CRISPR ethics: moral considerations for applications of a powerful tool

Abstract

With the emergence of CRISPR technology, targeted editing of a wide variety of genomes is no longer an abstract hypothetical, but occurs regularly. As application areas of CRISPR are exceeding beyond research and biomedical therapies, new and existing ethical concerns abound throughout the global community about the appropriate scope of the systems’ use. Here we review fundamental ethical issues including the following: 1) the extent to which CRISPR use should be permitted; 2) access to CRISPR applications; 3) whether a regulatory framework(s) for clinical research involving human subjects might accommodate all types of human genome editing, including editing of the germline; and 4) whether international regulations governing inappropriate CRISPR use should be crafted and publicized. We conclude that moral decision making should evolve as CRISPR science advances and hold that it would be reasonable for national and supranational legislatures to consider evidence-based regulation of certain CRISPR applications for the betterment of human health and progress.

I. Introduction

The CRISPR (Clustered Regularly Interspaced Short Palindromic Repeats)-Cas9 (CRISPR-associated protein 9) system (“CRISPR” or “the system”) is the most versatile genomic engineering tool created in the history of molecular biology to date. This system’s ability to edit diverse genome types with unprecedented ease has caused considerable excitement and alarm throughout the international biomedical community.

CRISPR appears to offer considerable promise in a wide variety of disease contexts. For example, around the world at least 15 clinical trials— focused on multiple myeloma, esophageal, lung, prostate, and bladder cancer, solid tumors, melanoma, leukemia, human papilloma virus, HIV-1, gastrointestinal infection, β-thalassemia, sickle cell disease, and other diseases— involving CRISPR applications have been developed^1–3. Moreover, as of May, 2018, in China at least 86 individuals have had their genes altered as part of clinical trials⁴.

While significant public support exists for therapeutic applications⁵, ethical (moral) and safety concerns about certain areas of CRISPR applications, such as germline editing, are apparent around the world⁶. Notably, such discussions commenced during the Napa Valley meeting of 2015 when a leading group of CRISPR-Cas9 developers, scientists, and ethicists met to examine the biomedical, legal, and ethical aspects of CRISPR systems⁷. From this meeting, more extensive deliberations were solicited, and the United States (U.S.) National Academies of Sciences, Engineering, and Medicine (NASEM or “The Committee”) invited the Chinese Academy of Sciences and the United Kingdom’s (U.K) Royal Society to participate in the International Summit on Human Gene Editing⁸. The goal of this meeting was to examine when, where, and how the technology might be applied in humans. This discussion continued in February of 2017 when a multidisciplinary committee of the NASEM published a comprehensive report examining numerous aspects of human genome editing⁹.

To date, the NASEM report provides perhaps the most influential, extensive analysis examining wide-ranging concerns about human genome editing¹⁰. Importantly, the Committee favored somatic genome editing, but did not permit genomic modification for any kind of enhancement^9,11. Also, though impermissible at present, the Committee concluded that human heritable genome editing, the modification of the germline with the goal of creating a new person who could potentially transfer the genomic edit to future generations, would be permissible under certain conditions: “In light of the technical and social concerns involved… heritable genome-editing trials might be permitted, but only following much more research aimed at meeting existing risk/benefit standards for authorizing clinical trials and even then, only for compelling reasons and under strict oversight…”⁹ Although by law U.S. federal funding cannot be used to support research involving human embryos^12–14, the NASEM report suggests that when technical and safety risks are better understood, then clinical trials involving germline editing in human subjects might begin⁹.

In this review, we aim to summarize fundamental ethical concerns about CRISPR use in general, but the list is not exhaustive. First, we briefly review CRISPR systems and their applications in editing genomes and epigenomes. Second, we describe how complexities of CRISPR science affect those of CRISPR ethics and vice versa. Third, we assess several key ethical considerations. Notably, although some of the ethical concerns are specific to CRISPR technology, many, such as research on human embryos, have been debated long before CRISPR¹⁵. Moreover, since CRISPR is still a maturing technology, novel applications in the future may raise new ethical quandaries meriting further attention and dissection. Fourth, it is important to point out that, though morality and law often overlap, significant differences exist. Even though law may affect ethics and vice versa, this discussion centers mostly on ethics. Finally, while discussing these issues, we assume no position on any topic; our account is merely descriptive. Therefore, we make no attempt to settle the aforementioned issues.

