ABOVE: Embryo models made out of stem cells enable researchers to study the mechanisms that lead to humans and other organisms. Nicolas Rivron

Descriptions of the embryo go back at least to the time of Aristotle, but it has only been since the late 19th century and early 20th century that advances in experimental approaches allowed scientists to interrogate how these cells develop.1,2 Mammalian embryos proved particularly difficult to study, though, because they develop inside the uterus; additionally, mammals live longer and are often larger compared to insects or amphibians. This limited many mammalian studies to small animals such as mice.

It is allowing people to do biochemistry with mammalian embryos that they couldn’t do otherwise. 

 —Alfonso Martinez Arias, University Pompeu Fabra

Starting in the 1970s, techniques from in vitro fertilization research expanded scientists’ abilities to culture and study human embryos outside of the uterus using donated embryos.3 However, human embryos cannot be cultured beyond 14 days post fertilization per international guidelines, common genetic manipulations are difficult, and their availability is limited.4 Human embryonic stem cells (ESCs) developed in the late 1990s gave researchers the ability to study human-specific embryonic gene expression and toxicological effects.5 However, these cells lacked the three-dimensional information that is critical to embryo development.

Learning what happens in early embryonic development is important for understanding human development and identifying the causes of early miscarriages and development disorders, prompting new approaches to this field of research.6,7 

Introducing the Stem Cell-Based Embryo Lineup

One group of researchers sought to determine if human pluripotent cells could aggregate and differentiate into the distinct germ layers—ectoderm, mesoderm, and endoderm—that give rise to the fetus. The team grew human ESCs in suspension to promote their aggregation into three-dimensional structures called embryoid bodies (EBs).8 These structures contained all three germ layers but lacked identifiable embryonic organization. Nonetheless, they proved that 3D human ESC models were possible and provided a format to study the effects of compounds on ESC maturation.9 

Illustration of a timeline of important milestones in early embryogenesis currently replicated in stem-cell embryo models. Embryonic stem cells (ESCs, gray) differentiate and divide into multiple cells of either trophectoderm (green) or epiblast (light blue) lineages. As growth continues, some cells divide and mature into primitive endoderm tissue. After implantation, the tissues of the developing embryo further organize. At 14 days post fertilization, at the start of gastrulation, the cells of the epiblast begin to mature into the three germ layers, the endoderm (pink), mesoderm (dark blue), and ectoderm (orange).
© NICOLLE FULLER, SAYO STUDIO

After a sperm fertilizes an egg, the totipotent cell divides into a zygote. As division continues, some cells divide into the growing mass, distinguishing cells that form the embryo (epiblast) from those that form the placenta (trophectoderm). Further cell divisions lead to the formation of the primitive endoderm, which forms a blastocyst. This structure implants around one week after fertilization. As development continues, the epiblast invades inward toward the primitive endoderm. This begins the differentiation into the three germ layers that will construct all other bodily tissues: ectoderm, mesoderm, and endoderm.

Fluorescent image of an embryo-like structure with yellow, cyan, and magenta labeled features.
Blastoid: Researchers generate embryo-like structures that include cells of the epiblast (yellow), future placenta (cyan), and future yolk sac (magenta).
Nicolas Rivron

Researchers could also differentiate human EBs into more mature cell types, for example cardiomyocytes, and use them to develop organoids, such as ones recapitulating cardiovascular tissue, but these models still did not replicate an embryo.10,11 Likewise, researchers could differentiate human ESCs directly into cells of the different germ layers, but these lacked spatial organization.12-14 By confining ESCs into a specific geometry, researchers generated 2D micropattern colonies that exhibited distinct germ layers organized in spatial layers.15 

“They are very good to study the interaction between signals that cells use to communicate with each other and [cell] fates,” said Alfonso Martinez Arias, a developmental biologist at Pompeu Fabra University. These include signaling pathways and physical force sensing.16,17 “They are teaching us a great deal about how, early in development, signals are read and interpreted between cells.”

