Efficient Generation of Induced Pluripotent Stem Cell-Derived Definitive Endoderm Cells with Growth Factors and Small Molecules

Abstract

Definitive endoderm (DE) differentiation leads to the development of the major internal organs including the liver, intestines, pancreas, gall bladder, prostate, bladder, thyroid, and lungs. The two primary methods utilized for in vitro differentiation of induced pluripotent stem cells (iPSCs) into DE cells are the growth factor (GF) and the small molecule (SM) approaches. The GSK-3 inhibitor (CHIR99021) is a key factor for the SM approach. Activin A and Wnt3a are utilized in the GF approach. In this study, both the GF and SM protocols were compared to each other. The results show that both the GF and SM protocol produce DE with a similar morphological phenotype, gene and protein expression, and a similar level of homogeneity and functionality. However, on both the gene expression and proteomic level, there is a divergence between the two protocols during hepatic specification. Proteomic analysis shows that hepatoblasts from the GF protocol have significantly differentially expressed proteins (DEPs) involved in liver metabolic pathways compared to the SM protocol. Well-validated DE differentiation protocols are needed to fully unlock the clinical potential of iPSCs. In the first step of generating DE-derived tissue, either protocol can be utilized. However, for hepatic specification, the GF protocol is more effective.

Keywords: definitive endoderm; growth factor; induced pluripotent stem cell cells; small molecule.

Figure 1

The differentiation of definitive endoderm and its applications. The definitive endoderm is represented in the black hexagon in the center image. Differentiation of the endoderm is separated into the anterior (green) and the posterior (blue) progenitors. On the right is the three major applications of definitive endoderm and its progenitors: (1) cell therapy, (2) disease modeling, and (3) drug screening.

Figure 2

Gene expression of endoderm and hepatoblast cells. Representative phase contrast micrographs of growth factor endoderm cells (GF-ENDs) (A,B), growth factor hepatoblast cells (GF-HEPBs) (C), small molecule endoderm cells (SM-ENDs) (D,E), and small molecule hepatoblast cells (SM-HEPBs) (F), showing morphological changes after differentiation (20×). qRT-PCR analysis for relative expression of endoderm genes in GF-END and SM-END cells (G), and hepatoblast genes in GF-HEPB and SM-HEPB cells (H). Columns show the combined mean ΔΔC_t values for each marker. Data represent relative expressions of transcripts normalized relative to GAPDH and undifferentiated controls. Data are represented as mean ± SEM for three biologically independent experiments (n = 3).

Figure 3

Protein expression of endoderm cells. Immunostaining of endoderm protein expression in growth factor endoderm cells (GF-ENDs) (A) and small molecule endoderm cells (SM-ENDs) (B). Percentage expression of proteins (C). Nuclei were stained with DAPI in all. Representative data from three independent experiments are shown.

Figure 4

Protein expression of hepatoblast cells. Immunostaining of hepatoblast protein expression in growth factor hepatoblast cells (GF-HEPBs) (A) and small molecule hepatoblast cells (SM-HEPBs) (B). Percentage expression of proteins (C). Nuclei were stained with DAPI in all. Representative data from three independent experiments are shown.

Figure 5

Comparison of GF endoderm and SM endoderm. Principal component analysis (PCA) showing the various cell types: CTR, iPSCs, GF endoderm, and SM endoderm as different shapes (A). Clustering is depicted for the individual groups with similar characteristics. The CTR is represented in red, iPSC in green, GF endoderm in light blue, and SM endoderm in dark blue. The PCA plot was generated using peptide abundance data of all peptides analyzed per cell type, with 10 replicates. A comparative heatmap of all the replicates per cell type (iPSC, CTR, GF endoderm, and SM endoderm) and the identified protein groups (B). Volcano plots of the differentially expressed proteins from the different cell types, GF and SM endoderm, against the iPSC were generated (C–E). The negative x-axis represents downregulation (blue) in the cell type compared to the control, and the positive axis represents upregulated (red) proteins in the different cell types compared to the control. A bar graph showing the number of significantly upregulated (red) and downregulated (blue) proteins in each cell type compared to the iPSCs (F). Proteomic pathway analysis of significant differentially expressed proteins compared to CTR (G,H). A dot plot was generated using the uniquely differentially expressed proteins in ShinyGO analysis, with KEGG pathway enrichment and fold enrichment based on the number of genes present in each pathway. The FDR cut-off was set at 0.05, and the number of pathways was set to 20 (G,H).

Figure 6

Comparison of GF endoderm and SM endoderm. Principal component analysis (PCA) showing the various cell types as different shapes: CTR, iPSC, GF hepatoblast, and SM hepatoblast (A). Clustering is depicted for the individual groups with similar characteristics. The CTR is represented in red, iPSC in green, GF hepatoblasts in light blue, and SM hepatoblasts in dark blue. The PCA plot was generated using peptide abundance data of all peptides analyzed per cell type, with 10 replicates. Comparative heatmap of all the replicates per cell type (iPSC, CTR, GF hepatoblasts, and SM hepatoblasts) and the identified protein groups (B). Volcano plots of the differentially expressed proteins from the different cell types, GF and SM hepatoblasts, against the iPSC were generated (C–E). The negative x-axis represents downregulation (blue) in the cell type compared to the control, and the positive axis represents upregulated (red) proteins in the different cell types compared to the control. Proteomic pathways analysis of significant differentially expressed proteins compared to CTR (F–H). A dot plot was generated using the uniquely differentially expressed protein in ShinyGO analysis, with KEGG pathway enrichment and fold enrichment based on the number of genes present in each pathway. The FDR cut-off was set at 0.05, and the number of pathways was set to 20 (F–H).

Figure 7

Comparison of TCA metabolites of both SM and GF hepatoblasts compared to the control. Diagrammatic representation of the targeted TCA cycle LC-MS workflow (A). A comparative heatmap of the metabolic profile of 15 iPSCs, SM hepatoblast, GF hepatoblast, and the CTR group replicated to the labeled metabolites (B). Volcano plots of the differentially expressed metabolites from the SM hepatoblast against the CTR group (C). The negative x-axis represents downregulated (blue) metabolites compared to the CTR, and the positive axis represents upregulated (red) metabolites compared to the control. Volcano plots of the differentially expressed metabolites from the GF hepatoblast against the CTR group (D). The negative x-axis represents downregulated (blue) metabolites compared to the CTR, and the positive axis represents upregulated (red) metabolites compared to the control.