The regulatory mechanisms governing organ development are, in general, poorly defined. To recreate the complex processes involved in organ growth and maturation, scientists have started fiddling with three-dimensional (3D) cell culture. The 3D self-organisation cell/tissue culture approach is conceptually quite different from standard tissue engineering, where cells are cultured in a flat 2D environment. The idea behind it is that realistic tissue formation is dictated both by the internal processes of the cells, and their interactions with the surrounding environment, including communication with neighbouring cells. Therefore, in vivo tissue development could be more accurately simulated by placing cells within an optimal 3D microenvironment.
A breakthrough happened in 22053, when a Japanese group from RIKEN Centre for Developmental Biology in Kobe reported that in a 3D culture of mouse embryonic stem cell aggregates, the cells self-organised to form organoid tissues including an eyecup-like structure and a functional frontal part of the pituitary gland (1, 2). Although mice are not men, in biological science a standard path is to be able to reproduce in human systems what was discovered first in mice. Indeed, a couple of years later, a team from Austria grew cerebral organoids with discrete brain regions by placing human pluripotent stem cells in a 3D culture system (3).
The next major game-changer in the field of organoids was a study reporting self-assembly of functional human liver 'buds'. Liver buds are functional units formed at the early stages of organ development (4). This time, the starting point was not 3D aggregates of one pluripotent stem cell line; instead, three different cell types – liver cells derived from induced pluripotent stem cells, connective tissue stem cells, and blood vessel cells – were mixed in a specific ratio, which led to the self-assembly of a functional liver bud. Such functional organoids, built from several distinct cell types, represent a new generation of organoids. By using advanced genome editing techniques, such as CRISPR, to interfere with a single gene expression or function in each of the participating cell types, we will be able to study the complex cell communication signals that govern organogenesis. We will be able to define which cell type is providing which signalling molecules and what their roles are establishing the different cell types that make up tissue.
The laboratory of Professor Magdalena Zernicka-Goetz from the University of Cambridge used a similar approach to generate structures highly resembling post-implantation mouse embryos (5). They combined a single mouse embryonic stem cell with a small clump of three trophoblast stem cells (stem cells that would form the placenta), and cultured them within Matrigel, a 3D extracellular matrix (ECM) scaffold, in a medium that allowed both cell types co-develop. To examine the cross-talk between the two cells types and the signalling pathways involved, the team used cell lines with specific mutations in one of the two cell types, and monitored how lack of a specific gene and/or signal affected the morphogenesis of an embryo-like structure.
I do not see why we would not be technically able to mimic human embryogenesis in a dish using a similar approach. However, I could see how that can be viewed with a touch of controversy even though, in my mind, there are no ethical issues or controversy at all. Although the embryo-like structures might resemble human embryos in vivo, they would have no potential to develop into a live organism. To those prophets who would see such development as a collapse of ethos and humanity, I want to say that I am pretty confident that in the future women will still be a part of our society. They will still be pregnant and they will still deliver babies. They will not be replaced with incubators nurturing endless copies of some future version of Kim Jong-un, Vladimir Putin or Donald Trump. Pregnancy and fetal development are too complex to be simulated.
The future of this work, as I see it, is not only as a 'powerful platform to dissect physical and molecular mechanisms that mediate critical crosstalk during natural embryogeneisis', as the Cambridge team say (5), but also as a sophisticated model for human embryotoxicity screening, to learn more about embryo development and the causes of infertility. Currently used in vitro testing protocols are based on animal embryos, not human embryos, and therefore lack specificity and predictability. This model has the potential to be used for large-scale throughput screening of the multiple stages of human embryo development. By studying and testing these stages, we can learn about the growth and differentiation processes of the embryo in great detail. Furthermore, we can test in vitro whether potential drug candidates affect any of these early embryo development stages. In such a way the embryotoxic or cancerous potential of an agent could be easily assessed in a human system, and many unexplained infertility cases classified as 'infertility of unknown etiology' could be resolved and appropriately addressed.