Manufacturing Uniform Cerebral Organoids for Neurological Disease Modeling and Drug Evaluation

Abstract

Human cerebral organoids are promising tools for investigating brain development and the pathogenesis underlying neurological disorders. To use organoids for drug effectiveness and safety screening, the organoids dispensed into each well must be prepared under precisely the same conditions as the cells. Despite decades of extensive research on approaches to improve organoid generation, various challenges remain, such as low yields and heterogeneity in size and differentiation both within and between batches. Here, we newly established uniform cerebral organoids (UCOs) derived from induced pluripotent stem cells by optimizing organoid size and performing real-time monitoring of telencephalic differentiation marker expression. These organoids exhibited morphological uniformity and consistent expression of FOXG1 during telencephalic differentiation, with high productivity. Moreover, UCOs faithfully recapitulated early corticogenesis, concomitant with the establishment of neuroepithelial populations, cortical plate neurons, and glial cells. Furthermore, UCOs systematically developed neural networks and exhibited both excitatory and inhibitory electrophysiological signals when exposed to neurotransmission blockers. Neurodevelopmental disease models derived from UCOs manifested neurite outgrowth defects, which could be ameliorated with targeted drug treatment. We propose UCOs as an advanced platform with low organoid variations and high reproducibility for modeling both brain development and neurological diseases.

Fig. 1.

Microwell-based aggregation of iPSC clumps makes organoids amenable to uniform growth. (A) Schematic diagrams representing the overall processes of EB formation and cerebral organoid culture using FOXG1-mCherry-expressed iPSCs in culture dishes, microwells, and 96-well ULA plates. DNA cassette for mCherry expression was inserted into the 3′ end of the last FOXG1 exon using the CRISPR/Cas9-mediated knock-in system. (B) Bright-field images showing EBs and organoids from the indicated conditions at each time point. Table showing the CV (%) values of the size of organoids in each group at days 15 and 30. The CV represents the percentage of the standard deviation divided by the mean value. (C) Dot plot showing the diameters of organoids in the indicated conditions. (D) Live images showing FOXG1-mCherry-expressed organoids of the indicated groups at days 15 and 30. The standard exposure time of the laser for mCherry fluorescence was 400 ms, and the low exposure time was 100 ms. Table showing the CV (%) values of the mCherry intensity of organoids in each group at days 15 and 30. The CV represents the percentage of the standard deviation divided by the mean value. (E and F) Dot plots showing the mCherry intensity (E) and relative FOXG1 mRNA expression, normalized by its expression in iPSCs (F), in the organoids of each group at days 15 and 30. All data are expressed as mean ± standard error of the mean (SEM), and significance of each group was calculated by comparing with the SA group. *P < 0.05, **P < 0.01, *** P < 0.001, and ns: not significant. Scale bars, 1 mm.

Fig. 2.

Wnt inhibition stimulates telencephalic differentiation and synchronizes organoid growth. (A) Schematic diagrams of the timeline and protocol for Wnt inhibition during neural induction following IWP-2 treatment at 3 doses (1, 2.5, and 5 μM) and various time points. Three types of protocols (M1-1, M1-2, and M1-3) were based on the 1,000-μm microwell culture condition described in Fig. 1A and Fig. S1B. (B) Live images presenting bright-field and FOXG1-mCherry fluorescence signals of organoids in each group at days 15 and 30. The exposure time of the laser in MI groups was 400 ms. (C) Dot plots showing the size of organoids in all groups at the indicated time points. (D and E) Tables showing the CV values (D) and QC-yield₁₀ (%) (E) of organoid size and mCherry intensity (day 30) in the indicated conditions. The QC-yield₁₀ was calculated as the percentage of the number of organoids within ±10% of the average values of all organoids. (F) Immunofluorescence images showing a cross-section of organoids stained with anti-SOX2 and anti-MAP2 antibodies at day 30. The nucleus was marked with 4′,6-diamidino-2-phenylindole (DAPI). (G) Dot plot showing the rosette thickness of organoids in the indicated groups. All quantitative data are expressed as mean ± SEM, and significance of each group was calculated by comparing with the SA group. *P < 0.05, **P < 0.01, ***P < 0.001, and ns: not significant. Scale bars, 1 mm.

Fig. 3.

