Of Science, CRISPR-Cas9, and Asilomar

On Thursday, March 19, 2015 Science published (on-line) a Policy Forum entitled  A Prudent Path Forward for Genomic Engineering and Germline Gene Modification.  The piece had 18 authors, including David Baltimore, Paul Berg, Alta Charo, George Church, George Daley, Jennifer Doudna, Ed Penhoet, Keith Yamamoto, and (as, with Alta, one of only two non-scientists and definitely as one of the lesser lights) me. The Policy Forum recommended that steps be taken to “strongly discourage…any attempts at germline genome modification for clinical application in humans, while societal, environmental, and ethical implications of such activity are discussed among scientific and governmental organizations.”

A bit more than forty years earlier, on Monday morning, February 27, 1975, David Baltimore opened the famous Asilomar conference on recombinant DNA. That conference had been called after a letter in Science from the leaders in the field called for a moratorium on recombinant DNA research until important safety issues could be worked out. The three and a half days of the Asilomar meeting produced safety guidelines which led the group to lift their (totally informal and non-binding) moratorium – and ultimately led the NIH Recombinant Activities Committee; to federal (and foreign) biosafety regulations; in some tellings the wildly successful application of recombinant DNA techniques to research and medicine; and, undoubtedly, the most famous story in modern scientific self-regulation. (An excellent, non-speculative, memoir of the Asilomar meeting by George Frederickson, head of the Institute of Medicine at the time of Asilomar and director of NIH a few months later, can be found here.)

The roughly 1400 words of the recent Science Policy Forum include the name “Asilomar” exactly once, in passing, even though two of its signers (and vocal advocates), David Baltimore and Paul Berg, were among the five organizers of the Asilomar meeting. They also include only limited discussion the statement’s origin and, more importantly, of its meaning. I was not one of the organizers of the meeting that led to the Policy Forum, though I did attend and participate in the meeting. Neither did I draft or wield a decisive editorial pen in the writing of the document, though I did have several opportunities to offer comments, which I did. (Some of them, I think, made it into the piece.) I can’t read the minds of the other authors, some of whom didn’t attend the meeting, but I can tell you a bit about my thinking, both on the document and on the underlying issues raised by advances in genomic editing.

A Little Background on the Document

On October 2, 2014, I got an email entitled “Invitation: Bioethics Workshop.” It came from Jennifer Doudna’s email account but was signed by Doudna, Mike Botchan, Jacob Corn, Ed Penhoet, and Jonathan Weissman (all of whom attended the meeting and signed the document). It also has a list of thirteen invitees, eight of whom ended up attending the meeting and signing the document. It invited me to a full day workshop on Saturday, January 24 in Napa Valley “to discuss the bioethical issues raised by the explosion in new genomic editing methods.” With some qualifications for calendar uncertainties, I said “yes!”

Coincidentally, I was flying home from Dulles airport in Washington, D.C. in late October and ended up sitting next to Mike Botchan. We didn’t know each other, but, after some conversation that established he was a molecular biologist (he was carrying around a new molecular biology textbook) and I was a law professor who worked on bioscience issues (I recognized that it was a molecular biology textbook and knew some of the authors), he surprised me by asking “Is your name Henry Greely?” We spent the next several hours happily talking about a wide range of issues in biosciences, including but not limited to genomic editing. Also, at some point someone decided that Alta Charo and I should do an overview of the legal and regulatory framework, so the two of us corresponded and talked about dividing up that territory; otherwise, I arrived at the meeting without a lot more information about it.

I have a nagging feeling that, at the meeting, we agreed the discussions would be confidential, so I won’t go into details. I don’t think my colleagues will be too upset if I let slip that the conversation on January 24 was lively – and, for me at least, great fun. Well before the end of the day, a consensus seemed to be emerging, one part of which was to move from “emerging” to “written down.” I got my first draft ((I think) of such a document on February 6 from Jennifer Doudna and, after quite a few revised versions (each with opportunities to comment), the final version was published on March 19. Changes were not usually discussed collectively, but most of the time they were understandable and unobjectionable, including the change to add a few authors who agreed with the document but had not attended the January meeting.

