Gene Editing, Law, and the Environment

Life Beyond the Human

Irus Braverman

• Publisher: Routledge
• Year: 2017
• Print ISBN: 9780367138462
• eBook ISBN: 9781351685887
• Edition: 1
• Topic: Law
• Subtopic: Environmental Law
• Index: Law

Technologies like CRISPR and gene drives are ushering in a new era of genetic engineering, wherein the technical means to modify DNA are cheaper, faster, more accurate, more widely accessible, and with more far-reaching effects than ever before. These cutting-edge technologies raise legal, ethical, cultural, and ecological questions that are so broad and consequential for both human and other-than-human life that they can be difficult to grasp. What is clear, however, is that the power to directly alter not just a singular form of life but also the genetics of entire species and thus the composition of ecosystems is currently both inadequately regulated and undertheorized. In Gene Editing, Law, and the Environment, distinguished scholars from law, the life sciences, philosophy, environmental studies, science and technology studies, animal health, and religious studies examine what is at stake with these new biotechnologies for life and law, both human and beyond.

Part I
Conserving Nature, Driving Evolution

Chapter 1
Rules for Sculpting Ecosystems

Gene Drives and Responsive Science

Kevin M. Esvelt

Introduction: A Relatively Accessible Technology that Allows One to Affect Many

Few dog owners believe that our beloved companions would be able to out-compete the wolves in Yellowstone. No matter how impressed we are by our canine companions, seeing a wolf pack in the wild makes the idea seem absurd. Even so, the dog-to-wolf comparison is representative of a general rule: there is little need to worry about engineered organisms surviving and proliferating in the wild. As Darwin put it, “Man selects for his own good, Nature for that of the being which she tends” (Darwin 1859).

When we tinker with an organism, whether by selective breeding or precision genome editing, we’re making changes to a system that evolved to optimize reproduction in its ancestral habitat. Since our changes are for our own benefit, they are highly unlikely to benefit the organism in that original context, and natural selection would consequently eliminate them. There may be exceptions, especially when we’ve changed the environment, but the default assumption is that we are not nearly as good as evolution in the wild.

Of course, this assumes that the normal rules of inheritance apply, that each gene has an equal chance of being inherited by offspring. In nature, some genes have evolved to break the rules: they have a better-than-even chance of being inherited. These “gene drives” can spread through populations even if they decrease the organism’s ability to reproduce: even though carriers produce fewer offspring, more of them will inherit the gene drive (Bu rt and Trivers 2009). With the advent of CRISPR genome editing, this is a trick we can duplicate (Es velt, Smidler, Catteruccia, and Church 2014).

Instead of simply replacing an existing DNA sequence with a new version, we can additionally encode instructions for the cell to perform the same replacement on its own (ib id.). Simply insert genes encoding the CRISPR system and guide RNAs directing it to cut the original sequence next to the edited version. When we introduce this DNA into the germ line cells of an organism, those that will go on to make sperm or eggs, CRISPR will precisely cut the target sequence. The cell repairs the damage by copying our sequence in its place. Once one copy is

Figure 1 Edited genes in sexually reproducing organisms normally exhibit Mendelian inheritance: when an engineered organism with two copies mates with a wild-type organism, the offspring are guaranteed to inherit one copy. But when these “heterozygotes” mate with a wild-type organism, the offspring have only a 50 per cent chance of inheriting the altered gene. If released into the wild, the edited gene will not spread. In contrast, offspring that inherit the edited gene and a CRISPR-based gene drive system exhibit genome editing in their reproductive cells: CRISPR cuts the wild-type sequence, causing the cell to copy the edited gene and the CRISPR components in their place. With two copies, all progeny are guaranteed to inherit the edited gene and the drive system. This process repeats in subsequent generations, causing the edited gene to spread through the population. Such a “global” drive system is self-sustaining: it could potentially spread to every population of the target species in the world.

Courtesy of Kevin M. Esvelt.

inserted, the cell can produce the CRISPR machinery on its own. It cuts and replaces the other copy of the gene, so now there are two. When that organism mates with a wild counterpart, all of the offspring are guaranteed to inherit a copy. And in those offspring, gene editing happens again: one copy becomes two. When those offspring mate with wild equivalents, all of their offspring are similarly guaranteed to inherit the edited gene and the gene drive, as will the next generation, and the next, and so on (Bu rt 2003). Think of it as a find-and-replace for the entire species (Figure 1). What might we accomplish with such a tool?

First, health. The natural world harbors many sexually reproducing species that profoundly harm human health, any of which might be altered or suppressed to block disease transmission. Imagine a world without malaria or schistosomiasis, dengue or yellow fever, and little risk of Lyme disease and other tick-borne infections.