UCOs recapitulate cortical development with various types of committed cells, similar to the human fetal brain. (A and B) Correlation matrix (A) and PCA plot (B) showing the correlation between samples of UCOs and SA organoids at days 40, 64, and 120 in the total transcriptome analysis. (C and D) Heatmaps showing the expression of NPC (C) and early neuron-specific (D) genes in UCOs and SA organoids at each time point. The bar colors indicate the scaled logarithm based on 2 of normalized expression (i.e., scaled Log₂Exp). (E) Heatmap showing the relative expression of region-specific markers (x axis) of the fetal brain during stages 4 to 5 (13 to 19 PCW) in the in vivo dissected region and the in vitro UCO/SA organoid samples (y axis). The color of the heatmap indicates the gene-wise scaled relative expression of the markers. CTX, cortex; A1C, primary auditory cortex; DFC, dorsolateral prefrontal cortex; IPC, posterior inferior parietal cortex; ITC, inferior temporal cortex; M1C, primary motor cortex; S1C, primary somatosensory cortex; MFC, medial prefrontal cortex; OFC, orbitofrontal cortex; STC, superior temporal cortex; V1C, primary visual cortex; STR, striatum; MID, midbrain; HIP, hippocampus; CERB, cerebellum; AMY, amygdala. (F) RRHO maps comparing the transition between in vivo developmental stages (stages 2 to 11 compared with stage 1) to UCOs/SA organoids (normalized to iPSCs). The bar colors signify the negative logarithm based on 10 of P values. The table shows the BrainSpan stages and their corresponding ages. PCW, post-conception weeks; M, months; Y, years. (G) Immunofluorescence images showing the edge regions of the UCOs/SA organoids at the indicated time points with staining for markers of NPCs (PAX6), IPs (TBR2), cortical layers (TBR1, CITP2, and SATB2), and glial cells (GFAP and S100β). The nucleus was stained with DAPI. (H to K) Bar graphs showing the percentages of cell populations with positive signals of indicated markers in the organoids at different time points. (L) Immunofluorescence images showing VGLUT1- or VGAT-positive glutamatergic or GABAergic neurons in UCOs at day 136. Insets show magnified views of the indicated neurons. Arrows indicate the colocalization of VGLUT1 and VGAT in the neuronal axon region. Arrowheads indicate VGAT puncta on the MAP2-positive axon. (M) Immunofluorescence images showing oligodendrocyte lineage marker expression (OLIG2, NG2, and MBP) on the edge of UCOs at day 136. (N) Immunofluorescence images showing a full section of UCOs/SA organoids marked with the anti-C-CASP3 antibody at day 136. Bar graph represents the percentages of C-CASP3-positive cell populations in UCOs/SA organoids at day 136. Quantitative data from immunofluorescence images are expressed as the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, and ns: not significant. Scale bars, 200 μm (G and M), 50 μm (L), 10 μm (L; magnified image), and 1 mm (N).

Fig. 4.

Enhanced neural network signals measured in UCOs with excitatory and inhibitory neuronal responses. (A) Schematic illustration of the procedures to attach UCOs/SA organoids to the 24-well MEA plates from day 100. Bright-field images show the organoids attached to the 16 electrodes of the wells. (B) Representative spike and burst raster plots of the UCOs/SA organoids at 2 and 4 weeks after attachment. Black bars represent spikes. Blue bars that are thicker than the black bars represent mean bursts (i.e., a cluster of spikes). Network bursting is marked with the spike histogram above the raster plots. (C) Bar graphs showing the number of spikes, number of bursts, weighted mean firing rates (wMFRs), and synchrony indices of UCOs/SA organoids at the indicated time points. (D) Raster plots of UCOs at 4 weeks showing the neuroelectrical signals before and after treatment with the indicated drugs. (E) Bar graphs showing the fold changes of the number of spikes, number of bursts, wMFRs, and synchrony indices after drug treatment in UCOs. The total duration of all raster plots was 300 s. Quantitative data from the MEA recording are expressed as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, and ns: not significant. Scale bar, 1 mm.

Fig. 5.

RTT-UCOs exhibited reduced neurite outgrowth of early neurons, which was recovered by treatment with blarcamesine. (A) Live images showing FOXG1-mCherry expression in WT- and RTT-UCOs at the indicated time points. (B) Bar graph of relative FOXG1 mRNA expression in the WT- and RTT-UCOs at days 15 and 30. (C) Immunofluorescence images showing MeCP2 expression in 2D neurons derived from WT- and RTT-UCOs at day 30 and neurite morphology marked with anti-TUBB3 antibody staining. Anti-doublecortin (DCX) antibodies were used to label newly generated neurons. (D and E) Bar graphs of the percentage of MeCP2-positive cell populations (D) and the neurite lengths (E) of 2D neurons derived from WT- and RTT-UCOs. (F) Schematic diagrams illustrating the overall neurite outgrowth assessment process via the live imaging of EGFP-expressed UCOs. (G) Live images showing real-time neurite outgrowth in WT- and RTT-UCOs over 4 d. (H) Neurite outgrowth curves showing gradually increasing lengths of outgrown neurites from 3D organoids. (I to K) Immunofluorescence images at day 4 and neurite outgrowth curves in the indicated conditions (NT, nontreated; Noc, nocodazole; Trof, trofinetide; Blar, blarcamesine). Significance of each group was calculated by comparing with the RTT group at indicated time point. Quantitative data from the fluorescence images are expressed as the mean ± SEM. *P < 0.05, **P < 0.01, and ***P < 0.001. Scale bars, 1 mm (A and G), 50 μm (C), and 100 μm [magnified images in (G) and (I) to (K)].