While the spatial organization of micropattern colonies improved developmental biologists’ window into the mechanisms of embryo biology, they were still 2D projections of a 3D structure. Following work done in mouse models, scientists developed methods to culture a pre-implantation embryo model, called a blastoid, from human ESCs.18-20 

Fluorescent image of a blastoid with yellow, magenta, and cyan labels.
Researchers culture human pluripotent stem cells to organize into embryo-like structures and visualize features such as junction proteins (yellow and magenta) and kinases (cyan) with fluorescent markers.
Nicolas Rivron

“This opens a whole range of possibilities to understand the questions that we are interested in,” said Nicolas Rivron, a developmental biologist at the Austrian Academy of Science who led one of the teams that developed a blastoid model. Because these blastoids only use ESCs derived from the epiblast, which becomes the embryo proper with germ layers, they model embryonic development’s pre-implantation period, up to seven days post fertilization.

To model the post-implantation period, researchers explored culturing ESCs with extraembryonic lineages, first trophoblast stem cells and finally those of the primitive endoderm.21,22 First published in 2023, these human models include extraembryonic tissues and resemble the post-implantation embryo.23 

While stem cell-based embryo models greatly extend the ability to study this period of development, researchers do not grow these models beyond the current 14-day limit established for donated embryos. However, gastrulation, the period beyond this cutoff, represents when the embryo begins to form the eventual body plan, making it an interesting stage to understand early organogenesis.

Because EBs can differentiate into unique tissues, scientists considered culturing ESCs to study gastrulation. Arias’s group pretreated ESCs with an activator of a key signaling pathway and grew the cells in low-adhering conditions.24 This promoted the cells to aggregate and form gastrulating embryo models, or gastruloids, which mimic development between the start of gastrulation at day 14 and to approximately 21 days post fertilization.25 Similar to EBs, gastruloids give rise to all three germ layers, but they lack the extra-embryonic tissue required for implantation and cannot develop brain tissue. 

“This is going to allow us to explore how [cell morphology and genes] are connected in order to build a normal being,” said Arias. “It is allowing people to do biochemistry with mammalian embryos that they couldn’t do otherwise.” For example, gastruloids provide researchers with new models to study the effects of neurodegenerative gene mutations and teratogens, which cause fetal abnormalities, on embryonic development.26,27 

“We’ve just spent the last 10 years validating [the gastruloid model], trying to see what it can teach us about the very early stages of development, and also recently, using it as a different way to access organoids,” Arias said. He added that while there are technical challenges standing in the way of this model reaching its full potential, he is optimistic that these will be resolved. 

The Many Paths to Stem Cell-Based Embryo Models

For several years, researchers studied human embryonic stem cells (ESCs) to understand the unique features of these pluripotent cells, but on their own, they poorly resembled the complex structures that they derived from. While human embryos donated from in vitro fertilization clinics offered insights into the early development period, they are technically and ethically challenging to work with. In recent years, scientists developed a variety of stem cell-based models of embryos to overcome these challenges.

An illustration of five stem cell-based embryo models used to study embryonic development at different stages. Embryoid bodies, at the top of the page, are a cluster of embryonic stem cells (ESCs) that differentiate portions into ectoderm (orange), mesoderm (dark blue), and endoderm (pink). Micropattern colonies in the top right have ESCs that differentiate into organized 2D layers: Trophectoderm (green), endoderm (pink), mesoderm (dark blue), and ectoderm (orange). In the bottom right, blastoids are 3D models of the pre-implantation embryo, with trophectoderm (green), primitive endoderm (yellow) and epiblast (blue) cells. Post-implantation embryos (bottom left) see greater organization of the trophectoderm (green), epiblast (light blue), and primitive endoderm (yellow). Gastruloids, top left, mimic stages of gastrulation and early organogenesis. ESC clusters elongate in culture and begin to differentiate into the three germ layers, endoderm (pink), mesoderm (dark blue), and ectoderm (orange).
© NICOLLE FULLER, SAYO STUDIO

(1) Embryoid Bodies: In a first example of a human 3D embryo model, researchers grew ESCs on non-adhesive dishes in media without a differentiation inhibitor or growth factor to develop into embryoid bodies (EBs) that have all three embryonic germ layers: ectoderm, mesoderm, and endoderm. EBs aid in drug screening and developing some organoids.


