Tiny lab–grown structures could give researchers a chance to see what they couldn't before, while avoiding ethical issues associated with studying actual embryos.

A team from the University of Michigan reported in the journal Nature Communications that they have coaxed pluripotent human stem cells to grow on a specially engineered surface into structures that resemble an early aspect of human development called the amniotic sac.

The cells spontaneously developed some of the same structural and molecular features seen in a natural amniotic sac, which is an asymmetric, hollow ball–like structure containing cells that will give rise to a part of the placenta as well as the embryo itself.

But the structures grown at U–M lack other key components of the early embryo, so they can't develop into a fetus.

It's the first time a team has grown such a structure starting with stem cells, rather than coaxing a donated embryo to grow, as a few other teams have done.

"As many as half of all pregnancies end in the first two weeks after fertilization, often before the woman is even aware she is pregnant. For some couples, there is a chronic inability to get past these critical early developmental steps, but we have not previously had a model that would allow us to explore the reasons why," said co–senior author Deborah Gumucio, professor of cell and developmental biology and professor of internal medicine.

The researchers have dubbed the new structure a post–implantation amniotic sac embryoid, or PASE. They describe how a PASE develops as a hollow spherical structure with two distinct halves that remain stable even as cells divide.

One half is made of cells that will become amniotic ectoderm, the other half consists of pluripotent epiblast cells that in nature make up the embryonic disc. The hollow center resembles the amniotic cavity – which in normal development eventually gives rise to the fluid–filled sac that protects and cushions the fetus during development.

Gumucio likens a PASE to a mismatched plastic Easter egg or a blue–and–red Pokémon ball – with two clearly divided halves of two kinds of cells that maintain a stable form around a hollow center.

The team also reports details about the genes that became activated during the development of a PASE, and the signals that the cells in a PASE send to one another and to neighboring tissues. They show that a stable two–halved PASE structure relies on a signaling pathway called BMP–SMAD that's known to be critical to embryo development.

Gumucio says that the PASE structures even exhibit the earliest signs of initiating a "primitive streak," although it did not fully develop. In a human embryo, the streak would start a process called gastrulation. That's the division of new cells into three cell layers – endoderm, mesoderm and ectoderm – that are essential to give rise to all organs and tissues in the body.

Besides working with genetic and infertility specialists to delve deeper into PASE biology as it relates to human infertility, the U–M team is hoping to explore additional characteristics of amnion tissue.

The team's work is overseen by a panel that monitors all work done with pluripotent stem cells at U–M, and the studies are performed in accordance with laws regarding human stem cell research. The team ends experiments before the balls of cells effectively reach 14 developmental days, the cutoff used as an international limit on embryo research – even though the work involves tissue that cannot form an embryo.