Two DNA-recording systems have reconstructed cellular family trees across mouse embryonic development, with one study tracing nearly every profiled cell in a single embryo back to one of the first two cells formed after fertilization. The companion papers — one in Science, one in Cell — tackle a problem that has constrained developmental biology for decades: mammalian embryos develop inside the uterus and cannot be observed continuously.

The Science study, led by co-senior authors Jay Shendure, PhD, professor of genome sciences at the University of Washington School of Medicine and a Howard Hughes Medical Institute investigator, and Chengxiang Qiu, PhD, a molecular and systems biologist at Dartmouth College, used DNA Typewriter, a lineage-tracing technology developed by Shendure and Junhong Choi, PhD, that writes developmental history directly into a cell's genome.

The numbers behind the experiment deserve scrutiny. The team attempted the procedure in 100 fertilized eggs and recovered 10 embryos for examination. One embryo carried the most informative record. That is a steep attrition rate, and managers evaluating the platform for their own labs should treat the headline result as a best-case sample rather than a routine yield.

How the recording works

The researchers redesigned the system's recording "tape" so its history would be easier to recover from individual cells, then inserted it into the genome of a fertilized mouse egg. As cells divided, inherited marks accumulated in strict order. "A burst of editing unequivocally marks the daughters of the first cleavage, which serve as inline replicates," the authors wrote in the Science paper.

Because daughter cells inherited marks already written before acquiring new ones, shared patterns revealed both common ancestry and the order of developmental branching.

"Most biological measurements are based on either live imaging, which is limited by the fact that most animal tissues are not transparent, or genomics, which are destructive and only measure a single timepoint," Shendure said. "Recording techniques like the ones in these studies enable measurements over time including in settings that we can't directly visualize."

What did the lineage record show?

The single best-recorded embryo allowed the team to trace nearly every profiled cell back to one of the first two cells formed after the fertilized egg divided. One founding cell produced more descendants, but the two branches generated diverse cell types in nearly equal proportions.

The ordered recording also let the researchers estimate when cell types diverged from shared developmental paths. Blood and retinal cells committed relatively early, while cells forming the skin's outer layer committed later. The results indicate that cell identity is not fixed in one coordinated step; different lineages become restricted on distinct schedules.

The method's technical advantages matter for anyone comparing lineage-tracing platforms. Earlier approaches often relied on DNA-cutting enzymes that could damage cells, erase previous records, exhaust recording capacity, or leave unordered marks whose chronology had to be inferred. "DNA Typewriter avoids all of this," said co-first author Haedong Kim, PhD, a postdoctoral scientist in genome sciences at UW Medicine. "It writes without fully severing the DNA, keeps recording relatively steadily, and writes everything in strict order, so we can record cell lineages at much higher resolution for a longer time."

What does the Cell paper add?

The companion Cell study, from Jonathan Weissman, PhD, and colleagues, used a prime-editing system called PEtracer to map lineage dynamics across more than 1.4 million cells from 16 mouse embryos. Together, the two papers demonstrate that mammalian cellular family trees can now be reconstructed at unprecedented scale.

"It's really proof of principle that we can do what was done with the roundworm in 1983, but for mammals like you and me," Weissman said.

What comes next

The work provides a framework for investigating how normal organs form and where development goes awry. More complete lineage records could support research into congenital malformations, neurodevelopmental and genetic disorders, and cancer, and could guide stem cell engineering. The measured results here — family trees from a handful of embryos — remain distant from the projections of predictive developmental models. Scaling the methods across additional embryos and experimental conditions, the authors argue, could support more quantitative and eventually predictive models of mammalian development.