Genome-Wide CRISPR Screens Identify Multiple Synthetic Lethal Targets That Enhance KRASG12C Inhibitor Efficacy

Abstract

Non-small lung cancers (NSCLC) frequently (∼30%) harbor KRAS driver mutations, half of which are KRASG12C. KRAS-mutant NSCLC with comutated STK11 and/or KEAP1 is particularly refractory to conventional, targeted, and immune therapy. Development of KRASG12C inhibitors (G12Ci) provided a major therapeutic advance, but resistance still limits their efficacy. To identify genes whose deletion augments efficacy of the G12Cis adagrasib (MRTX-849) or adagrasib plus TNO155 (SHP2i), we performed genome-wide CRISPR/Cas9 screens on KRAS/STK11-mutant NSCLC lines. Recurrent, potentially targetable, synthetic lethal (SL) genes were identified, including serine-threonine kinases, tRNA-modifying and proteoglycan synthesis enzymes, and YAP/TAZ/TEAD pathway components. Several SL genes were confirmed by siRNA/shRNA experiments, and the YAP/TAZ/TEAD pathway was extensively validated in vitro and in mice. Mechanistic studies showed that G12Ci treatment induced gene expression of RHO paralogs and activators, increased RHOA activation, and evoked ROCK-dependent nuclear translocation of YAP. Mice and patients with acquired G12Ci- or G12Ci/SHP2i-resistant tumors showed strong overlap with SL pathways, arguing for the relevance of the screen results. These findings provide a landscape of potential targets for future combination strategies, some of which can be tested rapidly in the clinic.

Significance: Identification of synthetic lethal genes with KRASG12C using genome-wide CRISPR/Cas9 screening and credentialing of the ability of TEAD inhibition to enhance KRASG12C efficacy provides a roadmap for combination strategies. See related commentary by Johnson and Haigis, p. 4005.

Figure 1.. Genome-wide CRISPR/Cas9 screens identify MRTX-849 synthetic lethal (SL) genes.

A, Schematic showing CRISPR/Cas9 screening strategy B, Volcano plots showing results of genome-wide CRISPR/Cas9 screens of KRAS^G12C;STK11 co-mutated non-small cell lung cancer (NSCLC) cell lines, analyzed by MaGeCK; orange circles indicate select SL genes (FDR <0.1) C, Circos plot showing overlap of SL genes in NSCLC lines (FDR<0.1). Outside arcs show SL genes within each line. Inside arcs show SL genes shared in multiple lines (dark orange) and those unique to individual lines (light orange). Purple lines show which cells share a given SL gene. The greater the number of purple links and the size of the dark orange arcs, the greater the overlap of SL genes between cell lines. D, Heat map showing select SL genes across the four cell lines. Color code indicates FDR for each gene in each line (scale at left). E, Bubble plot indicates enriched pathways (p < 0.05) of SL genes (FDR < 0.1). Datasets used for pathway analysis are color-coded as shown on right side. Size of circle indicates significance of each pathway assignment.

Figure 2.. Validation of YAP/TAZ/TEAD pathway genes.

A, Trypan blue-based proliferation assays (5 days) on indicated cell lines treated with TEAD1 or WWTR1 siRNA, as indicated, and/or MRTX-849 (at IC50), normalized to untreated (Control) cells, ****p<0.0001, *** p<0.001, 1-way ANOVA with Tukey’s multiple comparisons test. B, Proliferation assays on H2122 cells stably transduced with lentiviruses expressing either of two doxycycline-inducible TEAD1 shRNAs or control shRNA and treated with MRTX-849 (at IC50) of vehicle with or without prior doxycycline (Dox) treatment for 96 hr, ****p<0.0001, 1-way ANOVA with Tukey’s multiple comparisons test C, Proliferation assays on 72 hr Dox-induced H2030 and H2122 cell lines transduced with doxycycline-inducible dominant negative TEAD and treated with MRTX-849 (at IC50) or vehicle, as indicated, ****p<0.0001, 1-way ANOVA with Tukey’s multiple comparisons test. D-F, MRTX-849 dose-response curves (using modified MTS assay) for the indicated mouse cell lines stably overexpressing TEAD1(D), WWTR1 (E), YAP1 (F) or YAP1 mutants (G). IC50s were determined by GraphPad Prism.

Figure 3.. MRTX-849 treatment induces RHO/ROCK-dependent nuclear translocation of YAP.

A, H2030 cells were treated with MRTX-849 (at IC50) for 48 hrs, and the activity of the TEAD-responsive 8XGIITC-Luc reporter, normalized to a co-transfected Renilla luciferase construct, was determined. *p<0.05, Student’s t-test B, Heat map showing results of bulk RNA-seq of H2030 and H2122 cells treated with MRTX-849 (at IC50) for 48 hr in triplicate C, Bubble plot indicates pathways enriched (p < 0.05) in up-regulated genes (FDR < 0.1). Datasets (color-coded) used for pathway analysis are indicated at right with size of circles indicating significance. D, Immunofluorescence images showing YAP1 and DAPI staining of representative fields of H2030 cells treated with MRTX-849 (at IC50) for the indicated times. E, YAP1 and DAPI immunofluorescence of H2122, H23, and MiaPaca2 cells treated with MRTX-849 (at their respective IC50s) for 48 hrs. F, MRTX-849 treatment causes increased RHOA activity. H2030 cells were treated with MRTX-849 (at IC50) for 48h, and RHOA-GTP was quantified by ELISA. Luminescence at A490nm in treated samples normalized to DMSO-treated values is shown. *p<0.05, Student’s t-test G-H, ROCK inhibitor treatment impairs MRTX-849-induced YAP1 nuclear localization. The indicated NSCLC lines were treated with MRTX-849 (at IC50) with or without the ROCK inhibitor Y-27632 (10 μM), and YAP1 localization was assessed by immunofluorescence (with DAPI staining to identify nuclei). I, Trypan blue-based proliferation assays on H2030 and H2122 cell lines treated with MRTX-849 (at IC50) alone or with Y-27632 (10 μM) as indicated, normalized to untreated (Control) cells. ****p<0.0001, ***p<0.001, 1-way ANOVA with Tukey’s multiple comparisons test. All immunofluorescence images are representative of three independent experiments. Scale bar=20 μm

Figure 4.. Whole-genome CRISPR screens for MRTX-849 + TNO155 synthetic lethal genes.

