Page 59 - GTM-2-1

Global Translational Medicine ABE gene therapy for CVDs

off-target site (Figure 3). An important form of unwanted off–target effect in a Cas9/sgRNA-independent manner
editing on the target site is the indels. Because Cas9 nickases (Figure 3D) [15,67] . This type of editing most likely happens
are used in ABE, its likelihood to introduce indels is very in genomic loci that naturally unwind and expose single-
low. Among the key studies of ABE therapy for CVDs, the strand DNA to the freely available TadA in the nuclei, such
indel rate has be reported as 0.2% or lower [30,46,55] . as during DNA replication and gene transcription.

As mentioned previously, the more problematic form Because off-target effects are induced by complicated
of unwanted editing on the target sequence involves the mechanisms, the prediction and identification of off-target
bystander effect (Figure 3A). The three hallmark studies sites by a given ABE reaction are challenging (Table 2). The
using ABE for CVD therapy all cleverly chose the target most convenient and fast methods to nominate off-target
sites with only one adenosine in the editing window, sites are through computational prediction. Although these
circumventing this problem [30,46,55] . However, for most methods usually work fine in predicting sgRNA-dependent
other diseases, it will be inevitable to edit an adenosine off-target sites basing on sgRNA similarity [68,69] , or when
near other adenosines. Fortunately, a recent study reported sufficient prior data are available for machine-learning
a new version of ABE called ABE9, which exhibited a based prediction [70,71] , additional experimental validation
narrow editing window of only 1-2 nucleotides . By is still necessary to avoid false-positive nominations.
[59]
carefully choosing Cas9 variants with less PAM restriction Off-target effects can also be assessed experimentally by
and designing sgRNAs to put only the target adenosine in either using purified genomic DNA [72-75] or through cell
the editing window, it is promising to drastically reduce the culture [76,77] . However, these methods still cannot fully
likelihood of bystander effects in the future. characterize the off–target effects in animals or human
The undesired edits of ABEs on nucleic acid sites bodies where ABE is used as a therapy. Whole genome
distinct from the targeted site are often called the off-target sequencing (WGS) was a universal approach to detect
effect (Figure 3B-D). Because TadA was originally an the off-target effect both in cells and animal tissues .
[78]
RNA deaminase, a major ABE off-target effect was found The recent development of DISCOVER-seq (discovery
on RNA transcripts (Figure 3B), which was seemingly of in situ Cas off-targets and verification by sequencing)
independent from Cas9 and sgRNA . In addition, ABE and GUIDE-tag (GUIDE means genome-wide, unbiased
[64]
can deposit unwanted edits on sites with 1-2nt mismatches identification of DSB) techniques provides promising tools
to sgRNA (Figure 3C), which is known to be tolerated by to directly assess the off-target effects for in vivo genome
Cas9 [18,65,66] . Interestingly, ABEs can also exert the genomic editing [79,80] . However, whether these methods are sensitive

A C

B D

Figure 3. The types of undesired ABE editing. (A) Bystander effect. TadA edits other untargeted adenine in the editing window of the on-target site. (B) The
off-target effect on mRNA. TadA modified adenine in mRNA independent of Cas9/sgRNA. (C) sgRNA-dependent genomic off-target effect. The sgRNA
tolerates 1–2nt mismatches and guides ABE to modify an off-target site. (D) sgRNA-independent genomic off-target effect. TadA converts A to G in some
genomic DNA sites independent of sgRNA or Cas9. In all panels, the undesired adenine edits are in red, while the target adenine in green.

Volume 2 Issue 1 (2023) 8 https://doi.org/10.36922/gtm.232