are turned on and off, including the role of the epigenome,11 continues to be actively explored. Controlled gene expression and epigenetic alterations influence how tissues develop and differentiate and have clinical ramifications in such areas as cancer and embryonic development. Tools that enable investigators to alter DNA sequences in order to understand or improve their function are not new. Recent years, however, have seen the development of a suite of genome-editing tools that allow for easier, better controlled, and more accurate changes to DNA inside cells. These tools are based on exogenous enzymes that cut DNA at specific locations, combined with endogenous processes that repair the broken DNA, thereby enabling letters of the genetic code to be added, modified, or deleted. The speed with which this technology has been adopted in research laboratories and further adapted to tackle additional scientific challenges is a reflection of how powerful a technique the editing of genes and genomes will be for the scientific and clinical communities. The earliest applications of nuclease-based genome-editing methods employed targeted recognition of specific DNA sequences by proteins: homing nucleases (also known as meganucleases), ZFNs, and TALENs. However, the recent development of RNA-based targeting has greatly simplified the process of genome editing. The first publications on the subject, in 2012-2013, explained how the CRISPR/Cas9 system, derived from a natural bacterial defense mechanism against infecting viruses, can be harnessed to make controlled genetic changes in any DNA, including that of human cells (Cho et al., 2013; Cong et al., 2013; Jinek et al., 2012, 2013; Mali et al., 2013). This was a game-changing advance. These methods have rapidly been adopted by scientists worldwide and have greatly accelerated fundamental research that has included altering cells in the laboratory to study the functions of particular genes, developing models for studies of human diseases using stem cells or laboratory animals, creating modified plants and animals to improve food production, and developing therapeutic uses in humans. Genome editing has rapidly become an invaluable core technology in research laboratories and biotechnology companies, and is already moving into clinical trials (e.g., Cyranoski, 2016; Reardon, 2016; Urnov et al., 2010). 11The term “epigenome” refers to a set of chemical modifications to the DNA of the genome and to proteins and RNAs that bind to DNA in the chromosomes to affect whether and how genes are expressed. Copyright © National Academy of Sciences. All rights reserved. Human Genome Editing: Science, Ethics, and Governance 18 HUMAN GENOME EDITING PREPUBLICATION COPY—SUBJECT TO FURTHER EDITORIAL REVISION The Issues Individual-Level Concerns As with other types of medical interventions, whether genome editing can be used in patients will depend largely on understanding the safety and efficacy of the treatment and evaluating whether the anticipated benefits are reasonable with respect to the risk of adverse effects. Treatments based on genome editing are intended to make controlled modifications to specific portions of the DNA that affect the functions of their target(s) while avoiding changes to other portions whose alteration is not desired. The latter alterations, referred to as off-target events, could have consequences, many unnoticeable but others damaging, depending on their location and their effects. In general, human genome editing raises questions common to the process of researching and developing new treatments: which conditions or diseases are most suitable to address with these technologies, how to identify and evaluate off-target events and other potential side effects, and which patients are most appropriate for studies. As described in this report, regulatory systems for addressing the individual-level concerns associated with genome editing already exist in the United States and many other countries, but can be improved. Societal-Level Concerns The use of genome editing also has significant social dimensions that vary depending on the proposed application. The use of a genome-editing treatment whose effects are nonheritable and are restricted to an individual patient may not differ greatly from the use of a traditional drug or medical device. By contrast, making changes that may be inherited by future generations raises questions about the extent to which the long-term effects of proposed edits can be predicted and whether it is appropriate for humans to purposely alter any aspect of their genetic future (Frankel and Chapman, 2000; Juengst, 1991; Parens, 1995). In addition, identifying the increased range of applications made possible by genome editing may be yet another challenge to conventional conceptions of what constitutes a disease or disability. Societal-level concerns are particularly acute with respect to genome-editing interventions aimed at enhancing human capabilities. Such applications also raise questions about how to define and promote fairness and equity (President’s Council on Bioethics, 2003). Moreover, as with other genetic technologies, such genome-editing applications may raise concerns about coercive and abusive eugenics programs of the past, which were based on faulty