Tumor RAS-GTP PDX NF1 S1759C A. B. PDX BRAF D594N (LUN023) www.revolutionmedicines.com Introduction SHP2 (PTPN11) is a non-receptor protein tyrosine phosphatase and scaffold protein that functions downstream of multiple RTKs, integrating growth factor signals to promote RAS activation. Accordingly, autosomal dominant activating mutations in PTPN11 fuel pathogenic RAS/MAPK signaling and are drivers of human cancers. Genetic knockdown or pharmacological inhibition of SHP2 suppresses RAS/MAPK signaling and has been shown to limit the proliferation of cancer cells dependent upon a range of activated RTKs, pointing to a therapeutic opportunity for SHP2 inhibition in RTK-driven cancers 1 . Here, we describe the use of a potent and selective SHP2 allosteric inhibitor, RMC-4550. RMC-4550 suppresses RAS activation and proliferation in cancers that are driven by mutations downstream of RTKs, but still dependent on RAS nucleotide cycling 2 . We explore the functional consequences of SHP2 allosteric inhibition and demonstrate that RMC- 4550 decouples activated EGFR and the RAS-GEF, SOS1. Furthermore, we observe that RMC-4550 disrupts key protein-protein interactions in RTK- associated signaling complexes. As a convergent node for growth factor/RTK signaling, SHP2 may be a well-positioned target for disruption of the upstream signaling events that are necessary for pathogenic RAS/MAPK signaling in a genetically-defined subset of tumors. Figure 3. SHP2 inhibition with RMC-4550 disrupts GAB1 association and cellular localization Experimental Approach Figure 2. Mechanism of action of SHP2 inhibition is dependent on SOS1 Figure 4. RMC-4550 drives tumor growth inhibition in vivo in KRAS G12C , NF1 LOF , and BRAF class 3 models Results • KRAS G12C mutant cell lines were preferentially sensitive to SHP2 inhibition (Fig. 1A, hypergeometric p-value = 0.0043). This is consistent with prior work demonstrating that the KRAS G12C isoform retains intrinsic GTP hydrolysis capacity and therefore remains dependent upon upstream signals for activation 3 . • NF1 LOF and Class 3 BRAF mutations were also identified as markers of SHP2 sensitivity, (Fig. 1D,E) while Class 1 and 2 BRAF mutations were resistant. This is consistent with prior work demonstrating the RAS- GTP-dependence of NF1 LOF and Class 3 BRAF mutations 4,5 . • Decoupling of SOS1-mediated RAS-GTP loading from upstream signals using a constitutively active SOS-F mutant was sufficient to rescue pERK signaling from SHP2 inhibition (Fig.2). • SHP2 allosteric inhibition with RMC-4550 disrupts the interaction between SHP2 and GAB1(Fig. 3A). There is decreased membrane localization of SHP2 and GAB1 in RMC-4550-treated cells (Fig. 3B). • CDX and PDX studies demonstrated that well tolerated once daily, oral dosing of RMC-4550 drives dose-dependent tumor growth inhibition, with concomitant inhibition of tumor RAS-GTP and pERK in models of KRAS G12C NF1 LOF and BRAF D594N cancers (Fig. 4). Summary The observation that KRAS G12C , NF1 LOF , and class 3 BRAF mutations confer sensitivity to SHP2 inhibition in tumor cells establishes SHP2 inhibition as a novel and promising therapeutic strategy against tumors that harbor these oncogenic driver mutations. These data suggest dependence on RAS nucleotide cycling as a key determinant of SHP2 sensitivity. Mechanistically, SHP2 allosteric inhibition by RMC-4550 disrupts protein- protein interactions involved in signaling complexes that link RTK signals to RAS. This disruption is likely an important factor in the mislocalization of key proteins involved in RAS/MAPK signaling. A deeper understanding of