(2) Micropattern Colonies: To study the spatial organization of human embryos, researchers treated ESCs with a growth factor and grew them in geometrically confined wells with extracellular matrix coating. These ESCs  differentiated and assembled into radial layers of three embryonic germ tissues, or micropattern colonies. Researchers use these spatially organized colonies to study gene expression and the role of cell density in embryo development. 


(3) Blastoids: In 2021, four groups independently found that inhibiting growth and differentiation factors in naïve or primed  pluripotent stem cells (PSCs) and culturing the cells on non-adherent hydrogel generated a blastocyst-like structure, a blastoid, with cells of the trophectoderm, the epiblast, and the primitive endoderm. Blastoids resemble the pre-implantation period of embryo development.


(4) Post-Implantation Embryo Models: Scientists differentiated PSCs (naïve or primed stage) into lineages representing the primitive endoderm and trophectoderm and aggregated these with PSCs, modeling the epiblast, in defined ratios. These aggregates assembled into structures modeling the post-implantation embryo—an important milestone in developmental biology that could enable researchers to study this critical stage in the future.


(5) Gastruloids: To overcome limitations in studying embryonic development after 14 days, one group pre-treated ESCs with an activator of a proliferation pathway and then cultured the cells at a density of 300 cells in a low-adherence dish, prompting the ESCs to aggregate. Over 96 hours, they saw that these aggregates elongate, mimicking gastrulation, and cells begin to differentiate into those representing all three germ layers. These gastruloids allow scientists to study the earliest period of organ development that is prohibited in embryo research.

See full infographic: WEB | PDF

Limitations and Future Directions for ESC Embryo Models

“Manipulating gene expression in human embryos is challenging,” explained Marta Shahbazi, a stem cell biologist at the Medical Research Council Laboratory of Molecular Biology who works with both stem cell-based embryo models and mouse and donated human embryos. “To dissect mechanisms and tease apart contributions of different variables, the embryo models are much more powerful.”

These various models enable researchers to study gene expression, cell signaling, and epigenetic architecture during embryonic development that is difficult to do with embryos.28-30 Additionally, synthetic models of engineered extracellular matrix materials, placental organoids, and endometrial models extend the ability to study implantation ethically.31-33 

We definitely need more studies in which the models and the embryos are used side by side. 

 —Marta Shahbazi, Medical Research Council Laboratory of Molecular Biology

However, while these stem cell-based embryo models offer many research opportunities, they are not without their own caveats and limitations. “Embryo models tend to be first more difficult to implement in the lab, more difficult to reproduce, and some of them are really very inefficient,” Shahbazi said.

Although some models often reach the stage that they are developed to reproduce, others currently produce successful model embryos about one percent of the time, making them challenging to use experimentally. “This process has to be efficient,” Rivron emphasized. “If it’s not efficient, it’s not useful for science because if it’s one in one thousand, you cannot do experiments with this.”

Additionally, pluripotent cells can exist in different states, either naïve or primed, and the state of a cell population introduces variation in its gene expression and trajectory.34-36 These differences also exist between different iPSC and ESC lines, depending on the reprogramming protocol, prompting a wider analysis of these variations to aid researchers in choosing those that best resemble their intended model system.37-39 

“In terms of comparing the models to the embryos, for many of those models it’s not really straightforward to do such a comparison because they start to model stages that we cannot study in human embryos,” Shahbazi said. Some studies have detailed the transcriptional landscapes of donated embryos or blastoids cultured to various stages before and after the implantation stage, and even fewer have investigated gastrulation and early organogenesis from medical abortions.40-43 

Image of comma-shaped cell aggregates.
Scientists cultured mouse embryonic stem cells into structures that replicate gastrulation to study these later periods of development.
Chaitanya Dingare

“We definitely need more studies in which the models and the embryos are used side by side,” Shahbazi said. “These will really highlight to what extent they behave in the same way.” Researchers are also working on assembling biobanks, developmental cell atlases, and integrated reference tools to facilitate this work.44-46 

“In the next decade, we’re going to have to embrace the diversity of those models to understand which one is best at doing what,” Rivron said. “This zoo of embryo models is going to be very complimentary, and each one is important to develop certain aspects of knowledge.”

