Generation of human midbrain organoids from induced pluripotent stem cells

Background

Parkinson’s disease (PD) is a neurodegenerative disorder, affecting more than 1% of the population over 65 years of age. The majority of cases are idiopathic, while about 10% have been linked to genetic mutations. Classical hallmarks of PD are the loss of dopaminergic (DA) neurons in the substantia nigra pars compacta, accompanied by the presence of neuronal inclusions called “Lewy bodies”. Several cellular pathways have been implicated in PD pathogenesis, including mitochondrial dysfunction^1–3, perturbed neuronal activity^4,5 and dysregulated protein homeostasis due to lysosomal, autophagy and proteasomal defects^6–9. However, there is no treatment to halt the progression of the disease. To date, treatment of PD is limited to symptom management. It is therefore necessary to refine the models we use in fundamental research to understand the pathophysiology of PD and to develop more effective therapeutic strategies.

In 2006, Drs. Kazutoshi Takahashi and Shinya Yamanaka described the reprogramming of somatic cells into induced pluripotent stem cells (iPSCs)¹⁰. Since their discovery, this technology has opened up many new avenues of investigation, including research for PD. With their self-renewal abilities and potential to be differentiated into disease-relevant cells from all three developmental lineages, iPSCs provide a unique tool to study PD within a human neuron, without the difficulties in obtaining neurons from a human brain¹¹. iPSCs can be directly reprogrammed from skin, blood or urine of an individual without raising the ethical concerns previously triggered by the use of fetal stem cells. In 2009, Soldner et al. were the first to describe the generation of an iPSC cell-line from a patient with sporadic PD, and the subsequent differentiation of these cells into dopaminergic (DA) neurons¹². Taking advantage of iPSC technology, many studies have started to investigate the pathological mechanisms of PD in iPSC-derived DA neurons from patients, compared to neurons from healthy individuals^13–33.

Recently, the development of organoids has become a major technological advance in the stem cell field and represents a bridge between traditional 2D cultures and in vivo mouse models. In 2013, Lancaster et al. described a novel 3D model called a cerebral organoid, that recapitulated different areas of the human brain³⁴. Kept in culture, these organoids formed a complex self-organized neuronal tissue composed of a mixed population of neurons, astrocytes and oligodendrocytes. In contrast to neurospheres, the cells were organized in layers that, at early stages, included a ventricular zone composed of progenitors. Neurons within these organoids were functional and had spontaneous electrical activity in networks. Interestingly, brain organoids could be cultured for long periods to obtain morphologically and functionally mature cells, in contrast to neurosphere cultures^34–37.

Since 2013, different types of brain organoids have been generated based on adaptations of the initial protocol published by Lancaster. Earlier, the protocols involved no external addition of growth factors in the medium, resulting in self-differentiating cells. However, different laboratories now directly drive the stem cells towards specific fates. The key to efficient brain organoid generation is a defined combination of inductive signals and physical factors that drive pluripotent stem cells to form 3D brain organoids. The modulation of these factors gives rise to multiple types of brain organoids that can be used to model neurodevelopmental and neurodegenerative disorders that affect distinct regions within the brain. Protocols now exist for making human cerebral^34,38,39, forebrain-like (dorsal and ventral)^40,41, cerebellar⁴², cortical-like (dorsal and ventral)^43,44, hippocampal and choroid plexus-like tissue⁴⁵, midbrain^35,36,41, hypothalamic⁴¹, pallium and subpallium⁴⁶ brain organoids.

iPSC-derived human brain organoids recapitulate brain development and can be used to study normal neurodevelopment. Brain organoid development recapitulates the early to mid-fetal development, and the epigenomic signatures of the human foetal brain^41,47,48. So far, cerebral organoids have been used to study pathologies including microcephaly³⁴, Zika virus infection^49–52, and autism spectrum disorders^46,53.