By the way, thanks to California’s truly impressive drought, January 24 in Napa was warm, clear, and gorgeous, as was the 25th. The Carneros region doesn’t have the incredible natural beauty of Asilomar – though it is beautiful – but the weather was unbeatable.

How I Interpret the Document

All documents are ambiguous (at least, once lawyers read them). That’s particularly true of a collective writing enterprise done largely at a distance. I think there are three non-trivial possible ambiguities in this one. I will give my take on them, though without guaranteeing that all, or any, of my co-authors agree.

First, what does the document seek to “strongly discourage”?

[A]ny attempts at germline genome modification for clinical application in humans, while societal, environmental, and ethical implications of such activity are discussed among scientific and governmental organizations.

I read that as clearly encompassing “making babies” and to include “making babies” for enhancement purposes as well as disease prevention purposes. The “clinical application” there, I am confident, is in contradistinction to “research uses”, not to “non-clinical babymaking applications.” I think it just as clearly does not prohibit research on cells, cell lines, and tissues, even cells, cell lines, and tissues that could become part of the germline, such as human embryonic cell lines (hESCs), human induced pluripotent cell lines (hiPSCs), various more direct egg and sperm precursor cells, and even human eggs and sperm. To me, it is putting the modification into the germ line of what is or is intended to become a living human being that is our focus. (That is also what is widely banned, which the lawyers in the group argued should be – and was – noted in the recommendation.)

This does still leave open a question about ex vivo human embryos. Some people – including some reasonable people – consider them human beings. Other people – also including some reasonable people – do not, either before implantation or at some point between implantation and just after birth. My own view is that ex vivo embryos are not human beings and so the recommendation does not apply to them, though I would, personally, at this point still discourage researchers from experimenting on them without substantial societal discussion.

Second, is this only about human germline genomic modification? The paper, mirroring many of the authors, focuses almost exclusively on those issues, but only almost. There is a nod at the beginning to the fact that “in other organisms it provides methods to reshape the biosphere for the benefit of the environment and human societies.” As noted below, I think the implications of non-human uses are more pressing than human uses, but it is fair to say I did not win that argument. – at least in this paper (but I am stubborn). Issues around somatic cell genomic engineering are raised more directly, but, for my taste, are somewhat downplayed in this paper. Those uses, especially if they move beyond treatment to enhancement, will raise concerns.

Third, is this only about CRISPR-Cas9? That’s the easiest one to answer: no. It’s about recent and likely future advances in genome engineering. CRISPR-Cas9 is the current leader in truly revolutionizing the field by making such editing cheaper, faster, and more accurate. It is an amazing and deeply important accomplishment. And it is, understandably, the focus in a paper from this group. But should some other genome engineering technology end up supplanting (or even supplementing) CRISPR-Cas9, the issues discussed in this paper will apply, at least a priori, with equal force.

How I (Currently) See the Issues

So in this section I’m constrained neither by confidentiality nor by my lack of knowledge of other people’s thoughts. And as I am only mildly constrained by prudence, I’ll give you my current thoughts on the issues raised by much more effective genomic engineering, in three tranches: human germline genomic modification, human non-germline genomic modification, and non-human genomic modification. You may be surprised to know that I’ve listed them in what I see as their order of increasing importance.

            Human Germline Genomic Modification

Frankly, although the fuss has been about human germline genomic modification, I think that attention is misplaced. I don’t expect engineered human germline modification to be a big issue – as a practical matter – for a long time, if ever, for several reasons.

First, the safety issues are enormous. That’s not to say anything bad about CRISPR-Cas9 or other genome editing techniques, but the stakes are enormous . . . a human baby. Editing methods that are perfectly accurate and safe for most purposes need several more nines in that context. It is not just a matter of perfecting CRISPR-Cas9’s targeting ability or any other particular parameter; we’d want to see, in advance, whether other things, previously unanticipated, about the process were likely to cause problems in making babies. You’d have to be criminally reckless, or insane, to try to make a baby this way unless and until we’ve had a decade or more of preliminary research, with human tissues and with non-human animals (including certainly primates and maybe even some of the non-human apes), showing that it is safe. If the moral risk isn’t enough of a deterrent, the potential legal liability should be.

