SFK/FAK Signaling Attenuates Osimertinib Efficacy in Both Drug-Sensitive and Drug-Resistant Models of EGFR-Mutant Lung Cancer

Home , Selumetinib

Author Manuscript Published OnlineFirst on April 17, 2017; DOI: 10.1158/0008-5472.CAN-16-2300 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

SFK/FAK signaling attenuates osimertinib efficacy in both drug-sensitive and drug-resistant models of EGFR-mutant lung cancer

Eiki Ichihara1*, David Westover1*, Catherine B. Meador2, Yingjun Yan1, Joshua A. Bauer3,5,6, Pengcheng Lu4, Fei Ye4, Amanda Kulick7, Elisa de Stanchina7, Robert McEwen8, Marc Ladanyi9, Darren Cross8, William Pao1,2,6,**, Christine M. Lovly1,2,6,#

Departments of 1Medicine, 2Cancer Biology, 3Biochemistry, 4Biostatistics, 5Institute of Chemical Biology, High-Throughput Screening Facility, and 6Vanderbilt Ingram Cancer Center, Vanderbilt University School of Medicine, Nashville, Tennessee 37235, USA; 7Anti-Tumor Assessment Core, Memorial Sloan Kettering Cancer Center, New York, New York 10065, USA; 8AstraZeneca Oncology, Cambridge CB2 0RE, United Kingdom; 9Department of Pathology, Memorial Sloan-Kettering Cancer Center, New York, New York 10065, USA. *These authors contributed equally **Current affiliation: Roche Pharmaceuticals, Basel, Switzerland

Running Title: SFK/FAK inhibition enhances effects of osimertinib

Keywords: Epidermal growth factor receptor (EGFR), non-small cell lung cancer (NSCLC), osimertinib, Src family kinases (SFKs), tyrosine kinase inhibitor (TKI), therapeutic resistance.

Abbreviation List: EGFR: epidermal growth factor receptor; FAK: focal adhesion kinase; MAPK: mitogen activated protein kinase; NSCLC: non-small cell lung cancer; SFK: Src family kinase; TKI: tyrosine kinase inhibitor.

Financial Support: This study was supported in part by the National Institutes of Health (NIH) and National Cancer Institute (NCI) R01-CA121210 (CML), P01-CA129243 (ML, CML), and P30-CA008748 (EDS). EI was supported by the Uehara Memorial Foundation. DW was supported by a T32 training grant (2T32HL094296-06A1). CML was supported by a V Foundation Scholar-in-Training Award, an AACR-Genentech Career Development Award, a Damon Runyon Clinical Investigator Award, and a LUNGevity Career Development Award.

#Corresponding author: Christine M. Lovly, MD, PhD Vanderbilt University School of Medicine 2220 Pierce Avenue South, 777 Preston Research Building Nashville, TN 37232 Phone: (615) 936-3457 Fax: (615) 343-8668 Email: [email protected]

Word count: 4,592 Total number of figures and tables: 5

Downloaded from cancerres.aacrjournals.org on September 28, 2021. © 2017 American Association for Cancer Research. Author Manuscript Published OnlineFirst on April 17, 2017; DOI: 10.1158/0008-5472.CAN-16-2300 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Abstract

Mutant-selective epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKI), such as osimertinib, are active agents for the treatment of EGFR-mutant lung cancer. Specifically, these agents can overcome the effects of the T790M mutation, which mediates resistance to first and second-generation EGFR TKI, and recent clinical trials have documented their efficacy in patients with EGFR-mutant lung cancer. Despite promising results, therapeutic efficacy is limited by development of acquired resistance. Here we report that Src family kinases (SFK) and focal adhesion kinase (FAK) sustain AKT and mitogen-activated protein kinase (MAPK) pathway signaling under continuous EGFR inhibition in osimertinib sensitive cells. Inhibiting either the MAPK pathway or the AKT pathway enhanced the effects of osimertinib. Combined SFK/FAK inhibition exhibited the most potent effects on growth inhibition, induction of apoptosis, and delay of acquired resistance. SFK family member YES1 was amplified in osimertinib- resistant EGFR-mutant tumor cells, the effects of which were overcome by combined treatment with osimertinib and SFK inhibitors. In conclusion, our data suggest that concomitant inhibition of both SFK/FAK and EGFR may be a promising therapeutic strategy for EGFR-mutant lung cancer.

