This article from the March issue of Cosmos has been shortlisted for the 2020 Eureka Prize for Long-Form Science Journalism. The prize will be announced November 24.
A new device for bioengineering embryo-like structures is shedding fresh light on those earliest, most mysterious – and largely unobservable – moments of human development. Great leaps forward in safe, successful pregnancies and congenital defect prevention await, but so do a host of ethical questions.
I count six little ziggurats side by side, stolid and squat and obviously man-made. They are not going anywhere, but in between them things are on the move.
Circles of dots have begun to roil and rotate, angry blimps rising up among the static trapezoids, their contents swirling in a frenzy of disorder. Then, in each one, a ring solidifies and grows in the chaos.
It’s fleeting, and the jiggling hoop is soon churned back into the mass as the riot spreads unchecked.
These dots are human cells and their acrobatics are the beginnings of human life, though not as we know it. There is no womb, no pulsing maternal heart. Instead, the cells are born in an elaborate plastic chamber under constant video surveillance, and I am witness to their first two days of existence, compressed into a sparse 18 seconds by the wonders of time-lapse video.
The architect of this dazzling piece of cellular music hall is bioengineer Jianping Fu, whose field of expertise – mechanobiology – tries to understand how living cells change in response to physical forces.
Working from his University of Michigan lab on the outskirts of industrial Detroit, Fu is bringing the measured mindset of the machine-builder to the job of constructing life. He is one of an elite cohort of scientists trying to build a replica of the human embryo from the ground up to try to understand just how we are made – and where it can all go horribly wrong.
This area of science aims to alleviate the gamut of reproductive misery, from the anguish of infertility and miscarriage to gathering information about drugs like Thalidomide, which seem innocuous but have grave consequences if taken during pregnancy.
Fu’s bundles of cells, dubbed embryoids, could be used to screen drugs for toxicity in the womb. They might also unravel the mystery of why two out of every five pregnancies fail before 20 weeks. But these scientists, tinkering at the dawn of life, don’t know how far the cells can develop. There is talk their creations could one day provide a source of organs for transplant. The spectre of a baby in a dish looms ominously in the public imagination.
Which is why Fu’s bit of kit, written up in Nature in September and described by leading embryologist Ali Brivanlou as “a major advance in the knowledge of early human development”, is also an invitation for humanity to do some circle time on what it means to build a human.
The invention that’s created this excitement is a simple-looking chamber with three channels. In one, Fu places pluripotent stem cells, blue-sky building blocks that can become almost any human cell. They are immortal and can be frozen and thawed, forming a renewable resource that can be used for years.
Some of these are embryonic stem cells, originally derived from human IVF embryos. Fu also uses “induced pluripotent stem cells” – iPS cells for short – that come from adult skin cells re-programmed back to a primal state.
In the second channel Fu pours a liquid containing morphogens, the fertilisers that hurry stem cells on with the job of remaking themselves over and over. In the third, he lays a gel to support the growing masses, each one hived into its own mini-domain by evenly spaced support posts – the ziggurats.
In many respects there’s nothing novel about this. Researchers the world over are growing stem cells into structures that mimic the early embryo, before getting them to switch course in the first few weeks and become heart, brain or kidney cells.
These grow to make mini organs or “organoids”, about the size of a stunted chickpea, that are used to model diseases on the lab bench and test if drugs or gene tweaks might be a cure.
But Fu is aiming to make the whole shebang.
There’s no diverting cells off into hearts and brains – rather, he wants them to run their course and become something that resembles the early human embryo.
Fu has pushed these embryoids as far as anyone’s been yet and, true to his engineering pedigree, he’s done much of it by inventing a new 3D world for the stem cell colonies to grow in.
“Most excitingly… in this first set of experiments, is the fact that in a subset of those colonies we start to see some asymmetric tissues, asymmetric embryonic structures,” he tells me by phone.
Embryologists tend to get breathless about symmetry, or lack of it. After natural conception, when the sperm fuses with the egg to form the singlecelled zygote, there follows a cascade of dividing, with new cells budding outwards in a uniform ball which, at day three, becomes the 16-cell morula – Latin for mulberry, which it resembles.
But around day six there is a major break from the order of symmetry, as the front-to-back, belly-to-spine axis appears. Anyone aiming to make a human embryo needs to nail this. Fu didn’t just meet the milestone; he also produced the cells that become the amniotic sac – the fluid-filled bag in which the foetus floats.
“These asymmetric structures, basically they resemble the core of the peri-implantation human embryo,” he says.
