This new approach identified known and novel hits, including short insertions that could not be introduced by base editing. We also identified key parameters for prime editing screen design and demonstrated the use of synonymous edits to differentiate true and false positives.

While base editing scanning screens proved useful in identifying novel pathogenic mutations, they can only replicate 17% of known EGFR variants. We thus designed a prime editing library introducing almost 3000 patient-derived mutations and conducted a new EGFR activation screen.

Previous experiments were conducted in an EGFR WT cell line but drug resistance often emerges in patients due to secondary mutations. We thus repeated our screens in PC-9 cells harboring an hyperactive EGFR variant and identified both overlapping and context-dependent hits.

We then leveraged the same approach to identify variants impacting cell sensitivity to two clinically approved tyrosine kinase inhibitors. We found hits resistant to one or both molecules, suggesting that these data could be used to help prioritize drug treatments in the clinics.

After controlling for constructs impacting cell viability, we identified 19 hits that led to EGF-independent proliferation. 8 of these likely oncogenic variants were listed as VUS while 6 others were absent from databases, including unexpected C-terminal truncating mutations.

First, we delivered an sgRNA library tiling all EGFR exons along with a C-to-T or an A-to-G base editor to an EGFR WT cell line. We then selected for cells with constitutive EGFR activity by removing EGF from the culture medium before quantifying relative library distributions.