Introduction

Epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs), such as erlotinib, gefitinib, and afatinib, are standard therapy in patients with advanced non-small cell lung cancer (NSCLC) harboring somatic EGFR mutations (1-3). However, acquired resistance to these agents invariably occurs, often within a year after initiation of therapy (4-7). Disease progression is mediated by the secondary EGFR ‘gatekeeper’ mutation, T790M, in more than half of these cases (8-10).

Recently, mutant-selective third-generation EGFR TKIs, such as osimertinib (AZD9291) and rociletinib (CO-1686), have been developed with the goal of overcoming T790M-mediated resistance (11-14). In addition to increased potency against mutant EGFR, these TKIs also have less toxicity compared to first- or second-generation inhibitors because they “spare” wild-type EGFR. Recent clinical studies have demonstrated that osimertinib monotherapy induces a 61% response rate and durable benefit in patients with lung cancers harboring the EGFR T790M mutation (15), resulting in approval by the US Food and Drug Administration for the treatment of patients with advanced EGFR T790M mutation-positive NSCLC. As with other EGFR TKIs, however, the magnitude and duration of response to osimertinib is variable and resistance inevitably develops, suggesting that targeting EGFR alone will not achieve long-term benefits. In addition, although most tumors initially decrease in size during EGFR TKI treatment, the tumors typically reach a steady state, implying that there may be mechanisms that make tumor cells tolerant to EGFR inhibitors, even in EGFR TKI-sensitive tumors.

We hypothesized that additional signaling pathways active in tumor cells may attenuate the effects of osimertinib, thereby limiting its full anti-tumor activity. We found that signaling downstream of EGFR through the AKT and mitogen-activated protein kinase (MAPK) pathways remained active even in the presence of osimertinib. Sustained signaling through these pathways under continuous EGFR inhibition appears to be, in part, regulated by Src family kinase (SFK) and focal adhesion kinase (FAK) signaling. Concomitant inhibition of EGFR, SFKs, and FAK most effectively enhanced osimertinib activity and suppressed the development of resistance. We also found that amplification of the SFK, YES1, caused resistance to osimertinib. These findings may lead to new therapeutic options for patients with EGFR-mutant lung cancer.

Materials and Methods

Kinome-wide siRNA library screen: PC-9/BRc1 cells were transfected with the GE Dharmacon Protein Kinase siRNA library (GU003505). The siRNA library included nine 96-well plates containing siRNAs targeting 714 kinases and kinase-related genes (4 individual siRNAs per gene) in a SMARTpool format. PC-9/BRc1 cells were reverse-transfected in 96-well plates (4,000 cells/well in 10% FBS) containing 5 pmol of siRNA/well using DharmaFECT4 (Dharmacon, Lafayette, CO, USA). Each plate also contained two wells with 5pmol/well of non-targeting siRNA (siNT, Dharmacon). Twenty-four hours after transfection, cells were trypsinized and divided into replicate plates. After an additional 24 hour incubation, cells were treated with DMSO or osimertinib (5 nM). After 72 hours of treatment, cells were incubated with Alamar Blue (Invitrogen, Carlsbad, CA, USA) for 4 hours, and viability was measured by 3

fluorescence readings using an EnSpire 2300 (544 nm excitation, 590 nm emission) (PerkinElmer, Waltham, MA, USA). The viability measurements were normalized to the siNT controls by taking the mean of the controls divided by each well's viability value. The sensitizing index (SI) value was calculated for each siRNA to identify the gene targets that promote osimertinib sensitivity or resistance using the following formula, as previously described (16): SI = Expected total viability effect - Observed combined effect Expected total viability effect = (Rc/Cc*Cd/Cc), where Rc/Cc = the viability effect of siRNA without drug compared to siNT control Cd/Cc = the effect of drug on viability to siNT control Observed combined effect = Rd/Cc Rd/Cc = the viability effect of siRNA with drug compared to siNT control The Z-score was calculated as follows: Z-score= siRNA SI score - mean of all siRNAs in that plate)/standard deviation of all siRNAs in that plate. This experiment was performed in duplicate. Any genes with a Z score > 2 and an SI ≥ 0.05 were considered significant.

