CRISPR screens are feasible in TP53 wild-type cells

Abstract

A recent study by Haapaniemi et al (2018) reported that intact p53 signaling hampers CRISPR-based functional genomic screens. Brown et al report good performance of genome-scale screens in TP53 wild-type cells and reiterate best practices for CRISPR screening.

Figure 1. Performance of CRISPR–Cas9 functional genomic screens with respect to TP53 status

(A) Genome‐wide CRISPR–Cas9 screens were performed in five established human cell lines. Heatmap shows mean log₂‐fold change of all sgRNAs targeting the indicated p53 pathway genes. (B) Mean log₂‐fold change scatterplots for all 17,236 genes targeted by the TKOv1 library for screens described in (A). Bottom triangle indicates Spearman correlation between screen pairs. Red: cell lines with non‐functional p53. Green: cell lines with functional p53. (C) Mean log₂‐fold changes of all sgRNAs per gene are shown for the results of a CRISPR screen with TP53 wild‐type RPE1 cells. Gene sets comprising gold‐standard essential genes (essentials), ribosome or proteasome subunits, or gold‐standard non‐essential genes (non‐essentials) are highlighted. (D) Ratios of CERES scores for reference gene sets between TP53 wild‐type and mutant/non‐functional cell lines across 356 cancer cell lines in the 18Q4 Avana (DepMap) dataset for which TP53 functional status was available in Giacomelli et al (***FDR < 0.001, **FDR < 0.01, Wilcoxon rank sum test with multiple testing correction). (E) Precision–recall curves based on the mean log₂‐fold change of sgRNAs using gold‐standard essential and non‐essential genes across five RPE1 TP53 wild‐type genome‐wide CRISPR screens and two RPE1 TP53‐deficient (“null”) genome‐wide CRISPR screens. The CRISPR libraries and source of the screens are indicated. Also see main text for details and references.