ABSTRACT
Research directly on human embryos has gone through cycles of interest and neglect. The recent revitalization, including the making of ‘human developmental biology’, depended on fresh supplies of material and demand for medically relevant work. Human studies relied on mice but rejected simple extrapolation from this model mammal. Now, it is time to take stock while scanning the horizon for further change. Will research on human development be facilitated or frustrated? Will comparative approaches bring a greater variety of animal models into the picture? Will human stem-cell-based embryo models secure ever larger roles as exemplars of vertebrate development?
A Development meeting, ‘From stem cells to human development’ in September 2014, did much to establish a new field: human developmental biology. The prospect of using stem-cell models to discover mechanisms of development in our own species inspired organizers and attendees. The tenth anniversary, and the sixth meeting in what has become a biennial series, invites a look back in order to look forward. This ‘Perspective’ offers a retrospective to which, at the invitation of Development, developmental biologists will respond (Clark et al., 2024). I give the gist of an argument I have made with full documentation elsewhere (Hopwood, 2024). Then I table questions prompted by reflection on this history. What mix of species and other models, I wonder, will predominate in another decade or five?
Cycles of attention
Over the last couple of centuries, research on human development, always of special medical and anthropological interest, has pursued two main strategies. Being difficult to obtain, human embryos have mostly been studied indirectly via more accessible surrogates, such as chick, rabbit, dog, pig, macaque and, since the later 20th century, especially mouse. By contrast, some researchers insisted on making do with the human material they could find – though even that work has rested on other animals for knowledge and skills.
This direct approach has gone through cycles of attention and abandonment. In the early 19th century, leading German anatomists disparaged work on the few, often abnormal and poorly preserved human specimens from pregnancy losses. These, they argued, yielded less useful data about our own development than studies of chicks and domestic mammals. In the early 20th century, conversely, as the USA competed to become the powerhouse of science, the first embryological research institute was founded to study human embryos. The Carnegie Department in Baltimore accumulated anatomical information from preparations serially sectioned and reconstructed in plaster and helped to elucidate the menstrual cycle. By scheduling needed operations at known times after menstruation and intercourse, researchers began to collect conceptuses that filled the long-felt gap in the first fortnight after fertilization. Results from rhesus macaques, newly domesticated for the laboratory, provided a yardstick, but the focus was human development.
Some of the most thrilling discoveries came during and just after World War II. Two-celled humans were first seen by in vitro fertilization, it was claimed, and then, with more certainty, in the solution flushed through a fallopian tube removed during an operation. Yet, even directors of the Carnegie stronghold concluded that, as a morphological research programme, human embryology had run its course. They stopped hiring anatomists and hived off the collection. Embracing developmental biology as this took off in the 1950s, they inherited an experimentalist ethos and emphasized organisms that lent themselves to micromanipulation, biochemistry and genetics. During the 1980s and 1990s molecular methods did much to explain the generation of body plans. For many – including me when I worked on Xenopus – this was the royal road to understanding human development. Finding a surprising degree of conservation across the animal kingdom, such as the role of Hox genes in specifying the anteroposterior axis, reinforced that view.
Around the same time, however, in vitro fertilization began to revive interest in studying humans directly. The technique came out of the greater postwar involvement of science in reproduction, which made material available in several ways. But, from around 1970, researchers had to cope with a backlash against the liberalization of the anti-abortion laws. In principle, the achievement of a live birth from in vitro fertilization in 1978 and the making of a fertility industry gave unprecedented access to living preimplantation stages. In practice, this hinged on negotiating public support, notably through the 14-day rule, and supply remained a challenge. Winning that argument involved pressing the case that mice are an excellent model, but results have to be checked on human material before clinical use. Research showed that humans differ not just morphologically, but also in the timing and patterns of gene expression.
The renewal went way beyond IVF, as in vitro fertilization came to be known in the 1980s. From the early 1990s, anticipating the take-off of the internet, the National Institutes of Health supported curators and anatomists, who tend to default to humans, to digitize sectioned embryos in the Carnegie collection. An international consortium made slides visible online. From the mid-1990s, human molecular geneticists argued that, to investigate gene expression after the Human Genome Project, it would not do simply to extrapolate from mice. More ambitiously than the medical-benefit rationale, they insisted that molecular staining and then ‘omics’ would let humans grasp what ‘makes us unique, or at least not mice’. Geneticists and developmental biologists worked with gynaecologists in Newcastle and London as they persuaded the Medical Research Council and the Wellcome Trust to fund a biobank of post-implantation embryos and fetuses from terminations of pregnancy, the Human Developmental Biology Resource. Finally, IVF became a source of human embryonic stem cells, which cell culturists coaxed into impressively self-organized 3D structures, such as optic vesicles, around 2008. By allowing experimentation, these ‘organoids’ brought humans into the heart of developmental biology and aided a rapprochement with stem-cell science. This was the cue for the 2014 Development meeting.
