Login / Signup

Improving the on-target activity of high-fidelity Cas9 editors by combining rational design and random mutagenesis.

Daria S SpasskayaArtem I DavletshinStanislav S BachurinVera V TutyaevaDavid G GarbuzDmitry S Karpov
Published in: Applied microbiology and biotechnology (2023)
Genomic and post-genomic editors based on CRISPR/Cas systems are widely used in basic research and applied sciences, including human gene therapy. Most genome editing tools are based on the CRISPR/Cas9 type IIA system from Streptococcus pyogenes. Unfortunately, a number of drawbacks have hindered its application in therapeutic approaches, the most serious of which is the relatively high level of off-targets. To overcome this obstacle, various high-fidelity Cas9 variants have been created. However, they show reduced on-target activity compared to wild-type Cas9 possibly due to increased sensitivity to eukaryotic chromatin. Here, we combined a rational approach with random mutagenesis to create a set of new Cas9 variants showing high specificity and increased activity in Saccharomyces cerevisiae yeast. Moreover, a novel mutation in the PAM (protospacer adjacent motif)-interacting Cas9 domain was found, which increases the on-target activity of high-fidelity Cas9 variants while retaining their high specificity. The obtained data suggest that this mutation acts by weakening the eukaryotic chromatin barrier for Cas9 and rearranging the RuvC active center. Improved Cas9 variants should further advance genome and post-genome editing technologies. KEY POINTS: • D147Y and P411T mutations increase the activity of high-fidelity Cas9 variants. • The new L1206P mutation further increases the activity of high-fidelity Cas9 variants. • The L1206P mutation weakens the chromatin barrier for Cas9 editors.
Keyphrases
  • crispr cas
  • genome editing
  • copy number
  • gene expression
  • saccharomyces cerevisiae
  • genome wide
  • transcription factor
  • dna damage
  • machine learning
  • oxidative stress
  • cystic fibrosis
  • artificial intelligence
  • deep learning