However, as valuable as these models are becoming to answer many fundamental questions, donated embryos remain an invaluable source of information, especially to study features not easily replicated in models. For example, Shahbazi investigates the effect of aneuploidy on development and implantation. “This is very difficult if not almost impossible to model with the models,” she said. “The embryos come with their own alterations naturally occurring that we can study.”

Nonetheless, with greater access to stem cell-based embryo models, researchers can investigate more questions about developmental biology that may otherwise be unable to be interrogated. “For the first time, we have the possibility to understand where we come from and how we develop,” Rivron said. 

Nicolas Rivron is a co-founder of Dawn Bio and has two patents for his group’s blastoid method described in this story.

Aflonso Martinez Arias holds two patents for his group’s gastruloid method described in this story. 

  1. Artzt, K. Mammalian developmental genetics in the twentieth century. Genetics. 2012;192(4):1151-1163.
  2. Kim Y, et al. A new era of stem cell and developmental biology: From blastoids to synthetic embryos and beyond. Exp Mol Med. 2023;55:2127-2137.
  3. Edwards RG, et al. Fertilization and cleavage of in vitro of preovulatory human oocytes. Nature. 1970;227:1307-1309.
  4. Clark AT, et al. Human embryo research, stem cell-derived embryo models and in vitro gametogenesis: Considerations leading to the revised ISSCR guidelines. Stem Cell Rep. 2021;16(6):1416-1424.
  5. Thomson JA, et al. Embryonic stem cell lines derived from human blastocysts. Science. 1998;282(5391):1145-1147.
  6. Larsen EC, et al. New insights into mechanisms behind miscarriage. BMC Med. 2013;11(154):1-10.
  7. Ferrer-Vaquer A, Hadjantokakis A-K. Birth defects associated with perturbations in preimplantation, gastrulation, and axis extension: From conjoined twinning to caudal dysgenesis. Wiley Interdiscip Rev Dev Bio. 2013;2(4):427-442.
  8. Itskovitz-Eldor J, et al. Differentiation of human embryonic stem cells into embryoid bodies comprising the three embryonic germ layers. Mol Med. 2000;6:88-95.
  9. Flamier A, et al. A standardized human embryoid body platform for the detection and analysis of teratogens. PLOS One. 2017;12(2):e0171101.
  10. Yoon BS, et al. Enhanced differentiation of human embryonic stem cells into cardiomyocytes by combining hanging drop culture and 5-azacytidine treatment. Differentiation. 2006;74(4):149-159.
  11. Shkumatov A, et al. Matrix rigidity-modulated cardiovascular organoid formation from embryoid bodies. PLOS One. 2014;9(4):e94764.
  12. D’Amour KA, et al. Efficient differentiation of human embryonic stem cells to definitive endoderm. Nat Biotechnol. 2005;23:1534-1541.
  13. Chambers SM, et al. Highly efficient neural conversion of human ES and iPS cells by dual inhibition of SMAD signaling. Nat Biotechnol. 2009;27:275-280.
  14. Kattman SJ, et al. Stage-specific optimization of Activin/Nodal and BMP signaling promotes cardiac differentiation of mouse and human pluripotent stem cell lines. Cell Stem Cell. 2011;8(2):228-240.
  15. Warmflash A, et al. A method to recapitulate early embryonic spatial patterning in human embryonic stem cells. Nat Meth. 2014;11:847-854.
  16. Martin I, et al. Self-organization of a human organizer by combined Wnt and Nodal signalling. Nature. 2018;558:132-135.
  17. Etoc F, et al. A balance between secreted inhibitors and edge sensing controls gastruloid self-organization. Dev Cell. 2016;39(3):302-315.
  18. Rivron NC, et al. Blastocyst-like structures generated solely from stem cells. Nature. 2018;557:106-111.
  19. Kagawa H, et al. Human blastoids model blastocyst development and implantation. Nature. 2021;601:600-605.
