The capacity to sequence a human genome cheaply and the discovery that CRISPR/Cas technology can be used to make more precise genetic edits than previously possible do not change the decades-long debate on whether to genetically edit the human genome. Instead, they make urgent the need for regulation and education informed by comprehensive ethical analysis. Although there is a great deal of attention being given to genetically editing adult cells for purposes of treating health conditions in an individual, this discussion will focus on the most consequential and controversial question, which is whether we should be allowed to make changes to the germline—eggs, sperm, and embryos—that will manifest in a person and be passed on to that person’s children. If so, what kinds of changes should be allowed?
September 01, 2016
CRISPR Lit the Fire: Ethics Must Drive Regulation of Germline Engineering
Theoretically, laws in the United States create a social structure in which citizens are free to pursue their own versions of a good life to the extent that each pursuit does not unreasonably infringe on the rights of another. The goal is a morally neutral legal system. In reality, values are often implicit in laws that prescribe what the good life should be to some degree. Some countries that prioritize community and solidarity are more up front about espousing value-laden legal systems, whereas the United States tends to prioritize freedom and autonomy.
This distinction gets fuzzy when we consider laws that would govern intentional changes to the physical and mental states of human beings. Does enabling the freedom of parents to choose the genes of their children unreasonably infringe on the rights of those children to design their own preferred lives? Does prohibiting all germline engineering prevent parents from giving their children the genetic foundation for the most life opportunities? Do legal limits on the kinds of genetic changes allowed impose the lawmakers’ version of the good life on society? If we allow germline engineering, must we also require that some edits be made available to everyone, or even mandated? Answers to these and related ethical questions can be informed by the science of germline engineering and by the psychology of expectations.
State of Science
The explosion of genomic science and its clinical application in the past two decades has allowed us to: make informed reproductive decisions (two carriers for Tay-Sachs can choose whether to procreate with a 25 percent chance of having affected children1); tailor disease treatment (anticoagulant warfarin dosages can be adjusted based on the presence of three genes that affect its ability to thin the blood2); make lifestyle choices (individuals with predisposition to skin cancer can take extra care to avoid the sun3); and make future preparations (individuals with a higher chance of early-onset Alzheimer’s can take early action to fill out advanced directives and living wills4). Despite all this, we know little about how genes and noncoding genomic regions affect each other, or how the environment affects the genome. We are increasingly proficient in genetic language, but nowhere near fluent.
We have limited instructions for modifying human embryos, but we do have the tools. For years, biologists have been using transcription activator-like effector nucleases (TALENs) and zinc finger nucleases (ZFNs) to edit genomes; however, these technologies require engineering proteins to target specific DNA sequences, which is expensive and time-consuming.5 Clustered regularly interspaced short palindromic repeats (CRISPR) and CRISPR-associated protein (Cas), combined with the assistance of specific guide RNA (gRNA), can be directed to make specific cuts in DNA, which can then allow the deletion and/or insertion of desired sequences.6 This method allows for the editing of multiple sites simultaneously and is vastly cheaper, simpler, and more efficient than techniques involving TALENs and ZFNs.
The CRISPR/Cas system has already been used to engineer yeast to make biofuels, disable disease-causing parasites, and correct a mutation accounting for metabolic disease in mice.7 However, most labs are struggling to optimize the specificity of the gRNA required to direct CRISPR/Cas to the intended gene sequence without it cutting sequences elsewhere. There are multiple pathways for error: (1) intended specific gene cuts might have unintended consequences because of unknown relationships between the gene intended to be modified and other genomic regions; and (2) CRISPR might also make cuts in unintended locations in the genome, or might stay in cells after the intended cuts are complete and continue to edit similar off-target sequences.8 The consequences of these unintended actions within the genome are only limited by the imagination.
Saying we can precisely modify the genome of human embryos with the current level of editing technology is like saying a child with power tools can build a house. Yet, companies are already developing business models for gene therapy based on CRISPR, and many in the field are currently involved in or preparing for human clinical trials using this technology in the coming years. Current gene editing clinical trials are beginning with modifying genes in somatic cells that are not heritable. However, since April 2015, two different Chinese teams have published research on human embryos to edit disease-associated genes.9 Only some of the embryos in each of these studies were successfully edited, and both teams discovered significant off-target effects. Both studies used nonviable human embryos, and neither team intended to implant the embryos. The choice to experiment on human embryos is symbolic of the intention one day soon to modify the human germline.