Recently, human brain organoids have been used to investigate aspects of neurodegenerative disorders. Two groups generated cerebral organoids from iPSCs of Alzheimer’s disease (AD) patients carrying familial mutations for presenilin 1 or amyloid precursor protein duplication, and successfully recapitulated the aggregation of beta-amyloid protein and tau pathology (hyperphosphorylation and aggregation), the two neuropathological markers of AD, in a human model. Treatment of the 3D cultures with drugs targeting either amyloid-beta aggregation or tau phosphorylation decreased the pathological markers^54,55. These promising results demonstrated that human brain organoids represent a relevant model for drug discovery. The development of different types of brain organoids represents a major advance in the stem cell field. In particular, the development of midbrain organoids represents a new drug discovery tool for PD. Two groups published similar protocols to generate the human midbrain organoids, based on specific inductive signals introduced at specific stages in the 3D cultures to drive the stem cells towards a midbrain fate^35,36. They showed that the midbrain organoids are composed of functional midbrain neurons producing neuromelanin granules, a by-product of dopamine synthesis. Of the neuronal population, 30% was myelinated due to the presence of oligodendrocytes. Interestingly, Monzel et al., showed the presence of nodes of Ranvier and spontaneous saltatory transmission^35,36. Considering the mix of neuronal populations connected within the midbrain organoids, they represent an interesting model to discover new pathological mechanisms involved in PD.

In this paper, we provide a standardized protocol for a robust derivation of iPSCs lines into 3D midbrain organoids. This protocol is an adaptation of the Nature protocol paper initially published by Lancaster et al. in combination with discoveries from Jo et al. and Monzel et al.^34–36 in order to successfully produce high quality midbrain organoids. We also describe a cryosectioning protocol that we adapted to produce high-quality histological sections from midbrain organoids, overcoming difficulties resulting from the particular texture of cultured tissue as well as their small size, relative to rodent brains. Taken together, this article extensively explains the methods involved in generating these iPSC-derived organoids and their histological analysis.

Materials

The materials, reagents and equipment listed in this document can be substituted for comparable items. However, the performance of the protocol may not be the same and may need to be optimized or redeveloped upon significant modifications to the materials and/or methods. It is also important to note that there is a significant lot-to-lot variability for certain reagents. In order to monitor this variability, we recommend a systematic test of new batches.

Method

Protocol description for iPSC culture and maintenance

General principles for culturing and maintaining human iPSCs. Human iPSCs are sensitive to many stresses, including shear stress, heat shock, and changes in media formulation and must be manipulated with extreme care at all steps of the protocol.

Technical and safety considerations for manipulating IPSCs

• The iPSC colonies should not have been passaged more than 10 times after thawing.

• The cells need to be a minimum of passage 2 after thawing for generation of midbrain organoids.

• Do not work with colonies that present with differentiated areas (Figure 1a, Figure S1a). We recommend removing the differentiated areas during daily maintenance and to never exceed 5% of differentiated areas prior to organoid generation for optimal organoid formation.

• Sterile technique must be used at all times when working with cells or in preparing reagents and materials.

• Human iPSC lines are to be handled within a Class II biosafety laminar flow hood to protect the worker from possible biohazardous agents.

Figure 1. Midbrain organoid generation, timeline and steps of tissue formation.

a) Quality of iPSCs suitable for midbrain organoids formation. b) Timeline for midbrain organoid generation. PBMCs (peripheral blood mononuclear cells) from individuals were collected and reprogrammed into iPSCs. Commercial lines can be alternatively used that were reprogrammed from skin or PBMCs or other somatic sources. Uniform embryoids bodies are formed from iPSC colonies and then patterned into neuronal midbrain fate with inductive signals. EBs then sit 24 hours in tissue induction media post-embedding in Matrigel^® scaffold at day 7 to promote growth of tissue. The tissue formed was cultured on an orbital shaker for several weeks or months until their use in experiments. c) EB with smooth edge 48 hours after formation. d) Extrusion of buds on EB after midbrain patterning. e) Typical EB 24 hour after embedding in Matrigel^®. f) Day 1, 5 and 15 after transferring the tissue into final differentiation media. Scale bar= 500 μm.