II. CRISPR systems and their uses

Different CRISPR systems in genome editing

CRISPR is a natural bacterial defense system against invading viruses and nucleic acids. Over billions of years, multiple CRISPR-type immune systems have evolved. Naturally occurring CRISPR systems are typically classified by their repertoires of CRISPR-associated (cas) genes, which are often found adjacent to the CRISPR arrays^16,17. Although the characterization is yet to be finalized, two major classes of CRISPR-Cas adaptive immune systems have been identified in prokaryotes^18–20. This division is based on the organization of effector modules. Class 1 CRISPR-Cas systems employ multi-protein effector complexes and encompass 3 types (I, III, and IV). By contrast, Class 2 systems utilize single protein effectors and encompass 3 types (II, V, VI). Although various natural CRISPR-Cas systems have been repurposed for genome editing, due to its robust gene-editing efficiency and broader genome-targeting scope owing to its simple NGG PAM sequence requirement, the Cas9 from S. pyogenes (spCas9) is currently the most commonly used CRISPR-Cas9 protein. It is worth noting that multiple efforts are underway to discover novel Cas9 variants or re-engineer the existing Cas9 proteins, which will have less dependence on the stringent PAM-sequence requirement^21,22.

CRISPR goes beyond genome editing

The DNA-editing capacity of CRSPR-Cas9 is due to the ability of the WT Cas9 protein to cause double-stranded breaks at the target site that is determined by the custom-designed short guiding RNA (sgRNA)²³. The repair of DNA breaks frequently results in indels, due to the non-homologous end joining (NHEJ) repair mechanism. However, when a complementary template is available, homology-directed repair (HDR) machinery can use it and thereby achieve more precise gene editing. Notably, a single point mutation in either of the two catalytic domains of Cas9 results in a nickase Cas9 (nCas9), whereas mutations in both domains (D10A and H840A for spCas9) diminish Cas9’s catalytic activity, resulting in dead Cas9 (dCas9)²⁴. Interestingly, the application areas of modified Cas9 proteins are exceeding that of WT Cas9²⁵. Such uses are largely possible because the nCas9 or dCas9 can still be guided to the target sequence²⁶. Researchers employed these Cas9 variants for unique purposes. For example, tandem targeting of nCas9 has been utilized to improve targeting specificity^27,28. More recently, this enzyme has been used as the base platform for second generation genome-editing tools called “base editors”^29,30. Base-editing technology employs cytidine or adenine deaminase enzymes to achieve the programmable conversion of one base into another (C to T or A to G). Most importantly, the targeted base transition happens without DNA double-stranded breaks^29,31. We recently utilized this technology to edit the universal genetic code and introduced a “stop” codon in the genes³².

In addition to nCas9, researchers utilized the guidable capacity of dCas9 as a platform to recruit various effector proteins to a specific locus in living cells. Generally, these activities can be classified as epigenetic editing (to alter locus-specific epigenetic information), gene regulation (to turn the activities of single or multiple genes on or off), chromatin imaging (to label and monitor chromatin dynamics in living cells), and manipulation of chromatin topology (to alter 3D chromatin structure in the nuclear space)³³.

CRISPR research is progressing at a rapid pace. Recently, scientists have also uncovered new CRISPR-Cas systems (Cas13) that can target RNA instead of DNA^34,35. By enabling targeted RNA recognition and editing, these newer RNA-targeting CRISPR tools have their unique applications ranging from biomedical and biotechnological to the detection of nucleic acids^36,37. Although many ethical concerns are related to the catalytic activities of WT Cas9—partly because it permanently alters the genetic information—some of these activities of catalytically inactive dCas9, nickase-Cas9-based platforms, such as base editors and recently discovered RNA-targeting Cas proteins, may raise comparable moral issues depending on the duration of the exerted effect and the purpose of the experiments. Detailed discussion of such issues, however, is beyond the scope of this review.

