The impact of developmental biology on society is particularly acute when it comes to reproduction – research informs efforts to assist reproduction and understand what happens when pregnancy goes wrong. Recent developments in stem cells, culturing conditions, gene editing and sequencing are also revealing aspects of human embryonic development previously hidden from us. Here at Development, we are committed to promoting human developmental biology, through a dedicated section in the journal, regular meetings (you can book your place on our September ‘From stem cells to human development’ meeting now) and special issues (our upcoming one is accepting submissions until March 1st). As well as promoting the science, we like to encourage discussion of its impacts on society, and have published front section articles on various aspects of embryo ethics.
Given all of this I was very excited to attend the Progress Educational Trust (PET) meeting ‘Crossing Frontiers: Moving the Boundaries of Human Reproduction’, held in Camden, London, last month. The PET was established in 1992 with roots in an earlier campaign for the continuation of human embryo research, and through events and publications aims to ‘inform debate on assisted conception and genetics’. The day conference brought together representatives from research, law, policy, ethics and science communication, and an audience of medical students, patient group representatives, genome biologists, journalists, academics and members of the public with an interest in the topic.
Below are some of my highlights of what turned out to be an engaging and fascinating day. If you were at the meeting I would love to hear your thoughts on the debate, and for you to clarify any mistakes or misrepresentations I have (inevitably) made.
Making gametes in and out of the body
Allan Pacey and Richard Anderson kicked off proceedings by discussing the cells that kick it all off to begin with: sperm and eggs. Human reproduction can be challenging, and much of the challenge lies in making viable gametes (‘Being a good egg ain’t easy’, as Anderson put it). We’re also in a society where, as Anderson pointed out, the birth rate in women under 20 is exceeded by that of women over 40. Pair this with the dramatic reduction in fertility of eggs used for IVF as women get older, and we have a problem. (Unless, as a member of the audience later pointed out, one does not want to have children; she even asked whether as a society, given the planet’s population pressures, we should be doing more to encourage women not to have children.) Thousands of women are freezing their eggs, but this procedure comes with its own issues, not least the limited supply of eggs, and of course has its own costs.
One way these challenges could be met is to achieve gametogenesis in vitro: make gametes from stem cells. Azim Surani described how little we know about human gametogenesis and the incredible journey the gamete precursors, primordial germ cells, make in the embryo, but also how research performed on human embryos and animal models was changing things. Truly ‘closing the gap’ between the soma and the germline would have powerful implications: just imagine, a participant in one of the panel discussions said, making thousands of viable eggs from a single skin sample from a female donor over the age of forty, or fifty, or sixty; it would be quite a game changer. (I was a little dumbfounded by the possibility, discussed later over a coffee with someone who wanted to invest in the technology, of generating eggs from men, though I really don’t know where the science is in that regard – any insights would be appreciated in the comments ?)
Technological advances can call into question existing terminologies as well as forcing us to come up with new ones. Sue Avery described how this is nothing new: the advent of IVF in the late seventies had not only transformed fertility but also revolutionised our understanding of human development. Before then, we really didn’t have much of an idea of what an early human embryo looked like, and seeing the process of embryogenesis under a microscope raised the issue of how and when exactly we defined an embryo. Magdalena Zernicka-Goetz then introduced the more recent creation of ‘embryo-like structures’ from human stem cells, made possible by tinkering with culturing conditions and inspired by lessons from the embryo. This give a glimpse into the black box of implantation, the point at which the body plan is established, and which is a crucial nexus for pregnancy failure.
It is remarkable to think how these experiments (and earlier ones using mouse embryonic stem cells) subvert biology, in the sense of not requiring fertilisation to generate something that looks like an embryo. Hence the inevitable question, explored by Robin Lovell-Badge, of when ‘embryo-like’ becomes ‘embryo’, and of the moral and legal status of these structures. Lovell-Badge introduced ‘synthetic human entities with embryo-like features’ (SHEEFs), a term coined by John Aach, George Church and colleagues that encompasses ‘embryo-like structures’ as well as stem-cell derived organoids and gastruloids (embryos derived from fertilised eggs, whether through sex or assisted reproduction, are, in this parlance, rather jarringly termed ‘non-synthetic’).
The ethical questions raised by SHEEFs often include reference to the 14 day rule – the fortnight limit for the culturing of human embryos in vitro, around the time that the primitive streak appears. So should similar limits exist for SHEEFs? It makes little sense, it would seem, to give a defined time limit to a process that does not involve a day 0 (i.e. fertilisation). According to the 14 day rule, gastrulation marks the point at which you should not pass, but, Lovell-Badge asked, what if synthetic biology allowed us to bypass it all together? (I think the idea was to 3D print tissues to make something that looked or like a neural tube for instance without it ever having passed through gastrulation.) And then there are the inevitable questions of sentience: if a SHEEF develops the neural structures for pain recognition, but not those for consciousness, should we care?
For Lovell-Badge, we shouldn’t get too far ahead of ourselves – he pointed out that in terms of self-organisation/self-assembly, we’ve had teratomas in the lab for years, and they can often make incredibly complex structures like eyes or teeth. SHEEFs could thus have moral equivalence with teratomas. Furthermore, since they are derived from stem cells, SHEEFs currently hold the same status as any stem-cell derived structure (at least the UK, as far as I could tell), and so might not require a legalistic overhaul. It all brought me back to a session in last year’s ISSCR meeting in Boston (I wrote about it here), where in a session on the ethics of organoids Melissa Little argued that organoids were just extensions of stem cell differentiation protocols and did not require a radical rethink of stem cell guidelines, and Jeurgen Knoeblich argued that ethical debates about consciousness in cortical organoids were based on perceived rather than real risk.
