Download PDF

Robert Storm, PhD., Applications Scientist | Angeline Lim, PhD., Sr. Applications Scientist | Molecular Devices

Marine Meyer, Scientist | Doppl SA

Maria Clapés Cabrer, Product Manager | SUN bioscience

Introduction

Organoids are three-dimensional (3D), self-organizing in vitro cell culture organ models. These 3D structures are derived from stem cells and harbor key features of their native organs. Since the development of the first mouse intestinal organoids in 2009, a large variety of organoid models have been established.¹ However, they still rely heavily on the use of solidified extracellular matrix (ECM). This conventional culture method introduces a high level of heterogeneity, both in terms of morphology and distribution of the organoids, which complicates subsequent downstream readouts and image analyses.²

To overcome these challenges and closely follow organoid development, we used the innovative technology Gri3D®, a ready-to-use platform for high-throughput and reproducible 3D cultures.³ This platform enables the homogenous generation of a single microtissue in each microcavity in suspension-like conditions, without the need of a solid ECM. The organoids are positioned in predefined locations and on the same focal plane, thus allowing simultaneous tracking at high resolution over multiple days.

To demonstrate the utility of this high-throughput workflow screening, human rectal organoids were generated and cultured in Gri3D plates. For proof-of-concept, organoids were exposed to staurosporine and its effects on organoid morphology were monitored over time on the ImageXpress® Micro Confocal High-Content Imager. For image analysis, a deep learning-based model was developed using IN Carta image analysis software to segment label-free organoids grown in Gri3D plates. This approach allowed for robust automated image analysis of organoids for monitoring purposes. In addition, this approach can also be modified for end-point fluorescencebased viability analysis.

Using Gri3D plates combined with the ImageXpress Micro Confocal system and IN Carta Image Analysis Software, we followed the development and self-organization of human rectal organoids over time with transmitted light (TL) and fluorescence imaging and quantified the cytotoxic effects of staurosporine.

Materials and Methods

Gri3D technology

Gri3D is a ready-to-use platform for high-throughput and reproducible organoid culturing (Figure 1). Based on an array of ultra-dense U-bottom microwells in a hydrogel, single organoids are robustly generated in each microcavity, where they grow in suspension-like culture conditions without a solid-ECM.

Cell culture

Human rectal organoids were generated in Gri3D-96 imaging-bottom 500 μm microwells (SUN bioscience) starting from 100 cells per microwell in a single cell suspension and cultured for up to 7 days. For the live/ dead assay, organoids were treated with staurosporine for 24 hours and then stained with calcein AM (live cells) and ethidium homodimer-1 (EthD-1, dead cells) before image acquisition (Figure 2).

Figure 1. Gri3D technology. Gri3D is a 96-wellplate format plate for 3D cell culture. At the bottom of each well sits a cell-repellent polyethylene-glycol hydrogel were microcavities are imprinted. A pipetting port allows safe media exchange without organoid loss. Left: top view of Gri3D 500 μm microwells. Right: schematic of Gri3D well.

Figure 2. Workflow for high-content imaging and morphology analyses of organoids. Single cells were seeded in Gri3D microcavities. During development, organoids were imaged on an ImageXpress Micro Confocal system and their growth quantified using IN Carta software. Organoids were treated at day 5 and live/dead assay was carried out followed by image-based analyses at the end of compound exposure.

Image acquisition and analysis

Organoid development was followed over time with TL imaging on the ImageXpress Micro Confocal system (Molecular Devices) over 7 days. For the live/dead assay, fluorescence imaging in confocal mode (60 μm pinhole) with a 4X objective was performed.

For analysis of the images, the “export to IN Carta” function in MetaXpress® High-Content Image Acquisition and Analysis Software was used to export images to IN Carta image analysis software. Once imported into IN Carta, the SINAP module was used to carry out segmentation of all images. Each model was trained and verified before being used in the analysis protocol. Figure 2 summarizes the workflow for standardized organoid growth, imaging, and analyses.