III. CRISPR ethics and science: Uncomfortable bedfellows

Moral decisions, especially in biomedicine, are empirically informed and involve assessing potential risk-benefit ratios—attempting to maximize the latter while minimizing the former. To navigate ethical decision making, it is critical to consider the range of possible outcomes, the probabilities of each instantiating, and the possible justifications driving the results of any one. The ethical concerns about CRISPR genome engineering technology are largely due to at least three important reasons.

Firstly, there are concerns about the power and technical limitations of CRISPR technology. These include the possibilities of limited on-target editing efficiency^38,39, incomplete editing (mosaicism)^40,41, and inaccurate on- or off-target editing^42,43. These limitations have been reported in CRISPR experiments involving animals and human cell lines. However, the technology is evolving at an unprecedented pace. As more efficient and sensitive CRISPR tools are developed, many of these concerns may become obsolete. Yet for the sake of this review, we mention these limitations as one of the principal worries about CRISPR. The second concern is about the future of the modified organisms: whether they will be affected indefinitely and whether the edited genes will be transferred to future generations, potentially affecting them in unexpected ways. Combined with previously mentioned technical limitations and the complexities of biological systems, making precise predictions about the future of an edited organism and gauging potential risks and benefits might be difficult, if not impossible. Thus, uncertainty resulting from these factors hinders accurate risk/benefit analysis, complicating moral decision making.

Finally, the skeptical view is that even if the genome is edited as expected and the desired functional output is achieved at the time, the complex relationship between genetic information and biological phenotypes is not fully understood. Therefore, the biological consequence of editing a gene in germline and/or somatic cells may be unclear depending on the context. Many biological traits are determined by the complex regulatory actions of numerous genes. Hence, is it difficult, if not impossible, to “design” a biological phenotype at the whole-organism level. Across biological outcomes, whether in normal or in disease development, it is uncommon that a single gene is the only factor shaping a complex biological trait. Other genetic regulatory factors such as additional genes or distal regulatory elements (e.g., enhancer or repressor elements), as well as environmental and epigenetic factors, contribute to the emergence of a biological phenotype. To argue that modifying a gene changes a desired phenotype (under certain conditions) implies at least a reasonable understanding of other independent variables contributing to the phenotype’s instantiation. But this understanding is still far from complete in many normal and disease processes^44,45. Given the uncertainty regarding how gene expression and modification influence complex biological outcomes, it is difficult to appraise potential risk and benefit. This ambiguity creates a challenge on its own and is one of the sources obscuring efficient ethical deliberation and decision making.

Nevertheless, regulations governing cellular- and gene-therapy research may facilitate the safe development and oversight of some clinical trials involving CRISPR-based-editing applications. In the United States, for instance, cellular- and gene-therapy products, including many CRISPR applications, at this time are defined generally by the Food and Drug Administration (FDA) as biological products and are regulated by the FDA’s Center for Biologics Evaluation and Research (CBER)/Office of Cellular, Tissue, and Gene Therapies (OCTGT)^46–48. Even though the risks and benefits of many such therapies increasingly are better understood⁴⁹, questions regarding safety and efficacy remain. Thus, future advancements will continue to improve the benefits of this revolutionary technology, while minimizing the potential risks. Regardless of the uncertainty posed by novel CRISPR technologies and applications, in several locations around the world significant regulatory frameworks exist by which risks may be monitored and contained. However, wherever such infrastructure and oversight are lacking, safety and privacy risks—due to lack of regulatory oversight, monitoring, and the possibility of data breaches— might increase.

IV. Ethical concerns

To what extent should CRISPR experimentation be permitted in basic and pre-clinical biomedical research?

