Mapping how a single cell becomes an entire mouse
The lab of Zachary Smith (Max Planck Institute of Molecular Genetics) in collaboration with Jonathan Weissman (Howard Hughes Medical Institute Investigator at the Whitehead Institute of MIT) presents the first high resolution maps of the origin and outcome of vast numbers of cells in the developing mouse embryo — information that could enable scientists to better understand development and disease. The study has now been published in Cell.
Every one of the 37 trillion cells in our body can trace its origins to a single fertilized egg, but the path from egg to embryo has been all but impossible to understand. Much like we can draw a family tree tracking generations of our ancestors, scientists can also create family trees of cells, showing how they’re related to each other. While researchers have created these family trees for simple, transparent animals like roundworms, generating a comprehensive cell fate map for a mammal, which develops internally and is far more complex, has remained out of reach.
Now, a collaborative team between Dr. Zachary Smith’s lab at the MPIMG and Dr Jonathan Weissman ‘s lab at MIT have figured out how. Using a new CRISPR-based technology, the researchers have reconstructed cellular family trees spanning millions of cells in developing mice — the most complete lineage maps yet made for a mammal — and traced how those cells commit to their fates at the embryo develops from a simple cup of undifferentiated cells into a complex embryo.
“This has been an incredible collaboration,” Zachary Smith says, “it’s very rare in science to be able to tap into something truly fundamental, but with this new data we’re able to see how how virtually every cell relates to every other, where they came from and how they developed, and to do it all at once.”
New Tools to Study Development
A roundworm embryo is transparent, allowing researchers to watch every cell divide under a microscope, and was published as a seminal series of works in the early 1980’s. In contrast, a mouse embryo develops inside the mother, hidden from view, so scientists cannot watch development in real time over long periods and is much more complex.
Instead of observing this development, researchers have learned to make cells that can record it. The idea is to engineer cells to write their own history into their DNA: each time a cell divides, it adds a small, permanent mark to its genome. Those marks are inherited, so every cell carries a record of its ancestry that can be read out long after the divisions occurred. In 2019, Dr. Weissman teamed up with Dr. Smith, at the time a graduate student in the lab of MPIMG Director Alexander Meissner, to present an early proof of concept of this approach, but at the time the resolution was just enough to establish high level relationships and was difficult to control. The new technology, term PEtracer, instead uses prime editing to install these marks at more than a hundred sites in the genome.
Because the marks are read out by sequencing individual cells, the same experiment reveals both what a cell has become — its type and the genes it is expressing — and where it came from. In their new research, the Smith and Weissman labs applied PEtracer to study mouse development. The team engineered stem cells with these heritable genetic marks and injected them into an embryo. As the embryo developed in utero, these marks were incorporated into nearly every cell. Then, by analyzing each cell, the researchers were able to create massive family trees that showed the different types of cells and their relationships to each other over time. With the new technology, nearly 75% of cellular divisions were recovered during a period where the embryo grows from a few hundred to nearly one million cells.
Understanding Development and Diseases
The new maps and tools will allow researchers to better understand what happens during mammalian development, including how cells differentiate to grow tissues. They will also allow researchers to study disease, including where and when a developing embryo is most vulnerable to environmental or genetic stressors. Beyond the biological insights, the data will be used to train AI models of embryogenesis, with the ultimate goal of building a “virtual embryo” that can predict development outcomes. There are also a lot of “unknown unknowns” and new discoveries that could come from these tools, much like the map of the roundworm in the 1980s led to new biological discoveries no one could have anticipated.
---
* The text has been adapted from a HHMI press release. Read the full version here.












