti-mouse 488 | Invitrogen | A-11001 | Goat anti-Mouse IgG (H+L) Secondary Antibody, Alexa Fluor 488 |
Anti-rabbit 647 | Invitrogen | A-21245 | Goat anti-Rabbit IgG (H+L) Secondary Antibody, Alexa Fluor 647 |
Arl13b | proteintech | 17711-1-AP | ARL13B rabbit polyclonal antibody |
CELLSPIN system | IBS Integra Bioscience | 183001 | |
DAPI | Sigma-Aldrich, US | 32670 | 4′,6-Diamidino-2-phenylindole dihydrochloride; multiple suppliers |
DMEM/F-12 | Gibco, US | 31331093 | Dulbecco's Modified Eagle Medium: Nutrient Mixture F-12 |
Dorsomorphin | Sigma-Aldrich, US | P5499 | Compound C; multiple suppliers |
Embedding medium | AppliChem | A9011, 0100 | Mowiol; embedding medium; multiple suppliers |
Engelbreth-Holm-Swarm (EHS) matrix | Corning | 354277 | Matrigel hESC-qualified matrix; important: hESC qualified |
Fish gelatin | Sigma-Aldrich, US | G7765-250ML | Gelatin from cold water fish skin; multiple suppliers; autoclave after adding to PBS to dissolve and sterilize, store at 4°C |
Glycine | AppliChem | A1067,1000 | Glycine for molecular biology; multiple suppliers |
Inoculation loop with needle, disposable (1 µl) | Sigma Aldrich, US | BR452201-1000EA | multiple suppliers |
Insulin | Sigma-Aldrich, US | I3536-100MG | multiple suppliers |
L-glutamine | Gibco, US | 25030081 | L-glutamine (200 mM) |
Medium A | Stem cell technologies | #05850 | mTeSR1 (hiPSC medium) |
Medium B | Stem cell technologies | #05835 | Neural induction medium (NIM); neural differentiation medium |
Medium C | Gibco, US | 21103049 | Neural Basal Medium |
MEM | Gibco, US | 11140035 | MEM non-essential amino acids solution (100x) |
MycoAlert Mycoplasma Detection Kit | Lonza, Switzerland | #LT07-218 | Mycoplasma detection kit; multiple suppliers |
Nestin | Novus biologicals | NBP1-92717 | Nestin mouse monoclonal antibody (4D11) |
Paraformaldehyde (PFA) | AppliChem | A3813, 0500 | 4% in PBS, store solution at -20°C; caution: wear skin and eye protection and work under hood |
PBS tablets | Gibco, US | 18912014 | See manufacturer´s instructions; multiple suppliers |
Penicillin-Streptomycin (10.000 U/ml) | Gibco, US | 15140122 | Multiple suppliers |
Poly-L-lysine solution (PLL) | Sigma-Aldrich, US | P8920-100ML | Multiple suppliers |
pVim | MBL | D076-3S | Phospho-Vimentin (Ser55) mAb |
Reagent A | Stem cell technologies | # 05872 | Note to Protocol 1.1.1.2; ReLSR (Enzyme-free human ES and iPS cell selection and passaging reagent); please follow manufactorer´s protocol; alternative products from muliple suppliers available |
Reagent B | Sigma-Aldrich, US | A6964-100ML | Accutase solution is an enzymatic solution for single cell dissociation; multiple suppliers; protocol 1.1.2 "enzymatic cell dissociation solution” |
Research Cryostat Leica CM3050 S | Leica biosystems | CM3050 S | Multiple suppliers |
SB431542 | Selleckchem.com | S1067 | Multiple suppliers |
Spinner flask 250 ml | IBS Integra Bioscience | 182026 | |
Sucrose | AppliChem | A4734, 1000 | Multiple suppliers |
Superfrost ultra plus microscope slides | Thermo scientific, US | J3800AMNZ | Slides should be labeled with a "+" and positively charged |
Supplement 1 | Gibco, US | 17502048 | N-2 supplement (100x) |
Supplement 2 w/o Vitamin A | Gibco, US | 12587010 | B-27 supplement (50x), minus vitamin A; multiple suppliers |
Tissue-Tek Cryomold | Sakura, NL | 4565 | Multiple suppliers |
Tissue-Tek O.C.T. compound | Sakura, NL | 4583 | Multiple suppliers |
Triton X-100 | AppliChem | A1388,0500 | Multiple suppliers multiple suppliers |
TUJ1 | Sigma-Aldrich, US | T2200 | β-Tubulin III (rabbit polyclonal) |
TUNEL assay | Promega, US | G3250 | DeadEnd Fluorometric TUNEL system; multiple suppliers |
Tween 20 for molecular biology | AppliChem | A4974,0500 | Multiple suppliers |
waterproof sheet | BEMIS company, inc. | PM996 | Parafilm “M”; multiple suppliers |
Y-27632 | Selleckchem.com | S1049 | ROCK-inhibitor (Y-27632 2HCL); multiple suppliers |
β-mercaptoethanol | Gibco, US | 31350010 | 2-mercaptoethanol (50 mM); multiple suppliers |