Though it is less than a decade old, CRISPR-Cas9 has demonstrated unprecedented potential to revolutionize innovation in basic science. From viruses and bacteria^50,51, to simple model organisms, such as Drosophila melanogaster (fruit fly)⁵², Anopheles gambiae (mosquito)⁵³, Saccharomyces cerevisiae (budding yeast)⁵⁴, Hydra magnipapillata (hydra)⁵⁵, Caenorhabditis elegans (round worm)⁵⁶, Danio rerio (zebra fish)⁵⁷, and Arabidopsis thaliana (rockcress)⁵⁸, to larger animals such as pigs⁵⁹, cattle⁶⁰, and monkeys⁶¹, and even human zygotes^62,63, CRISPR experimentation has led to novel, important findings. Such benefits include at least the following: increased overall efficiency in gene editing compared with previous genomic engineering techniques like transcription activator-like effector nucleases (TALENs) and zinc finger nucleases (ZNFs)⁶⁴; significant insights into the evolutionary transformation of fish fins into tetrapod limbs⁶⁵; investigation into new organisms⁶⁶; genetic and epigenetic screens^67,68; the creation of novel cell lines⁶⁹; high-throughput screens and libraries⁷⁰; the elucidation of novel genomic and epigenomic regulatory pathways^71,72; insights into the development of butterfly coloring and patterning⁷³; the functional characterization of key genes and molecular signaling pathways^74,75; and drug-targeting screens^76,77. Data from such experimentation provide essential clues and understanding that promote biomedical discovery, advancement, and the basis for potential medical benefits.

One of the major controversies about CRISPR technology emerges from its potential application in human embryos. This controversy is not about CRISPR itself, but instead is largely due to the lack of clarity about the status of the human embryo. Although some in the scientific community maintain that it is ethically impermissible to experiment on human embryos after 14 days^78,79, it is impossible for any one party—whether it is a government, laboratory, funding agency, panel of experts, court, religious organization, or other group—to decide the status of a human embryo^80,81 and whether and precisely when it has “personhood”⁸²: Is the entity merely a ball of cells whose status is like that of human skin, which sheds regularly without further ado? Or does the entity hold complete personhood status— with irreducible, inalienable moral rights and to whom we owe important directed duties? Or is the embryo’s status characterized optimally as something in between? And if so, which moral rights might this entity hold, and which duties might we owe to it? Despite this perplexing complexity, decisions one way or the other must be executed, because proceeding with research or failing to do so has consequences: Banning or significantly limiting research on human embryos stymies progress at least by making unavailable or curtailing an option to investigate the therapeutic potentials of stem cells and the biology of totipotent cells, which currently are not known to be present in any other viable human tissue sources. Totipotent cells can divide indefinitely and have the capacity to develop into all tissue types. Depending on how the status of the embryo is appraised, however, the ban also could save it from potentially unjust, lethal research-related harms. Even if the research is justified because of its potential benefit to the embryo itself and/or to others, the embryo as such cannot give informed consent at the time of the research, since the entity is not sufficiently developed. But from the research it could experience potentially lifealtering consequences—good or bad— that may extend throughout the lifespan and future generations.

By contrast, promoting such investigation may facilitate the development of novel in vitro fertilization (IVF) techniques and advances for conditions such as spinal cord injury⁸³, Parkinson’s disease⁸⁴, burns⁸⁵, cardiomyopathy⁸⁶, and other ailments that might be ameliorated by approaches involving regeneration. Taken together, countries must continue to decide as the science progresses whether and how to legalize experimentation on human embryos. Current positions across the globe vary widely—from outright banning of the research to illegalizing its federal funding only (while still allowing private funding for research and the research itself) to authorizing federal monies for experimentation^4,6,87,88.

To what extent should CRISPR use be permitted in translational and clinical medicine?

CRISPR is significantly benefitting and improving immunotherapy⁸⁹, organoid engineering and development⁹⁰, in vivo drug target identification⁹¹, machine learning and artificial intelligence⁹², and disease-gene modification in viable human embryos³². The system offers nearly boundless potential to promote progress in combating HIV⁹³, hemophilia⁹⁴, cancer⁹⁵, Duchenne muscular dystrophy⁹⁶, amyotrophic lateral sclerosis⁹⁷, sickle-cell anemia⁹⁸, cystic fibrosis⁹⁹, infertility¹⁰⁰, and any number of novel diseases. The potential both for gaining knowledge and for developing treatments in humans seems nearly endless.

