E-Poster Presentation 34th Lorne Cancer Conference 2022

De novo design of CRISPR-Cas13b enables the silencing of tumour RNAs with single-base precision    (#304)

Wenxin Hu 1 2 , Mohamed Fareh 1 2 , Gurjeet Singh 1 2 , Shijiao Qi 1 2 , Michelle Haber 3 , Paul G Ekert 3 4 , Ilia Voskoboinik 1 2 , Joseph Trapani 1 2 , Mohamed Fareh 1 2
  1. Cancer Immunology Program , Peter MacCallum Cancer Centre, Melbourne, VIC, Australia
  2. Sir Peter MacCallum Department of Oncology, University of Melbourne, Melbourne, VIC, Australia
  3. Children’s Cancer Institute, Randwick, NSW, Australia
  4. Murdoch Children’s Research Institute, Parkville, VIC, Australia

Cas13 is a new CRISPR effector that targets single-stranded RNAs with high efficacy and specificity. Although Cas13 holds a great promise to suppress or manipulate tumour transcripts with single-base precision, the poor understanding of molecular mechanisms that govern its targeting restricts our ability to reprogram it for cancer discovery and therapeutics purposes.

            To comprehensively investigate Cas13b target recognition and cleavage processes, we developed innovative library screens in mammalian cells where > 200 tiled CRISPR RNAs (crRNAs) are designed to target several transcripts of interest with single-nucleotide resolution (Fareh et al, Nature Commun, 2021). We built in-house bioinformatic analysis pipelines to systematically investigate hidden parameters that govern Cas13 activity in these datasets. The single-nucleotide resolution data revealed that Cas13b is not constrained by protospacer-flanking sites (or PAM-like) that are commonly observed in other CRISPR systems, and therefore highlights its design flexibility. Notably, we revealed previously unknown RNA motifs within the spacer that are either highly enriched or depleted in extremely potent and unproductive crRNAs, respectively. Among these motifs, we found a two-nucleotide sequence at the 5’ end of the spacer that greatly enhanced Cas13b potency. To further validate these findings, we designed de novo crRNAs harbouring unnatural RNA motifs that we identified in the screen, which exhibited enhanced potency that largely outperformed conventionally designed crRNAs. Together, this single-nucleotide screen and bioinformatic analyses revealed how to reprogram Cas13b for efficient transcript silencing.

Finally, we leveraged these molecular features we discovered to reprogram Cas13b to silence several gene fusion transcripts that drive various tumours including acute lymphoblastic leukaemia. We demonstrated that targeting the breakpoint of fusion transcripts with de novo designed tiled crRNAs yielded very high silencing efficiency. Remarkably, these crRNAs failed to silence wild-type transcripts that lack rearranged motifs despite the extensive sequence homology with tumour-specific fusion transcripts.

Taken together, this study revealed key molecular features for efficient Cas13b reprogramming to silence tumour driver transcripts with single-nucleotide precision.