Abstract

Background

With lentivirally delivered CRISPR/Cas9 it has proved possible to adapt the infrastructure of short-hairpin RNA interference screens to genome-wide knock out screens . Proof of principle positive selection screens have identified genes whose knock-out leads to drug or toxin resistance, gain of metastatic potential or altered signalling in immune cells. Some authors have attempted to define the “essentialome” – the set of genes required for cell viability in culture. Extension of this work to identify genetic-context-dependent essential genes would open up new avenues to target "undruggable" cancer driver genes such as KRAS & TP53.

Method

Horizon has screened 30 colon and lung cancer cell lines with an sgRNA library targeting around 3000 genes (10 sgRNA/gene). Bayesian algorithms were used to define the set of fitness genes in each cell line. Comparison of these sets highlighted genes that may be selectively essential in certain contexts. An extended library of sgRNAs targeting hit genes was made in an improved vector bearing a modified tracRNA designed to have higher affinity for Cas9. This confirmation library was designed to contain sgRNAs positioned every 5-10 amino acids along putative target proteins.

Results

Primary screens confirmed previously reported synthetic lethal interactions such as the MDM2 and PPM1D dependence in TP53 wild-type cancers and SMARCA2 dependence in the majority of SMARCA4 mutant lung cancers. Additionally, putative novel interactions were identified, some of which were confirmed by follow up screens. Encouragingly, subsets of TP53 mutant colon cancers exhibited dependence on several potentially "druggable" targets for which we found highly active sgRNAs targeting pharmacologically modulatable domains.  Evidence for elevated dependence on the anti-apoptotic protein MCL1 was also observed in colon lines bearing point mutations in the F box factor responsible for MCL1 turnover, FBXW7.

Conclusion

Results to date are in accord with the view that CRISPR/Cas9 can identify novel synthetic lethal oncology targets.