Yet such knowledge and treatment acquisition are not without potential risk. With experimentation involving somatic cells, risk assessment seems at least comparable to that which arises in regularly practiced biomedical testing. The chief objective of phase-2 oncology trials, for example, is to evaluate the efficacy of a new drug or device¹⁰¹. Study participants may assume significant harms, including possibly irreversible side effects and death^102,103. In many countries, respect for autonomy permits assuming such risk with the requirement that informed consent occurs before enrollment in the research, regardless of whether participants are enrolling themselves or their dependents. If this risk is considered morally (and legally) permissible, then it would seem unjustified and unreasonable to not allow risk posed by CRISPR investigation. At the time of this writing, there is no empirical support to suggest that CRISPR experimentation would necessarily pose greater risk: the overall risk profile of CRISPR experimentation in human subjects remains unknown.

It could be argued, however, that heritable germline editing may present additional risk, because it involves not only the research participant, but also potentially his or her descendants. Of course, whether germline engineering technologies introduce risk beyond that which might be present in more common testing scenarios is an empirical matter. For instance, it is well-established that routinely used chemotherapies have mutagenic properties: alkylating agents, including cisplatin and cyclophosphamides, cause DNA adducts and crosslinks; antimetabolites, such as hydroxyurea, gemcitabine, and 5-fluorouracil, are nucleoside analogues and inhibit thymidine synthase; topoisomerases, such as etoposide, cause topoisomerase II inhibition, leading to double-stranded breaks in DNA; and anthracyclines, like doxorubicin, cause DNA intercalation¹⁰⁴. Therefore, significant exposure to any of these agents increases the probability of both incurring genetic mutations and passing on these unintended genomic alterations to future generations. Whether the risk level presented by such exposure is greater than, equal to, or less than that presented by CRISPR experimentation must be quantitatively determined by empirical evidence. It is also an empirical matter whether CRISPR introduces risk that is statistically significant beyond that which is incurred in the daily experience of a healthy individual with little-to-no exposure to mutagenic agents. Thus, to determine with confidence whether it is exceptionally risky to involve humans in CRISPR translational and clinical research, possible research-related risks must be compared with those in other potentially dangerous medical and every-day contexts. This is difficult, however, given that CRISPR technology is new and that robust, reliable data about CRISPR risk in human subjects are unavailable. Nevertheless, decision making about assuming risk in CRISPR studies and therapeutics should be considered according to legal infrastructure, national and possibly international regulatory agencies, and ultimately navigated by research participants and/or their legally authorized representative(s).

Important questions also arise about whether experiments involving heritable germline editing yield reliable, interpretable data. One objection is that such experiments are unlikely to be controlled and or predicted¹⁰⁵ because it could be impossible to analyze or understand the results from such experimentation until considerable time (decades or even generations) passes¹²². As previously noted, the central concern here is the uncertainty in the causal connection between gene expression/modification and the potential involvement of other factors shaping biological outcomes in the future.

Another risk, shared globally, is posed by the greater society. It is possible, for instance, that allowing CRISPR germline editing, even if only for medical purposes, might in some respect(s) lead to the return of eugenics, whose proponents believed that the human population can be improved by controlled breeding to increase the occurrence of desirable, heritable characteristics¹⁰⁶. Unfortunately, historically, this selective weeding of people with “bad” genes and breeding of those with “good” ones resulted in many atrocities, including the forced sterilization of individuals and the propagation of racially discriminatory policies—both of which were backed by state authorities and even educated elites in different societies. In the notorious case Buck v. Bell¹⁰⁷, for example, the United States Supreme Court (“the Court”) upheld a Virginia statute permitting the compulsory sterilization of individuals, such as Carrie Buck, who were considered “mentally unfit.” Buck was an economically disadvantaged woman who was labeled as “feebleminded” like her other family members of past generations. She was committed to the Virginia Colony for Epileptics and Feeble-Minded and was forcibly sterilized¹⁰⁸. Unfortunately, however, the evidence of the case strongly indicates that Buck, like the others in her family, was normal and that the high Court erred gravely¹⁰⁹. Its decision, authored by eugenics proponent Associate Justice Oliver Wendell Holmes, led to the sterilization of 50,000 Americans, set a precedent for the Nazi racial hygiene program, and is yet to be overturned¹²⁴. Hence, history reveals that egregious, systematic mistakes are always possible.

To what extent should CRISPR use be permitted for non-therapeutic purposes?

Important ethical questions also arise in non-therapeutic contexts such as enhancement of crops, livestock, gene drives, and human features¹¹⁰.

