TadA orthologs enable both cytosine and adenine editing of base editors

Abstract

Cytidine and adenosine deaminases are required for cytosine and adenine editing of base editors respectively, and no single deaminase could enable concurrent and comparable cytosine and adenine editing. Additionally, distinct properties of cytidine and adenosine deaminases lead to various types of off-target effects, including Cas9-indendepent DNA off-target effects for cytosine base editors (CBEs) and RNA off-target effects particularly severe for adenine base editors (ABEs). Here we demonstrate that 25 TadA orthologs could be engineered to generate functional ABEs, CBEs or ACBEs via single or double mutations, which display minimized Cas9-independent DNA off-target effects and genotoxicity, with orthologs B5ZCW4, Q57LE3, E8WVH3, Q13XZ4 and B3PCY2 as promising candidates for further engineering. Furthermore, RNA off-target effects of TadA ortholog-derived base editors could be further reduced or even eliminated by additional single mutation. Taken together, our work expands the base editing toolkits, and also provides important clues for the potential evolutionary process of deaminases.

Fig. 1. Editing capacity of ecTadA(VN)-derived ABEs.

a, b Illustration of ecTadA(VN) and –derived ABE structures generated via internal insertion strategy. c, d The editing frequencies are shown in heatmap format, with adenine editing efficiencies (A-to-G editing) shown in blue and cytosine editing efficiencies (C-to-T editing) in pink gradient color. Engineered ABEs were generated in combination with internal insertion with ecTadA(VN) (c) or NLS-ecTadA(VN)-linker (NL-ecTadA(VN)) (d). Data shown here represent means of results from n = 3 biologically independent experiments. NC negative control. VN, amino acid substitutions of A106V&D108N. Base editors generated by different fusion strategies were defined as X-deaminase, with X representing fusion sites, such as N for N-terminal and 535 for internal sites after the 535th amino acid. Source data are provided as a Source Data file.

Fig. 2. Engineering of TadA orthologs for functional ABEs, CBEs, and ACBEs.

a Phylogenetic analysis of TadA orthologs with a maximum-likelihood tree and six TadA groups were defined. The evolutionary distance scale of 0.5 was shown. b Editing signatures of selected TadA ortholog-derived base editors at sgRNA-1, sgRNA-18, sgRNA-PT14, sgRNA-HPE6, and sgRNA-S5 were shown in heatmap format, with adenine editing efficiencies (A-to-G editing) shown in blue and cytosine editing efficiencies (C-to-T editing) in pink gradient color. Data shown here represent means of results from n = 2 biologically independent experiments. c Phylogenetic tree showing the relationship between 25 functional TadA orthologs. Protein lengths and substrate specificities of derived base editors were presented simultaneously. NC negative control. VN, amino acid substitutions corresponding to A106V&D108N of ecTadA. Source data are provided as a Source Data file.

Fig. 3. Off-target effects, genotoxicity, and editing scopes of representative TadA ortholog-derived base editors.

a, b Box plots showing the number of RNA A-to-I (a) and C-to-U edits (b) induced by representative base editors or nCas9 control. n = 4 biologically independent experiments. Box plots here are defined by whiskers in terms of minima and maxima, and the center and bounds of the box by quartiles (Q1–Q3). c Quantification of γH2AX signaling in HEK293T cells transfected with representative base editors. The percentage of γH2AX positive population within GFP positive cells is shown. Data shown here represent means of results from biologically independent experiments. Data were presented as mean values ± SEM. Exact P values were as follows, P = 0.0037/TadA-8e, 0.01/F2G308(VN), 0.0084/Q57LE3(VN), 0.018/Q99W51(VN), and 0.0081/ecTadA(VN), respectively. d–k Editing scopes of representative TadA ortholog-derived base editors across 12 endogenous sites, including 1249-NL-ecTadA7.10 (d), 1249-NL-ecTadA(VN) (e) as control, 1249-NL-B5ZCW4(VN) (f), 1249-NL-Q57LE3(VN) (g), 1249-NL-Q99W51(VN) (h), 1249-NL-E8WVH3(VN) (i), 1249-NL-Q13XZ4(VN) (j), and 1249-NL-B3PCY2(VN) (k). * represents P < 0.05, ** represents P < 0.01 with two-tailed unpaired t-test. Adenine editing (A editing) was shown in red lines and cytosine editing (C editing) was shown in blue lines. Data were presented as mean values ± SEM. Source data are provided as a Source Data file.

Fig. 4. Minimization of RNA off-target effects for TadA ortholog-derived base editors.

a, b Deaminase mutagenesis screening to minimize RNA off-target edits of representative TadA ortholog-derived base editors. Heatmap format showing the on-target editing activity at sgRNA-1, with adenine editing activity in blue and cytosine editing activity in pink (a) and histogram showing RNA A-to-I conversion frequency at a specific site within mRNA transcript (b). NC, negative control. VN, FA, Rdel, VG, and VW, amino acid substitutions corresponding to A106V&D108N, F148A, R153 deletion, or V82G/W of ecTadA, respectively. RNA A-to-I conversion frequency of base editors containing Q13XZ4, B3PCY2, E8WVH3, Q57LE3, and Q9951 were depicted in orange, pink, blue, purple, and gray, respectively. Data shown here represent means of results from n = 3 biologically independent experiments. Data were presented as mean values ± SEM. c, d Box plots showing the number of RNA A-to-I (c) and C-to-U edits (d) induced by engineered TadA ortholog-derived base editors or nCas9 control. Box plots here are defined by whiskers in terms of minima and maxima, and the center and bounds of the box by quartiles (Q1–Q3). n = 4 biologically independent experiments. ** represents P < 0.01 and *** represents P < 0.001 with two-tailed unpaired t-test. Exact P values were as follows, P = 0.000241, 0.002087, 4.24e-05, and 6.33e-06, respectively. Source data are provided as a Source Data file.

Fig. 5. Simplified phylogenetic tree of tRNA-editing deaminases and AID/APOBEC deaminase family.

Substrate specificity of different deaminases was displayed in a phylogenetic tree, with TadAs showing all different substrate specificities of engineered orthologs in this study. DNA-specific C-to-U deamination was highlighted in orange, while both DNA and RNA-targetable C-to-U deamination was highlighted in yellow. Additionally, the shared substrate specificity of different deaminases was highlighted in the right panel, with A-to-I in RNA in green, C-to-U in DNA in blue, and C-to-U in RNA in purple. The red dash arrow indicated that AID/APOBEC might be derived from tRNA-editing deaminases.