A, Genome-wide CRISPR/Cas9 SL screens of H2122, H23, and H2030 cells in the presence or absence of MRTX-849 + TNO155 (at doses described in Results) were analyzed using MaGeCK. Select SL genes (FDR<0.1) are indicated by orange circles. B, Bubble plot shows pathways (p<0.05) enriched in SL genes (FDR< 0.1). Datasets used for analysis are color-coded at right; size of circle indicates significance level. C, Circos plot illustrating overlap of SL genes (FDR<0.1) between lines. D, Trypan-blue-based proliferation assays on H2030 and H2122 cell lines transfected with TEAD1 siRNA (where indicated) or scrambled control siRNA and treated with vehicle or MRTX-849 and/or TNO155 (at half the IC25 for both drugs) , as indicated. . ****p<0.0001, 1-way ANOVA and Tukey’s multiple comparisons test E, Proliferation of H2122 cells expressing two different TEAD1 shRNA and treated with MRTX-849 and/or TNO155, as indicated (IC25 dosage for both of the drugs). ****p<0.0001, 1-way ANOVA and Tukey’s multiple comparisons test F, Effects of dominant negative TEAD and MRTX-849 and/or TNO155 (IC25 dose) on proliferation of H2030 and H2122 cells. ****p<0.0001, 1-way ANOVA and Tukey’s multiple comparisons test. G, Representative YAP1 and DAPI immunofluorescence images (from 3 independent experiments) of H2030 cells treated with MRTX-849+TNO155 (each at respective IC0) for 48 hrs.

Figure 5.. TEAD inhibition enhances efficacy of MRTX-849 in KRAS^G12C-mutant cancers.

A, Trypan blue-based proliferation assays (6 days) on H2122, H2030, HCC-44, and H23 lines treated with MYF-03–176 (1μM) and MRTX-849 (at IC50 for each line) alone or in combination. ****p<0.0001, 1-way ANOVA and Tukey’s multiple comparisons test B-D, Proliferation assays on the indicated cell lines using VT104 (1μM) and MRTX-849 (at IC50 for each line) alone or in combination, ****p<0.0001, 1-way ANOVA and Tukey’s multiple comparisons test, ^#synergy by Bliss independent analysis. E-F, Relative change in tumor volumes after withdrawal of treatments at Day 30, *p <0.0001, 2-way ANOVA.

Figure 6.. G12Ci- and G12Ci/SHP2 inhibitor-resistant GEMM and patient samples induce pathways overlapping with SL genes:

A, Representative MRI images of KCL mice showing successive development of MTRX-849 and MRTX-849/SHP099 resistance B, (Left) Select enriched pathways (p< 0.05) for genes upregulated in MRTX-849-resistant KCL tumors (FDR < 0.1), (Right) GSEA demonstrating increased expression of YAP-TAZ signature genes in these tumors. C, Snapshot of RPPA showing increased YAP/TAZ levels in MRTX-849-resistant nodules; see also Supplementary Table S7. D, (Left) Select enriched pathways (p< 0.05) for genes upregulated in MRTX-849/SHP099-resistant KCL tumors (FDR < 0.1). (Right) GSEA demonstrating increased expression of YAP-TAZ signature genes in these tumors. E, Snapshot of RPPA showing increased YAP/TAZ levels in MRTX-849/SHP099-resistant nodules; see also Supplementary Table S9. F-G, Pathway analysis on sc-RNAseq of cells from fresh tumor biopsies of patients with G12Ci (AMG510)- or G12C/SHP2i (MRTX-849 +TNO155)-resistant NSCLC, as indicated.

Figure 7.. Validation of selected additional targets from screens:

A-B, Trypan blue-based proliferation assays (5-days) on H2030, H2122, and H23 cells transfected with RIOK2, VRK1, or scrambled siRNAs and/or treated with MRTX-849 (at IC50), as indicated. ****p<0.0001, ***p<0.001, **p<0.01, 1-way ANOVA and Tukey’s multiple comparisons test C, Trypan blue-based proliferation assays (7 days) on H2030 and H2122 cells treated with VRK-IN-1 (10 μM) and/or MRTX-849 (at IC50), as indicated. ****p<0.0001, ***p<0.001, **p<0.01, 1-way ANOVA and Tukey’s multiple comparisons test D-E, Same as A-B, but with ELP3, ELP5, or scrambled siRNAs, as indicated. F, Trypan blue-based proliferation assays on H2030 and H2122 cells transfected with scrambled or ELP5 siRNAs and treated with MRTX-849 and/or TNO155 (at IC25 of each drug in each line) as indicated. ****p<0.0001, ***p<0.001, **p<0.01, 1-way ANOVA and Tukey’s multiple comparisons test, G-I, Trypan blue-based proliferation assays on H2122 or H2030 cells stably transduced with doxycycline-inducible ELP3, RIOK2, ELP5, or control shRNA or control shRNA, exposed to Dox for 96 h, and treated with MRTX-849 (at IC50 for each line), as indicated, ****p<0.0001, 1-way ANOVA and Tukey’s multiple comparisons test