Genome-Wide CRISPR Screen for Essential Cell Growth Mediators in Mutant KRAS Colorectal Cancers

Abstract

Targeting mutant KRAS signaling pathways continues to attract attention as a therapeutic strategy for KRAS-driven tumors. In this study, we exploited the power of the CRISPR-Cas9 system to identify genes affecting the tumor xenograft growth of human mutant KRAS (KRAS^MUT) colorectal cancers. Using pooled lentiviral single-guide RNA libraries, we conducted a genome-wide loss-of-function genetic screen in an isogenic pair of human colorectal cancer cell lines harboring mutant or wild-type KRAS. The screen identified novel and established synthetic enhancers or synthetic lethals for KRAS^MUT colorectal cancer, including targetable metabolic genes. Notably, genetic disruption or pharmacologic inhibition of the metabolic enzymes NAD kinase or ketohexokinase was growth inhibitory in vivo In addition, the chromatin remodeling protein INO80C was identified as a novel tumor suppressor in KRAS^MUT colorectal and pancreatic tumor xenografts. Our findings define a novel targetable set of therapeutic targets for KRAS^MUT tumors. Cancer Res; 77(22); 6330-9. ©2017 AACR.

Figure 1. Genome-Wide CRISPR-Cas9 Screen of Isogenic KRAS WT/MUT Xenografts

(A) Schematic representation of genome-wide human GeCKO knockout screen in paired HCT116 cell lines with and without a KRAS G13D mutation. (B) Scatterplots of CRISPR-Cas9 guide scores (calculated as log₂ fold change of normalized read counts of individual sgRNAs in tumor xenograft samples compared with T0 cells normalized to the median log₂ fold change of the non-targeting controls in each sample) for HCT116^WT vs HCT116^MUT xenografts. The 1000 non-targeting sgRNAs and sgRNAs against 927 reference essential genes and 1580 essential fitness genes are shown in the separate panels. (C) Scatterplot of CRISPR-Cas9 guide scores of HCT116^WT vs HCT116^MUT xenografts. Common lethal sgRNAs (CRISPR-Cas9 guide score < −1 in both cell lines) are highlighted in red and KRAS synthetic lethal sgRNAs (CRISPR-Cas9 guide score < −0.45 in HCT116^MUT cells, CRISPR-Cas9 guide score > −0.45 in HCT116^WT cells, and [HCT116^MUT CRISPR-Cas9 guide score – HCT116^WT CRISPR-Cas9 guide score] < −0.45) are highlighted in green. (D) GSEA analysis of the top 10 KEGG pathways of targeted genes in the corresponding common lethal or KRAS synthetic lethal regions (FDR q-value threshold <0.05). Common lethal and KRAS synthetic lethal pathways are shown in red and green, respectively.

Figure 2. Validation of Individual KRAS Synthetic Lethal Metabolic Pathway Genes

(A) CRISPR-Cas9 guide rank score (derived from average of the two best sgRNA CRISPR-Cas9 guide scores within sgRNAs targeting a gene) for HCT116^MUT and HCT116^WT cells selected after 14 days in culture (in vitro) or after 14 days growth as tumor xenografts. Candidate genes are indicated in red. (B) Schematic of single sgRNA tumor xenograft experiments. Stable Cas9-expressing HCT116^MUT or HCT116^WT cells were infected with single sgRNAs targeting the candidate genes, and genomic DNA was deep sequenced to analyze indels and substitutions. Cells were injected into nude mice (n = 3–4 mice per sgRNA, 2 sgRNAs; total of 6–8 mice per cell line per gene) and tumor growth was measured for 24 days. (C) Tumor growth after injection of nude mice with HCT116^WT or HCT116^MUT cells transduced with non-targeting control sgRNAs (n = 6 mice) or SUCLA2-, NADK-, or KHK-targeting sgRNAs (n = 3–4 mice per sgRNA, 2 sgRNAs per gene; total of 6–8 mice per cell line per gene). Student’s t-test, *p < 0.05, **p < 0.005, ***p < 0.001. Error bars indicate ± SEM. (D) Metabolic pathways associated with genes identified in the CRISPR-Cas9 screen (GFPT1, SUCLA2, KHK, and NADK). (E) Tumor growth after injection of nude mice with HCT116^WT or HCT116^MUT cells. Mice were treated with the NADK inhibitor thionicotinamide (100 mg/kg) or vehicle by intraperitoneal injection every other day between days 12 and 24 (7 doses; n = 6 or 8 mice per group for WT and MUT, respectively). Student’s t-test, **p < 0.005, ***p < 0.001, ****p < 0.0001. Error bars indicate ± SEM. (F) Tumor volumes on day 25 of the experiment shown in (E). Each symbol represents a single mouse. P value determined by Student’s t-test. Bars indicate the mean and 95% confidence intervals (CIs). (G and H) Experiments were performed as described for E and F except xenografted mice were treated with a KHK inhibitor (25 mg/kg) or vehicle (n = 8 or 9 mice per group for WT and MUT, respectively) every other day between days 8 and 21. In G, Student’s t-test, **p < 0.005, ***p < 0.001, ****p < 0.0001. In G, error bars indicate ± SEM. In H, bars indicate mean and 95% CIs.