Cell culture: PC-9, HCC827, NCI-H1975, NCI-H3255, and HCC4006 cell lines were purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA) prior to 2009. The 11-18 cell line was generously provided by Koichi Hagiwara (17). PC-9/BRc1 were derived from PC-9 cells by chronic exposure to afatinib (18), HCC827, HCC4006, and 11-18 were cultured in RPMI 1640 medium (Mediatech, Corning, NY, USA) supplemented with 10% heat-inactivated FBS (Atlanta Biologicals, Flowery Branch, GA, USA) and penicillin (100 U/mL)/streptomycin (100 μg/mL) (Mediatech). Cells were grown in a humidified incubator with 5% CO2 at 37°C. All cell lines are tested for mycoplasma with the VenorGeM Mycoplasma Detection Kit (Sigma-Aldrich, St. Louis, MO, USA) approximately every 6 months. For experimental studies, the indicated cell lines were seeded at passage number ranging from 5 to 25. For TKI-resistant cells, passage 1 was defined as the first passage following confirmation of a resistant phenotype. After 25 passages, cell lines were discarded to prevent genetic and phenotypic drift.

Derivation of osimertinib resistant cell lines: To create resistant lines, parental cells were cultured with increasing concentrations of osimertinib, starting at the IC30 (19). At 80-90% confluence, cells were trypsinized and divided into 2 tubes. Half were frozen and the rest were reseeded into a new dish at a 20-30% higher dose of osimertinib. Fresh drug was added every 72 – 96 hours.

Inhibitors: Erlotinib was synthesized by the MSKCC Organic Synthesis Core. Osimertinib was either supplied by Astra Zeneca or purchased from Selleck Chemicals. All the other drugs were purchased from Selleck Chemicals (Houston, TX, USA).

Immunoblot and antibodies: The following antibodies were obtained from Cell Signaling Technology (Danvers, MA, USA) : EGFR (#2232), pEGFR Y1068 (#2234), HER3 (#4754), pHER3 Y1197 (#4561), pBRAF S445 (#2696), MEK (#9126), pMEK S217/221 (#9154), ERK (#9102), pERK T202/Y204 (#9101), PDK1 (#3062), pPDK1 S241 (#3061), AKT (#9272), pAKT S473 (#9271), pAKT T308 (#9275), ribosomal protein S6 (#2317), pS6 S240/S244 (#5364), FAK (#3285), pFAK Y397 (#8556), SFKs (#2108), pSFKs Y416 (#2101), YES (#3201), Integrin 4

Antibody Sampler Kit (#4749), PARP (#9542), E-cadherin (#3195), Vimentin (#3932), horseradish peroxidase (HRP)-conjugated anti-mouse (#7076) and HRP-conjugated anti-rabbit (#7074). The actin antibody (A2066) was purchased from Sigma-Aldrich. The BRAF antibody (sc-55522) was purchased from Santa Cruz Biotechnology (Dallas, TX, USA). For immunoblot, cells were harvested, washed in PBS, and lysed in RIPA buffer [50 mM Tris⋅HCl (pH 8.0), 150 mM sodium chloride, 5 mM magnesium chloride, 1% Triton X-100. 0.5% sodium deoxycholate, 0.1% SDS, 40 mM sodium fluoride, 1 mM sodium orthovanadate, and complete protease inhibitors (Roche Diagnostics, Indianapolis, IN, USA)]. Western Lightning ECL reagent (Perkin-Elmer) as used for signal detection. Phosphorylated bands were quantified using ImageJ software.

Xenografts: Five million cells were subcutaneously injected into the lower flank of athymic female mice. Treatment began 1 week following injection. Mice were randomized and dosed via oral gavage with osimertinib (5 mg/kg), dasatinib (15 mg/kg), or osimertinib (5 mg/kg) plus dasatinib (15 mg/kg). Mice were observed daily throughout the treatment period for signs of morbidity/mortality. Tumors were measured twice weekly using calipers, and volume was calculated using the formula length × width2 × 0.52. Body weight was assessed twice weekly. Mice were cared for in accordance with guidelines approved by the MSKCC Institutional Animal Care and Use Committee and Research Animal Resource Center.

siRNA experiments: siHER3, siSRC, siYES1 and control oligos were used at a concentration of 10 nM. Transfections were completed with DharmaFECT4 (Dharmacon) according to the manufacturer's protocol.

