By using self-organising, stem cell-derived embryo-like structures, researchers have created ‘hematoids’ – a new model that mirrors early human blood formation without relying on added growth factors. Here, Dr Jitesh Neupane discusses how this work has provided rare insights into the development of blood stem cells and offers potential new opportunities to study childhood blood disorders, drug toxicity and the origins of disease.
Current understanding of early human development comes from animal studies or laboratory-based in vitro models. Direct studies of human embryos are tightly restricted due to ethical and technical reasons, including the internationally recognised ‘14-day rule’, which limits research beyond the point of fertilisation and the stage at which the basic body plan begins to form.
Yet many crucial events, such as the formation of the three germ layer derivatives and the specification of primordial germ cells (PGCs), the precursors of sperm and eggs, occur within or shortly after this window.
Hematoids: the rationale
Our initial motivation to develop a stem cell-based three-dimensional (3D) embryo model was to understand how PGCs arise in humans.1,2 Using this lab-grown embryo model, we showed that human PGC-like cells can form in two distinct regions – the primitive streak and amniotic cells – suggesting a possible dual origin. We also identified a key role for the amnion-associated transcription factor islet-1 for the specification of PGC-like cells in vitro.
As these 3D structures developed, their transcriptomes resembled those of a post-gastrulating human embryo at around four weeks after fertilisation. They elongated, formed tissues representing all three germ layers (ectoderm, mesoderm and endoderm), developed contracting cardiomyocytes and endothelial cells and produced blood-forming cells, also known as haematopoietic stem cells (HSC), which later differentiated into maturing blood cells.
At around day 14, visible red pigmentation appeared, marking the onset of blood formation. We named these structures ‘hematoids’.
Hematoids are not embryos. They lack many embryonic and extra-embryonic tissues required for full development and cannot implant or form a placenta.
What intrigued us was that blood cells formed even in the absence of the yolk sac, which normally supports early blood formation in human embryos. This raised fundamental questions about where blood stem cells come from, what environments support them and why timing during development matters so much for blood formation.
Improving on existing laboratory systems
Over the past two decades, many approaches have been attempted to generate human blood stem cells from pluripotent stem cells. These include guided differentiation using growth factors, passage through endothelial intermediates or forced expression of key blood-related genes.
While these methods can produce blood progenitors, they have largely failed to generate true HSCs capable of long-term, multi-lineage engraftment in animal models. Many also rely on artificial cues or genetic manipulation, which limits their clinical relevance.
Recent breakthroughs have shown that engraftable blood stem cells can be produced from human induced pluripotent stem cells. For example, work from Thierry Jaffredo’s group in EFS Ile De France and Andrew Elefanty’s group in Murdoch Children’s Research Institute, Australia, showed that human induced pluripotent stem cells can produce HSCs capable of engrafting in immunodeficient mice and giving rise to multilineage haematopoietic cells.
These findings have moved the field much closer to the possibility of clinically functional PSC-derived HSCs. However, hematoids address a different limitation: most existing systems do not recreate the natural developmental environment in which human blood stem cells arise.
Hematoids are self-organising structures that generate their own supportive tissues. They form an intrinsic, embryo-like blood-forming niche that closely resembles the aorta-gonad-mesonephros (AGM) region, where definitive blood stem cells arise in the human embryo.
As hematoids develop without heavy reliance on added growth factors, they reflect how blood cells form naturally during early human development. For clinicians, this makes the model more reliable for studying the origin of haematopoietic cells during embryogenesis, with tremendous potential to investigate disease origins, test therapies, and understand how treatments may affect developing blood cells.
Key hematoid insights
This work provides a rare view into how human blood and immune cells first emerge. Early blood formation in the embryo occurs in the extra-embryonic yolk sac around week two. This phenomenon is known as primitive haematopoiesis, which produces short-lived blood progenitors but no long-term blood stem cells.
The true haematopoietic stem cells capable of sustaining blood production throughout life first asrise in the AGM region around week four, known as definitive haematopoiesis.
Hematoids recapitulate key features of definitive haematopoiesis, including the formation of an AGM-like niche, the expression of foetal haemoglobin and the generation of haematopoietic stem and progenitor cells that give rise to multiple blood lineages, including T-cells and natural killer cells.
