Spatiotemporal Gradient and Instability of Wnt Induce Heterogeneous Growth and Differentiation of Human Intestinal Organoids

Abstract

In a conventional culture of three-dimensional human intestinal organoids, extracellular matrix hydrogel has been used to provide a physical space for the growth and morphogenesis of organoids in the presence of exogenous morphogens such as Wnt3a. We found that organoids embedded in a dome-shaped hydrogel show significant size heterogeneity in different locations inside the hydrogel. Computational simulations revealed that the instability and diffusion limitation of Wnt3a constitutively generate a concentration gradient inside the hydrogel. The location-dependent heterogeneity of organoids in a hydrogel dome substantially perturbed the transcriptome profile associated with epithelial functions, cytodifferentiation including mucin 2 expression, and morphological characteristics. This heterogeneous phenotype was significantly mitigated when the Wnt3a was frequently replenished in the culture medium. Our finding suggests that the morphological, transcriptional, translational, and functional heterogeneity in conventional organoid cultures may lead to a false interpretation of the experimental results in organoid-based studies.

Keywords: Cancer; Computer Modeling; Tissue Engineering.

Graphical abstract

Figure 1

Morphological Heterogeneity Is Robust in Normal Organoid Cultures (A) A stitched micrograph of a Matrigel dome that embeds normal colonic organoids cultured for 6 days (left). A bull's eye (dotted circles) divided into three regions was used for the quantification. The core, intermediate, and edge regions are indicated with light, pale, and dark blue colors, respectively. High-magnification images focusing on the core (right top) and edge (right bottom) display significant differences in organoid size. (B) The size distribution of normal human intestinal organoids across the entire matrix dome area corresponding to the stitched image in “A” (left) and the mean organoid size in each region (right). A total of four Matrigel domes that contain normal intestinal organoids were analyzed to estimate the cross-sectional area of individual organoids using ImageJ. The color in each column is matched to the corresponding location in (A). (C) A stitched image of a Matrigel dome that contains CRC organoids cultured for 6 days under the same culture condition provided in (A) (left). High-power magnified images confirm that the CRC organoids in the core (right top) are morphologically similar to those in the edge (right bottom). (D) The size distribution of CRC organoids across the dome area corresponding to the stitched image in (C) (left), and the mean organoid size in each region (right). Similar to the normal organoid, four Matrigel domes that contain CRC organoids were analyzed to obtain the cross-sectional area of individual CRC organoids. A color bar at the bottom in both (A) and (C) shows the matched region in (B) and (D). Scale bars, 1 mm in the left images in both (A) and (C) and 200 μm in the right insets in both (A) and (C). Data are represented as mean ± SEM. One-way ANOVA with multiple comparison was performed to evaluate statistical significance of the differences. ∗p < 0.05; ns, not significant.

Figure 2

Spatiotemporal Gradient and Instability of Wnt3a Induce Heterogeneous Proliferation in Normal Intestinal Organoids (A) The temporal stability of Wnt3a in a complete medium incubated at either 4°C or 37°C (N = 3). (B) The concentration profile of Wnt3a incubated in the medium alone (Control), the medium containing organoids (+Organoids), Matrigel (+Matrigel), or both (+Matrigel + Organoids) at 37°C for 24 h. In each condition, supernatant (500 μL) in each well of a 24-well plate was collected at each time point for the quantification (N = 3). The difference between the “Control” and “+Organoids” is statistically significant at 12 h (p < 0.05). (C) A schematic depicting the method to collect core and edge organoids, separately. A 200-μL tip was first cut to generate an outer diameter of 3 mm using a sterile scalpel, and core Matrigel was harvested using a P200 micropipette. The remaining edge organoids were harvested separately afterward. (D) Visualization of the expression of phosphorylated β-catenin (green) in the organoids harvested from either the core or edge of the organoid cultured for 3 days. The F-actin (magenta) was counterstained to confirm the structure of an organoid. Scale bar: 25 μm. (E) Expression profile of the proliferative cells (Ki67, cyan) in normal (core and edge) or CRC organoids and its quantification (right; N = 8). Scale bar, 100 μm. Data are represented as mean ± SEM. Two-way ANOVA was performed in (A), and one-way ANOVA was performed in (E) to evaluate statistical significance of the differences. ∗p < 0.05, ∗∗∗p < 0.0001.