Figure S1. Additional informations.

a) Typical differentiated areas of NCRM1 iPSC cells cultured in mTeSR^TM (red arrows). Courtesy of Dr Carol X.-Q. Chen. b) SNCA_Tri reached approximately 4mm in diameter c) 100 day-old SNCA_Tri cryosections stained for TH (red), MAP2 (green) and Hoechst (blue).

Preparing Matrigel® hESC-qualified matrix coated culture dishes, thawing frozen human iPSCs cryovials, and daily maintenance of human iPSCs

• Start by thawing an aliquot of 150–200 µl Matrigel® hESC-qualified Matrix at 4°C, 15–20 minutes prior to use (depends on the dilution factor of each lot per manufacturer instructions).

• Prepare DMEM/F-12 + GlutaMAX^TM-I solution by removing 5 mL of DMEM/F-12 + GlutaMAX^TM-I from a 500 mL bottle and adding 5 mL of Antibiotic-Antimycotic. Place the bottle at 4°C until cold.

• Once the Matrigel® is thawed and DMEM/F-12 + GlutaMAX^TM-I is cold, prepare a coating solution diluted as per manufacturer instructions and mix well.

• Immediately use the diluted Matrigel® solution to evenly coat the tissue culture dishes (2 ml/ 60 mm dish and 5 ml/ 100mm dish) by swirling the tissue culture dish in each direction, multiple times. Incubate at 37°C for a minimum of one hour before use.

• Meanwhile, warm up mTeSR^TM 1 and DMEM/F-12 + GlutaMAX^TM-I medium at room temperature (RT) for a minimum of one hour. Do not use a water bath during this process because the media components would be degraded.

• Once the tissue culture dish is coated and mTeSR^TM 1 and DMEM/F-12 + GlutaMAX^TM-I medium are warmed, get the frozen cryovial of iPSCs from the liquid nitrogen storage container, and quickly thaw the vial by warming in a 37°C water bath. Continuously shake the cryovial until only a small frozen pellet remains.

• Using a 5 mL glass pipette, transfer the contents of the cryovial to a 5 mL solution of DMEM/F-12 + GlutaMAX^TM-I (with Antibiotic-Antimycotic), in dropwise manner and gently pipette up and down 1–2 times.

• Centrifuge the cells at 1200 rpm for 3 minutes at RT. After centrifugation, aspirate the medium, leaving the cell pellet intact. Using a 5 mL glass pipette, gently resuspend (1–2 times) the cell pellet in 3 mL of mTeSR^TM 1 medium containing Y27632 (1:1000 dilution).

• Take the coated tissue culture dish from 37°C and aspirate out the coating medium. Transfer the cells in mTeSR^TM 1 medium to the coated dish and place in a 37°C, 5% CO₂ incubator. Do not disturb the dish for 24h, to allow the cells to attach.

• Change the medium every day with mTeSR^TM 1 without Y27632, until the cell confluency reaches 70% and cells are close in contact.

• Daily, visually identify regions of differentiation under the microscope and remove them by aspiration under the sterile hood (Figure S1a).

Passaging of human iPSC lines

• When the cells reach 60–70% confluency, they are ready to be split.

• Aspirate out the medium from the dish and wash the cells with 3 mL DMEM/F-12 + GlutaMAX^TM-I (with Antibiotic-Antimycotic).

• Add gentle cell dissociation medium (5 ml for 100 mm dish and 1 mL for 60 mm dish) at RT until the cells at the edge of the colony begin to detach from each other. (Note: Do not leave cells for longer as the cell viability will be affected)

• Aspirate out the gentle cell dissociation medium from the dish and wash the dish with DMEM/F-12 + GlutaMAX^TM-I.