Expression constructs and transfections: Yes1 plasmid (#20665) was purchased from Addgene (Cambridge, MA, USA). The pWZL-Neo/Yes1 vector and the pVSV-G vector (Clontech, Mountain View, CA, USA) for production of the viral envelope were introduced into GP2-293 cells (80% confluence in a 10-cm dish) using Lipofectamine 2000 (Invitrogen). After 48 hours, viral particles released into the culture medium were concentrated by centrifugation at 15,000 × g for 3 hours at 4°C. The resulting pellet was then suspended in fresh RPMI-1640 medium and used to infect PC-9/BRc1 cells as previously described (20).

Vitronectin and fibronectin coating: Vitronectin and fibronectin (Sigma-Aldrich) were diluted in PBS and used to coat the culture surface with a minimal volume at 0.3 μg/cm2 and 3 μg/cm2, respectively. Plates were allowed to air dry for 1 hour at room temperature. Excess vitronectin or fibronectin was removed by aspiration.

Viability and proliferation assays: Cell viability was measured using the Cell Titer Blue assay (Promega, Madison, WI, USA). Cells were seeded in 96-well plates at a density of 3,000 cells per well and drug treated on the following day. Cell Titer Blue reagent was added 72 hours after drug addition, and fluorescence at 570 nm was measured on a Synergy MX microplate reader (Biotek, Winooski, VT, USA). Cell proliferation was assayed by cell counting experiments. Cells were seeded in 6-well or 12-well plates on day 0, and treatment was started on day 1. Cell numbers were counted using a Coulter Counter (Beckman Coulter, Brea, CA, USA).

Statistical analysis: Difference in cell viability and comparison of xenograft tumor volumes was performed using Student’s t-test in either R version 3.2 statistics package (R Core Team, 2015) or GraphPad Prism 5 (GraphPad Software, La Jolla, CA, USA). Fisher’s exact test in R version 3.2 statistics package was used to analyze whether different therapies influenced the likelihood of tumor recurrence in an in vivo model. A p-value of ≤ 0.05 was considered significant for all analyses. Power analysis to determine appropriate sample size for in vivo work was done with the following parameters: α = 0.05, power = 0.8.

Results

Kinome-wide siRNA screen identifies rational targets for combination therapy with osimertinib

To identify kinases that attenuate the effects of osimertinib in EGFR-mutant lung cancer, we performed a kinome-wide siRNA screen using PC-9/BRc1 (EGFR exon19 deletion/T790M) cells (18) (Table S1, Fig. S1A), and found 31 siRNAs that sensitized to osimertinib (Table S2). Among the top 10 hits, we identified MAPK1 (encodes ERK2), BRAF, and PDPK1 (encodes PDK1) as attenuating factors of osimertinib treatment (Fig. 1A, Table S2). In addition, PIK3C2A and PIK3CB, both of which encode subunits of phosphoinositide 3-kinase (PI3K), were among the top 30 hits. PI3K and PDK1 are components of the AKT pathway, while BRAF and ERK2 are components of the MAPK pathway. These data suggest that signaling downstream of EGFR, through the AKT and MAPK pathways, remains active and may be independently contributing to proliferation/survival even in the setting of continuous EGFR inhibition.

Based on these data, we examined changes in the AKT and MAPK pathways after osimertinib treatment, focusing on phosphorylation of key signaling kinases (Fig. 1B). EGFR phosphorylation was completely inhibited by osimertinib, while phosphorylation of AKT pathway components, including PDK1, AKT, and S6, was only partially blocked. Phosphorylation of BRAF, MEK, and ERK - components of the MAPK pathway - also persisted or became reactivated following osimertinib treatment. Similar results were observed in other EGFR-mutant NSCLC cell lines (Figs. 1C, S1B) and with other EGFR TKIs, including erlotinib and rociletinib (Figs. S1C, D). These observations suggest that downstream signaling through the AKT and MAPK pathways is still partially active in the presence of continuous EGFR inhibition.