Importantly, our research shows that blood stem cells are profoundly shaped by their developmental context, including timing, surrounding tissues and local signals. This offers the potential to study childhood blood disorders, including paediatric and developmental leukaemia, which often originate during early human development, with many initiating oncogenic events occurring in utero and propagating through foetal HSC and progenitor cells (HSCPs).
The model could also help explain why the same genetic mutation can lead to very different diseases depending on when it occurs during development.
Hematoids have the potential to enable direct interrogation of leukaemia-initiating events within authentic human embryonic and foetal haematopoietic niches, at the precise developmental stages when susceptibility to transformation is highest.
The system provides an unprecedented opportunity to examine how oncogenic mutations interact with foetal microenvironments to influence HSC emergence, lineage bias, and pre-leukaemic clonal expansion.
Although hematoids have the potential, they have not yet been shown to be models of leukaemia. However, hematoids can recreate early human blood development in a controlled laboratory setting and provide a platform to test how drugs or genetic changes affect blood stem cell formation before clinical symptoms appear.
In the future, this could help in identifying treatments that inadvertently damage developing blood or immune cells and guide safer treatment strategies earlier in drug development. However, this application has not yet been tested and will require substantial further research.
Generating transplant-ready cells
Allowing blood stem cells to form through natural developmental processes may produce cells that are more stable and functionally authentic than those generated using added growth factors and protein cocktails. However, this is still early-stage research and requires further investigation.
The blood stem cells produced by hematoids must be rigorously tested for long-term function through transplantation experiments. Although hematoid-derived HSPCs can give rise to myeloid and lymphoid cells, including erythroids, megakaryocytes, macrophages, monocytes, T-cells and natural killer cells, some blood cell types, such as B-cells, are currently missing, possibly due to limitations of the in vitro culture environment. But in vivo transplantation assays will resolve this issue, which are currently underway.
Furthermore, as hematoids rely on intrinsic self-organisation and the native haemogenic niche, rather than external stimulation, they produce relatively fewer blood stem cells.
Looking to the future of hematoids
Progress in this area of research depends on close collaboration between developmental biologists, stem cell researchers, geneticists and clinicians. Developmental biology provides the roadmap, stem cell science builds the model, and clinical expertise ensures that the research addresses real medical needs.
Without this multidisciplinary approach, it will not be possible to translate fundamental research into clinical tools that could eventually benefit patients.
Hematoids not only generate blood stem cells but also produce other tissues, including endothelial cells, cardiac cells, mature haematopoietic cells and immune cells.
Beyond haematology, this approach could be used in immunology, drug safety and toxicological studies. Any field that relies on understanding how human tissues form during development and how disease disrupts those processes could benefit from these models. Ultimately, hematoids have the potential for cell therapy and regenerative medicine.
If the long-term survival and differentiation of hematoid-derived blood stem cells into myeloid and lymphoid blood cell types could be validated in vivo using animal models, the most realistic near-term applications are disease modelling and drug screening, particularly for paediatric leukaemia and treatment-related toxicity tests.
Additionally, erythropoiesis remains a critical and often underexplored readout of developmental fidelity in stem cell-derived haematopoietic models.
The presence of foetal haemoglobin expression in hematoids suggests definitive erythroid identity. If the terminal differentiation, maturation and enucleation of erythroids derived from hematoid-HSPCs could be functionally validated, they could potentially be used for blood transfusion.
However, before any routine clinical use, the system would need to consistently produce functional blood stem cells capable of giving rise to myeloid and lymphoid cells, demonstrate safety in rigorous preclinical studies, and enable the efficient and scalable generation of clinically relevant cells.
While direct transplantation into humans remains further away, this work lays essential groundwork for safer, more developmentally informed therapies in future.
Author
Jitesh Neupane PhD
Stem cell and developmental biologist, Gurdon Institute and Department of Physiology, Development and Neuroscience, University of Cambridge, UK
References
- Neupane J et al. A post-implantation model of human embryo development includes a definitive hematopoietic niche. Cell Rep 2025;44(10):116373.
- Neupane J et al. The emergence of human primordial germ cell-like cells in stem cell-derived gastruloids. Sci Adv 2025;11(13):eado1350.