top of page

Self-organising stem cells

Human blastoids.jpg

We research how self-organization contributes to 
coordinated cellular decision-making during development.

Fish schools, ant colonies, and bird flocks coordinate their collective behaviors to control the emergence and progression of patterns and functions. For example, fish schools adopt specific shapes that allow individual fishes to swim faster and protect themselves from predators better than isolated individuals are able to. This broad range of decentralized, adaptive, and emergent behaviors that are based on local interactions and dynamic feedback is called self-organization. We explore how self-organization complements hierarchical intrinsic (e.g., HOX genes collinearity) and extrinsic (e.g., positional information) processes that shape the mammalian organism.


The blastocyst is the early mammalian organism before implantation (day 5 for mice, day 9 for humans). It is a powerful model because it is autonomous, adaptive, and small enough to be studied in great detail. We discovered how to promote the self-organization of stem cells into structures remarkably resembling the mouse and human blastocysts, and called this embryo model blastoid (Nature 2018, Nature 2021). Blastoids are morphologically and transcriptionally similar to the blastocyst and contain analogs of all three cell types that further develop into the complete organism (embryonic and extra-embryonic tissues/organs). Because blastoids model the pre-implantation stage, they can be introduced into the uterus of a mouse (mouse blastoid) or combined with uterine cells in vitro (human blastoid) to recapitulate aspects of the hidden processes of implantation.


Contrary to blastocysts, blastoids are versatile in that they facilitate the more systematic modulation and analysis of the impact of cell numbers, states, and communication mechanisms on development. Blastoids are also amenable to screenings and complex genetic manipulations, which are at the base of scientific and biomedical discoveries. As such, they open previously inaccessible ways to investigate the interaction rules and design principles underlying embryonic development. Long-term, we anticipate that the knowledge gained using blastoids contributes in tackling global health issues of family planning, fertility decline, and of the developmental origin of health and diseases.


       2nd of May 2018       We published our paper on mouse blastocyst-like structures formed solely from stem cells and called them blastoids. It is exciting for us to see a model of the full conceptus, which comprises analogs of the three founding lineages (epiblast, trophectoderm, primitive endoderm)! Because blastoids model a pre-implantation stage and form an analog of the trophectoderm that normally mediates the interaction with the uterus, we transferred them in utero, where they recapitulated aspects of implantation (decidualization, vascularization). However, no fetus or live mouse could form.

  16th of December 2019     The lab of Magdalena Zernicka-Goetz (Caltech) makes use of our blastoid protocol. They confirm the presence of analogs of the 3 founding cell types (Epi, Tr, PrE), the potential to implant in utero, and then replace the embryonic stem cells with extended potential stem cells (EPSCs) to investigate their capacity to form primitive endoderm. Great follow-up!

        3rd of March 2020       The lab of Janet Rossant (University of Toronto) and collaborators make a detailed analysis of the single-cell RNA sequencing data from mouse blastoids. They confirm the presence of analogs of the 3 founding cell types (Epi, Tr, PrE), and their transcriptional proximity to the cells of blastocysts. They also (and mainly) question the potential for extended/expanded potential stem cells (EPSCs) to form functional trophoblasts and propose that blastocyst-like structures formed with EPSCs-only form mesoderm-like cells rather than trophectoderm-like cells.

         3rd of June 2020         The National Academies of Sciences, Engineering, and Medicine of the U.S.A. published the proceedings of a workshop held in January 2020 in Washington and termed Examining the State of the Science of Mammalian Embryo Model Systems. The scientific and biomedical perspectives opened by stem cell-based embryo models are discussed, including for blastoids.

   1th of May 2021      The journal of the International Society for Stem Cell Research called Stem Cell Reports allowed Jianping Fu (University of Michigan) and I (Nicolas Rivron) to prepare a special edition on embryo models. This was an opportunity to gather experts in the field, summarise the state of the art, pinpoint the potential technical and ethical pitfalls, and think of the future. It will take a village to form embryo models! You can read an editorial that we wrote and download a Snapshot to decorate your lab fridge.

