Base editors (BEs) were recently developed by combining different nucleoside deaminase family members with the CRISPR-Cas9 system and have been used for targeted C-to-T/A-to-G base editing in various cells or living organisms. Since BEs catalyze the deamination of cytidine or adenosine at target sites without generating double-strand breaks (DSBs) and that the majority of human pathogenic variants are single mutations, BEs were thought to be promising tools to correct these disease-related point mutations.

In an article published in Nature Biotechnology, Dr. YANG Li from CAS-MPG Partner Institute for Computational Biology, Shanghai Institute of Nutrition and Health of Chinese Academy of Sciences, together with researchers from ShanghaiTech University, commented three recent studies (ZUO et al., 2019, Science; JIN et al., 2019, Science; Kim et al., 2019, Nature Biotechnology) about accuracy of base editing.

These three studies and several others have showed that cytosine base editors, specifically a commonly-used cytosine base editor3 (BE3), but not adenine base editors, induce elevated levels of genome-wide off-target substitutions, therefore raising concerns about the safety of base editing on cytosines.

The detected off-target editing, according to the authors, may be caused directly by the rat APOBEC1 moiety of BE3.

They further hypothesized that the use of native or engineered APOBEC deaminases with relatively low DNA binding or catalytic activity may help to reduce unexpected off-target substitutions.

Meanwhile, they found that deamination activity-reduced BEs might lead to inefficient on-target editing, as high on-target activity is essential for broad applications of base editing, especially for therapeutic-related applications in somatic cells.

In this scenario, it will be more prudent to adopt strategies other than using activity-reduced deaminases when developing high-precision BEs going forward.