Next, we determined if co-targeting EGFR and downstream pathways would result in a more complete inhibition of signaling and a greater effect on growth inhibition. Co-treatment of PC-9/BRc1 cells with osimertinib and either the pan-PI3K inhibitor, BKM120, or the PDK1 inhibitor, GSK2334470, inhibited AKT phosphorylation (with no further effect on ERK phosphorylation) compared to osimertinib alone (Fig. 1D). Osimertinib plus BKM120 or GSK2334470 was also associated with greater inhibition of proliferation compared to osimertinib alone (Fig. 1E). We examined vertical inhibition of the MAPK pathway by co-treating PC-9/BRc1 cells with osimertinib and either a pan-RAF inhibitor, TAK632, or an allosteric MEK inhibitor, selumetinib. Neither TAK632 nor selumetinib could completely inhibit ERK phosphorylation at any time point (Fig. S2A, B). However, combining either agent with 6

osimertinib completely inhibited ERK phosphorylation (Fig. 1F) and more effectively blocked cell proliferation compared with any agent alone (Fig. 1G). These data suggest that addition of AKT or MAPK pathway inhibitors to EGFR-targeting therapies represent rational combination therapeutic strategies for EGFR-mutant lung cancer.

SFKs sustain MAPK pathway signaling in the presence of EGFR inhibition

We next explored potential mechanisms that could explain sustained signaling through the AKT and MAPK pathways in the presence of osimertinib. To exclude the possibility of insufficient drug levels, we examined the effects of higher doses of osimertinib. However, increased osimertinib concentrations showed no further inhibition of EGFR, AKT, or ERK phosphorylation and did not abrogate pathway reactivation at 96 hours (Fig. 2A). In addition, maximum loss of viability in PC-9/BRc1 cells was reached at 100nM osimertinib (Fig. 2B). These data suggest that sustained AKT and MAPK signaling is not due to insufficient EGFR inhibition.

We sought to better characterize the population of cells that remained viable after 96 hour exposure to osimertinib. To assess whether osimertinib inhibited cell growth less potently in the remaining population, we generated concentration-response curves using PC-9/BRc1 cells that were pre-treated with either 100nM osimertinib or vehicle for 96 hours. We found no difference in EC50 for DMSO pre-treated cells compared to osimertinib pre-treated cells (3.46 vs. 3.44 nM, p = 0.96) (Fig. S2C). Likewise, sequencing of EGFR exons 19, 20, and 21 after 96 hour osimertinib exposure did not reveal any new mutations within the EGFR kinase domain in any of five cell lines tested (data not shown). Taken together, these data suggest that osimertinib treatment is not rapidly selecting for clones with primary osimertinib resistance. Alternatively, to explore potential ‘bypass’ signaling pathways, we profiled lysates from PC-9/BRc1 cells treated with osimertinib using a receptor tyrosine kinase array and found that phosphorylation of human epidermal growth factor receptor 3 (HER3) increased after drug treatment (Figs. S3A, B). However, HER3 knockdown had no effect on AKT or ERK phosphorylation (Fig. S3C), suggesting that HER3 does not act as a ‘bypass’ signal following osimertinib treatment in these cell models.

Next, we focused on a potential role for Src family kinases (SFKs), as SFKs are known upstream regulators of the AKT and MAPK pathways (21). Interestingly, immunoblot analysis revealed increased phosphorylation of SFKs after osimertinib treatment (Fig. 2C, 3A). Treatment with PP2, a selective SFK inhibitor (22), or dasatinib, a multi-kinase inhibitor that targets SFKs, attenuated SFK activation in the presence of osimertinib (Fig. 2C, lanes 3 and 4 vs. lane 2). Notably, PP2 or dasatinib treatment also led to more profound inhibition of ERK phosphorylation compared to osimertinib alone (Fig. 2C, lanes 3 and 4 vs. lane 2), suggesting that activity of the MAPK pathway is sustained by SFKs in the absence of EGFR signaling. Furthermore, co-treatment of PC-9/BRc1 cells with PP2 or dasatinib enhanced growth-inhibitory effects compared to osimertinib monotherapy (Fig. 2D). Combining osimertinib with bosutinib or saracatinib, two other clinically relevant TKIs with anti-SFK activity (23, 24), also resulted in improved cell growth inhibition compared to osimertinib alone (Figs. S4A-D).

