“Synthetic Embryos”: What's in a name?

"Synthetic Embryos" are neither Synthetic nor Embryos. So why are editors giving that name to stem cell-based models of human development?

One of the less convincing aspects of the last fortnight's flurry of announcements about advances in simulating early human development (see here) concerned their name. Headlines galore (in newspapers and scientific journals) referred to "synthetic embryos".

But embryo models, however impressive, are not embryos. To claim that the fundamental stages of embryo development that we learnt at school - fertilisation, cleavage and compaction - could now be bypassed to achieve the same result would be wrong. Nor are these objects "synthesised": indeed, their interest to us lies in the ways in which they organise themselves. The researchers merely place the stem cells in a matrix in appropriate conditions, then stand back and watch them do it. Scientists were therefore unhappy about this use of the term in news media, and relieved when the International Society for Stem Cell Research (ISSCR) stepped in with a press release:

"Unlike some recent media reports describing this research, the ISSCR advises against using the term "synthetic embryo" to describe embryo models, because it is inaccurate and can create confusion. Integrated embryo models are neither synthetic nor embryos. While these models can replicate aspects of the early-stage development of human embryos, they cannot and will not develop to the equivalent of postnatal stage humans. Further, the ISSCR Guidelines prohibit the transfer of any embryo model to the uterus of a human or an animal."

Although this was the ISSCR’s first attempt to put that position to the public, it had already made that recommendation to the research community two years previously. Its 2021 Guidelines for Stem Cell Research and Clinical Translation had recommended researchers to "promote accurate, current, balanced, and responsive public representations of stem cell research". In particular:

"While organoids, chimeras, embryo models, and other stem cell-based models are useful research tools offering possibilities for further scientific progress, limitations on the current state of scientific knowledge and regulatory constraints must be clearly explained in any communications with the public or media. Suggestions that any of the current in vitro models can recapitulate an intact embryo, human sentience or integrated brain function are unfounded overstatements that should be avoided and contradicted with more precise characterizations of current understanding."

An academic paper in Stem Cell Reports, introducing the Guidelines, was more specific:

"[...] to best reflect the state and the envisioned applications of these structures made from stem cells, the use of the umbrella term 'embryo model' or 'stem cell-based embryo model' is encouraged, while the use of the term 'synthetic embryo' or 'artificial embryo' or 'embryoids' should be avoided."

Although the ISSCR Guidelines do not comprise a body of law, they do act as an informal international convention, and researchers and institutions around the world adopt its principles in their conduct and self-governance. Consequently, scientists working on stem cell based models of early development, like their colleagues working in areas such as brain organoid research, have stuck to the ISSCR principles. Few have blundered into using the misleading moniker, "synthetic embryo".

Why, then, were newspapers and other media outlets using it? Clearly, eyeballs are more likely to swivel towards "Synthetic Human Embryos" than "Integrated Models of Early Human Development. It's hard to believe that some of our very best science journalists were using it so as to provoke. They really were not.

I have a lot of sympathy with the position of the science writers and editors incurring the scientists' ire. First, why should journalists have known of the ISSCR's recommendations on the use of the term "synthetic embryo"? A journalist who found Recommendation 4.1 of the ISSCR Guidelines would probably not have found them specific enough to address the point, and the academic introduction containing the missing detail is hard to find. In May 2021, I highlighted the section quoted above from Stem Cell Reports. If I had trouble finding my copy, with the specialist resources at my disposal, a journalist might more easily have discovered that the expression "synthetic embryology" had been used by scientists when comparing and contrasting their work with that of synthetic biologists. Seeing that might make it seem legitimate to refer to the products of "synthetic embryology" in the same fashion. An academic introduction to some research guidelines is many steps short of a press release, and journalists should not be blamed for being unaware of it. The Society's clarification should lead more science writers to avoid the expression, which is a good start.

My second reason for being sympathetic to the use of the terrible term is that no suitable alternative has been provided, other than in the Stem Cell Reports paper, which recommends the umbrella terms “embryo models” or "stem cell based embryo models".  That longer expression has leaked into the public domain as a result of the R-SCBEM project that is working on a UK research code.  However, as one very respected journalist told me that, "realistically SCBEM [stem cell based embryo models] will not make it into the intro of a news story on a subject that most readers only heard of for the first time last week". This is where the ISSCR statement doesn't help. It tells the media what not to call stem cell based models of human development, but does not offer a term that should be used. This is not to say that there are a shortage of scientific names, which are useful within the scientific sphere. However, these are far from media friendly and are not intuitive to (or dare I say going to generate interest for) their broad spectrum of readers. Until the ISSCR provides a digestible alternative, editors are likely to keep slapping "synthetic embryo" above their journalists' articles.

When asked why she had used the term “synthetic embryo”, the journalist I contacted remarked that, "We’re still working out the right language and it’s something we’re discussing and will no doubt evolve along with the science".

It is absolutely in the public’s interest (and in the interest of science), that scientific research is explained in terms that the public understands. There is, therefore, a need, I think, for the scientific community to supply a name to the media or endure the penalties of misinformation. When the research was first proposed, I once suggested that a word such as "simbryo" be used, advising against proposals to name the envisaged entities by qualifying the word "embryo" with a prefix (e.g. "ABC embryo") on the basis that, although such terms might be useful within the scientific community, using the word ‘embryo’ might be misleading once it entered the public arena. In late 2021, an elegant proposal came from Jesse Veenvliet of the Max Planck Institute's "stembryogenesis lab" in Dresden. You can read his rationale for "stembryo" here. As Veenvliet's proposal was adopted in the editorial introduction to special edition of the journal Developmental Biology entitled, without irony, Unifying Synthetic Embryology, it carries some weight.

In such an intensely competitive field of research, disagreement among researchers, even as to names, is inevitable. In consequence, however, journalists and their audiences are confronted by a slew of terms which may or may not be synonymous or overlapping, with no agreed term for the overall class of stem cell based embryo models. We cannot blame them if they make up snappy titles of their own. Few, however, would disagree that it would be better for the scientific community to provide such a term. But it isn't happening.  (For the present, researchers themselves do not always use the same terminology when discoursing between each other, leading to calls for - scientific - naming conventions.) What we need is an institution to fulfil the role occupied, in relation to the naming of celestial bodies such as stars, planets, and minor planets, by the International Astronomical Union. In the absence of an equivalent body for developmental biology, it falls to the ISSCR to step up to the plate, and quickly.  As I noted above, it has already suggested a name, if obscurely, to researchers.  Perhaps it should be more explicit in recommending “embryo models” to the media?

Perhaps the ISSCR is having pause for thought.  It would be understandable for members of the public to ask why the ISSCR should tell the media not to call objects “synthetic embryos” on the basis that they are not embryos, while simultaneously asking scientists to call the same objects “embryo models”. Of course, the qualifier “model” is critically different from “synthetic”, which denotes an actual, possibly viable embryo.  Even so, adopting the live-wire e-word in relation to something which is not an embryo concerns me somewhat, susceptible as it is to misunderstanding and misinformation.  Might it reinforce mistaken presumptions that are then magnified in public discourse?  Maybe stembryos, simbryos or TABLEs (Things A Bit Like Embryos) would be less confusing?  In any event, care should be taken over the public communication of these important objects.  Until we hear from ISSCR, - if we do -, my advice to journalists is to use “embryo models”.