Scientists
at the Francis Crick Institute have discovered a set of simple rules that
determine the precision of CRISPR/Cas9 genome editing in human cells. These
rules, published in Molecular Cell, could help to improve the
efficiency and safety of genome editing in both the lab and the clinic.
Despite
the wide use of the CRISPR system, rational application of the technology has
been hindered by the assumption that the outcome of genome editing is
unpredictable, resulting in random deletions or insertions of DNA regions at
the target site.
Before
CRISPR can be safely applied in the clinic, scientists need to make sure that
they can reliably predict precisely how DNA will be modified.
"Until
now, editing genes with
CRISPR has involved a lot of guesswork, frustration and trial and error,"
says Crick group leader Paola Scaffidi, who led the study. "The effects of
CRISPR were thought to be unpredictable and seemingly random, but by analysing
hundreds of edits we were shocked to find that there are actually simple,
predictable patterns behind it all. This will fundamentally change the way we
use CRISPR, allowing us to study gene function with greater precision and
significantly accelerating our science."
By
examining the effects of CRISPR genome editing at 1491 target sites across 450
genes in human cells, the
team have discovered that the outcomes can be predicted based on simple rules.
These rules mainly depend on one genetic 'letter' occupying a particular
position in the region recognized by the 'guide RNA' to direct the molecular
scissors, Cas9 .
Guide
RNAs are synthetic molecules made up of around 20 genetic letters (A,T,C,G),
designed to bind to a specific section of DNA in the target gene. Each genetic
letter has a complementary partner—A binds to T and C binds to G—which stick
together a bit like Velcro. The guide RNA is like the 'hook' side of Velcro,
designed to stick to the 'loop' side on the target gene.
Guided
by the RNA molecule, the Cas-9 enzyme scans along the genome until it finds the
region of interest. When the RNA guide matches the correct DNA sequence, it
sticks like Velcro and Cas9 cuts through the DNA. The DNA is broken three
letters from the end of the target sequence, and bits of genetic code are then
inserted or deleted, seemingly haphazardly, when the cell attempts to repair
the break.
In
this study, the researchers found that the outcome of a particular gene edit
depends on the fourth letter from the end of the RNA guide, adjacent to the
cutting site. The team discovered that if this letter is an A or a T, there
will be a very precise genetic insertion; a C will lead to a relatively precise
deletion and a G will lead to many imprecise deletions. Thus, simply avoiding
sites containing a G makes genome editing much more predictable.
"We
were amazed to discover that the rules that determine the outcome of CRISPR
human genome editing are so simple," says Dr. Anob Chakrabarti, Wellcome
Trust clinical Ph.D. fellow in the Crick's Bioinformatics and Computational
Biology lab and joint-first author of the study. "By bearing these rules
in mind when designing our guide RNAs, we can maximise the chances of getting
the desired outcome of a specific gene edit—which is particularly important in
a clinical context."
The
team also discovered that how 'open' or 'closed' the target DNA is also affects
the outcome of gene editing. Adding compounds that force DNA to open
up—allowing Cas9 to scan the genome—led to more efficient editing, which could
help when modifications need to be introduced in particularly closed genes.
"The
good news is that regardless of the tissue of origin—which influences the
degree of DNA 'openness' at specific genes—target regions containing an A or T
at the key position show common editing," says Paola. "This means
that, if we carefully select the target DNA, we can be pretty confident that
we'll see the same effect in different tissues."
Josep
Monserrat, Crick Ph.D. student in the Cancer Epigenetics lab and joint-first
author of the study, says: "We hadn't previously appreciated the
significance of DNA openness in determining the efficiency of CRISPR genome editing. This could be another
factor to consider when aiming to edit a gene in a specific way. We are excited
to observe that distinct cell types share common editing at precise target
regions, and hope translation of our findings will be beneficial across
disciplines."