Even in such game–changing reproductive advances as in vitro fertilization or mitochondrial replacement therapy, what has remained necessary is that the gametes come from the father’s testes and the mother’s ovaries, respectively. But a new lab technology rapidly advancing in mouse studies could upend that biological imperative by, at its hypothetical endpoint, creating embryos from sources such as repurposed skin cells.

The implications of “in vitro gametogenesis” (IVG), three experts write in a perspective essay in the journal Science Translational Medicine, could be helpful for infertility patients and for research, but also deeply vexing for society and policymakers.

IVG holds the promise of radically advancing fertility and the ability to intervene against disease at the pre– or post–embryonic stage, wrote Dr. Eli Adashi of Brown University, I. Glenn Cohen, professor at Harvard Law School, and Dr. George Daley, dean of Harvard Medical School. But it also could lead to ethical nightmares – if people become empowered to create and choose among scores of embryos in the pursuit of ideal children, for example.

“There’s something troubling about an inexhaustible supply of gametes that can be fertilized into an inexhaustible supply of embryos,” said Adashi, professor of medical science and former dean of medicine and biological sciences at Brown University.

IVG arises from the ability of scientists to manipulate stem cells, either derived from embryos or induced from normal adult body tissues. Working with mice in the lab, scientists have reported key successes. They have created fertilizable egg cells from embryonic stem cells and have generated sperm–like cells as well. They’ve also created fertilized embryos from stem cells. In that case, the scientists coaxed stem cells into an early stage of gamete development and then finished their development within the reproductive organs of mice before accomplishing fertilization.

So far it is not feasible – technically or legally – to create a human baby via IVG, and depending on many remaining technical, economic and regulatory hurdles, Adashi said, it might never be. But it does now seem at least prospectively plausible.