• Add 5mL DMEM/F-12 + GlutaMAX^TM-I to the dish and gently detach the colonies by scrapping with a cell scraper. Using a glass pipette, transfer the detached cell aggregates in a Falcon tube.

• Centrifuge the Falcon tube with cells at 1200 rpm for 3 minutes.

• After the centrifugation is done, check the pellet and aspirate out the supernatant (do not let the pellet dry so leave 100 μL-200 μl DMEM/F-12 + GlutaMAX^TM-I after aspiration).

• Using a glass pipette, resuspend the pellet gently to break the cell aggregates in mTeSR^TM 1 medium with 10μM Y27632.

• For seeding cells, add 10% of the original cell suspension to a new coated dish when passaging between similar size dishes. For a small to large dish passage-take 20% of the original cell suspension and from a large to small dish, take 5%.

• Incubate the dishes at 37°C until the colonies reach 70% confluency.

Freezing of human iPSC lines

• As described in the previous section, detach cells with gentle cell dissociation medium, and determine cell number with a cell counter prior to pellet the cells by centrifugation.

• Gently aspirate supernatant. Gently resuspend the cell pellet in FBS in half the volume needed to result in 1 mL of cell aggregate mixture per cryovial containing 1.5-2 millions of cells.

• Add an equal volume of 20% DMSO/FBS to the cell aggregate mixture to obtain a final DMSO concentration of 10%. Mix well with three up and down.

• Transfer 1 mL of cell aggregate mixture to each cryovial.

• Place cryovials in a cryo-box and transfer cryo-box to –80°C for overnight storage.

• The following day, transfer cryovials to liquid nitrogen.

Generating midbrain organoids

Eight days are necessary to generate midbrain organoids from iPSC colonies. Then, the organoids are grown with constant agitation for maturation. The timeline is described in Figure 1b.

Seeding of iPSCs – Day 0

• Start with one 10 cm dish of iPSCs at 70% confluency, cultured in mTeSR^TM 1 to obtain 96 midbrain organoids from a 96-well ultra-low attachment plate.

• Dissociate cells using Accutase by removing medium from cells, wash cells once with DMEM/F-12 + GlutaMAX^TM-I + Antibiotic-Antimycotic and add 5 ml Accutase at RT. Incubate cells for 3 minutes at 37°C, then stop reaction with DMEM/F-12 + GlutaMAX^TM-I + Antibiotic-Antimycotic. Transfer the medium with the cells into a 15 mL Falcon tube and centrifuge for 3 min, at 1200 rpm in a regular Falcon centrifuge. Remove the supernatant.

• Using a 1mL tip combined with a 200 µl tip resuspend cells in 5 ml of neuronal induction medium (Table 3) by pipetting gently up and down 3 times. Count cells. Note: The 200 µl tip is at the end of the 1 ml tip.

• Plate 10,000 cells/well in an ultra-low attachment 96 well U-bottomed plate with a multichannel pipette and add neuronal induction medium (Table 3) to a total of 200 µl/well with a multichannel pipette.

• Centrifuge the plate for 10 min at 1200 rpm at 37°C.

• Incubate cells for 48 hrs at 37°C in a regular cell incubator.

Note: We tested 96-well plates U-bottom and V-bottom ultra-low attachment (data not showed). While the formation of EBs were similar in both type of plates, we noticed a better embedding (Day 7) in U-bottom 96-well plates rather than in V-bottom 96-well plates.

Change medium – Day 2

Observation: EBs should have smooth and round edges (Figure 1c)

• Change media to neuronal induction medium WITHOUT ROCK inhibitor using a multichannel pipette.

• Incubate cells for 48 hrs at 37°C in a regular cell incubator.

Change medium – Day 4

Observation: EBs should reach a diameter of >300 µm (typically 400–600 µm) with smooth and round edges

• Change medium to midbrain patterning medium (Table 4) using a multichannel pipette.

• Incubate cells for 48 hrs at 37°C in regular cell incubator.