Second, conservation. Invasive species are a major cause of extinctions, particularly on islands. Gene drive systems might be used to suppress or locally eliminate these invasive populations, potentially saving many native species that would otherwise become extinct.

Finally, eco-friendly agriculture. Imagine a world where crops are grown without pesticides, because the pest species have been altered to dislike the crops’ taste.

In short, gene drives offer a way to solve ecological problems using biology, not bulldozers and poisons. The technology is far from theoretical: CRISPR-based drive systems have been demonstrated in yeast (Di Carlo et al. 2015), in fruit flies (Gan tz and Bier 2015), and in two species of malarial mosquitoes (Ga ntz et al. 2015; Hammond et al. 2016). Ho wever, this capability changes one of the bedrock assumptions undergirding our understanding of the living world. Recall the example of the dogs and wolves: our default expectation for selectively bred or engineered organisms is that natural selection will eliminate the engineered genes. But with a CRISPR-based gene drive system, this changes: we must assume that it will likely spread in the wild. Even if the organism cannot cross oceans on its own, it may do so by hitching a ride. And even if not, the odds are good that someone, somewhere, will move it for their own reasons. A single researcher now has the power to alter ecosystems, making decisions that could—if not countered—affect everyone.

Risks, Human Fallibility, and the Wise Engineering of Complex Systems

What do we do if a gene drive goes wrong? Suppose that someone were to release a drive system accidentally, or without authorization, or even malevolently. Because CRISPR can be programmed to cut nearly any DNA sequence, we can build a second gene drive to override any change spread by an earlier one (Esve lt, Smidler, Catteruccia, and Church 2014). This second gene drive can restore the pheno-type, the original traits of the organism, if not yet the exact DNA sequence. But no matter how we alter an ecosystem—whether with physics, chemistry, engineering, or biology—we can’t guarantee that it will return to its original state, even if we remove the source of the change. That’s true for a forest fire, an oil spill, and a gene drive. Any of them could force it past an ecological tipping point, an S-curve of no return. And because scientists should hold themselves morally responsible for all the consequences of our work, we should carefully consider whether, when, and how to conduct experiments capable of directly impacting the shared environment.

If we decide we do want to engineer very complex systems, how should we go about it?

Rule one: be humble. If we decide to act, we should always aim to make the smallest possible change likely to solve the problem. This is especially true for evolved systems, because natural selection favors evolvability, which is fostered by having many different weak interactions between components. That means changing one component is likely to affect something unintended. Since we can’t reliably predict these consequences, it’s best to make as small a change as possible.

Rule two: start local. No matter how minimal the change, don’t make that change everywhere in the world at once. Try it out in a small and contained area, observe the effects, and then decide whether scaling up is warranted.

The problem with the CRISPR-based gene drive system just described is that it has everything it needs to copy itself. Once released into a susceptible population, it is a self-scaling system that could have global effects. How do you run a field trial of a self-propagating system? Oddly, a National Academies of Sciences, Engineering, and Medicine report on gene drives suggests it is possible, albeit citing guidelines that predated CRISPR-based gene drives (2016). But consider the difficulty. Suppose we were to pick an isolated is land for the trial. It must be far enough away that storms cannot carry a fallen tree harboring a pregnant female organism to the mainland; models of much less potent drive systems have suggested that there is a nontrivial chance of spread in the event of a containment breach (Marshall 2009). Any boat or air traffic must be carefully monitored, lest the organism surreptitiously hitch a ride to populations elsewhere. And the island must be placed under military cordon to keep everyone unauthorized away. Why a military cordon? Because people will intervene for fun or profit. Gene drives are of considerable public interest as technologies go, so the field trial is likely to be public knowledge. Genetic engineering is controversial, and it would look very bad if the drive system escaped, meaning some people will have an incentive to release it. Conversely, if the drive system is likely to offer a benefit, there will be a tremendous temptation for people to move it.

An instructive example is rabbit calicivirus, which was deliberately imported and studied under quarantine by Australian scientists as a possible method of controlling the continent’s invasive rabbit population. In 1995, the virus escaped quarantine and spread throughout Australia (Schwensow et al. 2014). Remarkably, farmers in New Zealand, which boasts what may be the most strictly enforced biological border-control system in the world, successfully launched an illegal operation to smuggle rabbit calicivirus into their country (O’Hara 200 6). The government was forced to issue a pardon in exchange for information as to the extent of the introduction; many farmers were reported to show no remorse.

The lesson is that people will predictably intervene to foil the best-laid plans—including efforts to prevent organisms from moving past barriers. For a highly newsworthy technology such as gene drives, it is difficult to imagine scenarios in which outsiders do not interfere.