Risk, Safety, and Harm
The most immediate reason to regulate germline engineering is that it is currently not safe. Scientific consensus holds that our understanding of gene editing technology and other genetic intervention are too limited to endeavor modifying germline cells for clinical applications.10 But with increasing excitement and growing research investment in the editing system, this consensus will likely break. The question becomes: when, if ever, and for which clinical applications of human germline cells will the expected benefits offset the risks?
Demonstrating low enough risk to establish the first human trials will be an enormous task. It is unknown whether nonhuman animals and nonviable human embryos would be affected by germline gene editing in the same way as would viable human embryos. These questions bear on the extent to which we can extrapolate experimental data from the former to the latter. The next logical step would be experiments involving the use of viable human embryos and finally clinical trials with modified germline cells. Putting aside the moral distress clinical use of modified human germline cells will cause for many, potential physical risks to resulting persons remain unknown unless all off-target effects and other unintended consequences can be traced and eliminated. Because of the vastness of the human genome, which contains more than three billion base pairs, and the multigene complexity of most inheritable diseases, the likelihood of completely eliminating unintended effects from a genetic edit seems low.
Even targeting genes associated with a Mendelian (single-gene) disease can cause problems: the target gene might be critically linked to other biological pathways, or the introduced CRISPR system might cut at incorrect locations within the genome.11 An edit that appears to have had a successful and isolated impact might not demonstrate unintended consequences until some point after the baby is born, or might not manifest until future generations. Again the question is which genetic predispositions identifiable in an embryo are so deleterious that unavoidable risks to the first human subjects will be worth taking?
As already demonstrated by the aims of the two different Chinese teams, the first human editing endeavors will likely be focused on eliminating disease. This stems from the notion that, all other things equal, a human life is better without certain deleterious genetic predispositions. If we accept this premise and can identify these predispositions, the conclusion that we should attempt to edit them out of the human genome does not necessarily follow, because in most cases we can instead choose to select against affected human embryos containing these predispositions. Why should we introduce the additional risks of unintended editing consequences described above, if we can instead use methods like preimplantation genetic diagnosis (PGD) and in vitro fertilization (IVF) techniques—methods proven safe over the past 25 years—to ensure the chosen embryo does not have abnormal genes, or lack important genes?
The response might be that we would use editing of germline cells only to modify genes correlated to diseases, similar to the Chinese teams’ targeting genes causing susceptibility to HIV and a gene associated with beta thalassemia. However, if targeting such genes individually would substantially lower the risk of disease, then why not use embryo screening? In the case of modifying genes to increase HIV resistance, education and pre – and post-exposure therapies might be more effective and less risky disease prevention methods.
According to the “someone else” paradox, it is possible that accurate embryo editing to eliminate disease does not prevent harm, because the editing process brings a different person into existence.12 If individuals are to some degree the product of their genes (research currently says who we are is determined by about 50 percent genes, 50 percent environment), then it is possible that changing an embryo’s genes eliminates one future person (who would have existed pre-edits) and creates a new one, who will be born. The person who is born never had the predisposition to disease, so that person is not helped by the editing process. If anything, the embryo with disease genes who is edited out of existence might be harmed by not getting a chance at life, albeit with disease genes. To give this argument full consideration, we should think about how much editing is required to create a new person. Is one gene enough? What affects must this gene have? Our best tool to understand the extent to which a gene defines an individual is the lived experience.
What Is a Disease?
There is a remarkably intricate relationship between one’s genetics and one’s environment that leads to the unique nature of each individual.13 No two people with the same genetic-based condition will be affected by it in the same way. This is obvious for spectrum conditions like autism or Down syndrome, which result in a vast range of physical manifestations with which some will lead independent employed lives and others will require constant care. But even for nonspectrum genetic conditions, the experience will be different based on the individual’s complete genome and environment.
