Limits...
Identification of oncogenic driver mutations by genome-wide CRISPR-Cas9 dropout screening

View Article: PubMed Central - PubMed

ABSTRACT

Background: Genome-wide CRISPR-Cas9 dropout screens can identify genes whose knockout affects cell viability. Recent CRISPR screens detected thousands of essential genes required for cellular survival and key cellular processes; however discovering novel lineage-specific genetic dependencies from the many hits still remains a challenge.

Results: To assess whether CRISPR-Cas9 dropout screens can help identify cancer dependencies, we screened two human cancer cell lines carrying known and distinct oncogenic mutations using a genome-wide sgRNA library. We found that the gRNA targeting the driver mutation EGFR was one of the highest-ranking candidates in the EGFR-mutant HCC-827 lung adenocarcinoma cell line. Likewise, sgRNAs for NRAS and MAP2K1 (MEK1), a downstream kinase of mutant NRAS, were identified among the top hits in the NRAS-mutant neuroblastoma cell line CHP-212. Depletion of these genes targeted by the sgRNAs strongly correlated with the sensitivity to specific kinase inhibitors of the EGFR or RAS pathway in cell viability assays. In addition, we describe other dependencies such as TBK1 in HCC-827 cells and TRIB2 in CHP-212 cells which merit further investigation.

Conclusions: We show that genome-wide CRISPR dropout screens are suitable for the identification of oncogenic drivers and other essential genes.

Electronic supplementary material: The online version of this article (doi:10.1186/s12864-016-3042-2) contains supplementary material, which is available to authorized users.

No MeSH data available.


sgRNAs depleted in the whole genome screen. a Scatterplot representing fold changes of the 57 096 targeting sgRNAs in the HCC-827 cell line at day 14 and day 21. Fold changes at day 14 or day 21 were calculated compared to the control time point at day −10. All time points were measured in duplicates and median fold changes are shown. Dark green colored dots represent the 1 000 non-targeting control sgRNAs and grey colored dots represent the 57 096 targeting sgRNAs. Genes of interest were annotated by the software Spotfire and visualization was further enhanced by red colored dots. b Same as in a) but for CHP-212 cell line. c Scatterplot of fold changes of 1 571 kinases in the HCC-827 cell line at time points day 14 and day 21 versus control time point day −10. d Same as c) but for HCC-827 cells. e, f Scatterplots for Q1 and RSA down of 57 096 targeting sgRNAs in the HCC-827 (e) and CHP-212 cell lines (f) at time point day 14. Dark green colored dots represent the 1 000 non-targeting control sgRNAs. g, h Scatterplot for Q1 and RSA down of the 1 571 sgRNAs for kinases are shown for the HCC-827 (e) and CHP-212 (f) cell lines
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC5016932&req=5

Fig2: sgRNAs depleted in the whole genome screen. a Scatterplot representing fold changes of the 57 096 targeting sgRNAs in the HCC-827 cell line at day 14 and day 21. Fold changes at day 14 or day 21 were calculated compared to the control time point at day −10. All time points were measured in duplicates and median fold changes are shown. Dark green colored dots represent the 1 000 non-targeting control sgRNAs and grey colored dots represent the 57 096 targeting sgRNAs. Genes of interest were annotated by the software Spotfire and visualization was further enhanced by red colored dots. b Same as in a) but for CHP-212 cell line. c Scatterplot of fold changes of 1 571 kinases in the HCC-827 cell line at time points day 14 and day 21 versus control time point day −10. d Same as c) but for HCC-827 cells. e, f Scatterplots for Q1 and RSA down of 57 096 targeting sgRNAs in the HCC-827 (e) and CHP-212 cell lines (f) at time point day 14. Dark green colored dots represent the 1 000 non-targeting control sgRNAs. g, h Scatterplot for Q1 and RSA down of the 1 571 sgRNAs for kinases are shown for the HCC-827 (e) and CHP-212 (f) cell lines

