Human PSCs determine the competency of cerebral organoid differentiation via FGF signaling and epigenetic mechanisms

Abstract

Various culture methods have been developed for maintaining human pluripotent stem cells (PSCs). These PSC maintenance methods exhibit biased differentiation; for example, feeder-dependent PSCs efficiently yield cerebral organoids, but it is difficult to generate organoids from feeder-free PSCs. It remains unknown how PSC maintenance conditions affect differentiation. In this study, we identified fibroblast growth factor (FGF) signaling in feeder-free PSC maintenance as a key factor that determines the differentiation toward cerebral organoids. The inhibition of FGF signaling in feeder-free PSCs rescued organoid generation to the same level in feeder-dependent cultures. FGF inhibition induced DNA methylation at the WNT5A locus, and this epigenetic change suppressed the future activation of non-canonical Wnt signaling after differentiation, leading to reliable cerebral organoid generation. This study underscores the importance of PSC culture conditions for directed differentiation into cerebral organoids, and the epigenetic status regulated by FGF signaling is involved in the underlying mechanisms.

Keywords: Cell biology; Genomics; Omics; Stem cells research; Transcriptomics.

Graphical abstract

Figure 1

Cerebral organoid generation failed in feeder-free PSC cultures and was rescued by FGF signaling inhibition (A) Overview of the organoid generation protocol. (B and C) Bright-field images of undifferentiated PSCs (B) and organoids on Day 18 (C). FF-PSCs did not form round aggregates, which were rescued by PD17 treatment. Scale bar, 200 μm (B) and 500 μm (C). (D) Immunostaining of organoids on Day 18 for the forebrain markers FOXG1 and LHX2. FF-PSC-derived aggregates lacked the expression of the forebrain markers. Scale bar, 200 μm. (E and F) Fold changes in FOXG1 and LHX2 expression (relative to the gene expression in OnF-PSCs; normalized to ACTB; n = 3). Curves with error shading indicate the loess regression with SEM (F) Bright-field images of organoids on Day 36. FF-PSC-derived aggregates had multiple cysts, whereas aggregates had smooth neuroepithelium-like dome structures under the other conditions. Scale bar, 500 μm. (G) Schematic diagrams of the morphological assessment criteria based on the existence or distribution of the neuroepithelial dome structures: Score A, domes all around the aggregate; Score B, domes more than halfway around; Score C, at least one dome; Score D, no dome structures. (H) Percentages of organoids with each morphological score. OnF, n = 33; FF, n = 26; FF + PD17, n = 26. (I and J) Immunostaining of organoids on Day 36. Organoids derived from OnF- and FF + PD17-PSCs had apico-basal polarity of the neuroepithelium (I) and CP-like structures (J). Scale bar, 100 μm.

Figure 2

Long-term culture of cerebral organoids derived from feeder-free PSCs with FGF inhibition (A and B) Immunostaining of FF + PD17-PSC-derived organoids on Days 80 and 120 for SOX2, CTIP2, and SATB2 (A) and for TBR1 and RORB (B). The laminar structures were organized by Day 120. Scale bar, 100 μm. (C) UMAP plot of single cells distinguished by celltype. DP, dividing progenitor; RG, radial glia; IPC, intermediate progenitor cell; ImN, immature neuron; SP, subplate; CFuPN, corticofugal projection neuron; CPN, callosal projection neuron; GE, ganglionic eminence. (D) Violin plots showing the expression of selected genes associated with cortical progenitors and neuronal subtypes. (E) VoxHunt spatial brain mapping onto E13.5 mouse brains based on the ISH data from the Allen Developing Mouse Brain Atlas. FF + PD17-PSC-derived organoids exhibited a high correlation with the cerebral cortex. OB, olfactory bulb; Pall, pallium; SPall, subpallium; POTel, preoptic telencephalon; TelH, telencephalo-hypothalamic transition area; Hyp, hypothalamus; D, diencephalon; M, midbrain; PPH, prepontine hindbrain; PH, pontine hindbrain; PMH, promedullary hindbrain; MH, medullary hindbrain; SpC, spinal cord; TelA, telencephalic vesicle (alar plate); TelR, telencephalic vesicle (roof plate).

Figure 3

Feeder-free PSCs preferentially differentiated into the neural crest lineage (A) Bulk RNA-seq on Day 0 and 6. (B) PCA plot of RNA-Seq datasets on the basis of the top 1000 genes with the highest variance. (C) Volcano plots comparing FF-PSCs and FF + PD17-PSCs on Day 0 and their derivatives on Day 6. (D) Heatmap visualizing the relative transcriptional similarity (score from 0 to 1 determined using the KeyGenes algorithm) of organoids on Day 6 to major ectodermal lineages. NC/CP, neural crest and cranial placode; NE, neuroectoderm; NNE, non-neural ectoderm. (E) Heatmap of neural- and neural-crest-related marker gene expression. (F and G) Fold changes in SOX10 expression (relative to the gene expression in OnF-PSCs; normalized to ACTB; n = 3). Curves with error shading indicate the loess regression with SEM (G) Immunostaining of organoids on Day 18 for the neural crest marker SOX10. Scale bar, 200 μm. (H) Immunostaining of organoids on Day 36 for the peripheral neuron marker BRN3A. Scale bar, 50 μm.

Figure 4

Genome-wide methylome analysis identified epigenomic targets to determine organoid competency (A) Violin plot of CpG methylation rate. (B) Scatterplot of CpG methylation rate values for FF-PSCs versus FF + PD17-PSCs. There is a strong correlation (Pearson’s correlation = 0.994) between the two. (C) Volcano plot highlighting PD17-induced differentially methylated regions (DMRs; methylation difference >35% and SLIM adjusted p<0.01). (D) Percentage of hypo- and hypermethylated DMRs overlapping with candidate cis-regulatory elements (cCREs). PLS, promoter-like signature; pELS, proximal enhancer-like signature; dELS, distal enhancer-like signature. (E) Venn diagram showing the number of DMR-associated genes that were differentially expressed between FF- and FF + PD17-PSCs on Day 0 or 6 (fold change >1.5 and Benjamini–Hochberg adjusted p<0.001). (F) DNA methylation profiles of the WNT5A gene. Red box indicates the DMR. Coverage plots of ChIP-seq experiments show the enrichment of H3K4me1 (enhancer mark), H3K27ac (active mark), and H3K27me3 (repressive mark) at the WNT5A DMR. (G) Boxplot of WNT5A expression on Day 0 and 6. WNT5A was upregulated on Day 6 in FF-PSC cultures and inhibited by PD17 treatment.

Figure 5

Non-canonical Wnt signaling hindered the cerebral organoid generation (A) Overview of the experiments for WNT5A activation. (B) Fold changes in FOXG1 and SOX10 expression with WNT5A activation in FF + PD17-PSC cultures (normalized to ACTB; n = 3; mean ± SEM). (C) Heatmap of non-canonical Wnt signal-related marker gene expression. (D) Overview of the experiments for Wnt inhibition. (E) Fold changes in FOXG1 and SOX10 expression with Wnt blockade in FF-PSC cultures (normalized to ACTB; n = 3). +++ denotes the addition of 3x concentration of IWR1e. IWP2 blocked neural crest differentiation and rescued forebrain identity.