Change medium and embedding in Matrigel® – Day 7

Observation: Neuroectoderm buds should have started to extrude before the embedding (Figure 1d)

• Transfer aliquots of Matrigel® reduced growth factors to ice and let it reach 4°C.

• Remove the medium from each well with a multichannel pipette and add 30 µl of reduced growth factor Matrigel® with the manual repeater-pipette and sterile distritips.

• Incubate the plate for 30 min at 37°C in a regular cell incubator.

• Add 200 µl of tissue induction medium (Table 5) and incubate for 24 hours at 37°C in a regular cell incubator.

• Autoclave a box of 1mL cut tips if not already done.

Transfer the organoids to final differentiation medium – Day 8

Observations: Neuroepithelium should be more developed (Figure 1e)

• Add 3 mL of final differentiation medium per well of an ULTRA LOW ATTACHMENT 6 well plate.

• Transfer the midbrain organoids with a cut 1000 µl pipette tip into the 6 well plate (5 organoids/well).

• Put the 6 well plates on the orbital shaker (set at 70 rpm for the orbital shaker listed- note: speed setting would change based on the shaker diameter) in the 37°C regular cell incubator and change the medium every 2–3 days using final differentiation medium (Table 6). When the midbrain organoids reach 1 mm in diameter, increase the final differentiation media volume to 5 mL to provide enough nutrients (Figure 1f).

Note: By putting 5 organoids per well, we limit the number of media changes to three times a week once the hMOs attain the maximum size of 4mm.

Observations: From day 1 in final differentiation media to day 50 the organoids should grow from 600 μm to approximately 4 mm (Figure 2a.b).

Figure 2. Midbrain organoid characterization.

a) Difference in size expected from day 1 to day 50 in final differentiation media. The hMOs derived from NCRM1 line will grow from 600μm to 4mm. Scale bar= 4mm. b) hMOs derived from XCl-1 line achieved similar growing efficiency and reached approximately 4mm. Scale bar= 4mm c) Quantitative real-time PCR depicting normalized expression level of midbrain (EN1, Nurr1, LMX1B, LMX1A, MAOB, Calb1, TH, COMT, DDC), noradrenergic (DBH), and serotonergic (GBX2) markers in XCl-1 hMOs 66 days, compared to endogenous GAPDH and actin controls (n=3, mean+/-SEM).

Immunostainings, images acquisition, single cell RNA sequencing and statistical analysis

Immunostaining and Fontana Masson staining. Cryosections were rehydrated in PBS for 15 min and surrounded with a hydrophobic barrier using a barrier pen. The sections were then blocked for one hour at room temperature in a humidified chamber, with blocking solution (5% of normal donkey serum, 0.05% BSA, and 0.2%Triton X-100 in PBS. They are then incubated overnight at 4°C with primary antibodies diluted in blocking solution (See Table 9). The following day, cryosections were washed three times in PBS, fifteen minutes each, and then incubated in secondary antibodies diluted in blocking solution (See Table 9) for one hour at room temperature. Then we washed the sections three times in PBS for fifteen minutes each. Hoescht (diluted 1/5000 in PBS) was incubated 10 min on sections, followed by a wash in PBS for 10 min. Finally, we mounted the section with an aqua-mounting media and visualized the staining under a confocal microscope (Figure 3).

Figure 3. Composition of hMOs.

a) Cryosection of 30 day-old XCl-1 derived hMO. The immunofluorescence staining for tyrosine hydroxylase reveals dopaminergic neurons (red) and nuclei (blue). We observe multiple ventricle-like structures or “rosettes”. b) A typical rosette in XCl-1 derived hMOs is composed of neural progenitor cells, including FoxA2 positive progenitors (squares), and differentiated cells MAP2 positive (triangles). c) Cryosections of 30 day-old SNCA_Tri derived hMOs stained for neurons (MAP2), dopaminergic neurons (TH) and nuclei (Hoechst). Scale bar = 1 mm. We observe the dopaminergic neurons stained with tyrosine hydroxylase TH (red), and the neurons stained for MAP2 (green). The cells negative for MAP2 and TH are progenitors. Higher magnification of dopaminergic neurons revealed by TH staining. Scale bar = 100μm d) Single cell RNA sequencing of two 47 day-old SNCA_Tri background derived hMOs. Unsupervised clustering yielded 8 clusters representing different cell types. e) Heat map of dopaminergic marker expression levels within each cluster f) Violin plot of the expression levels of the dopaminergic marker TH in each cluster and table showing the percent of cells expressing TH in clustering of neurons.

