Date: 27th July 2020
Base editors (BEs) are potent tools for precise genome editing, and can be used to correct single disease-causing mutations. Cytosine base editors (CBEs) enable efficient cytidine-to-thymidine (C-to-T) substitutions at targeted loci, and do so without creating a double-stranded break. However, they can lack precision and ‘bystander’ Cs lying adjacent to the ‘target’ C can also edited. Now scientists have engineered new CBEs that can precisely modify a single targeted C, whilst minimising the editing of bystanders, increasing the accuracy of base edits in disease sequence models up to 6,000-fold compared with current base editors.
Base editors comprise deaminases fused to either Cas9 nickase or catalytically-dead Cas9. CBEs are important genetic tools and could potentially correct more than 5,000 pathogenic single-nucleotide polymorphisms (SNPs) associated with human-inherited diseases. However, nearly 38% of the SNPs that are caused by T-to-C disease point mutations lie in the context of CC, therefore there is a need to develop new CBEs that can accurately edit just the target nucleotide.
Now, a team of scientists, led by Zue Gao and Erwei Zuo, from Rice University, US, the University of Chinese Academy of Sciences and the Chinese Academy of Agricultural Sciences, have identified a human deaminase as a candidate for developing sequence-specific BEs in mulitple C contexts.
By building on previous work demonstrating that the cytodine deaminase APOBEC3G (A3G) exhibited preferential deamination of the third C in the 5′-CCC-3′ motif, the team here sort out to determine whether this motif preference could be preserved in a CBE. By replacing the deaminase in a popular CBE, called BE4max, with the new A3G – the team created A3G-BE.
Initial tests in human cells lines at one loci showed that the A3G-BE edited up 42% of the cognate Cs and only 1-3% of bystander Cs. In contrast, BE4max edited up to 62% of both cognate and bystander Cs without obvious selectivity. Further experiments at nine different loci containing the dinucleotide CC motif determined that A3G-BE could precisely edit the second C in the sequence context of 5′-CC-3′ dinucleotides.
However, at several of the loci A3G-BE showed relatively low base editing efficiencies, so seeking to enhance editing they sought to improve catalytic activity, solubility and ssDNA binding. To achieve this team engineered mutations into key functional residues of A3G-BE, creating a series of variants. By screening the variants at one of the low efficiency sites they were able to identify 2 variants with enhanced editing efficiency but which retained sequence specificity.
Armed with these newly, engineered precise editors the team wanted to test the A3G-BEs in a disease-relevant context. To achieve this they sought to generate SNPs of reported human pathogenic diseases in non-disease cell lines in which the wild type sequences lie within the preferential 5′-CC-3′ motif. The team chose to create models for cystic fibrosis, hypertonic myopathy, and transthyretin amyloidosis. While all A3G-BEs showed significant success at precisely editing the targets over BE4max, the cystic fibrosis model performed particularly well – with all of the A3G-BEs inducing more than 50% of perfectly modified alleles, while BE4max averaged just 0.6%.
These data suggested that A3G-BEs could be used to accurately model diseases however, could they be used as potential therapeutic agents? To answer this the team used the CBE variants to correct 3 disease model cell lines harbouring the disease-relevant sequence for hereditary pyropoikilocytosis, cystic fibrosis, and holocarboxylase synthetase deficiency. The A3G-BEs had higher targeting precision than BE4max for reversing pathogenic SNPs for all three diseases and in the case of holocarboxylase synthetase deficiency, one A3G-BEs variant perfectly corrected only the target C nucleotide in more than 50% of the sequences, with up to a 6,496-fold higher correction frequencies than BE4max.
Off-targeting edits are a concern for any potential therapeutic tool, and whole-genome off-target characterisation for the most active A3G-BE variant showed that it induced minimum DNA off-target SNVs across the genome while maintaining highly efficient and selective editing at the on-target position.
Base editors are emerging as a powerful tool for genome editing. However, it is imperative for translation into the clinic that the use of these tools enables precise and correct edits together with the minimum off-target edits. The work here demonstrates the next step towards enhancing CBEs to recognise a specific CC motif, and permitting accurate and predictable edits of the target nucleotide whilst eliminating bystander editing.
We are currently seeing huge advances and investment into base editing technology. Whilst they are yet to be tested in clinical trials, that day inevitably marches ever closer. Verve Therapeutics announced earlier this month preclinical proof-of-concept data in non-human primates through the successful use of base editing as a new treatment approach for coronary heart disease. They have also been used in vivo to repair a single nucleotide mutation in TMC1 deaf mice which was partially able to restore hearing. It has also been demonstrated that multiplexed precise base editing works efficiently in primates, enabling the editing of up to three target sites simultaneously. Whilst safety and efficiencies are top of the list of concerns for any new technologies entering the clinic, a recent machine learning model, BE-Hive, has been developed which accurately predicts base editing outcomes and determine which BE is ‘best-in-class’.
This really is an exciting and important time for base editors and the work described here will strengthen their applications and drive towards the clinic. Furthermore, with the team here identifying 540 human pathogenic SNPs that could be precisely correct by A3G-BEs it is hoped these CBEs will make a significant contribution towards treating genetic disease.
For more information please see the press release from Rice University
Lee, S., N. Ding, Y. Sun, T. Yuan, J. Li, Q. Yuan, L. Liu, J. Yang, Q. Wang, A. B. Kolomeisky, I. B. Hilton, E. Zuo and X. Gao (2020). “Single C-to-T substitution using engineered APOBEC3G-nCas9 base editors with minimum genome- and transcriptome-wide off-target effects.” Science Advances 6(29): eaba1773.