Genome-wide CRISPR screening reveals novel therapeutic targets in RIT1-driven lung cancer

ABSTRACT

In recent work, we performed CRISPR/Cas9 screening in RIT1 (Ras-like in all tissues)-mutant cancer cells. We found that RIT1-mutant cells are vulnerable to loss of mitotic regulators, and mutant RIT1 synergizes with YAP1 (yes-associated protein 1) in oncogenesis. These findings can be leveraged to identify targeted therapies for RIT1-mutant cancer.

Lung cancer is the most common cause of cancer-related deaths worldwide.¹ Approximately 75% of lung cancer cases are caused by mutations in the Receptor Tyrosine Kinase (RTK)/RAS signaling pathway.² Mutations within this pathway – in genes such as EGFR (epidermal growth factor receptor) and KRAS – are mutually exclusive, meaning that tumors are often driven by one key oncogene.² In the past decade, targeted therapies have revolutionized cancer treatment by offering a means to kill cancer cells that harbor specific oncogenic mutations. Targeted therapies are more effective and less toxic than nonselective, cytotoxic chemotherapy. Despite this progress, however, there are still oncogenes that cannot be targeted with currently available inhibitors.

RIT1 (Ras-like in all tissues) is a small GTPase mutated in approximately 2% of lung adenocarcinoma cases and amplified in another 14% of cases.² Out of the currently identified oncogenic RIT1 mutations, the M90I variant (RIT1^M90I) is most prevalent.³ RIT1 mutations are known to promote anchorage-independent growth and xenograft tumor formation.³ Despite being classified as an oncogene, very little is known about how RIT1 alterations cause cancer, and few cell lines are available to study RIT1 mutations (Figure 1a). RIT1 is structurally similar to other GTPases, notably KRAS; however, RIT1 does not phenocopy KRAS.^3,4 Instead, context-dependent factors dictate RIT1’s ability to activate signaling through the RTK/RAS pathway.^3,4 As such, there are currently no targeted therapies available for RIT1-driven diseases. To address this gap in the field, we sought out to better understand RIT1 biology to not only gain more insight into oncogenic mechanisms underlying RIT1-driven cancer, but also to identify druggable targets. In addition to lung cancer, RIT1 mutations and/or focal amplifications are present in myeloid malignancies and the developmental disease Noonan syndrome.^5,6

Figure 1.

Isogenic CRISPR screening identifies genetic vulnerabilities in RIT1-driven lung cancer. A. Only one RIT1 (Ras-like in all tissues)-mutant lung cancer cell line is commercially available, thereby making RIT1 mutations difficult to study. Cell line numbers obtained from the Dependency Map (DepMap) portal. B. Left, in PC9 lung adenocarcinoma cells, a mutation in EGFR (epidermal growth factor receptor) renders cells sensitive to the EGFR inhibitor erlotinib. Right, expression of RIT1^M90I in PC9 cells confers erlotinib resistance and restores cell survival. C. This erlotinib resistance phenotype was used to perform CRISPR/Cas9 screens in isogenic PC9 cells expressing lung cancer driver oncogenes, including RIT1^M90I. D. From these CRISPR screens, we found that RIT1^M90I weakens the Spindle Assembly Checkpoint (SAC), rendering cells vulnerable to inhibitors of SAC components such as the Aurora kinases. We also found that RIT1^M90I synergizes with YAP1 (yes-associated protein 1) to induce transcription of YAP1 targets, suggesting that YAP1 inhibitors should be investigated for the treatment of diseases caused by mutation of RIT1. Figure created with Biorender.com