  20. Yanagida A, et al. Naive stem cell blastocyst model captures human embryo lineage segregation. Cell Stem Cell. 2021;28(6):1016-1022.e4.
  21. Harrison SE, et al. Assembly of embryonic and extraembryonic stem cells to mimic embryogenesis in vitro. Science. 2017;356(6334):eaal1810.
  22. Weatherbee BAT, et al. Pluripotent stem cell-derived model of the post-implantation human embryo. Nature. 2023;622:584-593.
  23. Oldak B, et al. Complete human day 14 post-implantation embryo models from naive ES cells. Nature. 2023;622:562-573.
  24. van den Brink SC, et al. Symmetry breaking, germ layer specification and axial organisation in aggregates of mouse embryonic stem cells. Development. 2014;141(22):4231-4242
  25. Moris N, et al. An in vitro model of early anteroposterior organization during human development. Nature. 2020;582:410-415.
  26. Galgaczi S. et al. Huntingtin CAG expansion impairs germ layer patterning in synthetic human 2D gastruloids through polarity defects. Development. 2021;148(19):dev199513
  27. Mantziou V, et al. In vitro teratogenicity testing using a 3D, embryo-like gastruloid system. Reprod Toxicol. 2021;105:72-90.
  28. Martyn I, et al. A wave of WNT signaling balanced by secreted inhibitors controls primitive streak formation in micropattern colonies of human embryonic stem cells. Development. 2019;146(6):dev172791.
  29. Yoney A, et al. WNT signaling memory is required for ACTIVIN to function as a morphogen in human gastruloids. eLife. 2018;7:e38279.
  30. Zhou F, et al. Reconstituting the transcriptome and DNA methylome landscapes of human implantation. Nature. 2019;572:660-664.
  31. Shao Y, et al. Self-organized amniogenesis by human pluripotent stem cells in a biomimetic implantation-like niche. Nat Mater. 2017;16:419-425.
  32. Turco MY, et al. Trophoblast organoids as a model for maternal–fetal interactions during human placentation. Nature. 2018;564:263-267.
  33. Shibata S, et al. Modeling embryo-endometrial interface recapitulating human embryo implantation. Sci Adv. 2024;10(8):eadi4819.
  34. Theunissen TW, et al. Systematic identification of culture conditions for induction and maintenance of naïve human pluripotency. Cell Stem Cell. 2014;15(4):471-487.
  35. Nichols J, Smith A. Naïve and primed pluripotent states. Cell Stem Cell. 2009;4(6):487-492.
  36. Hough SR, et al. Single-cell gene expression profiles define self-renewing, pluripotent, and lineage primed states of human pluripotent stem cells. Stem Cell Rep. 2014;2(6):881-895.
  37. Osafune K, et al. Marked differences in differentiation propensity among human embryonic stem cell lines. Nat Biotechnol. 2008;26:313-315.
  38. Liu X, et al. Comprehensive characterization of distinct states of human naive pluripotency generated by reprogramming. Nat Meth. 2017;14:1055-1062.
  39. Bock C, et al. Reference maps of human ES and iPS cell variation enable high-throughput characterization of pluripotent cell lines. Cell. 2011;144(3):439-452.
  40. Zhou F, et al. Reconstituting the transcriptome and DNA methylome landscapes of human implantation. Nature. 2019;572:660-664.
  41. Karvas RM, et al. 3D-cultured blastoids model human embryogenesis from pre-implantation to early gastrulation stages. Cell Stem Cell. 2023;30(9):1148-1165.e7.
  42. Tyser RCV, et al. Single-cell transcriptomic characterization of a gastrulating human embryo. 2021;600:285-289.
  43. Xu Y, et al. A single-cell transcriptome atlas profiles early organogenesis in human embryos. Nat Cell Biol. 2023;25:604-615.
  44. Dawood Y, et al. Human embryonic and fetal biobanking: Establishing the Dutch Fetal Biobank and a framework for standardization. Dev Cell. 2023;58(24):2826-2835.
  45. Haniffa M, et al. A roadmap for the Human Developmental Cell Atlas. Nature. 2021;597:196-205.
  46. Zhao C, et al. A Comprehensive human embryogenesis reference tool using single-cell RNA-sequencing data. BioRxiv. 2024.