Back

Single-base precision design of CRISPR-Cas13b enables systematic silencing of oncogenic fusions

Hu, W.; Kumar, A.; Qi, S.; Sadras, T.; Casan, J. M.; Ma, D.; Brown, L. M.; Haber, M.; Voskoboinik, I.; Trapani, J. A.; Ekert, P. G.; Fareh, M.

2022-06-23 molecular biology
10.1101/2022.06.22.497105 bioRxiv
Show abstract

Precision oncology programs can rapidly identify oncogenic gene fusions in individual patients1-3. However, despite their established oncogenic status, the vast majority of gene fusions remain undruggable due to the lack of specific inhibitory molecules4, 5. Here, we establish PspCas13b, a poorly characterized programmable RNA nuclease, as a versatile tool to silence various oncogenic fusion transcripts. Our Single-Base Tiled crRNA screens (SiBTil), unbiased computational analysis, and comprehensive spacer-target mutagenesis revealed key determinants of PspCas13b activity. De novo design of crRNAs harbouring basepaired or mismatched guanosine bases at key spacer positions greatly enhances the silencing efficacy of otherwise inefficient crRNAs, expanding the targeting spectrum of this enzyme. We also reveal the interface between mismatch tolerance and intolerance, which unlocks an unexpected single-base precision targeting capability of this RNA nuclease. Notably, our de novo design principles enable potent and selective silencing of various gene fusion transcripts and their downstream oncogenic networks, without off-targeting of non-translocated variants that share extensive sequence homology. We demonstrate that PspCas13b targeting the breakpoint of fusion transcripts enables efficient suppression of ancestral and single-nucleotide mutants (e.g. BCR-ABL1 T315I) that often drive clinical cancer relapse. Collectively, this study provides new design principles for PspCas13b programming to specifically recognise and degrade any undruggable fusion oncogenic transcript, thus providing a new conceptual framework for personalized oncology.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.