Fontana Masson stainings (Figure 4b) were performed with an Abcam staining kit (#ab150669) following provider’s instructions on regular paraffin sections of hMOs.

Figure 4. Dopaminergic neurons release by-products of dopamine synthesis.

a At day 35 in final differentiation media, midbrain organoids are treated with 100 μm dopamine for 10 days. Under dopamine treatment, brown/black areas appeared. b) Fontana Masson staining confirmed the presence of neuromelanin granules with dopamine treatment. Black dots are extracted with colorimetric selection from GIMP software and quantified by ImageJ⁶³. c) Statistical analysis of relative neuromelanin granules number in GraphPad Prism, unpaired t-test, p*** <0.001, 4n. d) Dopaminergic neuron revealed by TH staining with DAB chromogen and counterstained with Romanowsky-Giemsa revealing neuromelanin in dark green (arrow)⁶⁴.

Imaging. iPSCs colonies (Figure 1a, Figure S1a) were imaged with an inverted microscope Motic AE2000 and the Moticam BTO camera associated, while the EBs images (Figure 1c–f) were taken with a transmitted light microscope EVOS XL Core. The hMOs (Figure 2 and Figure 4a, Figure S1b) were imaged with a ZEISS Stemi 508 stereomicroscope combined with a ZEISS Axiocam ERc 5s camera. The fluorescence images (Figure 3, Figure S1c) were acquired with a Leica TCS SP8 confocal. Fontana Masson stainings (Figure 4b,d) were acquired with a clinical microscope Olympus BX46 and an Olympus DP27 digital color camera associated.

Raw fluorescent images were opened in ImageJ software (version 2.0.0-rc-69/1.52i) with a red, blue, green or yellow color associated to each channel, before all images were merged to create a merged image. Black dots from Fontana Masson stainings pictures were extracted with colorimetric selection from GIMP software (version 2.8.22) and quantified by ImageJ (version 2.0.0-rc-69/1.52i) following the method described in 63. Briefly, using GIMP software the pixels associated with neuromelanin staining were colored extracted, and quantitating the number of extracted pixels using Image J. A histogram of the image was created, which separates the total number of pixels in the image into 255 color categories spanning the visible spectrum. The peak corresponding to the brown‐black colour i.e. neuromelanin was determined by cutting and summing the appropriate counts from each channel of the melanin peak.

Single Cell RNA sequencing. After dissociation, the single cell suspension in PBS with 1% BSA was put on ice. Cell viability was determined with live dead staining kit and approximately 5000 cells were loaded per lane in 10X Genomics Chromium 2 single cell sequencing chip. The samples were processed following the 10X protocol to prepare cDNA libraries for next-gene sequencing. The sequences were aligned to the human genome (CRCh38) and de-multiplexed to match RNA sequences with cell barcodes using 10Xcell Ranger. The R package Seurat used to analysis the single cell libraries (R notebook appended). The sequences quality was confirmed by checking the number of unique RNA sequences for each cell (nFeature_RNA) and the total number of RNA sequences in each cell (nCount_RNA), where were both in the expected range. The percent of total RNA that was from mitochondrial RNA was calculated, very few cells had over 12% mitochondrial RNA indicating that most cells were alive with intact mitochondria at the time of sequencing. The two hMOs single cell libraries were combined and Louvain nearest neighbor network detection with a resolution of 0.2 was used to cluster cells after Principal Component Analysis for dimensional reduction. The resulting clusters were annotated using a combination of 1) comparing the topmost differentially expressed genes (DGE) distinguishing each cluster and 2) the expression levels of accepted cell type markers. The DGE were determined between cluster X and all other clusters. The gene marker lists for each cell type can be found in the R notebook. To distinguish between similar clusters the DGE between cluster X and Y were calculated, identifying the markers distinguishing these two clusters. The number of cells expressing TH compared to the total number of cells and the number of cells in each cluster was calculated to get the proportion of TH positive cells.