Consider two individuals with the same genetic-based total deafness. One is born to deaf parents in a thriving deaf community. The child is taught American Sign Language from an early age, forms strong social bonds, receives encouragement, develops confidence, and lives a satisfying life. The second is born to hearing parents, who at the advice of a physician provide the child with cochlear implants, which allow a sharper sense of sound vibrations but do not grant true hearing.14 The child does not learn sign language, and cannot hear well. The child grows up feeling isolated and incapable, not quite part of the deaf community or the hearing, which leads to lifelong demotivation and depression. The outcomes of these hypotheticals could easily be reversed—child one grows up feeling unsatisfied and depressed, child two motivated and happy—but the point is twofold: (1) a single, albeit significant, genetic condition is not determinative of quality of life; and (2) what constitutes a “condition” or “disease” is debatable.
Many of our greatest achievements and most satisfying moments spring from our greatest challenges. Many of these challenges are partly genetic—everything from running a marathon despite being blind to shopping for proper-fitting pants despite being a normal enough size. Finding joy and meaning often comes from succeeding in a society built for the ephemeral “average” person, and also from being involved in the evolution of acceptance—whether teaching that difference is good, learning it, or actively transforming parts of your community to be more accommodating. With this in mind, what genetic conditions can we agree are so deleterious that they should be edited out? An often-cited example is Tay-Sachs, which, beginning around six months of age, involves a progressive deterioration of nerve cells leading to blindness, deafness, and inability to swallow, and usually results in death by the age of four.15 Many—maybe most—believe that an individual’s life, and society, would be better without Tay-Sachs.
How would we go about determining the set of conditions that are deleterious enough to edit out and worth the technical risks described earlier? Predispositions to cancer or heart disease are strong candidates. We might design a matrix of condition variables, including longevity, pain, ability to relate to others, and ability to comfortably experience basic functions like eating, sleeping, breathing, moving around, and sexual pleasure. If, for a given genetic-related condition, the combination of these drops below a certain threshold, it might meet the criteria for inclusion on the “Edited List.” But we return to the problem that one person’s experience of a genetic condition will not be another’s. Even two young children with Tay-Sachs will have different experiences, some of which could be satisfying and joyful. If we say the average experience of Tay-Sachs warrants editing it out of existence, we are claiming that a given affected individual’s positive experiences and the value he or she would bring to others are not worth the negative experiences and the burden that individual would pose to society. This might be true. Who is willing and has the right to make these determinations?
Once we start making these determinations, our priority shifts from accommodation and acceptance to efficiency and comfort. This shift carries implicit expectations, which can be upset in various ways. If access to germline engineering of deleterious conditions is limited by geography or resources, those without access will be disadvantaged by conditions allowed to manifest and also by a likely decrease in attention to treatment and accommodation of those conditions. If we can edit genetic blindness out of society, why continue making signs in braille and training seeing-eye companions?
Assuming we can ensure equal access to the list of acceptable gene edits across society, our expectations become calibrated to longer, healthier lives. As mentioned, it is possible that the edits will have unforeseeable health consequences due to our poor understanding of gene-gene and gene-environment interactions, which would run counter to the intentions of germline engineering. Even if unforeseeable health effects are kept to a minimum, our new expectations will be a society without cancer, heart disease, Tay-Sachs, and all other similarly determined genetically based deleterious conditions. A century ago, without proper sanitation or penicillin, living past age 50 was an achievement. Today, it is a shameful tragedy when an 80-year-old dies due to avoidable and curable diseases like pneumonia or staph infection. In the engineered future unplagued by the Edited List disorders, society might feel the same about a person who dies at age 90 from secondary complications associated with Down syndrome, like heart valve problems, simply because it could have been avoided with editing. At that point, does Down syndrome (a chromosomal disorder) get added to the List? The Edited List will have to be regularly reevaluated to reflect expectations shifting toward a healthier and longer living society. What constitutes “healthy” will shift, as our condition variables expand. What is to prevent the addition of high verbal capacity, being taller than six feet, and being able to do 30 pull-ups to the matrix of basic life experiences? If nothing prevents these additions, then is anything wrong with establishing new health baselines?