Certain areas of CRISPR applications, such as enhancement of crops and livestock, may have greater impact on society and humanity at large. In 2016 the United Nations Food and Agriculture Organization estimated that 795 million people in the world were undernourished¹¹¹. And according to the World Health Organization, 2 billion people are unable to obtain key nutrients like iron and vitamin A¹¹². Abundant evidence demonstrates that CRISPR-Cas systems could be used to improve nutrient content in foods^113–122. Why not decrease malnutrition by maximizing access to foods of higher quality? In principle, CRISPR has the potential to fortify foods efficiently for individuals who are suffering from a lack of basic nutrients. Promoting benefit in this way carries moral weight at least comparable to any other ethical concern raised herein, especially given the very large number of those who are nutrient-deprived. Nevertheless, with this benefit arise worries about “accessibility” to the product(s): Who will benefit from it? Will such food products be available to all who need it, and at what price?

Another CRISPR application that has unprecedented potential to directly benefit and save millions of lives is “gene drive technology”¹²³. In using gene drives, researchers employ CRISPR to speed up genetic recombination such that a “gene of interest” is rapidly distributed to the entire population much faster than in a typical Mendelian inheritance rate. Therefore, this application has the potential to edit the genome of an entire population or even the entire species. Using CRISPR-mediated gene drives, investigators have demonstrated that a gene allele providing a parasite-resistance phenotype in mosquitos could quickly spread through the population in a non-Mendelian fashion^124,125. This highly cost-effective technology has many potential benefits and applications for public health, species conservation, agriculture, and basic research¹²⁶: Gene drives may provide a fundamental tool to fight against deadly diseases such as malaria^127–129, dengue fever¹³⁰, Chagas and Lyme disease¹³¹, and schistosomiasis¹³². This application also may control and/or alter a wide variety of animals (e.g., rodents and bats), invasive plant pests, and reservoirs^135–134. Thus, the technology has unprecedented power that may save millions of lives each year. However, it is also important to consider the expected and unexpected risks of this technology: Once applied, it will eventually affect every individual of the entire species. Knowing this, researchers are developing and incorporating key safety “offswitch” features such as novel ways to (i) control, (ii) inhibit, and (iii) reverse/eliminate gene drive systems from the population in case of an unexpected or emergency event^135–138.

Furthermore, is it morally permissible to use CRISPR to enhance human features such as height, muscle mass, vision, or cognitive factors like learning aptitude and memory? It becomes particularly difficult to answer this question because deviation from a “norm” may become hallmark of human diseases for many of these features. The real challenge, then, is to define what is a “norm” and when a biological phenotype is a disease deserving treatment. “Medical necessity” often becomes ambiguous. Therefore, the boundary between “therapy” and “enhancement” can be murky³. For example, a gene-editing approach may allow a reduction in bad cholesterol, thus leading to a healthier life style. Whether this hypothetical scenario, which may benefit both the individual and society in the long run, can be classified as enhancement or a medical need is unclear.

Also, in certain regions of the world, such as the United States, positive liberties are granted morally and are constitutional rights. Will medical enhancement be considered as a form of free speech?¹³⁹ Alternatively, what authority should may stop individuals from exercising their right in this context?

Who should have access to CRISPR technology and/or its products?

Benefits from CRISPR innovation raise concerns and controversies about fairness and distributive justice across all layers of society. These matters are not specific to CRISPR technology, but may apply to all other technologies arising from academic research. Like many novel biomedical advancements, new CRISPR applications are expected to be profitable for patent holders. At least the initial prices of CRISPR-based products, such as gene therapy, are likely to be costly¹⁴⁰. To this end, an ethical question is whether the high price-tag will make the CRISPR-product available to only a special class of people across societies. Since much of the funding for CRISPR characterization and development was provided by grants from government funds and thus taxpayers’ money^141–152, it is ethically problematic to deny potentially life-saving benefits of the technology to the very individuals who funded much of the development in the first place. Moreover, even if it were affordable for some, there may be economic harms associated with high-price purchases. For instance, those needing the technology and its applications to maintain a reasonable quality of life, or even life itself, might be forced to make painful economic choices about whether to spend funds on treatment, food, or other essential living necessities. While this problem is not unique to genomic engineering advancements, allowing price gouging to continue unaddressed is unhelpful and potentially allows physical, psychological, and economic harms to continue. Encouraging the establishment of anti-pricegouging laws, where possible, could ameliorate some of these concerns¹⁵³.

