Editors' ChoiceGenome Editing

After the CRISPR-Cas9 cut: Predicting generated mutations

See allHide authors and affiliations

Science Translational Medicine  19 Dec 2018:
Vol. 10, Issue 472, eaaw0525
DOI: 10.1126/scitranslmed.aaw0525


Analysis of synthetic guide RNA targets yields prediction algorithms for expected mutations following genome editing.

The development of clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein 9 (CRISPR/Cas9)–based methods greatly increased the precision and efficiency of gene editing. Gene editing involves guide RNA (gRNA) targeting of specific sequences, Cas-9 DNA break, and DNA repair commonly resulting in mutations. Predicting the possible resulting mutations could allow the design of safer gene therapies based on use of CRISPR/Cas9. Now, Allen and colleagues developed an elegant, high-throughput system, able to predict repair-induced mutagenesis by measuring edits that occurred in more than 40,000 gRNAs on synthetic targets.

In their screening system, both gRNA and its target (including flanking sequences) were cloned together into standardized vectors and lentivirally incorporated into cells; the resulting mutations were then analyzed. Because this system allows efficient insertion of different gRNAs and flanking sequences, with easy implementation in multiple cell types, they studied mutational results of thousands of gRNA-target conditions. Their results show that mutational profiles were reproducible and specific to gRNA-targets. In K562 chronic myelogenous leukemia cells, single nucleotide insertions and deletions were most common, with longer deletions occurring more than rare long insertions. They also noted that the mutations varied with cell line and Cas9-associated effector proteins.

Most importantly, authors identified local sequence properties guiding specific mutations. For example, the system predicted that dinucleotide repeats at cut sites resulted in mutagenesis contracting the repeated sequence. Experimental results are in agreement with this finding; indeed, preclinical studies in animal models using Cas9-mediated treatment for treating diseases defined by repeats, such as Fragile X syndrome, have previously shown similar outcomes. With predictive data, authors generated and tested an accurate prediction tool, now available at their website. Future studies may require evaluation of gRNA-target flanking sequences larger than those studied here, as well as extensive validation in nonsynthetic targets. However, a predictive tool to understand the mutation and resulting phenotype after a CRISPR-Cas9 cut is critical to safer and more accurate genome editing treatments.

Highlighted Article

View Abstract

Navigate This Article