Mentions: To answer the question of whether a negative selection CRISPR-Cas9 screen can identify genes playing a non-redundant role in the oncogenic driver pathways, we deployed two different approaches: (1) we analyzed fold changes for all 57 096 sgRNAs; and (2) we focused on the 1 751 sgRNAs which target kinases. Fold changes for time points day 14, 21, and 28 were calculated as change in frequency of the respective sgRNA compared to the control time point at day −10. We compared fold changes of all sgRNAs from day 14 versus day 21 and found that most of the 1 000 non-targeting control sgRNAs overlaid with the majority of all targeting sgRNAs (Fig. 2a, b). This indicates that many data points fall into the background variability of the CRISPR screen (Fig. 2a, b). As a threshold level we used the fifth percentile of the fold changes of the depleted sgRNA (Fig. 2). This includes 1 450 sgRNAs of all genes from a total of 57 096 sgRNAs (Fig. 2a) and 22 sgRNAs from kinase sgRNAs (Fig. 2c) for the HCC-827 cell line. For the CHP-212 cell line, 1 462 sgRNAs are among the 5 % most depleted sgRNAs for all genes and 24 sgRNAs for kinases (Fig. 2b, d). We provide the lists of 1 000 most depleted genes for the HCC-827 and CHP-212 cell lines (Additional files 4 and 5: Tables S3 and S4). We found EGFR scoring high among the strongest depleted genes for the EGFR-mutant HCC-827 cell line being well above the threshold level (Fig. 2a). Similarly, NRAS and MAP2K1 – a kinase downstream of NRAS [17] - were among the most depleted genes for the NRAS-mutant cell line CHP-212 within the threshold level (Fig. 2b). RAF1, another kinase downstream of NRAS, was found to be below the threshold level (Fig. 2b). These data show that oncogenic drivers and pathways can be differentiated from the multitude of key survival genes in our CRISPR-Cas9 screen, indicating that an oncogenic driver mutation causes a strong dependency.Fig. 2


Identification of oncogenic driver mutations by genome-wide CRISPR-Cas9 dropout screening
sgRNAs depleted in the whole genome screen. a Scatterplot representing fold changes of the 57 096 targeting sgRNAs in the HCC-827 cell line at day 14 and day 21. Fold changes at day 14 or day 21 were calculated compared to the control time point at day −10. All time points were measured in duplicates and median fold changes are shown. Dark green colored dots represent the 1 000 non-targeting control sgRNAs and grey colored dots represent the 57 096 targeting sgRNAs. Genes of interest were annotated by the software Spotfire and visualization was further enhanced by red colored dots. b Same as in a) but for CHP-212 cell line. c Scatterplot of fold changes of 1 571 kinases in the HCC-827 cell line at time points day 14 and day 21 versus control time point day −10. d Same as c) but for HCC-827 cells. e, f Scatterplots for Q1 and RSA down of 57 096 targeting sgRNAs in the HCC-827 (e) and CHP-212 cell lines (f) at time point day 14. Dark green colored dots represent the 1 000 non-targeting control sgRNAs. g, h Scatterplot for Q1 and RSA down of the 1 571 sgRNAs for kinases are shown for the HCC-827 (e) and CHP-212 (f) cell lines
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC5016932&req=5