As noted earlier, CRISPR may be used to fortify foods. Those residing in some of the most impoverished areas of the world are positioned to benefit the greatest from these products. How might such individuals gain access to these foods? How could companies benefit such that reaching out to these populations might be desirable and/or lucrative? Will such foods be able to safely and successfully surpass rogue governments?

Limiting human genome editing? Somatic vs. germline editing

As noted previously, an obvious application of CRISPR technology is cell and gene therapy. To date, gene therapy mostly involves the use of genome-engineering technologies to edit somatic cells to treat genetic diseases. As mentioned above, clinical trials involving CRISPR-based gene therapy are already under way. Although clinical gene and cell therapies have had major road blocks in the past, due to unanticipated injuries and death^154,155, significant safety improvements have been implemented over the last decade¹⁵⁶. With the advances of CRISPR technology and safer delivery approaches, therapeutic applications of gene therapy are on the rise¹⁵⁹. In the United States and elsewhere, federal regulations provide the needed legal and ethical frameworks, through the institutional review board system, to potentially minimize and manage potential risks^157–161

At present there is considerable excitement about such experimentation in the United States. In January of 2018, for example, the U.S. National Institutes of Health (NIH) launched the Somatic Cell Genome Editing program, seeking “to improve the delivery mechanisms for targeting gene editing tools in patients, develop new and improved genome editors, develop assays for testing the safety and efficacy of the genome editing tools in animal and human cells, and assemble a genome editing toolkit containing the resulting knowledge, methods, and tools to be shared with the scientific community”¹⁶². Heritable genome editing, by contrast, is perhaps the systems’ greatest discussed controversy. Recently, professional scientific and medical societies, industry organizations, and CRISPR pioneers together have released greater than 60 statements and reports about whether such editing in humans is morally permissible⁶. Most statements hold that heritable germline experimentation should be prohibited currently, although statements from the Netherlands¹⁶³, the United Kingdom¹⁶⁴, Spain¹⁶⁵, and the United States⁹ suggest that editing could be permissible if certain requirements were satisfied. The NASEM Committee’s report on germline editing, for example, specified that the following provisions must be met for human heritable germline research to commence: “the absence of reasonable alternatives; restriction to preventing a serious disease or condition; restriction to editing genes that have been convincingly demonstrated to cause or to strongly predispose to that disease or condition; restriction to converting such genes to versions that are prevalent in the population and are known to be associated with ordinary health with little or no evidence of adverse effects; the availability of credible preclinical and/or clinical data on risks and potential health benefits of the procedures; ongoing, rigorous oversight during clinical trials of the effects of the procedure on the health and safety of the research participants; comprehensive plans for long-term, multigenerational follow up that still respect personal autonomy; maximum transparency consistent with patient privacy; continued reassessment of both health and societal benefits and risks, with broad ongoing participation and input by the public; and reliable oversight mechanisms to prevent extension to uses other than preventing a serious disease or condition”⁹. Though fears about misuse in this context abound, it is also important to point out that there are reasonable arguments supporting heritable germline editing in research, such as the protection of defective embryos,¹⁶⁶ the elimination of certain disease that might be obliterated optimally early in embryonic development, and the exercise of free speech.¹³⁹

Should international regulations governing CRISPR use be crafted and promulgated?

Though ethics statements are important, by themselves, they provide little force. Typically, if ethics guidelines are infringed, the consequences suffered by the violator(s) are fairly minimal, compared to those arising when in violation of certain laws. Violations of ethics statements may lead to loss of funding, the retraction of publication, job loss, and mistrust among the scientific community. By contrast, punishments by law may lead to heavy fines and potentially incarceration. Given the significant potential promise, the dark history of eugenics, the potentially serious transgenerational risks, and the theoretical potential for misuse, it is reasonable for the global community to consider instantiating national and supranational regulations, if not revising older agreements such as the Geneva¹⁶⁷ and United Nations Conventions on Biological and Toxin Weapons¹⁶⁸ to reflect changes in genomic engineering technologies. While doing so likely will not eliminate all risks, it is arguably one of the few options available to reasonably control and/or minimize them.