Figure 3. Focused Secondary Validation sgRNA Screen

(A) Boxplot of CRISPR-Cas9 guide scores in the primary GeCKO screen of sgRNAs targeting candidate genes selected by RIGER overlap for inclusion in the validation mini-library. P values determined by Student’s t-test. (B) Schematic of secondary mini-library screen. Lentiviruses were transfected with a pooled plasmid library representing 2500 sgRNAs targeting ~250 genes chosen from the primary GeCKO screen, with 9 sgRNAs per gene target. HCT116^WT and HCT116^MUT cell lines (n = 3) were infected with the lentiviruses and injected into nude mice (n = 3 mice per transduction replicate, total n = 18 mice). (C) Scatterplots of CRISPR-Cas9 guide scores for HCT116^WT vs HCT116^MUT xenografts, showing 230 non-targeting control sgRNAs (gray), sgRNAs targeting 25 essential genes (red), and sgRNAs targeting 150 candidate KRAS lethal genes (green). Individual sgRNAs targeting KRAS and MAPK1 are highlighted in dark green. (D) CRISPR-Cas9 guide rank scores for HCT116^MUT and HCT116^WT xenografts. KRAS and MAPK1 are highlighted in green. (E) Heatmap from STAR output. Genes that scored as significantly depleted at FDR < 0.25 only in HCT116^MUT xenografts in all 3 transduction replicates. STAR output from averaged values from 3 transduction replicates using a 25% threshold. Genes are ranked by the difference between HCT116^MUT and HCT116^WT STAR scores. Genes with no sgRNAs meeting the 25% threshold are given a value of 0. (F) CRISPR-Cas9 guide rank scores for HCT116^MUT and HCT116^WT xenografts. Candidate KRAS synthetic lethal genes are highlighted in green.

Figure 4. Identification and Validation of INO80C as a Candidate KRAS-Dependent Tumor Suppressor Gene

(A) CRISPR-Cas9 guide rank scores (derived from average of all sgRNA CRISPR-Cas9 guide scores targeting a gene) for HCT116^MUT and HCT116^WT xenografts. INO80C, NF2, and RALGAPB are highlighted as indicated. (B) Heatmap from STAR output. Genes that scored as significantly enriched at FDR<0.1 only in HCT116^MUT xenografts in all 3 transduction replicates. STAR output from averaged values from 3 transduction replicates using a 25% threshold. Genes are ranked by the difference between HCT116^MUT and HCT116^WT STAR scores. Genes with no sgRNAs meeting the 25% threshold are given a value of 0. (C) CRISPR-Cas9 guide rank score (derived from average of all sgRNA CRISPR-Cas9 guide scores targeting a gene) Capan-2 xenografts. INO80C and NF2 are highlighted as indicated. (D) Tumor growth after injection of nude mice with HCT116^WT and HCT116^MUT -dCas9 cells (upper) or DLD-1^{KRAS WT} and DLD-1^{KRAS MUT} -dCas9 cells (lower) transduced with non-targeting control sgRNAs (n = 6 mice) or INO80C-targeting sgRNAs (n = 4 mice per sgRNA, 2 sgRNAs per gene; total of 8 mice). Student’s t-test, *p < 0.05, **p < 0.005, ***p < 0.001, ****p<0.0001. Error bars indicate ± SEM. (E) Tumor growth after injection of nude mice with H358-dCas9 cells (upper) or Capan-2-dCas9 cells (lower) transduced with non-targeting control sgRNAs (n = 9 mice) or INO80C-targeting sgRNAs (n = 5 or 6 mice per sgRNA, 2 sgRNAs; total of 11 mice). Student’s t-test, *p < 0.05, **p < 0.005, ***p < 0.001, ***p<0.0001. Error bars indicate ± SEM.