Mouse study finds a second cell source for the placenta, challenging the textbook view

Researchers at the California Institute of Technology have found that the mouse placenta does not arise from a single cell population, as textbooks have long held. A small number of cells from the far side of the embryo also migrate to help form its foundations, and without them the placenta's precursor structure does not develop. The work, led by Magdalena Zernicka-Goetz, appears in Nature Cell Biology.

Mapping implantation cell by cell

Implantation, when the embryo attaches to the uterine wall about six days after fertilisation, is one of the least understood stages of development. It is also when most pregnancies are lost. To study it, the team, whose first author is Dong-Yuan Chen, measured gene activity in more than 22,000 individual cells from mouse embryos across six closely spaced stages. They combined this with live imaging of embryos as they implanted, and with experiments in which particular cells were removed or tracked.

The result is a detailed, cell-by-cell record of the transition. The authors have released the full dataset through an interactive website.

A second origin for the placenta

Before implantation, the embryo is a hollow ball, the blastocyst, with a small inner cluster of cells that becomes the body. The outer layer, the trophectoderm, forms the placenta and other support tissues. It has a polar region, which touches the inner cluster, and a mural region, which lines the rest of the ball. The accepted model says the placenta comes only from the polar side.

The Caltech team identified a handful of mural cells, often two to four per embryo, at the border between the two regions. These "border mural" cells are among the first to contact the uterine wall. They activate genes associated with placenta-forming cells and then move towards the ectoplacental cone, the structure that builds much of the mouse placenta. Their descendants remain on its outer edge as the placenta begins to form.

When most of the mural cells were removed from early embryos in culture, the embryos did not form the early placental structure. In the uterus, very few of these embryos developed at all.

Signs of a similar population in humans

Human and mouse embryos implant differently, so the findings cannot be transferred directly. Even so, the researchers filmed human blastocysts in culture and saw a comparable group of border cells moving towards the embryo. Some of these cells switched on the same marker genes found in mice. Analysis of published datasets also revealed cells with this signature among the invasive cells that anchor the human placenta in the uterine wall. This suggests, but does not establish, that a similar population contributes to the human placenta.

Other findings and implications for embryo models

The data revealed further changes around implantation. The cells destined to form the body pass through a brief state between two known stages of stem-cell development. During it, the gene Wt1, better known for roles in kidney development and a childhood kidney cancer, switches on sharply and then off again. According to the team, none of the stem-cell lines currently used to model this transition fully captures it. Molecular signs of the head-to-tail body axis also appeared earlier than previously reported.

The results may also explain a limitation of stem-cell-derived embryo models. These structures are widely used to study early development and infertility, but they cannot implant properly or make a functioning placenta. The authors suggest that they lack correctly formed mural cells.

What comes next

Zernicka-Goetz says the group hopes to develop methods to generate these cells in embryo models, with the aim of improving implantation. If that succeeds, such models could help to clarify why human implantation fails so often. The proposal that early contact with the mother primes the border cells to migrate and invade remains a hypothesis. The mouse findings are the strongest part of the evidence, while the human data rely on cultured embryos and existing datasets.

Source: Uncovering the hidden origins of the placenta (Phys.org, 13 November 2026).

Image: AI-generated illustration.