Implantation. It strikes at the heart of this project. Somewhere between six and 12 days after conception, the human embryo nestles into the fleshy wall of the uterus. This is essential if the pregnancy is to continue. But when it goes wrong, it’s disastrous. Nearly three quarters of all pregnancies that miscarry by 20 weeks are a failure to implant. And the two months after implantation are when the embryo is most vulnerable to the effects of drugs or maternal infections that cause birth defects.
Given the gravity of these issues, you’d think an avalanche of research would have shed light on them. In fact, we know so little about the period it’s called the ‘black box’ of embryology.
The implanted embryo, key to the enigma of early life, is mostly invisible to science. The first reason is structural: the embryo is sequestered in a uterus beyond our gaze. The second is ethical: scientists have grown human IVF embryos for 13 days in the lab, but no one has gone beyond two weeks.
Endorsed by the UK’s Warnock Committee in 1984 and then by the US National Institutes of Health’s Human Embryo Research Panel in 1994, the 14-day rule bans lab research on embryos beyond that point.
It is around this time that a groove called the primitive streak appears in the embryo: among other things this marks a cut-off beyond which twinning is impossible. This is the defining line of the individual human which, some believe, is morally significant. The 14-day rule is enforced by legislation in at least 12 countries.
The reasoning is clear, but the effect is to keep the black box in shadow; science hasn’t had a way to illuminate it.
Now the spotlight falls on Fu’s plastic channels, properly called microfluidic devices. In Fu’s earlier experiments, only around 5-10% of his embryoids reached that asymmetry waypost. But in his lab just a short drive from the factory where Henry Ford christened the first of his Model Ts, Fu has rewritten the rules of embryoid production.
“With this microfluidic system, now we can generate such human embryo-like structures with very high efficiency, up to about 95%,” he says.
Fu’s system isn’t just meeting quality benchmarks, and this is where the parallels with Motor City’s early denizens become a little more pronounced. “It is a scalable system,” he says. “I would even say that now… it becomes a manufacturing system, depending on the needs, right? Depending on how many you need.”
Fu’s tone is measured, matter of fact. But on this moonshot to the dark side of our collective beginning he carries the interests of Everyman, so be reassured that he tells his tale with deep concern for our wellbeing.
What if we could develop a screening mechanism, that’s an adjunct to pharmaceutical development, that enabled us to provide a safeguard for a new drug?
When he says the system “will be very useful for high throughput screening”, that clinical parlance describes the potentially vast benefits to real embryos of testing hundreds of drugs on embryoids first. His work could also show why so many pregnancies miscarry – and improve our measly IVF success rate of 20%.
Indeed, his precision-tooled temperament may be a prerequisite for the kind of slog needed here, to get an embryoid to trace the footsteps of a real human embryo, in something the scientists call recapitulation. It is only by mimicking the true-life course that a dish-bound doppelganger can give test results that are valid.
And that is a big ask, not least because there is no gold standard; the ‘black box’ means there is no definitive embryo library to provide reference points for the journey. Then there are the endless variables in embryoid research: the physical layout of the device, when to add morphogens and chemical signals, and countless other tiny details.
Which is where Fu took his Enterprise into deep space. His embryoids reached the early phase of another critical developmental stage called gastrulation. When the furrow-like primitive streak appears at 14 days, it signals that the cells below, a little like the settling of ploughed earth, are falling into three layers: the ectoderm, mesoderm and endoderm, respective precursors of skin and brain, muscle, and internal organs. Their emergence heralds the laying down of the body plan.
But Fu’s embryoids also showed something else undeniably human. Budding off from those furiously dividing balls were the barely discernible outlines of germ cells. These define the male and female sex – they are the cells that go on to produce sperm or eggs.
Megan Munsie is Deputy Director of the Centre for Stem Cell Systems at the University of Melbourne and has been entangled with the stem cell story for two decades, first as a researcher and now as an expert on policy and ethics. The appearance of these sex cells left a deep impression.
“This is a part of human development that we just really can’t follow,” she says. “The start of the germ lineage I think is absolutely fascinating.”
Nor was it the only aspect of Fu’s achievement that gave her pause.
“When we looked across his images I found them quite extraordinary,” she says. “In almost all the clusters in the wells the patterning was consistent, and as you can imagine in this area of biology there is often quite a lot of inconsistency.”
That cookie cutter reliability is essential if you want to screen a bunch of drugs for harmful effects; the substrate has to repeat faithfully across each test. One example, of course, stands out.