In the last ten years, some labs have been racing to turn stem cells into the nearest equivalent to a whole conceptus that they can. Others have concentrated on inventing embryoids and organoids that recapitulate certain processes, including gastrulation and organogenesis, with a view to analysing mechanisms in dishes. We see the strands of innovation come together in the reliance of claims of successful mimicry on comparison (‘benchmarking’) with embryos from fertilized eggs: staged sections digitized from the Carnegie collection and molecular datasets generated from material supplied through IVF clinics and biobanks. Where rather separate communities had studied pre- and post-implantation development, these stages were increasingly arranged in one series as human developmental biology was consolidated as a field. That may have encouraged the frequent calls, prompted by the extended culture of IVF embryos in 2016, somehow to open a ‘black box’ obscuring weeks 2-4 post-fertilization.
This renaissance of research on human development used fresh supplies of material from reproductive medicine to meet a more urgent demand for clinical relevance or, as it came to be called, ‘translational science’. Yet change depended also on model organisms as exemplars of what could be done, as sources of knowledge and as test-beds for techniques, especially those for analysing small samples. Almost every advance in humans was trialled on mice. According to a win-win argument often deployed, similar results in humans and mice confirmed the suitability of the model, whereas divergent ones pointed to the need for further human studies.
The continual interplay has functioned well, but will not necessarily carry on as before. Funders, less willing to support research on mice, have invested more in humans. At a more granular level, however, the choice is not and never has been either human or model organism, but about the balance between direct and indirect approaches, while the favoured models have changed and the possibilities they offer have been transformed.
New politics of species choice?
Like at the Carnegie Department in the 1940s, things can feel most exciting to insiders just when, in hindsight, we see that the writing was on the wall. So it is wise to ask, while the field is buzzing, where we are in the latest cycle, if a cycle it proves to be. That is also because all kinds of challenges remain. Highlighting successes risks downplaying how difficult and uneven access to material still is and how much this differs by country. Fragile arrangements for research on human embryos and stem-cell models need strengthening through public consultation and improved regulation.
Questions crowd in. How to earn, keep or bolster public support, maybe by learning from the UK campaign of the 1980s? How will the accumulating data about similarities and differences go beyond a vast catalogue to produce holistic understanding from which fresh insights might emerge? To what extent could and should research tackle not just natural pathologies and technical reproducibility but also the normal variation that was a bane of anatomical research on human development and the subject of a dated project in ‘racial embryology’? Could artificial intelligence offer an acceptable way of filling gaps in the developmental series? Will advances be mainly in translating to the clinic, in drug and toxin testing, or could – conceivably, at the same time – new uses of known developmental mechanisms or even new processes be discovered? Having stressed genes, cells and embryos, will research build on studies of implantation and placentation to bring more fully into the picture not just pregnant bodies, but also social worlds?
In light of the historical cycles of attention, it makes sense to ask in particular how the politics of species choice could change. Will the wheel turn back to the other animals on which developmental biologists, including those who study humans, continue to work? This could happen because opportunities open up in other species or close down in ours. In the unlikely worst case, effects on regulation and infrastructure of general realignments or failures of specific campaigns would block access to our embryos, and even stem-cell models, and force a return to model organisms. Or might successful lobbying, medical demand and commercial interests lead research to concentrate ever more on humans as the 21st century staggers on?
Recent shifts in species politics offer clues, at least for the next few years. Growing recognition of the peculiarity of the mouse has encouraged labs to use other mammals, including primates, to distinguish divergent features from conserved. Mammalian reproduction and embryology have a long tradition of comparative studies. But the latest iteration is currently more like the macaque work at the Carnegie Department, with its strong human focus, than evo-devo, with its broad curiosity about the origins of various taxa.
The other main shift represents a more radical departure, though it has parallels across the biomedical and biological sciences: claims that humans are now a – or even the – model. Pragmatically, developmental biology textbooks have begun to list humans as models and to dedicate chapters to us, while acknowledging the irony that we were once, and in many ways still are, the species to be modelled. How much further will this modest expansion go towards the human embryology texts for medical students that have long sweetened the bread-and-butter anatomy with the jam of developmental mechanisms? The pushes of non-medical funders and the pulls of studying biodiversity and improving animal breeding, not to mention the continuing advantages of model organisms, will limit any convergence.
Programmatically, the promise of quasi-autonomous experimentation on stem-cell-based models is prompting some researchers to talk up these human systems as the best models of such fundamental processes as somitogenesis. They point to the wealth of well-characterized mutations and sequence data, while accepting the need for embryos as normal standards (or ‘ground truth’). Mice, with their genetic tools, have their advocates still. But it is an extraordinary turnaround that investigators of human development are bidding to provide models in this strong sense.
Do these innovations presage broader changes? What, in any case, will those be and whence will they come? Answers, this history leads one to expect, will vary by stage of development studied, approach taken, home discipline and country, but perhaps there will be common themes. Over to you.
Acknowledgements
I thank the Editors at Development for commissioning and commenting on this piece.
Footnotes
Funding
The Leverhulme Trust funded N.H. for research on IVF.
References
Competing interests
The author declares no competing or financial interests.