Quarantined field trials will only be safe if we are willing to establish a military cordon around the island and sink any boat that comes near. The quarantine must remain intact until the drive system is no longer capable of spreading—whether because it is a suppression drive that manages to eliminate the local population, or because it’s overwritten by a second drive system that cannot spread through unaltered organisms. Such an extreme quarantine, although possible, would be high-risk.

Gene Drive Risks: Primarily Social

Still, what exactly is the nature of the concern? What risks are posed by gene drive systems? Are ecosystems so fragile that an accidental release from a laboratory would cause problems?

The answer is no. Most gene drive systems, especially those that might be accidentally released, are thought to pose very little, if any, ecological risk. It is easy to build a drive system, but it takes careful engineering to build one that can’t be blocked by natural DNA sequence variation in the population, or even by the occasional incorrect copying event following CRISPR cutting (Noble et al. 20 16a). The first drive system created in fruit flies is costly enough to the organism that mutants of the wild-type sequence—which preserve the original function of the target gene and cannot be cut—will quickly outcompete the drive system. If the drive system were to escape, a wave of yellow fruit flies would likely spread around the globe, quickly followed by a reversion to the usual color, thanks to natural selection favoring the resistant mutants—which are themselves generated by the drive system at a nontrivial rate. Ecological effects are therefore highly unlikely.

What about extinction? There has been a great deal of misleading press about “the extinction invention,” written by journalists who focus on the potential to build population suppression drives that spread infertility or bias the population toward one sex (Regalado 2016). But models that take into account drive-resistant mutations clearly show that these will prevent the population from diminishing to the point of no return (Noble et al. 2016a; Marshall et al. 2016). If the goal is the removal of the organism, this will require frequent releases of organisms carrying suppression drives and most likely more than one version. A single unauthorized release is exceedingly unlikely to suffice. Moreover, the process would take dozens of generations, which is more than enough time for anyone so inclined to build organisms that deliberately contain mutations blocking the suppression drive; if desired, these could themselves be spread by their own drive system to immunize the population. Somewhat surprisingly, personal conversations with ecologists have revealed that they are among the most sanguine of scientists when it comes to potential impacts of CRISPR-based gene drive.

That’s not to say there is no ecological risk. Rather, the magnitude will differ tremendously by the species and nature of the change. Notably, gene drive systems will directly impact only a single target species, unlike the many other ways we are already impacting ecosystems globally. In the context of the Anthropocene and the sixth great mass extinction (see Chapter 2, this volume), gene drives are small potatoes, especially because phenotypic changes can be overwritten.

The same logic implies few security risks. Again, drive-mediated changes would be slow to spread, requiring many generations of vertical parent-to-offspring transmission. They’d be easily detected by sequencing the genome, because CRISPR gene drives have a uniquely recognizable DNA signature that is probably impossible to hide. And, of course, these changes could be overwritten by a subsequent immunizing reversal gene drive. A weapon that is slow, easily detected, and readily countered does not present a major threat, certainly not relative to many other available technologies.

So what’s the problem?

In the modern era, perception is paramount. Suppose one or more organisms with a global drive system escape from a field trial or from a laboratory and are lucky enough to find mates. Because most offspring will inherit the drive system, there’s a reasonable chance that there will soon be enough that stochastic chance is unlikely to eliminate the drive. If not overwritten, it will likely spread to every population of that species in the world (Esvelt, Smidler, Catte ruccia, and Church 2014).

Imagine the headline: “Scientists accidentally convert an entire wild species to GMOs. Is CRISPR to blame?” The damage to public trust in scientists and governance would be severe and long-lasting. At a minimum, it would be the end of hopes to use gene drives against malaria and schistosomiasis. A mere decade-long delay could keep us from preventing millions of deaths and billions of infections. Judging by past ethical and safety lapses in fields such as gene therapy, a decade would let us off lightly. Add to this situation a dash of widespread public suspicion of genetic engineering, thanks in part to companies such as Monsanto, season it with recurrent media stories every time the drive system spreads to a new city, country, or continent, and bake it with additional incitement from loud anti-technology activists, and you have the recipe for a decades-long delay for gene drives and serious potential damage to more mundane CRISPR-based applications—a very large fract...

Table of contents

• Cover
• Title
• Copyright
• Dedication
• Contents
• Acknowledgments
• Notes on Contributors
• INTRODUCTION Editing the Environment: Emerging Issues in Genetics and the Law
• PART I Conserving Nature, Driving Evolution
• PART II Technologies of Governance
• PART III Human-Nonhuman Boundaries, Worked and Reworked
• AFTERWORD Governing Gene Editing: A Constitutional Conversation
• Index