Approaches to benchmark and characterize in vitro human model systems

Abstract

In vitro human models, such as gastruloids and organoids, are complex three-dimensional (3D) structures often consist of cells from multiple germ layers that possess some attributes of a developing embryo or organ. To use these models to interrogate human development and organogenesis, these 3D models must accurately recapitulate aspects of their in vivo counterparts. Recent advances in single-cell technologies, including sequencing and spatial approaches, have enabled efforts to better understand and directly compare organoids with native tissues. For example, single-cell genomic efforts have created cell and organ atlases that enable benchmarking of in vitro models and can also be leveraged to gain novel biological insights that can be used to further improve in vitro models. This Spotlight discusses the state of current in vitro model systems, the efforts to create large publicly available atlases of the developing human and how these data are being used to improve organoids. Limitations and perspectives on future efforts are also discussed.

Keywords: Cell atlas; Human development; Model systems.

Fig. 1.

Different in vitro model system types. Schematic of how various in vitro model systems mentioned in this Spotlight (2D versus 3D and patient-derived versus hPSC derived) are created to mimic human development and organogenesis. Both hPSC and patient-derived models can be grown in 2D and 3D, as well as using bioengineering approaches. Both culture sources have become common research tools, including models, such as 2D monolayers of epithelial cells representing the lining of the intestine (Kozuka et al., 2017), and complex 3D brain organoids that can be grown in bioreactors (Lancaster et al., 2013). 2D cell culture models have been studied and developed for decades, and are widely available; however, lack of the ability to recapitulate 3D organization of cells in vivo limits utility. 3D models allow cells to organize and interact in ways that more closely mimic their in vivo counterparts. 3D models can create complex structures with many cell types and can be grown in a variety of formats, including suspension, synthetic and extracellular matrices. 3D systems still have disadvantages, including heterogeneity in organoid size and composition, leading to challenges in adapting organoids to more high-throughput approaches (Quadrato et al., 2017). Not all hPSC or patient-derived cell lines are the same, and they often require their own inherent optimization, which can limit scalability. Engineered systems, including ‘organ-on-a-chip’ and microfluidic systems, enable greater complexity and control over spatiotemporal signalling cues or mechanical cues such as flow. Engineered systems can also include co-culture systems in which multiple distinct cell types are cultured together. Engineered systems can be 2D or 3D.