Proofreading the editors
Editas’ UDiTaS assay uncovers off-target genome edits, large and small
The threat of off-target edits has spurred CRISPR companies to build in quality control systems from the earliest stages of drug development. Editas Medicine Inc. thinks its Uni-Directional Targeted Sequencing methodology, dubbed “UDiTaS,” will detect and characterize every type of on- and off-target edit created by its therapeutic candidates.
The method exploits transposon biology to amplify and sequence sites prone to off-target editing by CRISPR drug candidates, regardless of what kinds of sequence changes that editing caused. “We can now quantitate and measure in one tube what all the editing outcomes are at a single cut site, and that wasn’t a really efficient process before we came up with this technology,” said Editas CTO Vic Myer.
Editas is developing therapies that combine the nucleases Cas9 or Cpf1 with gene-targeted guide RNAs (gRNAs) to treat eye diseases, and has research programs for cystic fibrosis, Duchenne muscular dystrophy (DMD), sickle cell disease, β-thalassemia, and α-antitrypsin (AAT) deficiency. The company is also partnered with Juno Therapeutics Inc. to engineer T cells to treat cancer.
On Wednesday, Editas’ lead product EDIT-101 received the European Medicines Agency (EMA)’s Orphan Medicinal Product designation. EDIT-101, a CRISPR-based product that targets CEP290, is in preclinical testing to treat Leber congenital amaurosis type 10 (LCA10).
While CRISPR therapeutics are designed to target specific sequences, the system’s biggest vulnerability is the potential to produce cuts elsewhere in the genome, particularly in regions that share homology with the targeted site. When cells repair these breaks with their DNA repair machinery, off-target sites can acquire a wide range of mutations -- ranging from small insertions or deletions (indels) to large chromosomal rearrangements -- that can trigger adverse events like oncogenesis, even when they are rare.
But Myer said finding these edits isn’t as simple as getting a full readout of every edited genome, because the typically low frequency of off-target edits makes them