Therapeutic Germ Line Alteration: Has CRISPR/Cas9 Technology Forced the Question?

Public acceptance of human gene therapy has rested on the assurances that therapeutic genetic alteration in humans would (1) be restricted to therapy for disease rather than for enhancement, and (2) be restricted to somatic cells, incapable of transmitting genetic alterations to future generations of humans. In situations where systemic gene therapy protocols were thought to present a very small (but finite) risk of inadvertent germ line transmission, gene therapy trials have been allowed to proceed.^1,2 This decision has been at least partially based on the very low efficiency of germ line alteration with viral vectors then in clinical use. However, the recent publication by Liang et al. demonstrates that high frequency gene editing is feasible in human embryos using a CRISPR/Cas9-mediated approach.³ The actual performance and publication of these experiments has itself created substantial controversy. Whether one agrees with the performance of these experiments or not, they have clearly forced researchers, clinicians, and ethicists to face the issue of therapeutic germ line alteration with renewed energy.

The 2000 American Association for the Advancement of Science (AAAS) report by Frankel and Chapman articulated the issues that remain to be resolved with therapeutic human germ line alteration, which included a clear exclusion of genetic enhancements, a high standard for safety and efficacy of the methodology (so as to avoid creating new gene defects), and a need for extensive societal discussion of the moral acceptability of changing the inherited human genome. They recognized that such germ line manipulations “might change attitudes toward the human person, the nature of human reproduction, and parent-child relationships.”⁴ The specifics of the report by Liang et al. indicate that their approach is clearly intended as a therapy for β-thalassemia, an important genetic disease worldwide.³ The report also indicates major safety and efficacy limitations with the current generation of CRISPR/Cas9, despite the high frequency of editing events at the β-globin (HBB) locus (approximately 52% hit rate in HBB, 14.3% repaired with the intended ssDNA repair template). Unexpectedly, they also observed that the endogenous δ-globin (HBD) gene served as a competing repair template, such that in 25% of repair events, the endogenous HBD gene was used for template repair instead of the intended ssDNA repair template. Investigators also observed off-target alterations at other genomic sites frequently in the tripronuclear (TPN) human zygotes, a rate consistent with observations in cancer cells.³ This frequency of off-target hits and of use of unintended endogenous DNA repair templates raises the strong possibility that the current version of CRISPR/Cas9 technology could cause unintended genetic defects in human embryos, not all of which might be readily detectable prior to implantation. These findings yield the clear conclusion that the safety and efficacy of the current technology is not sufficient to warrant therapeutic intervention with viable human embryos.

Certain more substantive questions remain unanswered. If (and more likely when) modifications of CRISPR/Cas9 overcome its current limitations, would society find it acceptable to treat a genetic disease in a manner that could be inherited? The answer from the NIH-Recombinant DNA Advisory Committee (RAC) remains “no,” as indicated by the November 2013 version of the NIH Guidelines for Research Involving Recombinant or Synthetic Nucleic Acid Molecules: “RAC will not at present entertain proposals for germ line alterations but will consider proposals involving somatic cell gene transfer.”⁵ However, it may be necessary to more explicitly clarify guidelines on performing gene editing experiments in nonviable and previable human embryos. Furthermore, a more fully elaborated rationale may be required for allowing genetic therapy of disease in a single generation while prohibiting it in the multi-generation context. These larger questions can no longer be considered theoretical.