Figure 3

Repeated Replenishment of Wnt3a Reduces the Gradient of Wnt3a in a Matrigel Dome (A) The concentration profile of Wnt3a when the organoid culture medium was replenished every 3 h (Wnt_rep) versus every other day (Control) for ~6.5 days. Data are represented as mean ± SEM. (B–D) (B) A graphical geometry of the microenvironment of a conventional organoid culture applied for the computational simulation. The diameter and height of Matrigel dome were set to 7 and 1.2 mm, respectively, based on the actual measurements. The heatmap profiles of the Wnt3a concentration inside the hydrogel dome during the conventional culture (C) versus Wnt-replenished condition (D). Concentration profiles at 0, 1, and 12 h are presented with the angled (upper) and vertical cross-cut views (lower) in each condition. Modular color scales show the concentration range used at each time point.

Figure 4

Reduced Intra-heterogeneity of the Wnt Gradient Contributes to Enhancing the Organoid Growth and Epithelial Function of the Core Organoids (A) The size distribution pattern of the organoids cultured in the conventional (Control) or Wnt-replenished condition (Wnt_rep) for 6 days. The cross-sectional area of 20 organoids along different lines crossing the center (position at 0 mm) of the Matrigel circle provided in Figure S7 was quantified using ImageJ. All the data points provided show statistical significance (p < 0.05) except the point indicated by a gray arrow (N = 23). (B) The abundance of aged organoids at day 6 in cultures in the conventional (Control) or Wnt-replenished cultures (Wnt_rep) quantified by the assessment of phase contrast image. N = 12. (C) A hierarchical heatmap profiling the expression of 59 genes related to physiological epithelial functions of the organoids cultured in either conventional (Control) or Wnt-replenished condition (Wnt_rep) followed by the locational collection of the core versus edge (N = 3). (D) The volcano plots displaying the significantly (p < 0.05) upregulated genes (colors coded with green and pink, whereas non-significant genes were colored with gray) when the organoids were cultured in the Wnt-replenished condition compared with the conventional cultures. Once cultured in each condition, organoids were collected from the core (top) and edge (bottom), respectively, and then used for the qPCR assessment. (E) The detailed expression profile of the genes identified in (D) with statistical significance (N = 3). The green and pink columns indicate the significantly upregulated genes in the core and edge organoids, respectively. (F) Visualization of MUC2 expressions in the organoids cultured in either conventional (Control) or Wnt-replenished condition (Wnt_rep) for 3 days followed by the incubation with the differentiation medium for additional 3 days. Scale bar, 100 μm. (G) Quantification of the averaged MUC2 expression per organoids presented in “F” via image analysis (Control-core, N = 5, Control-edge, N = 11, Wnt_rep-core, N = 5, Wnt_rep-edge, N = 4). (H) Quantitative estimation of the total MUC2 expression of the organoids in a well, cultured in different conditions (Control versus Wnt_rep) by multiplying the integrative MUC2 expression per organoid and the actual number or organoids in each region (core versus edge). The “Adjusted” MUC2 intensity was a theoretical estimation of the total MUC2 expression when the core organoids cultured in a conventional condition were assumed not undergoing the heterogeneous Wnt3a gradient inside the Matrigel dome. The numbers of organoids applied for the calculation, 75 (core) and 297 (edge) in both “Adjusted” and “Control”; 75 (core) and 220 (edge) in “Wnt_rep.” Three stitched images were used for quantification. Control, N = 7, Wnt_rep, N = 5, Adjusted, N = 14. Data are represented as mean ± SEM. One-tailed and two-tailed unpaired t tests were performed in (A) and (B), respectively, and two-way and one-way ANOVA was performed in (G) and (H), respectively, to evaluate statistical significance of the differences. ∗p < 0.05, ∗∗p < 0.001.