Statistical analysis. The statistical analysis was performed in GraphPad Prism v7.0. For the quantification of neuromelanin granules, we performed a normality test followed by a parametric unpaired t-test, p***<0.001.

Results and observations

We observed that good quality of iPSCs is a primary determinant in successfully generating high-quality of hMOs. iPSCs colonies are maintained daily and passaged in order to present no differentiated area (Figure 1a). If the colonies present with less than 5% of differentiated areas after the precautions described, we removed the differentiated areas prior to the generation of hMOs to ensure an optimal quality of hMOs (Figure S1a). The process to generate hMOs from iPSCs, at 70% confluency, takes eight days (Figure 1b). From this point, embryoid bodies formed from the iPSCs, were differentiated toward a midbrain fate in stationary culture and embedded in Matrigel^® to promote the formation of the tissue. Indeed, we observed the progressive appearance of the tissue within the EB (Figure 1c–f). Once the EBs presented with multiple bud extrusion (Figure 1e), they were transferred to shaking culture to promote the growth of the tissue (Figure 1b and 1f). After 50 days of shaking culture in final differentiation media, the hMOs grew to approximately four millimeters (Figure 2a.b) and presented with several midbrain markers (Figure 2c). The presence of midbrain and dopaminergic markers were assessed by quantitative real-time qPCR. Compared to iPSC line sample, we observed an enrichment of several common midbrain markers (EN1, Nurr1, LMX1B, LMX1A, TH, MAOB, Calb1, DDC, COMT) in 50 days-old hMOs (Figure 2c). Conversely, we did not detect any enrichment for dopamine beta-hydroxylase as a marker for noradrenergic neurons (Figure 2c). Interestingly, by using SHH and FGF8 signals we also induce the formation of serotonergic neurons as detected by GBX2 marker (Figure 2c). This finding is consistent with recent reports with hMOs generated using other protocols⁶⁵.

Immunostainings on cryosections of hMOs thirty-day old hMOs, showed a typical cytoarchitecture (Figure 3a) with multiple rosettes³⁶. The center of the rosettes was composed of neural progenitors cells, including Dopaminergic progenitors that were positive for FoxA2 (Figure 3b), negative for MAP2 (Figure 3b squares) while the outside layer is composed of more matured cells, including neurons that were MAP2 positive (Figure 3b). As expected for hMOs, we detected the presence of tyrosine hydroxylase (TH) cells (Figure 3c). This observation was confirmed by single cell RNA sequencing revealing dopaminergic lineage markers (Figure 3d–f). Unsupervised clustering yielded 8 clusters representing cell types that would be expected to be found in the human brain (Figure 3d). The cell types mostly group together by unsupervised clustering. The cluster annotated as ‘mixed’ contains many cell types but few or no neurons. Radial glia are cells differentiated from stem cells into cells with the potential to become neurons or glia, Radial Glia-1 higher activation of ribosomal pathways than Radial Glia-2 which is further along the differentiation pathway. Neuronal cluster 1 contains interneurons, while cluster 2 contains high levels of dopaminergic markers (Figure 3e). The expression of the dopaminergic marker TH in Neurons -1, Neurons – 2 and NPCs is 15%, 34% and 11.5% respectively. The average expression across the three clusters of neurons is 20% (Figure 3f).