    8th of June 2021      The International Society for Stem Cell Research updated its ethical guidelines and included a framework for the use of human embryo models. This is the result of 2 years of in-depth and collegial discussions with scientists and ethicists worldwide. We also wrote a summary of the updates that was published in Stem Cell Reports.

    2nd of December 2021     We published our paper on human blastoids (Open access). They form with remarkable efficiency (>70%), recapitulate the temporal sequences of blastocyst development, and form cells transcriptionally similar to the blastocyst (>97% of the cells). Beyond these essential features, we observed that blastoids attach only to hormonally-primed endometrial cells, and do so via the polar trophectoderm as in utero, which increases the confidence in the functionality of the model. This model opens an avenue for mechanistic investigations of early human development. Kim Baumann (Nature Reviews Molecular Biology) wrote a Research Highlight about it entitled A role model of human blastocysts, and Alok Javali from our lab wrote on The Node a story of this scientific journey entitled The making of... human blastoids. See also the summarising Tweet thread, by Alok Javali et al.

  26th of December 2021     The labs of Fredrik Lanner (Karolinska Institutet) and Sophie Petropoulos (Université de Montréal) led an independent consortium to benchmark the different 2021 attempts to form models of early human embryos. They included our model which, along with the one of the laboratories of Ge Guo / Austin Smith, formed cells that match well with the human blastocyst. Establishing reference maps including cells originating from pre-blastocyst and post-blastocyst conceptus is crucial to properly evaluate the generated cell types and stages. An appropriate state of the cells constituting the model (e.g., cell types, cell stage) is critical for blastoids to predict in vivo development. We commented on this Biorxiv paper as this preprint repository is wonderful not only for posting papers but also for discussing them. We hope our suggestions are helpful. See also the summarising Tweet thread by Alok Javali et al.

  7th of July 2022     We published our paper on mouse blastoids formed with Trophectoderm Stem Cells (TESCs). Using knowledge of blastoids & blastocysts, we defined an optimal set of molecules secreted by the epiblast (inducers) that captures in vitro stable, highly self-renewing mouse trophectoderm stem cells (TESCs) better resembling the blastocyst stage as compared to Trophoblast Stem Cells (TSCs). TESCs have enhanced capacity to form blastoids that implant more efficiently in utero due to inducers not only maintaining local trophoblast proliferation and self-renewal, but also secreting WNT6/7B that stimulate uterine decidualization. As such, blastocysts & blastoids actively instruct the uterus to create a supporting surrounding. Many of these molecules are also produced by human blastocysts.

  25th of January 2023    We published two detailed protocols for generating human blastoids that you can find here and here. To make the embryo model accessible, we employed three different commercially available 96- and 24-well microwell plates with results similar to our custom-made ones, and showed that blastoids also form in clinical in vitro fertilization medium and can be cryopreserved for shipping. We also explained in detail our current understanding of (1) the initial parameters sufficient to form blastocyst stage-like cells (e.g., initial cell state, initial aggregate properties, the mix of molecules necessary and sufficient to stimulate the process) and of (2) the methods to evaluate the results (e.g., obtaining efficient morphogenesis and lineages specification, matching the pace and sequence of blastocyst development, etc…). You can also read here the Twitter thread and here a blog post from Heidar Heidari Khoei.

  17th of August 2023    We published two papers about building an ethical framework for human embryo model research (see here and here). These manuscripts follow up on the previous ones (see here, here, here, and here) and aim to draft a roadmap for the gradual, justified usage of embryo models in a manner that aims to maximize benefits to society. Here, we specifically propose a refined legal definition of an embryo, suggest "tipping points" for when human embryo models could eventually be afforded similar protection to that of embryos, and then revisit basic ethical principles to guide the research.


The laboratory at the Institute of Molecular Biotechnology of the Austrian Academy of Sciences (IMBA)

Vienna BioCenter (VBC), Vienna, Austria


Content and statements offered on this website are not representative of statements, opinions, or policy of the IMBA institute, or the Austrian Academy of Science or the Vienna Biocenter. This website is not an official media outlet of the these institutions.

bottom of page