NEW YORK – Researchers at Massachusetts General Hospital led by Keith Joung have found that adenine base editors (ABEs) and cytosine base editors (CBEs) that exhibit RNA off-target editing activity can also self-edit their own transcripts, leading to heterogeneity in base-editor coding sequences.
ABEs and CBEs are known to efficiently and precisely edit point mutations in DNA with minimal off-target DNA editing. However, recent studies have also shown that despite their reputations for generating low levels of off-target DNA edits, base editors may create unexpected problems.
In March, an international team reported that the cytosine base editor 3 (BE3) induced SNVs with frequencies more than 20-fold higher than the spontaneous mutation rate. In April, Joung's team reported that both ABEs and CBEs could cause extensive transcriptome-wide off-target RNA editing in human cells. And in May, a Broad Institute team led by David Liu reported that they had found ABEs to generate low but detectable levels of widespread adenosine-to-inosine editing in cellular RNAs.
In a new study published yesterday in Nature Biotechnology, Joung and his colleagues described engineered ABE variants that they devised in order to reduce off-target RNA-editing activity. These SECURE-ABE variants have comparable on-target DNA-editing activity to ABEs and are also among the smallest Streptococcus pyogenes Cas9 base editors described to date, the researchers said.
The team also tested CBEs with cytidine deaminases other than APOBEC1, and found that while the human APOBEC3A-based CBE induced substantial editing of RNA bases, an enhanced APOBEC3A-based CBE, human activation-induced cytidine deaminase-based CBE, and the CBE Target-AID based on a cytidine deaminase from the sea lamprey Petromyzon marinus all induced less editing of RNA.
To engineer SECURE-ABE variants, the researchers used a protein truncation strategy to reduce the RNA-recognition capability of the optimized ABEmax fusion. They generated a smaller ABEmax variant lacking the Escherichia coli tRNA-specific adenosine deaminase (TadA) domain, which they called miniABEmax, and used RNA-seq to compare the transcriptome-wide off-target RNA-editing activity of miniABEmax to ABEmax in HEK293T cells. They assayed each of these editors and a nickase Cas9 control with three guide RNAs: two targeted to endogenous human gene sites and one to a site that does not occur in the human genome.
The researchers found that miniABEmax expression induced a total number of edited adenines that was not consistently lower than what they observed with ABEmax, but that the overall distribution of individual RNA adenine-editing efficiencies induced by miniABEmax were generally shifted to somewhat lower values.
Through further engineering efforts, the researchers generated 34 miniABEmax variants bearing various substitutions at 26 amino acid residue positions in E. coli TadA and screened each editor for on-target DNA-editing and off-target RNA-editing activities in HEK293T cells. They found that 23 of the 34 variants induced editing comparable to that observed with miniABEmax and ABEmax, and that 14 of the 34 variants showed reduced editing activity on at least three of the six RNA adenines that they examined as compared to miniABEmax.
Based on their DNA- and RNA-editing profiles, the investigators continued experimenting with two miniABEmax variants (K20A/R21A and V82G). To characterize their transcriptome-wide off-target RNA-editing profiles, the researchers performed RNA-seq with each of these variants and the HEK site 2, ABE site 16, and NT gRNAs. In contrast to what they observed with miniABEmax, the K20A/R21A and V82G variants both induced substantially reduced numbers of edited adenines relative to ABEmax. Further, the distribution of individual RNA-adenine-editing efficiencies for the two variants was shifted predominantly lower with both variants relative to ABEmax and miniABEmax.
To more fully characterize the on-target editing efficiencies of miniABEmax(K20A/R21A) and miniABEmax(V82G), the team then tested each variant in different sequence contexts with gRNAs for 22 genomic sites in HEK293T cells. MiniABEmax(K20A/R21A) and miniABEmax(V82G) retained efficient absolute on-target modification activity, but these efficiencies were typically reduced as compared to ABEmax.
The researchers' analysis of ABE activity with 22 gRNAs also identified a new and unexpected imprecise C-to-G base-editing activity within the editing windows of some DNA on-target sites. This C-to-G on-target DNA editing was observed with ABEmax and miniABEmax(V82G) using the HEK site 2, ABE site 7, and FANCF site 1 gRNAs. This unwanted editing was consistent across replicates and reached frequencies as high as 14.6 percent.
Having previously shown that off-target RNA editing occurs with a CBE harboring the rAPOBEC1 enzyme (BE3), the team wanted to determine whether CBEs harboring other cytidine deaminases such as human APOBEC3A (hA3A), enhanced A3A (eA3A; an engineered A3A with more precise and specific DNA-editing activity), human activation-induced cytidine deaminase (hAID), or the Petromyzon marinus cytidine deaminase CDA1 (pmCDA1) might also induce unwanted edits.
They found that hA3A-BE3, eA3A-BE3, and hAID-BE3 induced mean editing efficiencies of 91 percent, 82 percent, and 32 percent, respectively, at the RNF2 on-target site, and that Target-AID (with a pmCDA1 deaminase at its C-terminal end) showed a mean editing efficiency of 87.1 percent. RNA-seq experiments found that hA3A-BE3 induced tens of thousands of C-to-U edits distributed throughout the transcriptome. The eA3A-BE3 editor showed a dramatically reduced number of RNA edits relative to hA3A-BE3. And hAID-BE3 and Target-AID induced numbers of RNA C-to-U edits that were comparable to what was observed in the negative control.
Given their abilities to edit the endogenous human cell transcriptome, the investigators hypothesized that CBEs and ABEs might also self-edit their own transcripts, potentially generating sets of heterogeneous base-editor proteins. In order to assess this theory, they used an analysis pipeline to quantify self-editing events and observed C-to-U edits at 83 to 125 different C positions distributed throughout the BE3 transcript with standard expression and 149 to 177 different C positions distributed throughout the BE3 transcript with overexpression. Absolute numbers of missense mutations created by these edits ranged from 25 to 44 and from 55 to 64 among replicates with BE3 standard expression and overexpression, respectively.
Importantly, even when overexpressed, the two SECURE-BE3 variants the researchers engineered did not induce any detectable C-to-U edits in their own transcripts.
Similarly, ABEmax and miniABEmax both induced A-to-I changes at dozens of positions throughout their own transcripts with editing efficiencies ranging from 7 percent to 69.8 percent among replicates performed with three different gRNAs. Nearly all of the edits induced by the ABEs are expected to induce missense mutations, the researchers said.
"The work described here extends our understanding of the off-target RNA-editing activity of DNA base editors, expands the options available to minimize these unwanted effects, and further expands the toolbox of base editors that can be used without inducing high-level RNA editing," the authors concluded. "It will be interesting to directly compare all of these variants and perhaps to combine mutations from them to create base editors with even more optimized on-target DNA-, off-target DNA- and off-target RNA-editing profiles. Our description of self-editing by DNA base editors provides yet another strong motivation to use expression and/or delivery strategies that limit the duration of activity."