Concomitant inhibition of FAK and SFKs is necessary to sustain MAPK and AKT pathway inhibition in the presence of EGFR blockade and to achieve greater inhibition of proliferation

PP2, bosutinib, and saracatinib had minimal effect on AKT phosphorylation. Dasatinib, however, was better able to inhibit AKT phosphorylation (Fig. 2C, lane 3 vs. lane 4; Fig. S4C). Moreover, dasatinib was consistently more efficacious than PP2, bosutinib, and saracatinib at increasing apoptosis (determined by expression of cleaved PARP) and inhibiting cell proliferation (Figs. 2C, 2D, S4C, S4D). These data suggest that there are additional factors that attenuate the effects of osimertinib through incomplete AKT inhibition, and that these factors can be inhibited by dasatinib.

Focal adhesion kinase (FAK) is known to interact with SFKs. Notably, FAK activity can be inhibited by dasatinib, likely through an indirect mechanism (25). Analogous to SFKs, we also noticed that autophosphorylation of FAK increased after osimertinib treatment (Fig. S5A). Therefore, we examined the expression of FAK and SFKs in multiple EGFR-mutant lung cancer cells by immunoblot (Figs. 3A, S5B) and found that SFK and/or FAK phosphorylation was increased to varying extents after osimertinib treatment in all cell lines evaluated.

We next determined if PF573228, a selective, ATP-competitive inhibitor of FAK (26) (Fig. S5C), could influence sensitivity of EGFR-mutant cells to osimertinib. PC-9/BRc1 cells were treated with osimertinib alone or in combination with PF573228 and/or PP2. Addition of PF573228 abrogated the osimertinib-induced increase in FAK phosphorylation (Fig. 3B, lane 3 vs. lane 2). Osimertinib combined with the SFK inhibitor, PP2, inhibited phosphorylation of SFKs and ERK (Fig. 3B, lane 4 vs. lane 2). Osimertinib/PF573228 and osimertinib/PP2 combinations both had modest effects on phosphorylated AKT or cleaved PARP. However, the triple combination of osimertinib, PF573228, and PP2 completely inhibited phosphorylation of FAK and SFKs, leading to complete inhibition of AKT, increased cleaved PARP (Fig. 3B, lane 5 vs. lanes 3 and 4), and enhanced growth inhibitory effects (Fig. 3C). Similar results were observed for triple combinations that included bosutinib or saracatinib (Figs. S5D). Osimertinib combined with dasatinib, which inhibits SFKs and can indirectly inhibit FAK activity (25, 27), showed effects similar to the osimertinib/FAK inhibitor/SFK inhibitor triple combination (Fig. 3B, lane 6 vs. lane 5; Figs. S5E). These data suggest that activation of both SFKs and FAK plays a role in attenuating the effects of osimertinib, and that inhibiting both SFKs and FAK leads to enhanced growth inhibitory effects in EGFR-mutant lung cancer cells.

The combinatorial effect of dasatinib with osimertinib was also examined in PC-9/BRc1 xenografts (Figs. 3D, S6A, S6B). There was no significant body weight loss in either group (Fig. S6A). Both osimertinib monotherapy and osimertinib combined with dasatinib inhibited tumor growth to almost undetectable levels (Fig. S6B). Five weeks after treatment cessation, 3 out of 7 mice in the osimertinib + dasatinib treatment arm showed no tumor regrowth (“tumor cure”), while there were no tumor cures in the osimertinib monotherapy arm (Fig. 3D), though the difference in tumor regrowth cases between the two arms was not statistically significant (p = 0.077).

Given that integrins are reported to activate SFK/FAK signaling (28), we examined the expression of several integrins in the absence and presence of osimertinib treatment. Expression of integrins αv, β3 and β5 was increased after treatment with osimertinib (Fig. 3E). Moreover, after 24 hours of treatment with osimertinib plus SB273005, an antagonist of integrin αvβ3 and αvβ5, we observed greater inhibition of AKT and ERK compared to osimertinib alone (Fig. 3F, lane 3 vs. lane 4). Next, since others had previously reported that cell-cell contacts influence signaling along the integrin-FAK pathway (29, 30) we investigated whether cells plated at higher density responded differently to osimertinib compared to cells plated at lower density. Interestingly, we observed an increase in SFK and FAK activity following osimertinib exposure that was more pronounced in the high density group (Fig. S7A).These effects are consistent with the hypothesis that integrins are activating downstream signaling through SFK/FAK. To further investigate the necessity of integrins toward SFK/FAK-mediated AKT and MAPK signaling, we determined the effect of the integrin ligand, vitronectin, on integrin complexes αvβ3 and αvβ5. In this experiment, PC-9/BRc1 cells were cultured in serum-free media since serum contains a large amount of extracellular matrix components, including vitronectin. Consistent with our hypothesis, cell culture dishes coated with vitronectin attenuated the effects of osimertinib, whereas fibronectin coating did not (Fig. S7B). Collectively, these data suggest that integrins activate FAK/SFKs, thereby attenuating the effects of osimertinib.