Second, the medical demand should be small. In what cases would a couple need germline genomic modification in order to have a healthy (to be more precise, a “not known to be in a particular way unhealthy”) child? I can only think of three:

  1. A would-be parent who is homozygous for a dominant disease (two copies, let’s say, of the Huntington’s allele),
  2. A would-be parents couple who both have the same autosomal recessive disease (let’s say cystic fibrosis), and
  3. A woman with a disease caused by her mitochondrial DNA sequence.

The third example was expressly excluded from the Science paper. It is the subject of other discussions and, more importantly, of a different, simpler, and likely safer intervention through mitochondrial transfer. How many would be parents are in the first or second situation? And how many of them will survive long enough, and be healthy enough to want to have children? We don’t know but the answer must be “very few.”

There is another potential category – people who are at risk of having a child with a well-understood genetic disease (just about everyone at risk for having a child, in fact), and who are concerned enough about it to do something (other than effective contraception) to forestall it, but who, for whatever reason, will not use preimplantation genetic diagnosis (PGD) plus embryo selection or fetal genetic testing plus abortion. Just how many such people are there, who would be willing to do germline genomic modification but not the much better-established and well-tested PGD? There are none in the Vatican; “unnatural” reproduction is wrong whether or not embryos or fetuses are destroyed. There may be a few people in this world who hold that combination of positions, but I suspect only a few. That few should shrink even more if, as I suspect, even germline genome modification requires testing of the modified embryos and possibly discarding those where, for one reason or another (off target modifications?), it didn’t work.

Third, the non-medical demand will also be small, at least for a reasonable time. I think this is the real fear of most people – genetically engineered superhumans. But it turns out that, after hundreds of billions of dollars spent, we know surprisingly little about the genetics of disease. We know almost nothing about the genetics of “enhancement.” I cannot think of a single non-disease trait where we can say confidently that one non-pathogenic allele is highly likely to confer a substantial advantage over another. That will change, of course, but how much and how fast? My guess, in both cases, is not very.

And fourth, germline genome modification in humans will continue, for a long time, to be controversial. How many would-be parents, how many companies, how many clinics, want to take on that controversy for such a small payoff?

So why was human germline genomic modification the focus of the Science piece – and an issue that Science was so willing to publish so quickly? Well, it is extremely controversial. The public concern, stoked by some organizations, of “playing God” by “reshaping the human race” is not, or at least seems not to be, trivial. That probably would have been enough to ensure a focus on germline modification at this point.

But I think the answer is, at least, in part, deeper. Some people feel “Science” (the field, not the magazine) promised that we would not try human germline modification. That was, of course, an easy promise to make when it was impossible, like promising not to try human reproductive cloning when it seemed impossible. And like the cloning promise, which at the time, around 2000, seemed to help smooth the way for human non-reproductive cloning/somatic cell nuclear transfer, swearing off germline modification made non-germline modification more acceptable. (Ironically, even human non-reproductive cloning turned out to be much harder than anticipated and was not achieved until 2013.)  The addition of this “inside science” concern, I suspect, made the topic irresistible.

Human Non-Germline Genomic Modification

Non-germline (somatic cell) genomic modification in humans – your basic gene therapy – is, in fact, a likely huge use for genomic editing technologies. About 35 years after gene therapy was first tried by Martin Cline, it is finally approaching clinical use. In fact, one gene therapy, called Glybera, has already been approved in Europe. Others are in phase 2 and phase 3 trials around the world. I would be surprised if FDA did not approve a few in the U.S. in the next year or two.

CRISPR-Cas9 and its equivalents can make the genomic modifications faster, cheaper, and more accurately. That’s not the only issue with gene therapy certainly, but this revolution should accelerate the already incoming tide of gene therapy.

Why was this discussed only in passing in Science? I think because it is not very controversial. The issues of somatic cell gene therapy, a.k.a. human non-germline genomic modification, have been discussed for many years and, apart from questions of safety, efficacy, hype, and research ethics, none has seemed very important. Changing the genes of one person, who will die without passing those on to anyone else, just hasn’t raised deep questions.

I think that’s appropriate, though I’d note those questions of safety, efficacy, hype, and research ethics continue to be important. But it is ironic because genomic editing should pay off, in a noticeable and possibly very large way, much sooner for somatic cell gene therapy than in germline modification.

