The recipe for mammalian life is simple: take an egg, add sperm and wait. But two new papers demonstrate that there’s another way. Under the right conditions, stem cells can divide and self-organize into an embryo. In studies published in Cell1 and Nature2 this month, two groups report that they have grown synthetic mouse embryos for 8.5 days, longer than ever before. The embryos developed distinct organs — a beating heart, a gut tube and even neural folds.
The process is far from perfect. Just a tiny fraction of the cells develop these features, and those that do don’t entirely mimic a natural embryo. But the work still represents a major advance that will help scientists to see organ development in unprecedented detail. “This is very, very exciting,” says Jianping Fu, a bioengineer at the University of Michigan in Ann Arbor. “The next milestone in this field very likely will be a synthetic stem-cell-based human embryo,” he says.
The two research teams achieved the feat using similar techniques. Magdalena Zernicka-Goetz, a developmental and stem-cell biologist with laboratories at the University of Cambridge, UK, and the California Institute of Technology in Pasadena, has been working on this problem for a decade. “We started with only embryonic stem cells,” she says. “They can mimic early stages of development, but we couldn’t take it any further.” Then, a few years ago, her team discovered3 that, when they added stem cells that give rise to the placenta and yolk sac, their embryos developed further. Last year, they demonstrated4 that they could use this technique to culture embryos until day 7. In their latest paper, published in Nature today, Zernicka-Goetz’s team describes how they grew embryos for another 1.5 days.
Embryos in glass
Zernicka-Goetz’s team used a technique developed by Jacob Hanna, a stem-cell biologist at the Weizmann Institute of Science in Rehovot, Israel, who has also been working on this problem for years. Last year, Hanna’s team reported5 that they had developed a device for culturing natural mouse embryos for an unprecedented length of time outside the uterus. This incubator, which kept the embryos going from day 5 to day 11, takes aspects of a previous technology — in which the embryos reside in glass vials that rotate on a Ferris-wheel-like system — and adds ventilation. The ventilation system controls the pressure and the mixture of oxygen and carbon dioxide entering the vials.
After Hanna’s paper was published, his team shared part of their incubator with other developmental and stem-cell biologists. “The brain of this machine, we shared with everyone who asked for it,” he says, including Zernicka-Goetz and her colleagues, who tweaked it slightly for their experiments. In a paper published in Cell on 1 August, Hanna’s team describes how they used the system to grow embryos for 8.5 days. Full gestation in mice is about 20 days.
That period is long enough for the brain regions to develop, the heart to start beating and the neural and gut tubes to form. These synthetic embryos look a lot like natural embryos that form when mouse sperm meets egg, but they “were not 100% identical”, Hanna says. “You can see some defects and some changes in the organ size.”
Each team grew their embryos by combining three cell types, and Hanna’s team also managed to create all three types from naive embryonic stem cells. “It offers a way to simplify the process,” Hanna says. “You can start everything from one population.” Zernicka-Goetz’s team reported a similar accomplishment in a preprint published6 on bioRxiv. (In their Nature paper, the researchers relied on placenta precursor cells from a cell line to create the embryos.)
Zernicka-Goetz’s team also conducted an experiment in which they knocked out a gene called Pax6, which has a key role in brain development. When they eliminated this gene, the mouse heads didn’t develop correctly, mimicking what occurs in natural embryos that lack that gene. The result demonstrates “that the system is actually functional”, says Zernicka-Goetz.
“These two papers, they empower one another,” says Martin Pera, a stem-cell biologist at the Jackson Laboratory Center for Precision Genetics in Bar Harbor, Maine. “Two very skilled groups can really produce rather similar results independently.”
For researchers, these synthetic models have many advantages over natural embryos created from eggs and sperm. Because they grow outside the uterus, they’re much easier to observe. They’re also easier to manipulate using genome-editing tools. “We can perturb, we can manipulate, we can knock out every possible mouse or human gene,” Fu says. That could make them useful for uncovering the role of different genes in birth defects or developmental disorders. Zernicka-Goetz plans to use this model to understand why some pregnancies fail.
Hanna hopes to use the technique to develop synthetic human embryos that can be a source of organs and tissues for people who need them.
What about humans?
But translating this work into humans won’t be easy. Researchers have coaxed human stem cells to become blastocysts — a hollow, a rapidly dividing ball of cells — and even to mimic some aspects of gastrulation — when the early embryo organizes into distinct layers composed of different cell types. But reaching the stage of organ formation in human cells, which happens about a month after fertilization, presents a significant technical challenge. Still, Ali Brivanlou, a developmental biologist at The Rockefeller University in New York City, is optimistic. “The field is not too far away.”
And the more advanced these embryos become, the greater the ethical concerns. One key question is whether these synthetic structures should be considered embryos. The International Society for Stem Cell Research, based in Skokie, Illinois, has long advised against culturing human embryos past day 14 (equivalent to day 6 in a mouse) — roughly when the ‘primitive streak’ appears, the structure that marks the beginning of gastrulation. In 2021, the society removed the limit and issued guidelines stating that such research should have a compelling scientific rationale, and should use the minimum number of embryos necessary to achieve the scientific objective.
Still, Pera sees a need for a continued conversation about the ethics of such models. Researchers have been working on human embryo models for years without much opposition. But he worries about a backlash as researchers begin to develop human embryo models that start developing organs. “The reaction to that could jeopardize this whole field of research,” he says. “It’s important that people know what is being proposed and that it’s done with some kind of ethical consensus,” adds Pera. “We have to go cautiously.”