Results

Tracking organoid growth over time

Automated image analysis is an integral part of most high-throughput imaging platforms. The ability to monitor cells and organoids in real time is dependent on robust image analysis of label-free TL images. Challenges associated with analysis of TL images include low contrast, high background noise, and the presence of debris and artifacts. To overcome these issues, the SINAP module—a tool that uses deep-learning algorithms to improve accuracy and reliability of high-content screening assays— was used in the image analysis workflow to identify and measure the area and form factor of individual organoids (Figure 3A). Single cells cultured on the Gri3D plate rapidly aggregate to form a cyst. Organoid development was followed over 7 days of culture and the increase in organoid area was accompanied by a decrease in form factor after 4 days, indicating that growth is associated with the development of more complex organoid morphology, also known as budding (Figure 3B).

Figure 3. Growth of human rectal organoids on Gri3D over 7 days. A.) TL (single plane) images of organoids over 7 days of culture. B.) Quantification of organoid area and form factor over time. One- way ANOVA Dunnett’s multiple comparisons, *P < 0.05, P**** < 0.0001, ns: non-significant. n=20. D: Day. Scale bar: 500 μm.

Organoids grown in Gri3D are more uniform than in solid-ECM drops

To verify the robustness of Gri3D plates, organoids were grown in parallel—either in the microcavities or embedded in solid-ECM. Organoids cultured on Gri3D were in predefined locations, the microwells, and in a single plane where they were randomly distributed in solid-ECM drops. Organoids in Gri3D were larger, less heterogeneous in size, and more complex in morphology compared to those grown in Matrigel drops (Figure 4).

Figure 4. Comparison between human rectal organoid cultures grown on Gri3D or within solid-ECM drops. Output of TL images analysed with IN Carta software on A.) Gri3D or B.) solid-ECM drops (N=20). Human rectal organoids were cultured in parallel for 4 days starting from the same cell suspension. Box plots of C.) area and D.) form factor. Form factor equals to 1 for a perfect circle. Each single dot represents an organoid. Two-tailed T-test, ****P < 0.0001. Scale bars: 500 μm.

Response of organoids to staurosporine treatment

To assess the feasibility of using the Gri3D platform for endpoint assays, such as cytotoxicity assessment, human rectal organoids were exposed to staurosporine for 24 hours. Organoids were then stained with calcein AM and EthD-1. Images were analyzed using IN Carta software. Each organoid was identified using SINAP and intensity measurements for calcein AM (for live cells) and EthD-1 (for dead cells) were retrieved. Compared to the vehicle control, organoids treated with 10 µM staurosporine showed significantly less calcein AM signal and increased EthD-1 area per organoid (Figure 5).

Figure 5. Response of human rectal organoids exposed to staurosporine. Live/dead assay was performed on 7-day-old organoids grown on Gri3D after 24 hours exposure staurosporine. A.) calcein AM intensity per segmented organoid. B.) Ethidium homodimer-1 (EthD-1) to calcein AM intensity ratio. C.) Dead area per organoid. Error bars show standard deviation. Each dot represents an organoid. D.) Maximum projection images of organoids after live/dead assay. Green: calcein AM, live; red: EthD-1, dead. One-way ANOVA Dunnett’s multiple comparisons, **P < 0.01, P**** < 0.0001, ns: non-significant. Scale bar: 500 μm.

Conclusions

• Organoids grown in Gri3D show less variability and increased reproducibility compared to conventional solid-ECM drops.
• The arrayed and single focal plane positions of organoids in Gri3D allow efficient imaging and monitoring of individual organoids over time.
• The combination of Gri3D technology, the ImageXpress Micro Confocal imaging system, and AI-enabled image analysis allows the robust phenotypic characterization of organoids.
• We provide an approach for high-throughput imaging of organoids enabling phenotypic-based screening workflows.
• This innovative approach has high potential to solve key challenges related to 3D cultures and compound assessment— at large scale—using patient-derived samples.

References

1. Kim, J., Koo, BK. & Knoblich, J.A. Human organoids: model systems for human biology and medicine. Nat Rev Mol Cell Biol 21, 571–584 (2020).
2. Fatehullah, A., Tan, S. & Barker, N. Organoids as an in vitro model of human development and disease. Nat Cell Biol 18, 246–254 (2016).
3. Brandenberg, N. et al. High-throughput automated organoid culture via stem-cell aggregation in microcavity arrays. Nat. Biomed. Eng. 4, 863–874 (2020).

Download PDF