Non-Human Genomic Modification

To me, the biggest likely change in our world from CRISPR-Cas9 and other genomic editing methods won’t be in humans but in the non-humans we use the methods to modify. As it gets cheaper and easier to modify genomes, non-human genomes offer freedom from a lot of regulation, liability, and political controversy, while offering plenty of opportunities to improve the world, become famous, or make money – with combinations of all of the above.

Want to end malaria? Come up with a modified version of Aedes aegypti that can’t transmit yellow fever, dengue fever, or chikungunya viruses to humans and will outcompete and eventually eliminate the wild type. Want to make a really economical biofuel? Take an algae and modify its genome in thousands of ways to optimize it for producing hydrocarbon fuel. Want to bring back the passenger pigeon? Use CRISPR-Cas9 to modify the genomes of existing band tail pigeons to match, more or less, the genomes sequenced from specimens on the extinct passenger pigeon. What to corner the market in high-end gifts? Start playing around with horse genomes adding in bits and pieces from other species in an effort to produce actual unicorns. What to make a splash as an artist? Use CRISPR-Cas9 to make a warren of truly glow-in-the-dark rabbits.

In fact, on the same day Science published the moratorium call on-line, it published on-line an article on one very successful “gene drive” system, using CRISPR-Cas9, that could spread a chosen genetic variant very quickly through an entire population. (See the news story in the next day’s magazine here.)

It is these kinds of uses of genomic engineering that could reshape the biosphere. As the ability to make carefully engineered genomic changes becomes more widely accessible, the possibility of insufficiently controlled or considered experiments increases dramatically. And so, of course, does the chance of more controlled interventions. I would like to see much more focus on this issue, of great practical importance, instead of so much attention on the sexier issue of germline genome modification in humans.

What Would I (As Currently Advised) Do?

Perhaps not surprising, I would do what the Science piece called for. While reminding people that making babies this way is illegal or heavily regulated in most of the world, I would call for a moratorium on even trying it until both further scientific research (mainly on its safety) and public discussion and study (mainly on ethics) had been attained. I would hope the U.S. National Academies or a similarly august body would do an in-depth study of the issues, with similar studies done in other countries and by other bodies. And I would call for a meeting, another Asilomar – and preferably, says the romantic, and Californian, in me, at Asilomar – at which the details of the moratorium could be worked out. I would hope that that meeting would pay more attention to the huge issues raised by our accelerating ability to modify non-human germlines – in the lab, on the farm, and in the wild, as well some of the non-germline human issues.

In the long run, I believe the permissibility of using germline genomic modification to make babies will be, and should be, a political issue. Right now, I suspect I would opt for regulating it on a safety/benefit basis, allowing it only when the potential benefits outweighed the risks. But I might change my mind, either because of newly discovered facts or well-made arguments. Importantly, though I do not think that my view should govern. The people, through their governments, should govern. If South Dakota, or Germany, wished to ban it and California, or Singapore, wished to encourage it – preferably, in all cases, after free, open, and active debate – so be it.

But to get to those decisions, we need to lay the groundwork, with scientific research, ethical argument, and venues for discussion. Although all politics is local, I do admire the way the U.K. handled the mitochondrial transfer issue. First they had lots of ethical and scientific discussion, including an excellent report by the Nuffield Council. Then they had an actual Parliamentary debate and vote (a free vote, too, not bound by the party whips). When Parliament approved it, it didn’t become “approved” and freely available, but went back to the Human Fertilisation and Embryology Authority, which licenses specific clinics to perform the procedure under specific conditions. Now, whether a process like that is feasible in the United States can be (easily) doubted, but some combination of expert and public involvement and debate, politically open and legitimate decision-making, and controlled implementation seems like a good idea.

I hope, and believe, that the Science piece has a reasonable chance to make a small but useful contribution to that end. And so I am pleased to have been part of it and thank, and congratulate, the organizers of the January Napa meeting, especially Jennifer Doudna, for their work!

