Identifying and targeting collateral lethal vulnerabilities in cancers
Abstract/Summary
Genomic deletions targeting major tumor suppressor genes frequently include adjacent passenger genes,
encoding cell essential housekeeping functions. These cancer cells survive due to co-expressing functionally
redundant paralogs residing in non-deleted regions of the genome. As such, these “collateral deletions” in tumor
suppressor loci can confer cancer cell-specific vulnerabilities through targeted extinction of the remaining
paralog. Our “collateral lethality” concept was first demonstrated in GBM with deletion of the 1p36 tumor
suppressor locus encompassing ENO1, resulting in profound sensitivity to ENO2 depletion or pharmacologic
inhibition (Muller et al. 2012). Subsequently, we demonstrated that deletion of mitochondrial malic enzyme 2
(ME2) in the SMAD4 locus engendered lethality upon shRNA-mediated depletion of the remaining mitochondrial
malic enzyme activity encoded by the ME3 paralog (Dey et al. 2017). To systematically and comprehensively
identify collateral lethal targets in cancer, we first analyzed the Broad Institute’s Cancer Dependency Map
(DepMap), a publicly accessible database hosting essentiality scores of 17386 genes from a pooled genome-
scale CRISPR knockout study conducted in 1054 cancer cell lines of diverse lineages (Dempster et al. 2019;
Ghandi et al. 2019). The computational efforts yielded multiple collateral lethal candidate pairs including
REEP3/4, PTDSS1/2, INTS6/INTS6L, PRPS1/2, LDHA/B, and CSTF2/CSTF2T via a two-class comparison
method that regressed cell line sensitivity vectors against whole-genome CCLE expression and copy number
data to predict paralog-depletion based sensitivity. While computational analysis of DepMap data can identify
some candidate collateral lethal pairs, small sample sizes for many collateral deletions have stymied robust
conclusions. Building on these computational methods, here we will apply the CRISPR/Cas12a polygenic
knockout platform to systematically identify collateral lethal pairs anchored in deletion events targeting common
tumor suppressor gene loci such as TP53, CDKN2A/B, ARID1A, PTEN, SMAD4, RB1 and NF1 across cancer
types. This conceptual and experimental framework, coupled with a cell/organoid/tumor validation platform,
seeks to identify and stringently validate collateral lethal target sets that can then be channeled into a drug
discovery pipeline with the goal of expanding precision cancer treatments.