A Meaningful Life
The expectation of unlimited modifiability might encourage shifts in baseline health before reproductive gene editing is made technologically safe (for scientific uncertainty reasons described earlier), and might also affect the human capacity for satisfaction. Knowing that we designed the efficiency and comfort afforded by engineering certain conditions out of society, we will always believe we have the power to reach the next level of efficiency and comfort. This expectation of control might undermine the acceptance of one’s circumstances, which can lead to dissatisfaction and conflict. If the option is available to edit one’s children to be taller, how can we expect parents not to take advantage when they know that, statistically, taller people have greater success? What happens to the parent-child relationship when the selected child is not as tall as expected, or not as successful? Can we expect the child, who assumes that success comes with height, to work as hard? Can the child find as much satisfaction in accomplishments, knowing that some were due to an intentional genetic modification?
The counterargument says that some people are just genetically luckier—do we see more relationship and expectation conflicts with naturally taller, or smarter, or more attractive people? Likely not. But the difference is the intention attached to the genetic modification. Individuals would not engineer their children or themselves if they did not have the expectation for an outcome. Some do attach expectations to unintentional natural genetic differences—why won’t my brilliant daughter go into computer programming?—but we can preempt unreasonable expectations that would attach to germline engineering by regulating it. Putting the height example aside, expectations would attach to engineering a child not to have autism (assuming we can identify all genetic foundations of autism). If the resulting child is not autistic, but still manifests with related symptoms like impaired social interaction and communication skills, or with another condition that presents challenges, how will the manifestation of traits targeted for removal affect the parents and the parent-child relationship?
Flourishing might be defined for many as a combination of appreciating one’s present circumstances—being satisfied with the present—and setting and achieving attainable goals.16 For those who might not have the capacity to set goals, then flourishing might just be finding moments of satisfaction and joy. If we move into the realm of germline engineering, regulations and laws must first provide adequate protection against potential inheritable risks. Allowing modification of human embryos without sufficient risk-benefit analysis does not comport with any ethical or legal system. Regulations should also embody careful attention to how swiftly the baseline of health is allowed to shift, and the extent to which this shift prevents us from appreciating our present well-being—and accepting difference.
Endnotes
1. Alexandra A. Gason et al., Tay Sachs Disease Carrier Screening in Schools: Educational Alternatives and Cheekbrush Sampling, 7 Genetics in Med. 626 (2005).
2. Eric A. Millican et al., Genetic-Based Dosing in Orthopedic Patients Beginning Warfarin Therapy, 110 Blood 1511 (2007).
3. Jonathan L. Rees, The Genetics of Sun Sensitivity in Humans, 75 Am. J. Hum. Genetics 739 (2004).
4. Ellen J. Steinbart et al., Impact of DNA Testing for Early-Onset Familial Alzheimer Disease and Frontotemporal Dementia, 58 Archives Neurology 1828 (2001).
5. Arthur L. Caplan et al., No Time to Waste—the Ethical Challenges Created by CRISPR, 16 EMBO Rep. 1421 (2015).
6. Id.
7. Heidi Ledford, CRISPR, the Disruptor, 522 Nature 20 (2015), available at http://www.nature.com/news/crispr-the-disruptor-1.17673.
8. Jocelyn Kaiser, The Gene Editor CRISPR Won’t Fully Fix Sick People Anytime Soon. Here’s Why, Sci. (May 3, 2016), http://www.sciencemag.org/news/2016/05/gene-editor-crispr-won-t-fully-fix-sick- people-anytime-soon-here-s-why.
9. Chinese Scientists Defy Ethics, Double Down on Editing Human Embryos, Genetic Engineering & Biotechnology News (Apr. 13, 2016), http://www.genengnews.com/gen-news-highlights/chinese-scientists-defy-ethics-double-down-on-editing-human-embryos/81252608.
10. David Baltimore et al., A Prudent Path Forward for Genomic Engineering and Germline Gene Modification, 348 Sci. 36 (2015).
11. Id.
12. Stephen Holland, Bioethics: A Philosophical Introduction (2003).
13. Stephen B. Manuck & Jeanne M. McCaffery, Gene-Environment Interaction, 65 Ann. Rev. Psychol. 41 (2014).
14. Karamomoko Shirai et al., [Subjective Evaluation of Cochlear Implants through a Questionnaire Survey in 100 Long-Term Adult Cochlear Implant Users], 117 Nihon Jibiinkōka Gakkai Kaihō [Japan Otolaryngology Soc’y Bull.] 1329 (2014).
15. Gason et al., supra note 1.
16. Positive Psychology in Practice (P. Alex Linley & Stephen Joseph eds., 2004).