Hank Greely

Director, Center for Law and the Biosciences

7 Responses to Of Science, CRISPR-Cas9, and Asilomar
  1. Really great points. I hadn’t considered what the size of target market for these treatments would be. After reading this, I completely agree that it seems like the cases where germline editing applies and PGD does not is likely to be extremely small. So while I agree that its important to make sure therapeutic attempts wait until the appropriate research and if it receives public sanctioning, it hardly seems like we’ve reached a kind of watershed moment for human society with regards to this technology. I agree that the non-human applications are much more concerning.

    Ecosystem engineering is much more relevant today and likely to have larger consequences in the near-term. Not to mention these applications can cross borders so have international political concerns. Humans have already made a mess of things with invasive species across the globe – both intentionally and unintentionally released. Take a look at New Zealand’s struggles to protect the native flora and fauna.

    The mosquito genetics community has been talking about population editing for years and has discussed ways to do it safely and responsibly (see for example http://www.sciencedirect.com/science/article/pii/S1471492204002909). Unfortunately the cheapness of CRISPR editing methods and its ease of use may completely bypass these intentional, soft law regulations. The gene drive paper in flies is particularly concerning because it illustrates how easy it would be for a rogue or careless researcher completely flout all the proposed safeguards through reckless or sloppy science.

    With human editing the use cases are likely to remain small, their societal effects minor, and we reserve the option to regulate more or less with time. The transpolitical, self-perpetuating nature of gene drives means it would be a lot harder to put the proverbial genie back in the bottle.

    Anyways thank you for the post. It’s interesting to hear the thoughts behind the article and also many of the things that you thought were left unsaid.

  2. Pingback: Cells Weekly - April 12, 2015 - Stem Cell Assays

  3. Pingback: Germ-line gene editing: Time for a new Asilomar? Perhaps, but with some caveats… New op-ed in the Huffington Post | Bioethics & Society at King's College London

  4. Pingback: Germ-line gene editing: Time for a new Asilomar? New op-ed by Dr Silvia Camporesi in the Huffington Post | Department of Social Science Health & Medicine at King's College, London

  5. I’m also excited about the possibility of food production! McMammoth is going to be a great dollar menu item! Joking aside, we will be unable to stop the spread of genetically modified humans, just like we can’t control genetically modified plants. There is always “across the border” for research.

  6. This is a huge issue. Hank lays it out clearly.

    As I see it, we should discuss and investigate it in depth, and go slowly, when and if we do decide to– but we should not ban it, because we do not yet know how valuable it might turn out to be.

  7. Thank you for this insightful summary.

    You assert the following:

    “Importantly, though I do not think that my view should govern. The people, through their governments, should govern. If South Dakota, or Germany, wished to ban it and California, or Singapore, wished to encourage it – preferably, in all cases, after free, open, and active debate – so be it.”

    I agree that such issues are a proper province for debate in democratic governments.

    My question is inspired by my experience from a similar domain of rapidly expanding complexity amplified by the explosive growth of computational power; that of financial derivatives.

    By way of background, the Bank of International Settlements in Geneva, the central bank of central banks, states there are currently over 620 trillion dollars in derivatives outstanding. Although industry voices from the International Swaps and Derivatives Association (ISDA) assert this figure is overstated because of duplicative accounting, prominent central bankers such as Andy Haldane, chief statistician of the Bank of England, emphatically disagree.

    Irrespective of which view is correct, what is clear is that the amount of derivatives extant dwarfs that of the estimated 15 trillion currently on the balance sheets of the global central banks. A review of the central bank literature reveals the extreme complexity of properly regulating these derivatives. Governments and central banks do not have the resources nor expertise to properly track the myriad of new products, how they should be properly valued, and what level of capital should be set aside in reserve for each asset class. Synthetic ETFs are but the latest example.

    Ben Bernanke and Paul Volcker have each asserted that continued recapitalization of this derivatives overhang poses a serious threat to stability, despite lessons learned from the crash of 2008.

    In an era of dramatic governmental funding cuts worldwide and an ongoing crusade against governmental regulation of any sort, to what extent will democratic societies be able to influence future trajectories in the development of germline gene modification in the hands of private actors?

    Is this an issue worth addressing sooner, rather than later?

    The experience and stories of central bankers in wrestling with derivatives seems both relevant and sobering here.

Comments are closed.