Stepping back from visions of Gattaca
In the ‘What next for genome editing?’ section, Andy Greenfield began with a plea to avoid adopting tedious tropes when we discuss ethics – people immediately bring up Orwell, Huxley, Gattaca, designer babies and eugenics, when the more interesting and pressing ethical questions concern what researchers are developing right now. Not only are these more interesting, they are also positive, in terms of promoting how research is helping human health. I liked this positive, contemporary angle: it’s not that we shouldn’t have discussions about future advances, it’s that they can become a bit like a session in the Black Mirror writing room, where imagination outpaces reality, and we fail to deal with what is feasible today.
Many of the ethical concerns in the meeting hinged around the differences between gene editing of the soma and of the germline. Greenfield made the point that we already perform germline therapy, in the form of preimplantation genetic diagnosis (PGD), which imposes non-random transmission of genetic material and leads to a change in future generations. Much ethical discussion of potential future gene editing and the human population could thus be refocused on PGD. Philippa Taylor was stridently against any form of germline editing – you can fix the patient (somatic editing is in line with medicine as she sees it), but the patient’s offspring have no say in the issue, and as they in a sense represent the human race, that is another issue entirely. Here, Greenfield’s point seemed pertinent: we already prevent the birth of individuals whose genotypes are deemed unfavourable, so there is not much ethical distance between PGD and a safe and efficacious form of gene editing, which, he stressed, we do not yet have. So, for those who oppose germline editing, the question might be turned on its head – do you support PGD, a practise which has become routine? Unless I missed it, I’m not sure opponents of germline editing addressed this question in the session.
Update 19/01: I heard back from Philippa Taylor on Twitter:
“To answer q about my view-yes we’ve always opposed PGD for ethical reasons & I said so in panel discussion”
So I did miss it – apologies to Philippa, and thanks for the clarification. I should further clarify here that not everyone at the event who opposed germline editing did so for the same reason (many would be in favour of PGD, I imagine). It wasn’t a simple case of two sides.
Greenfield proposed that not only should we continue gene editing, we also need to continue literal editing, in terms of proper science communication as a prerequisite for a mature, informed debate about the science. Sandy Starr agreed – not only is there a societal duty to do so, but human embryo research is dependent on fertility patients as donors of ‘non-synthetic’ embryos, and so is dependent on their continued support. Starr described how, with help from patient groups and practitioners, the PET generated a report on the basic understanding of genome editing. Starr argued getting your terms and definitions right is crucial. For instance, he insists on ‘genome editing’ rather than terms like ‘gene engineering’, as it best describes what is going on (I have undoubtedly strayed from his terminology in this article!). Practitioners and science communicators should be able to describe technologies like CRISPR clearly, but also to emphasise that not all editing is CRISPR-based, and indeed that CRISPR is not the end of the story. But how to define a genome in the first place? It seems a conceptually simple and complex at the same time – there are, proposed someone from the floor (I think it might have been Philip Ball), more than 7 billion human genomes, so it is not one steady state. Another audience member pointed out that one place to increase scientific literacy was in the classroom, but unfortunately the ethical implications of science tend to be taught separately from the science itself.
The final session took quite a broad look at the future of reproduction. Anna Smajdor explored how artificial gametes like those discussed by Azim Surani earlier in the day had the potential to ‘democratise’ reproduction. Biological boundaries were, with technological advances, dissolving in front of our eyes. (Again, currently, while artificial gametes might be realistic, we do not have artificial wombs, so would still need women involved in the process as far as I can tell.) Echoing an above-mentioned audience member, Smajdor wondered whether we focus too much on the positives of reproduction, and argued (in what she later admitted was a bit of a ‘rhetorical flourish’) that sexual reproduction is incompatible with Western liberal values, citing the inverse correlation between increasing indices of female emancipation, for instance education and career opportunities, and reduced child rearing.
Guido Pennings then argued that wide scale (in terms of the genome) and widely adopted germline editing was inevitable, to the demise of PGD. He imagined a future where embryos were screened in a more statistical manner over a range of genetic indicators and altered accordingly, and where parents would be faced not with the ability of creating an ideal embryo, but a good enough one. This drew some critique from the audience – the more we understand genomes, the more complex we understand them to be, and the more we appreciate that so many traits are multifactorial. Thus the dream of a checklist of genetic changes to be edited was not realistic. We are also not anywhere near demonstrating safe and efficacious germline editing in humans, and presumably will not be for some time, so PGD is not going anywhere for now.
This discussion reminded me of the reaction to the Mitalipov paper last year – multiple commentators pointed that, since PGD does very well in selecting against disease-carrying embryos, why would you edit those embryos you would normally discard after PGD? So should we be putting more effort into PGD rather than germline editing?
There really were so many interesting ideas floating about, arguments and counter arguments, ideas I winced at initially but later gave me cause for thought, and those I wholeheartedly supported until another came up to question my assumptions. As Andy Greenfield said, some of the most exciting debates concern what we are doing right now, and it must be a very exciting time for anyone involved in human development and reproduction.
The day was extensively tweeted using #PETconf17
Phillip Ball has written a great summary of the meeting for Prospect.
And Azim Surani’s work was picked up by the Guardian.