“Thalidomide, who was expecting that? What if we could develop a screening mechanism, that is an adjunct to pharmaceutical development, that enabled us to provide a safeguard for a new drug?” asks Munsie.
Barely a kilometre from Munsie’s office, across Melbourne’s verdant and expansive Royal Park, Andrew Elefanty is using stem cells to try to make bespoke blood for people whose bone marrow has failed, or is in overdrive making blood cells that don’t work properly – the leukaemias.
It is the kind of medicine that shows just how personalised the embryoid project could become.
In order to create the customised blood, Elefanty’s team at the Murdoch Children’s Research Institute takes blood samples from volunteers. The immature red cells are removed and, with the help of genes delivered in a virus, programmed back to a pluripotent state – they become iPS cells.
In crimson culture fluid in blue-capped flasks, these iPS cells are then urged by the team towards the embryoid stage. But, unlike Fu, the team halts development at the point of gastrulation, directing the cells to become mesoderm alone.
Why? Because it is that layer, as Elefanty shows me in a stunning photo, that fashions the seminal version of our biggest blood vessel, the aorta, in whose walls, at this rudimentary time, the red blood cells are made.
Duplicate Elefanty’s process with the blood of someone with leukaemia and the hope is that, one day, you could make healthy blood for them that’s a perfect match.
But I’m confused about something.
Since stem cells were first coaxed into tiny 3D brain-like structures in 2008, scientists have produced a cornucopia of organoids. All of these – mini brains, hearts, kidneys and even blood cells – can be made directly from pluripotent stem cells. Why, then, would you want to push them through the embryoid pathway first, which seems to be taking the long way round?
“The way that works the best, if you like, is if you try to direct them via a trajectory where you do actually try to replicate some of the steps during embryonic differentiation,” says Elefanty. “That’s the road map that you’re following.”
The whole process is artificial, but there’s something of a “nature knows best” adage at its heart. Elefanty is leveraging nature’s navigational toolbox, whose items have been checked off by the rigorous oversight of evolution, to make a better blood cell.
The embryoid pathway could also be a way forward for organ replacement.
Writing in Nature in 2018, stem cell pioneer Martin Pera and colleagues noted that mini brains, livers and kidneys made from stem cells are pretty basic, and that maybe we can do better.
“Initiating organ development in an environment as similar as possible to the developing embryo might… generate structures that more closely resemble mature, functional organs, for drug screens or even for transplantation,” they write.
Usable organs, however, both in terms of sophistication and size, would need the embryoid to be pushed much further along the developmental path than it has been. Elefanty points out two largish hurdles to that ever happening.
“If you don’t have the right cell type you can’t make a whole embryo. And secondly, what people can’t underestimate is the difficulty in reproducing the environment that is the same as the implanted embryo,” he says.
Here’s the rub. Fu’s whirling cell clusters are the right ones to make the headline act of the embryo, but are missing the supporting entourage: the trophoblasts that become the placenta, and the hypoblasts that go on to make the yolk sac, needed to nourish the early embryo.
These are critical kit in their own right, but they also play a key role in what is called patterning, which plots the layout of the body – for example, getting your liver and spleen in the right spot or crystallising the geography of the brain’s bumps and folds.
Here, however, the incoming breakers of science seem relentless.
In 2018, Japanese researchers cultured human trophoblast stem cells for the first time, raising the prospect of an artificially grown placenta. Another announcement last June, however, may have leapfrogged that altogether.
A group led by Xuefei Gao from the University of Hong Kong created something called “expanded potential stem cells”. With some clever chemical coaxing they heightened the potency of human embryonic stem cells and iPS cells – stem cells on steroids if you like. Crucially, expanded potential stem cells can make those missing bits – the placenta and yolk sac.
The final step would be replicating implantation, and in October 2019 the Salk Institute’s Juan Carlos Izpisua Belmonte led a team that created expanded potential stem cells from mice, some by reprogramming cells from the critters’ ears, to make what they say “could potentially be… fully functional synthetic embryos in vitro”.
Equipped with placenta and yolk sac, these are primed to become fully fledged foetuses – and Belmonte successfully implanted the structures in a mouse uterus. Only 7% took and, after a week, they were badly malformed. But it is early days.
As the science rushes onwards it is hard enough to understand it, let alone pass judgment on how far it should all be allowed to go. There are ground rules in place. The 14-day rule is as bedrock as ethics gets, exerting global influence. But, for the purposes of regulation, is an embryoid an embryo?
