Target validation by combining Oncolines® profiling and gene dependency screens

  • Large-scale gene dependency screens performed on hundreds of cancer cell lines provide insight into the impact on cell viability caused by individual knockout (CRISPR) or knockdown (RNAi) of thousands of genes.1-5

  • Correlating gene dependency scores and response of the same cell lines to a drug of interest can help validate the drug’s target.

  • 2D cell viability assays were performed for the SHP2 inhibitor vociprotafib and CHEK1 inhibitor SCH900776 in 102 cancer cell lines (the Oncolines® panel).

  • A significant correlation was calculated between de IC50 values of vociprotafib and the CRISPR gene dependency scores of the SHP2-encoding gene PTPN11 (Figure 1A).

  • Among the > 17,000 genes in the CRISPR dataset, PTPN11 is the second most strongly correlated gene, while the genes encoding the SHP2-interacting proteins GRB2 and SOS1 also rank high (Figure 1B).

  • RNAi gene dependency scores may improve target validation for a pan-essential gene, because complete knockout of these genes using CRISPR results in high dependency for all cell lines (Figure 2).

  • Although the CRISPR gene dependency scores of the pan-essential gene CHEK1 are not significantly correlated with SCH900776 cell line response (Figure 3A), the correlation is significant when using the RNAi dataset (Figure 3B).

References

1. Meyers et al. (2017) Computational correction of copy number effect improves specificity of CRISPR-Cas9 essentiality screens in cancer cells. Nature Genetics 49, 1779-1784.

2. Dempster et al. (2019) Extracting biological insights from the project Achilles genome-scale CRISPR screens in cancer cell lines. BioRxiv, 720243.

3. McFarland et al. (2018) Improved estimation of cancer dependencies from large-scale RNAi screens using model-based normalization and data integration. Nature Communications 9, 4610.

4. Dempster et al. (2021) Chronos: a cell population dynamics model of CRISPR experiments that improves inference of gene fitness effects. Genome Biology 22, 343

5. Pacini et al. (2021) Integrated cross-study datasets of genetic dependencies in cancer. Nature Communications 12, 1661.

Figure 1 | CRISPR gene dependency correlation with vociprotafib
(A) A scatterplot comparing the IC50 values of the SHP2 inhibitor vociprotafib and the CRISPR gene dependency scores of the SHP2-encoding PTPN11 gene across the 102 Oncolines® cell line panel. (B) A volcano plot of the Pearson correlation between drug IC50 values and CRISPR dependency scors of  > 17,000 genes. The horizontal line indicates significance after multiple testing correction.
Figure 2 | Difference in gene dependency patterns between CRISPR and RNAi
The gene dependency pattern for PTPN11 (left) and CHEK1 (right) in both the CRISPR (purple) and RNAi (blue) dataset for the 102 cell lines. If the gene has a dependency score smaller than -0.5 it is considered to be essential for cell proliferation.
Figure 3 | CRISPR and RNAi gene dependency correlation with SCH900776
A volcano plot showing the correlations of the > 17,000 genes in the CRISPR dataset (A) and the > 5,800 genes in the RNAi dataset (B) with SCH900776 based on the Pearson correlation. The horizontal line indicates significance after multiple testing correction.