Fusion of Mu Gam protein for improved base editors
in BE4 was achieved through fusion of Gam protein derived from Phage Mu. The
Gam protein fusion product in the form of BE3-Gam and BE4-Gam vectors helps in
binding with DSBs leading to reduction in indels and an improvement in product
purity. Thus fusion of Mu Gam protein protein to BE3 and BE4 gave one of the
best base editors for conversion of C:G to T:A15 (Figure 9).
Figure 10. Components of
plasmids BE3-GAM (with one copy of UGI) and BE4-Gam (with two copies of UGI).
Delivery of base editors as RNPs
In an important recent study, It was argued that DNA
delivery method using plasmid vectors results in uncontrolled Cas9 and sgRNA expression even after
the on-target locus has been edited, thus providing opportunity for off-target
editing. Therefore, delivery of exogenous DNA plasmid
vectors was replaced by DNA-free lipid mediated delivery of ribonucleoproteins
(RNPs) prepared through association of BE proteins with gRNA. Delivery of RNPs when
tried in mice inner ear and zebrafish embryos exhibited much higher target
specificity and reduced off-target editing21.
Adenine base editors (ABEs)
Initially, base editing was restricted to C®T conversion involving
the use of naturally
occurring DNA cytosine
deaminase. However, no
DNA adenine deaminase occurs in nature, although RNA adenine deaminases occur
for modification of tRNAs, so that synthetic DNA adenine deaminases had to be developed from RNA
adenine deaminase through
protein engineering using directed evolution. Once DNA
adenine deaminase was successfully synthesized in the laboratory, base adenine
base editors (ABE) could be developed for conversion of adenine into
inosine (I) (Figure
11); the latter mimics guanine (G) during DNA replication,
so that T:A base pair could be edited into
C:G base pair (Figures 11, 12). The detailed structure of an ABE is
depicted in Figure 13.
Figure 11. Conversion of adenine
into inosine using synthetic adenine deaminase.
Conversion of T:A base pair into C:G base pair using adenine base editor (ABE).
13. An adenine base editor (ABE) showing different componenets
including an adenine deaminase (red), gRNA (green), and Cas9 nickase (blue)11.
generations of of ABEs were developed, so that ultimately seventh generation improved ABEs included the following four ABEs: ABE 6.3, ABE 7.8, ABE 7.9 and ABE 7.10911.
Of these, ABE7.10 was the most active editor, with high level of editing
efficiency (53%). The editing window, however, was still 4-7 in ABE7.10 and 4-9
in the other three seventh generation ABEs. These ABEs did not display any significant A to non-G conversion at
because the removal of inosine through BSE from DNA is not as common as that of
uracil (U)9 due
to the presence of UNG11. ABEs also
performed better than many BEs in terms of off-target editing and
frequency of indels
produced during editing. In an actual study, ABE7.10 modified only
4/12 off-targets with a frequency of 1.2% indels, in comparison with
off-targets with a 14% indel rate in other BEs.
Base editing in RNA
and ‘REPAIR’ technology
October, 2017, along with the report of ABEs by the group led by David Lui11,
base editors were also developed for editing of RNA transcripts. These RNA base
editors were developed by another group led by Feng Zhang24 of the
Broad Institute, who has also been visible in recent years due to CRISPR patent
battle25.. In most studies on
base editing, target site occurs in
genomic DNA, but this technology has been limited by the requirement of
canonical PAM-NGG site at the target locus, although base editors for sites
with non-canonical PAM were also developed (discussed earlier). Keeping this in
view, RNA base editors were developed using
the programmable type VI CRISPR-associated RNA-guided RNase Cas13b; the
technology has been described as REPAIR RNA editing for programmable A to I
(G) replacement12. The naturally occurring ADAR (adenosine
deaminase acting on RNA) was used with disabled Cas13 (dCas13) and guide RNA
(gRNA) for endogenous A®I editing at specific
transcript targets in mammalian cells. The specificity of dCas13b was further improved
through mutagenesis to generate REPAIRv1 and then REPAIRv2, which had very high level of specificity. Hopefully,
REPAIRv2 or other improved form of REPAIR will be utilized in future for a
variety of purposes, particularly for teansient alteration in the transcripts, where s
permanent alteration of the genome is not required (e.g., temporary relief from