In Australia it probably is, even though no one is pushing the experimental agenda at anywhere near Fu’s level.
Dianne Nicol is Chair of Australia’s Embryo Research Licensing Committee, which would make any such adjudication should the question arise. It hasn’t – yet. She originally trained as a developmental biologist and is now a Professor of Law at the University of Tasmania.
Nicol directs me to The Research Involving Human Embryos Act 2002, which defines a human embryo as something made by fertilisation of a human egg by a sperm. But it goes beyond that, to things with a human nuclear genome that have “the potential to develop up to, or beyond, the stage at which the primitive streak appears”.
“That does seem to me to include these embryoids that have a capacity to develop to the primitive streak. And if that’s the case then we have a regulatory environment to deal with them,” she says, careful to stress she’s speaking in her capacity as legal academic, not Committee chair. Should the Committee decide similarly, research like Fu’s in Australia would require a licence.
The Act specifies that any research on human embryos must be done under a licence issued by the Committee, a second layer of scrutiny after review by an institutional ethics committee. The decision whether or not to grant such a licence would, almost certainly, take cues from philosophers who specialise in stem cell ethics.
One job for those professional thinkers is to look at why the primitive streak is a moral line in the sand, beyond the question of individuation.
The standard answer is that it heralds the arrival of ectoderm, which prefigures the nervous system, which transmits pain that will, way on down the track, be perceived by a conscious brain. So the primitive streak seems to mark a point beyond which harm could be inflicted.
I asked Insoo Hyun, a Professor of Bioethics and Philosophy at Case Western Reserve University in the US, if the moral weight attached to the primitive streak is justified.
“I think, from a secular point of view, it is very hard to defend that, because we’re not even at the point where we’ve got functioning neurons or any kind of capacity for experience or pain,” he says.
Hyun also notes that many versions of the 14-day rule are nuanced, specifying that culture of embryos cannot go beyond 14 days or the appearance of the primitive streak, whichever comes first.
That points to 14 days being relevant only because it coincides with primitive streak formation. But the whole embryoid project could rejig the developmental order – scientists could, theoretically, make the primitive streak happen earlier or later.
Adding to the regulatory imbroglio are the tectonic shifts we’ve seen in the ways of concocting human life, from test tube babies to reproductive cloning.
In 1996, biology superstar Dolly the sheep was the first mammal to be cloned from an adult cell. Fashioned by putting DNA from a sheep’s udder cell into a sheep embryo – a process called somatic cell nuclear transfer – Dolly was proof of concept for reproductive cloning, which remains universally banned in humans.
Along with those reproductive advances, Hyun tells me, the laws that define embryos have shifted focus. “You’re seeing embryo definitions in legislation that have less and less to do with how it was created and more and more to do with what they can become,” he says.
You’re seeing embryo definitions in legislation that have less and less to do with how it was created and more and more to do with what they can become.
Which suggests, he adds, that the lawmakers are preoccupied with one big issue when it comes to reproductive technology: “Does the thing in question have the power to make a baby if transferred into the womb?”
When the moral rightness of something hinges on what it could become, you have what philosophers call an “argument from potential”. But such arguments, says Hyun, can get hoisted on their own petard as science advances.
What would happen if, following Belmonte’s work, you could program a human skin cell into an expanded potential stem cell that could ultimately make a baby under the right conditions? In that brave new world, would a flake of dandruff meet a requirement for moral protection?
And what of social potential? If regulation itself prohibits transfer of research embryos into a womb, should that allay concerns about what they might become?
Good ethics, of course, starts with good facts, and Elefanty reminds me that our capabilities are limited.
“There is still a considerable degree of concern over making embryos in a dish which will make viable organisms and so forth because that’s still got a degree of playing God associated with it,” he says.
“I think it is partly moot because I think the technology is nowhere near there.”
Elefanty could well be right. And no one is even remotely suggesting we could, or should, grow a baby in a lab. But the science is moving at speed, and so it may be prudent to get our ethical ducks in a row sooner rather than later.
Just in case.
This article appeared in the March 2020 issue of Cosmos magazine. You can subscribe to the magazine here.
Read science facts, not fiction...
There’s never been a more important time to explain the facts, cherish evidence-based knowledge and to showcase the latest scientific, technological and engineering breakthroughs. Cosmos is published by The Royal Institution of Australia, a charity dedicated to connecting people with the world of science. Financial contributions, however big or small, help us provide access to trusted science information at a time when the world needs it most. Please support us by making a donation or purchasing a subscription today.