79cellular assay to evaluate the sgRNA activity, allowing us to avoid the use of weak
sgRNAs (details below). Then, several groups performed large library-based screens
in a variety of cell lines and organisms and graciously created the web tools avail-
able to the public [ 15 – 20 ]. We further combined our sgRNA selection strategy with
these published scoring systems to increase the targeting efficiency. Another impor-
tant consideration when selecting sgRNAs is the potential for off-target effects [ 6 ,
7 , 21 ]. Undesired genome editing can be avoided by using an engineered Cas9
which has a higher fidelity, as well as by referencing published scoring tools to
computationally select specific target sequence. Below, we describe the experience
acquired from 200 genome editing projects and the key factors that determine
successful targeting.
4.5 Guide RNA Activity is the Key
The principle of genome editing relies on successful production of a sequence-
specific DNA break by a programmable nuclease. The break then triggers DNA
repair responses inside the cell via either non-homologous end joining (NHEJ) to
create sequence disruption or homology-directed repair (HDR) to create intended
DNA replacement. From our experience, the efficiency of guide RNA-mediated
DNA break is the major factor that determines the success of the genome editing
events. The activity requirement is particularly obvious for the large KI projects.
Given the laborious and time-consuming nature of genome engineering in mice, it
is crucial to confirm the activity of guide RNA before use. It is best to validate the
activity of guide RNAs directly in mouse zygotes via injection or electroporation
and in vitro culture to blastocyst stages for DNA-editing analysis. However, when
this approach is not readily accessible, the validation can also be done in cultured
cells though chromatin accessibility, and epigenetic states in certain genomic regions
are expected to be different from those in mouse zygotes. Nevertheless, the expenses
of cultured cell-based validation are lower, and a larger number of guide RNAs can
be tested per batch in cultured cells.
In the beginning of our CRISPR service, we picked mouse kidney epithelialmK4 cells to validate sgRNA activity by the T7E1 cleavage assay because mK4
cells are highly transfectable [ 22 ]. To establish the minimum sgRNA activity
required for efficient gene targeting in mouse embryos, we cloned 29 sgRNAs into
a pX458 vector (addgene #48138), which expresses both sgRNA and Cas9 protein.
After transfection into mK4 cells, the sgRNA’s relative activity was obtained by
comparing with a Tet2 sgRNA (target sequence: GAAAGTGCCAACAGATATCC)
and the result ranged from 26 to 195% (Fig. 4.1a). We then injected these sgRNAs
individually into fertilized mouse zygotes, followed by embryo transfer to produce
live offspring. After genetic editing analysis of these offspring, we determined that
for a specific sgRNA to be able to induce genetic modifications in mouse zygotes,
this guide must exhibit at least 85% of Tet2 sgRNA activity in a cell-based T7E1
assay. Based on this observation, we further surveyed 204 sgRNAs targeting 74
4 A Transgenic Core Facility’s Experience in Genome Editing Revolution