Efficient multitool/multiplex gene engineering with TALE-BE

Abstract

TALE base editors are a recent addition to the genome editing toolbox. These molecular tools are fusions of a transcription activator-like effector domain (TALE), split-DddA deaminase halves, and an uracil glycosylase inhibitor (UGI) that have the distinct ability to directly edit double strand DNA, converting a cytosine (C) to a thymine (T). To dissect the editing rules of TALE-BE, we combined the screening of dozens of TALE-BE targeting nuclear genomic loci with a medium/high throughput strategy based on precise knock-in of TALE-BE target site collections into the cell genome. This latter approach allowed us to gain in depth insight of the editing rules in cellulo, while excluding confounding factors such as epigenetic and microenvironmental differences among different genomic loci. Using the knowledge gained, we designed TALE-BE targeting CD52 and achieved very high frequency of gene knock-out (up to 80% of phenotypic CD52 knock out). We further demonstrated that TALE-BE generate only insignificant levels of Indels and byproducts. Finally, we combined two molecular tools, a TALE-BE and a TALEN, for multiplex genome engineering, generating high levels of double gene knock-out (∼75%) without creation of translocations between the two targeted sites.

Keywords: TALE; base editors; cell engineering; gene editing; t-cells.

FIGURE 1

(A). Schematic representation of a TALEN and a TALE-BE. (B). Average of the highest C-to-T conversion frequencies of 37 base editors (among all edited bases within the target spacer region) versus the frequencies of indels created by the matching TALEN (N = 2, independent T-cells donors). (C). Repartition of TC/GA in the spacer of the 37 TALE-BE. (D). C-to-T conversion frequencies of the target C (top) or G (bottom) at different positions within the 15 bp TALE-BE spacer (box represent 5%–95% percentile, whiskers represent min and max) with cartoon of DNA double helix drawn between top and bottom strands to aid visualization of edited positions within the spacer. Colors of cartoon are selected to match those of the box plot, with darker colors indicating more editing. (E). Average of the highest C-to-T conversion frequencies of 37 base editors versus indels frequencies generated within the spacer. (F). Editing purity (median) within the cell population. Left: C-to-T conversion within the C-to-A/G/T population. Right: C-to-T conversion within the C-to-A/G/T + Indels population. For all panels: N = 2, independent T-cells donors.

FIGURE 2

(A). Scheme of the strategy to generate artificial base editor target sites. In a first step a pool of ssODN encoding various base editor spacer sequences is inserted into TRAC locus. In a second step the TALE-BE is transfected. Two days post transfection the genomic DNA is collected, and the inserted sequence is analyzed by NGS. (B). Mean C-to-T conversion frequencies of the target Cs (top) or Gs (bottom) at different positions within the 15 bp TALE-BE spacer. (C). Schematic representation of the ssODN pool collection with spacer length ranging from 5 to 39 bp. (D). heatmap of C-to-T conversion when the TC was present on the top strand in function of the spacer length. (E). heatmap of C-to-T conversion when the TC was present on the bottom strand in function of the spacer length. For all panels: N = 2, independent T-cells donors.

FIGURE 3

(A). Schematic representation of target spacer sequence for TALE-BE targeting CD52. Top: TALE-BE (ex2 SA-1, two and 3) designed to edit CD52 exon2 splice acceptor site. The conserved G of the splice site, targeted by the TALE-BE) is depicted in red. Bottom: TALE-BE targeting CD52 signal peptide sequence in exon 2 (SP). TALE-BE spacer sequence (with targetable Cs or Gs numbered) and peptide sequence are depicted. (B). CD52 negative cell frequency in the TALE-BE treated (targeting CD52 exon2 splice acceptor site) or mock electroporated PBMC populations, 6 days post electroporation, measured by flow cytometry. (C). Editing (E) frequencies (C-to-T conversion) of the conserved G of the exon two acceptor splice site (Editing) and the indel (I) frequencies within the target locus, measured by NGS 6 days post transfection. (D). CD52 negative cell frequency in the TALE-BE treated (targeting CD52 signal sequence in exon2) or mock electroporated PBMC populations, 6 days post electroporation, measured by flow cytometry. (E). Editing frequencies (C-to-T conversion) at different position within the TALE-BE target spacer (CD52 signal sequence in exon2) and indel frequencies within the target locus, measured by NGS at Day 6 post transfection. (F). Frequencies of peptide species created by the TALE-BE targeting the CD52 signal sequence in exon 2. The first 16 most abundant species are presented. Mutation relative to the native signal peptide are in red. (N = 2, independent T-cells donors). (G). Editing purity within the cell population. Aggregate of the four TALE-BE targeting CD52. Left: C-to-T conversion within the C-to-A/G/T population. Right: C-to-T conversion within the C-to-A/G/T + Indels population. For all panels: N = 2, independent T-cells donors.