Precise in vivo functional analysis of DNA variants with base editing using ACEofBASEs target prediction

Abstract

Single nucleotide variants (SNVs) are prevalent genetic factors shaping individual trait profiles and disease susceptibility. The recent development and optimizations of base editors, rubber and pencil genome editing tools now promise to enable direct functional assessment of SNVs in model organisms. However, the lack of bioinformatic tools aiding target prediction limits the application of base editing in vivo. Here, we provide a framework for adenine and cytosine base editing in medaka (Oryzias latipes) and zebrafish (Danio rerio), ideal for scalable validation studies. We developed an online base editing tool ACEofBASEs (a careful evaluation of base-edits), to facilitate decision-making by streamlining sgRNA design and performing off-target evaluation. We used state-of-the-art adenine (ABE) and cytosine base editors (CBE) in medaka and zebrafish to edit eye pigmentation genes and transgenic GFP function with high efficiencies. Base editing in the genes encoding troponin T and the potassium channel ERG faithfully recreated known cardiac phenotypes. Deep-sequencing of alleles revealed the abundance of intended edits in comparison to low levels of insertion or deletion (indel) events for ABE8e and evoBE4max. We finally validated missense mutations in novel candidate genes of congenital heart disease (CHD) dapk3, ube2b, usp44, and ptpn11 in F0 and F1 for a subset of these target genes with genotype-phenotype correlation. This base editing framework applies to a wide range of SNV-susceptible traits accessible in fish, facilitating straight-forward candidate validation and prioritization for detailed mechanistic downstream studies.

The inceptionist’s guide to base editing – de novo PAM generation to reach initially inaccessible target-sites

“…S3). Injection of the kcnh6a-inception mix containing the adjusted inception guide RNA revealed 27 % of heart phenotypes (Table S2) comprising 2:1 atrioventricular block and reduced ventricular contractility similar to earlier reports (Cornean et al, 2022) (Fig. 2H, Movie S1).…”

“…To address the efficiency of a knock-out via inception, we targeted the well described oculocutaneous albinism 2 ( oca2 ) gene responsible for the pigmentation of the retinal pigmented epithelium (RPE) in the Japanese rice fish medaka ( Oryzias latipes ) (Cornean et al, 2022, Lischik et al, 2019). The loss of pigmentation depends on bi-allelic editing of oca2 which we use as proxy to determine the knock-out efficiency via an established analysis pipeline (Thumberger et al, 2022).…”

Boosting targeted genome editing using the hei-tag

“…To investigate if the addition of the hei-tag provides an easy and straightforward alternative route for increasing the activity of a nuclear protein of interest, we selected a C-to-T base editor version with intermediate efficiency (BE4-Gam Komor et al, 2017 ) to introduce non-sense or severe miss-sense mutations into the pigmentation gene oca2 . We employed our tool ACEofBASEs ( Cornean et al, 2022 ) to design and evaluate sgRNA target sites that introduce non-synonymous codon mutations and/or pre-mature STOP codons upon editing of the respective open reading frame (ORF). We compared three different sgRNAs ( OlOca2 T1 , T3, and T4 ) employing the original BE4-Gam and the hei-tag fused variant (heiBE4-Gam).…”

“…Base editors are composed of a modified Cas9 that only nicks one DNA strand and does not introduce a double-strand break (Cas9 nickase or Cas9n) and a nucleotide deaminase for precisely targeted nucleotide editing ( Anzalone et al, 2020 ). To increase the efficiency of base editors, several iterative rounds of optimization of the employed deaminases and linkers have been undertaken, yielding optimal performance with the newest variants ( Carrington et al, 2020 ; Cornean et al, 2022 ; Rosello et al, 2021 ; Zhao et al, 2020 ). To investigate if the addition of the hei-tag provides an easy and straightforward alternative route for increasing the activity of a nuclear protein of interest, we selected a C-to-T base editor version with intermediate efficiency (BE4-Gam Komor et al, 2017 ) to introduce non-sense or severe miss-sense mutations into the pigmentation gene oca2 .…”

“…All sgRNAs for medaka ( OlOca2, Rx2, Rx3, Cryaa ) and zebrafish ( DrOca2 ) were designed using the CCTop target predictor with standard parameters ( Stemmer et al, 2015 ). The sgRNAs used for base editing ( OlOca2 T1, T3, T4 ) were designed or evaluated using ACEofBASEs ( Cornean et al, 2022 ) and selected for introducing a pre-mature STOP codon. The following target sites were used [PAM in brackets]: OlOca2 T1 ( GAAACCCAGGTGGCCATTGC [AGG]), OlOca2 T2 ( TTGCAGGAATCATTCTGTGT [GGG]), OlOca2 T3 ( GATCCAAGTGGAGCAGACTG [AGG]), OlOca2 T4 ( CACAATCCAGGCCTTCCTGC [AGG]) DrOca2 T1 ( GTACAGCGACTGGTTAGTCA [TGG]), DrOca2 T2 ( TAAGCACGTAGACTCCTGCC [AGG]), Rx2 ( GCATTTGTCAATGGATACCC [TGG]), Cryaa ( GGGAGAAGTGCTTGACATCC [AGG]), Rx3 ( AGCAGAGCGCGCAAAGAACC [AGG]).…”