Fig2: sgRNAs depleted in the whole genome screen. a Scatterplot representing fold changes of the 57 096 targeting sgRNAs in the HCC-827 cell line at day 14 and day 21. Fold changes at day 14 or day 21 were calculated compared to the control time point at day −10. All time points were measured in duplicates and median fold changes are shown. Dark green colored dots represent the 1 000 non-targeting control sgRNAs and grey colored dots represent the 57 096 targeting sgRNAs. Genes of interest were annotated by the software Spotfire and visualization was further enhanced by red colored dots. b Same as in a) but for CHP-212 cell line. c Scatterplot of fold changes of 1 571 kinases in the HCC-827 cell line at time points day 14 and day 21 versus control time point day −10. d Same as c) but for HCC-827 cells. e, f Scatterplots for Q1 and RSA down of 57 096 targeting sgRNAs in the HCC-827 (e) and CHP-212 cell lines (f) at time point day 14. Dark green colored dots represent the 1 000 non-targeting control sgRNAs. g, h Scatterplot for Q1 and RSA down of the 1 571 sgRNAs for kinases are shown for the HCC-827 (e) and CHP-212 (f) cell lines
Mentions: To answer the question of whether a negative selection CRISPR-Cas9 screen can identify genes playing a non-redundant role in the oncogenic driver pathways, we deployed two different approaches: (1) we analyzed fold changes for all 57 096 sgRNAs; and (2) we focused on the 1 751 sgRNAs which target kinases. Fold changes for time points day 14, 21, and 28 were calculated as change in frequency of the respective sgRNA compared to the control time point at day −10. We compared fold changes of all sgRNAs from day 14 versus day 21 and found that most of the 1 000 non-targeting control sgRNAs overlaid with the majority of all targeting sgRNAs (Fig. 2a, b). This indicates that many data points fall into the background variability of the CRISPR screen (Fig. 2a, b). As a threshold level we used the fifth percentile of the fold changes of the depleted sgRNA (Fig. 2). This includes 1 450 sgRNAs of all genes from a total of 57 096 sgRNAs (Fig. 2a) and 22 sgRNAs from kinase sgRNAs (Fig. 2c) for the HCC-827 cell line. For the CHP-212 cell line, 1 462 sgRNAs are among the 5 % most depleted sgRNAs for all genes and 24 sgRNAs for kinases (Fig. 2b, d). We provide the lists of 1 000 most depleted genes for the HCC-827 and CHP-212 cell lines (Additional files 4 and 5: Tables S3 and S4). We found EGFR scoring high among the strongest depleted genes for the EGFR-mutant HCC-827 cell line being well above the threshold level (Fig. 2a). Similarly, NRAS and MAP2K1 – a kinase downstream of NRAS [17] - were among the most depleted genes for the NRAS-mutant cell line CHP-212 within the threshold level (Fig. 2b). RAF1, another kinase downstream of NRAS, was found to be below the threshold level (Fig. 2b). These data show that oncogenic drivers and pathways can be differentiated from the multitude of key survival genes in our CRISPR-Cas9 screen, indicating that an oncogenic driver mutation causes a strong dependency.Fig. 2

View Article: PubMed Central - PubMed

ABSTRACT

Background: Genome-wide CRISPR-Cas9 dropout screens can identify genes whose knockout affects cell viability. Recent CRISPR screens detected thousands of essential genes required for cellular survival and key cellular processes; however discovering novel lineage-specific genetic dependencies from the many hits still remains a challenge.

Results: To assess whether CRISPR-Cas9 dropout screens can help identify cancer dependencies, we screened two human cancer cell lines carrying known and distinct oncogenic mutations using a genome-wide sgRNA library. We found that the gRNA targeting the driver mutation EGFR was one of the highest-ranking candidates in the EGFR-mutant HCC-827 lung adenocarcinoma cell line. Likewise, sgRNAs for NRAS and MAP2K1 (MEK1), a downstream kinase of mutant NRAS, were identified among the top hits in the NRAS-mutant neuroblastoma cell line CHP-212. Depletion of these genes targeted by the sgRNAs strongly correlated with the sensitivity to specific kinase inhibitors of the EGFR or RAS pathway in cell viability assays. In addition, we describe other dependencies such as TBK1 in HCC-827 cells and TRIB2 in CHP-212 cells which merit further investigation.

Conclusions: We show that genome-wide CRISPR dropout screens are suitable for the identification of oncogenic drivers and other essential genes.

Electronic supplementary material: The online version of this article (doi:10.1186/s12864-016-3042-2) contains supplementary material, which is available to authorized users.

No MeSH data available.