Uncovering the hidden origins of the placenta

When an embryo attaches to the wall of the uterus roughly six days after fertilization, it begins to build the body and placenta. Called implantation, this is the stage of life scientists know least about. It is also the stage when most pregnancies are lost.

Caltech biologists have now produced the most detailed picture yet of the moment of implantation in mouse embryos, revealing a surprising origin for the placenta. It was previously believed to grow from a single group of cells, but by filming embryos as they implanted and monitoring gene activity in more than 22,000 individual cells during and after implantation, the team discovered a second source: a few cells from the opposite side of the embryo that travel to help build it. When these cells were removed, the structure that gives rise to the placenta failed to form.

Human embryos appear to have similar cells, which means the findings could help explain why some pregnancies fail very early. They also point to what is missing from today's lab-grown embryo models—embryo-like structures made of stem cells—which cannot implant or make a working placenta.

The research was conducted in the laboratory of Magdalena Zernicka-Goetz, Bren professor of biology and biological engineering, and is described in a paper published in Nature Cell Biology on Nov. 12.

A second source for the placenta

In the new study, the team combined several approaches to capture implantation as it happens. They identified which genes were switched on in more than 22,000 single cells from mouse embryos at six closely spaced stages. At the same time, they filmed living embryos as they implanted and tested what happened when specific cells were removed or tracked. The result is a cell-by-cell road map of implantation, and it overturns a long-held textbook view of where the placenta comes from.

Before implantation, the mammalian embryo is a hollow ball called a blastocyst. Inside sits a small cluster of cells that will become the body. The outer layer, the trophectoderm, will form the placenta and other support tissues. Scientists divide this outer layer into two regions: the "polar" side touching the inner cluster, and the "mural" side lining the rest of the ball. The accepted model holds that the placenta comes only from the polar side.

The Caltech team found otherwise. At the border between the two regions, a handful of mural cells, often just two to four per embryo, behave differently from their neighbors. These "border mural" cells are among the first to touch the uterine wall. They switch on genes associated with placenta-forming cells, then travel upward to join the ectoplacental cone, the structure that goes on to build much of the mouse placenta. These cells' descendants stay in place, on the outer edge of the cone, as the placenta begins to form.

When the researchers removed most of the mural cells from early embryos grown in the lab, these embryos never formed the early structure that becomes the placenta. In the uterus, very few went on to develop at all.

"For decades, textbooks have told us the placenta comes from one place," says Zernicka-Goetz. "Watching embryos implant showed us that a few cells from the other side of the embryo make the journey too, and that without them the placenta's foundations don't form properly. Our movies of implantation show that these cells are the first embryo-mother contact cells. We hypothesize that this is the contact that primes them to migrate and invade to form the placenta from the top."

Clues to implantation failure

Previous studies have shown that human embryos implant differently from mouse embryos. Even so, when the team filmed human blastocysts in culture, they saw a similar group of border cells moving toward the embryo. A few of these cells switched on the same marker genes the team had found in mice. By analyzing published datasets, the researchers also found cells with this signature among the invasive cells that anchor the human placenta into the uterine wall. This raises the possibility that a comparable population helps build the human placenta.

The road map also revealed other changes that had been hard to see in prior experiments. Just as the embryo implants, the cells that will form the body pass through a brief, distinct state between two known stages of stem-cell development. A gene called Wt1, best known for its role in kidney development and a childhood kidney cancer, switches on sharply at this moment and then switches off again. None of the stem-cell lines that laboratories currently use to model this transition fully capture it. The team also detected the first molecular signs of the head-to-tail body axis earlier than previously reported.

Scientists can now grow embryo-like structures from stem cells, which are powerful tools for studying early development and infertility. But these models cannot implant properly or make a real placenta. The new results suggest one reason: they lack correctly formed mural cells. "We now hope to be able to establish methods to create these cells in our embryo models with the hope to improve their implantation," Zernicka-Goetz says. "If we succeed, these studies might shed new light on why human embryo implantation fails so often."