V. Conclusions and future directions

CRISPR technology continues to mature, and existing systems are being engineered to contain innovative capabilities; excitingly new CRISPR systems with novel functions are still being discovered. The potential benefits of such revolutionary tools are endless. However like any powerful tool, there are also potential associated risks raising moral concerns. To make truly informed decisions about areas of ethical controversy, well-controlled, reproducible experimentation and clinical trials research are warranted. Currently, this is difficult because many international laws discourage or ban such research and/or inhibit its funding for certain types of investigation. Thus, widespread data about benefits and risks are unavailable. It is critical, however, for countries to examine their reasoning behind these prohibitions to ensure that they are not simply arising out of fear and without reasonable justification.

Going forward, many support establishing an organization that will decide how best to address the aforementioned ethical complexities. Recently, a group of European scientists founded the Association for Responsible Research and Innovation in Genome Editing (ARRIGE) to examine, and provide guidance about, the ethical use of genome editing^169,170. Furthermore, Janasoff and Hurlbut recently advocated for the development of an international, interdisciplinary “global observatory for gene editing”¹⁷¹. Briefly, they argued that deliberations about moral issues in gene editing should not be dominated by the scientific community, but instead should include a “network of scholars and organizations similar to those established for human rights and climate change. The network would be dedicated to gathering information from dispersed sources, bringing to the fore perspectives that are often overlooked, and promoting exchange across disciplinary and cultural divides”¹⁹⁸.

As the technology evolves, so will discussions about ethical and legal frameworks circumscribing its uses. The above-mentioned platforms present interesting ideas for furthering debates and potential resolutions. The research and ethical guidelines from national and international organizations, where diverse disciplines of society contribute, will be critical for federal funding agencies and IRBs to enforce and regulate, to minimize the potentials risks and maximize the potential benefits of CRISPR technology. However, it is likely that the enforcement of ethical research laws and guidelines ultimately will be assumed by legal systems, principal investigators, and institutional review boards.

Table 1.

Risk/Benefit Considerations in CRISPR Technology

	Benefit(s)	Risk(s)/Harm(s)
Basic and pre-clinical research	new model organisms and cell lines increased gene-editing efficiency high throughput screens novel drug targets access to totipotent cells identification of novel signaling, regulatory, and developmental pathways development of novel gene-editing approaches (base editing and RNA targeting) knowledge advancement	experimentation involving human embryos is controversial and illegal in some countries potential for privacy and confidentiality breaches
Translational and clinical medicine	immunotherapy organoids novel drug targets artificial intelligence modification of pathological genes novel therapeutics and fertility applications procreative liberty ability to “fix” single base changes knowledge advancement potential for equitable access	serious injury, disability, and/or death to research participant(s) and/or offspring blurry distinction between therapeutic and enhancement applications, leading to potential subtle or obvious exacerbation of inequalities misapplications eugenics potential for inequitable access and exacerbation of inequalities
Non-therapeuticapplications	enhancement to augment select faulty or human characteristics fortification of crops and livestock successful control of pests, invasive species, and reservoirs (gene drives) disease/infection control (e.g., malaria, dengue fever, Lyme and Chagas disease, schistosomiasis) ecosystem alteration to protect endangered species (gene drives) safety crop cultivation knowledge advancement	eugenics exacerbation of racism and inequality theoretical risk for damage to ecosystems theoretical risk of misuse
Access to CRISPRtechnology	inexpensive (technology itself) widely available profit, economic growth innovation	price gouging prohibitively expensive applications
Regulations for clinicalresearch involvinghuman subjects	established framework in some countries to manage research risk legal mechanisms for redress already exist, depending on location	lack of appropriate supervisory infrastructure, oversight, and/or regulatory framework in many nations unclear how to supervise the research even in some countries with regulatory oversight over regulation might hinder progress
National andinternationalregulations, law, andpolicy	prevention against misuses of technology safeguard against potential unethical conditions	potential to encroach on societal autonomy limit discovery and progress difficult enforcement lack of uniformity