the mechanisms of SHP2 allosteric inhibition has important implications for informed patient selection strategies and broadens the potential utility of SHP2 allosteric inhibition as a clinical therapeutic. Functional characterization of SHP2-mediated RAS-MAPK pathway inhibition in cancer cells bearing oncogenic mutations dependent upon nucleotide cycling of RAS Christopher J. Schulze 1 , Franziska Haderk 2,3,4 , Carlos Stahlhut 1 , Robert J. Nichols 1 , Golzar Hemmati 2,3,4 , David Wildes 1 , Christos Tzitzilonis 1 , Kasia Mordec 1 , Abby Marquez 1 , Jason Romero 1 , Tientien Hsieh 1 , Gert Kiss 1 , Elena S. Koltun 5 , Adrian L. Gill 5 , Mallika Singh 1 , Mark A. Goldsmith 1,5 , Jacqueline A. M. Smith 1 , and Trever G. Bivona 2,3,4 1 Department of Biology, Revolution Medicines, Inc., Redwood City, CA, USA 2 Departments of Medicine and 3 Cellular and Molecular Pharmacology, 4 Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, San Francisco, CA, USA, 5 Department of Chemistry, Revolution Medicines, Inc., Redwood City, CA, USA A. C. D. A. C. C. D. • RMC-4550 was used to profile the sensitivity of RAS/MAPK-driven cancer cell lines to SHP2 inhibition • Proliferation assays were conducted using 3D spheroid cultures. Signaling assays of RAS-GTP (ELISA) and pERK (Alphalisa) were conducted to verify the on-target effects of SHP2 inhibition • Mechanistic experiments were performed in HEK293 cells, with inducible expression of genetic constructs for SOS-F experiments • Cell line-derived xenograft (CDX) and patient-derived xenograft (PDX) studies were designed to translate in vitro findings to relevant mouse models and test drug-like properties of RMC-4550 pERK NF1 LOF Class 3 BRAF CDX MIA-PaCa2 (KRAS G12C/G12C , pancreas) Cell Line Tissue Mutation Growth IC 50 (μM) pERK IC 50 (μM) NCI-H358 Lung KRAS G12C/+ 0.1 0.028 NCI-H1838 Lung NF1 l183fs 0.1 0.004 SK-MEL-113 Skin NF1 -/- 0.03 0.030 A-375 Skin BRAF V600E/V600E >10 >10 NCI-H1755 Lung BRAF G469A/+ >10 >10 NCI-H1666 Lung BRAF G466V/+ 0.3 0.006 NCI-H508 Caecum BRAF G596R/+ 0.06 0.004 Project DRIVE RMC-4550 References 1. Chen, Y.-N.P., et al. Allosteric inhibition of SHP2 phosphatase inhibits cancers driven by receptor tyrosine kinases. Nature 535, 148-152 (2016). 2. Nichols, R.J., et al. Efficacy of SHP2 phosphatase inhibition in cancers with nucleotide-cycling oncogenic RAS, RAS-GTP dependent oncogenic BRAF and NF1 loss. bioRxiv (2017). 3. Patricelli, M.P., et al. Selective Inhibition of Oncogenic KRAS Output with Small Molecules Targeting the Inactive State. Cancer Discovery 6, 316- 329 (2016). 4. Nissan, M.H., et al. Loss of NF1 in cutaneous melanoma is associated with RAS activation and MEK dependence. Cancer Research 74, 2340-2350 (2014). 5. Yao, Z., et al. Tumours with class 3 BRAF mutants are sensitive to the inhibition of activated RAS. Nature, 1-15 (2017). * Proliferation and signaling assay data are representative of ≥ 2 independent experiments each performed in duplicate. Figures show mean +/- S.D. B. RAS-GTP E. Figure 1. KRAS G12C NF1 LOF and Class 3 BRAF Mutations are markers of SHP2-sensitivity SOS-F EGF-Induced pERK Activation B. SOS-F construct C. SOS-F RAS-GTP D. A. 3X-FLAG SHP2 Co-IP 3X-FLAG GAB1 Co-IP B. Cellular fractionation DMSO, no EGF DMSO, +EGF RMC-4550 10 uM, +EGF RMC-4550 1uM, +EGF RMC-4550 0.1 uM, +EGF 0 1 2 3 normalized RAS-GTP signal SOS-WT SOS-F (-) 5 15 30 (-) 5 15 30 0 50000 100000 150000 200000 Duration of EGF stimulation (min.) Cellular pERK signal (RLU) PDX NF1 S1759C (TH77) Tumor pERK CDX MIA-PaCa2 E. -10 -8 -6 -4 0 50 100 150 log M [RMC-4550] Cellular pERK (% Control) SOS-F SOS-WT