UCL Great Ormond Street Institute of Child Health


Great Ormond Street Institute of Child Health


Why don’t all embryos, and stem cell organoids, develop the same morphology?

Supervisors: Dr Gabriel Galea, Dr Dagan Jenkins

Variability in organoid self-organisation is a major limitation to the clinical translation of stem cell-based therapies. Some of this variability is genetically determined (1), but even structures derived from the same iPSC line are often unpredictable. In vivo embryonic development is itself variable and inherently stochastic at all levels of organisation. At the molecular level, gene expression and protein degradation rates differ between individual cells. At the cellular level, the migration paths of embryonic cells differ between embryos undergoing equivalent morphogenetic processes (2). At the tissue level the biomechanical stresses withstood differ, producing variability in the rate and configuration of morphogenesis (3). We are accustomed to a spectrum of post-natal morphologies which range from the conventional to the pathological. Some unexplained variabilities are trivial, such as one foot being half a shoe size bigger than the other. However, some deviations from normal development are catastrophic. An example of this is failure to closure the embryonic neural tube, which causes severe CNS malformations such as spina bifida (4). Neural tube closure is a binary, yes/no event, yet its achievement is variable between genetically identical embryos. As a result, spina bifida is commonly partially penetrant. So, how do most embryos robustly compensate for developmental stochasticity inherent to morphogenesis while their genetically-identical siblings do not? And can we harness this understanding to harmonise organoid differentiation?

A good example of this is spina bifida caused by conditional deletion of a membrane co-receptor called Vangl2. We have previously shown that variability in the number of developing spinal cord cells (“neuroepithelium”) in which Vangl2 is deleted largely predicts the occurrence of spina bifida in mice by stopping a critical apical constriction behaviour (5). However, even when Vangl2 deletion is triggered in all spinal cells, only half of the affected embryos develop spina bifida, giving us a clinically-relevant model through which to understand whether gene-level molecular stochasticity ultimately predicts organism-level phenotype variability.

The proposed studentship will apply cutting-edge techniques to relate early sources of developmental stochasticity at the molecular, cellular and organoid/tissue level to late morphological outcomes.

1. At the molecular level, to quantify stochasticity in Vangl2 expression using super-resolution imaging of single molecule mRNA FISH of neuroepithelial cells derived from human iPSCs in vitro and mouse cells in vivo.
2. At the cell level, to relate Vangl2 transcript levels to functionally relevant behaviours including apical constriction.
3. At the tissue level, to establish whether an average Vangl2 expression threshold predicts tissue-level readouts including biomechanical tension quantified by retraction from laser ablation in organoid systems in vitro and development of spina bifida in vivo.

The student will learn to perform AiryScan super-resolution imaging, derive neuroepithelial structures from human iPSC cells, and apply in vivo transgenic strategies to conditionally delete genes in mouse embryos in vivo. In addition, depending on their interests, they could use Crispr/Cas9 engineering to modulate iPSC Vangl2 expression levels.

1. Strano et al, cell Reports, 2020
2. McDole et al, Cell, 2018
3. Maniou et al, PNAS, 2021
4. Nikolopoulou et al, Development, 2017
5. Galea et al, Nature Communications, 2021