Finally, we treated hMOs at day 35 with 100μM dopamine for 10 days to look at dopamine synthesis by-products with a focus on neuromelanin granules⁶⁶. This experiment was not necessary for maturation of hMOs but allowed us to confirm the presence of dopaminergic neurons, as well as the ability of dopamine synthesis. We observed the appearance of brown/black areas, suspected to be neuromelanin granules accumulations, also known as by-products of dopamine synthesis (Figure 4a)³⁶. Silver stains were shown to label neuromelanin granules in the substantia nigra⁶⁷. Thus, we confirmed the presence of neuromelanin granules in hMOs by Fontana Masson staining (Figure 4b) and observed a significant increase of neuromelanin granules after dopamine treatment (Figure 4b-c) and its localisation in dopaminergic neurons (Figure 4d). Finally, we also generated hMOs from an iPSC line of patient with PD carrying triplication for synuclein (SNCA_Tri, Figure S1 b–c), and observed that they reached approximately 4 mm in size. Additionally, staining on 100 day-old SNCA_Tri revealed the presence of dopaminergic neurons too (TH) confirming the value of this protocol for future PD research studies. Raw images and data are available as underlying data.

Concluding remarks

In our group, we have successfully generated hMOs from patient-derived iPSCs with similar dopaminergic neuron yield than previously published³⁶. However, there are various challenges that are associated with the generation of organoids. (i) It is important to note that the quality of the iPSCs remains the most crucial step in the formation of organoid tissue. Differentiated iPSCs would either avoid proper formation of EBs or lead to the formation of non-homogenous EBs that would contribute to variable material for experimentations. iPSC lines are very sensitive and require delicate culture techniques to avoid differentiation. This can be achieved by choosing a range of iPSC passage number suitable for generating good EBs as well as spending extensive effort to remove any cell with differentiation sign. (ii) Batch-to-batch reproducibility is difficult to achieve. It can be controlled by optimizing chemical and physical parameters of media and incubation. (iii) Optimal concentrations of components in the media need to be carefully chosen. There are various chemical factors that contribute to the generation of organoids, and therefore require careful standardization. (iv) Generation of uniform EBs, is the key factor in the generation of organoids. Uniform, smooth and continuous edges of EBs are essential to develop uniform organoids. The primary step for assessment is the neural induction that results in formation of embryoid bodies. The shape (spherical with smooth edges) of EBs at this stage is the defining factor of organoid formation. The EBs generated by our protocol have consistent shape. Although, to facilitate tissue induction and further develop a 3-dimensional structure, Matrigel® is used as a scaffold. Since this scaffold is present only in the early days of organoids, the shape until then is consistent due to this physical parameter in place whereas, once the organoids outgrow the scaffold and are capable of differentiating independently in the medium, the shape can vary slightly from one organoid to another (so far, we have observed slight variations in shape but not to a great extent). As the organoids are generated in a controlled environment, their shape and size is fairly consistent. (v) The speed of the shaker is crucial in the final differentiation of organoids and maintaining 3D organization. (vi) The first protocols for cerebral organoids generation required used of paraffin, one-by-one Matrigel® embedding , and one-by-one transfer into final plates³⁴. This procedure was time consuming and could led to tissue damaged or contamination by the multiple transfers steps involved. By using our midbrain generation protocol, we enabled a scaled-up production, without touching directly the tissue at any step. This allowed us to generate easily big batches of 500 hMOs derived from multiples iPSC lines for comparison studies. (vii) Cryosectioning organoids is a challenging analysis step. Since the organoids form a structure distinct from that of brain tissue, some protocol adaptations were necessary to consistently generate high quality sections. Furthermore, due to tissue organization and small sizes, sections have to be optimized for each stage of organoid maturation. Other methods such as clearing techniques can be useful to overcome this challenge. So far, we have overcome the challenges and generated numerous high-quality midbrain organoids. This method manuscript is aiming to help the community to generate hMOs for Parkinson’s disease studies.

Data availability