In recent work, we performed genome-wide CRISPR/Cas9 knockout screens in PC9 lung adenocarcinoma cells stably expressing a lung cancer oncogene (RIT1^M90I, KRAS^G12V, or EGFR^T790M/L858R).⁴ These screens took advantage of the observation that each introduced oncogene confers resistance to EGFR inhibition (Figure 1b-c).^3,4,7 We used this drug resistance phenotype to probe the requirements for oncogene-driven survival. We computationally analyzed the screen data to identify genetic dependencies (genes that, when knocked out, confer a growth disadvantage) and cooperating factors (genes that, when knocked out, confer a growth advantage). Interestingly, we found that RIT1-mutant cells were more dependent than KRAS-mutant cells on several genes implicated in the Spindle Assembly Checkpoint (SAC), including Aurora kinases.⁴ The SAC is a mitotic surveillance mechanism which ensures that all chromosomes are properly aligned before anaphase begins.⁸ These findings suggest that RIT1-mutant cells are uniquely sensitive to loss of SAC genes. To investigate this further, we performed a small molecule inhibitor screen and found that RIT1-mutant cells were more sensitive than KRAS-mutant cells to Aurora kinase inhibitors.⁴ The drugs alisertib (an Aurora kinase A inhibitor) and barasertib (an Aurora kinase B inhibitor) selectively inhibited the growth, proliferation, and colony formation of RIT1-mutant cells.⁴ Together, these data identify the SAC as a therapeutic vulnerability in RIT1-mutant cells.

We next sought to investigate the mechanism underlying RIT1’s dependency on SAC genes. We performed live-cell, time-lapse imaging microscopy and analyzed chromosomal abnormalities in RIT1-mutant HeLa H2B-GFP cells. We found that RIT1-mutant cells progressed through mitosis significantly faster and accumulated more mitotic abnormalities than parental cells.⁴ These data suggest that RIT1 weakens the SAC, thereby rendering cells vulnerable to additional SAC perturbation through genetic knockout or pharmacological inhibition. We tested this hypothesis in vivo and found that depletion of Aurora kinase A abrogated tumor growth in a xenograft model of RIT1-mutant lung cancer.⁴ Together, these findings highlight the potential efficacy of Aurora kinase inhibitors for the treatment of RIT1-driven diseases (Figure 1d).

In addition to exploring genetic dependencies of RIT1-mutant cells, we also investigated positively selected gene knockouts from our CRISPR screen. We found that loss of genes within the Hippo signaling pathway cooperated with RIT1^M90I to promote growth and proliferation.⁴ The Hippo pathway is a well-studied tumor-suppressive signaling pathway that regulates the stability of the YAP1 (yes-associated protein 1) transcription factor.⁹ Suppression of Hippo signaling leads to stabilization and nuclear translocation of YAP1 to induce transcription of target genes.⁹ Therefore, we hypothesized that RIT1^M90I and YAP1 synergize to promote oncogenesis (Figure 1d). In vivo xenograft experiments support this hypothesis: in mice, we found that RIT1^M90I and YAP1 alone are weak oncogenic drivers of xenograft tumor growth, but co-expression of RIT1^M90I and YAP1 leads to robust and rapid tumor formation.⁴ To further investigate this RIT1^M90I + YAP1 synergy, we analyzed human tumor samples in The Cancer Genome Atlas (TCGA) and an independent clinical cohort. TCGA analysis revealed that RIT1-mutant and -amplified tumors exhibited loss of Hippo pathway genes, indicating a potential cooperative role of Hippo suppression and RIT1 activation in human lung cancer.⁴ Additionally, we performed immunohistochemistry analysis of human tumors and found that RIT1-mutant tumors showed significantly higher nuclear-localized YAP1⁴. Together, this human tumor data strengthens our findings and provides further evidence to support the hypothesis that RIT1^M90I and YAP1 synergize to promote the growth and proliferation of non-small cell lung cancer tumors. These results suggest that YAP1 inhibitors should be explored as possible therapies for diseases characterized by RIT1 mutation or amplification (Figure 1d).

In summary, our work offers key insight into better understanding RIT1 biology and answering critical questions: how do RIT1 alterations cause cancer, and how can we better treat RIT1-mutant tumors? Our findings can be leveraged in future work to test the utility of Aurora kinase inhibitors and YAP1 inhibitors for the treatment of RIT1-driven diseases.

Funding Statement

This work was supported in part through National Cancer Institute R00CA197762 and R37CA252050 to A.H.B, donations from the Smith family to A.H.B., National Cancer Institute Cancer Center Support Grant P30 CA015704 New Investigator support to A.H.B. A.K.R. was supported in part by National Institute of General Medical Sciences T32GM007270.

Disclosure statement

No potential conflict of interest was reported by the author(s).