Based on these findings, the potential mechanism whereby SFK/FAK signaling attenuates the effects of osimertinib in EGFR-mutant cells is shown schematically in Fig. S7C. Under continuous inhibition of EGFR signaling by osimertinib, integrins can activate SFK/FAK, leading to sustained AKT and ERK phosphorylation. AKT or MAPK pathway inhibitors enhance the effects of osimertinib by directly inhibiting the respective downstream signaling, whereas dasatinib enhances the effects of osimertinib by inhibiting SFK/FAK, leading to simultaneous inhibition of both the AKT and MAPK pathways.

Osimertinib plus dasatinib combination therapy is superior to osimertinib in combination with either AKT or MAPK pathway inhibitors

To determine the optimal combination therapeutic strategy, we compared the growth inhibitory effects of BKM120 (PI3K inhibitor), selumetinib (MEK inhibitor), or dasatinib (SFK inhibitor) with or without osimertinib in multiple EGFR-mutant lung cancer cell lines. Monotherapy with each drug inhibited tumor cell proliferation to varying degrees (Fig. S8A). Addition of BKM120, selumetinib, or dasatinib enhanced the effects of osimertinib (Fig. 4A). However, the dasatinib-osimertinib combination was consistently the most effective at inhibiting tumor cell proliferation. We also determined the effects of the various combination therapies on apoptosis (Figs. 4B, S8B). The selumetinib combination only modestly increase cleaved PARP in each of the cell lines evaluated, which is consistent with previously published data (31). In contrast, the greatest induction of cleaved PARP was consistently observed with the dasatinib combination. Moreover, the dasatinib combination was the most effective at delaying the emergence of resistant clones (Figs. 4C, S8C, S8D).

Amplification of YES1, a SFK, mediates acquired resistance to osimertinib 9

Finally, we explored the role of SFK/FAK signaling in the setting of acquired resistance to osimertinib. For this purpose, we established five osimertinib (AZD9291) resistant cell lines (Table S1) using a dose escalation method (18). Two of the five cell lines (H1975/9291 and SH450/9291) exhibited an epithelial to mesenchymal transition phenotype denoted by loss of E-cadherin and gain of vimentin expression (Fig. S9A). In PC-9/BRc1/9291 cells, whole exome sequencing revealed focal amplification (15.7 gene copy number) of YES1 on chromosome 18, which encodes the SFK, YES1. Immunoblot analysis showed considerably increased expression of YES1 protein in PC-9/BRc1/9291 cells compared to the ‘parental’ PC-9/BRc1 cells (Fig. 5A). Treatment of PC-9/BRc1/9291 cells with osimertinib alone inhibited EGFR phosphorylation, but did not inhibit downstream ERK, AKT or S6 phosphorylation (Fig. 5B). Combination therapy with osimertinib and the SFK inhibitors, PP2 or dasatinib, inhibited both EGFR and downstream effector signaling. Combination therapy also resulted in increased apoptosis as shown by cleavage of PARP. Importantly, pharmacologic or genetic inhibition of YES1 restored sensitivity to osimertinib in PC-9/BRc1/9291 cells (Figs. 5C, D). Moreover, ectopic expression of YES1 rendered the PC-9/BRc1 ‘parental’ cells (from which the PC-9/BRc1/9291 cells were derived) less sensitive to osimertinib (Fig. 5E). Consistent with these results, dasatinib combined with osimertinib decreased the rate of tumor growth in a PC-9/BRc1/9291 xenograft model (Fig. 5F). Immunoblotting confirmed SFK phosphorylation was inhibited by the osimertinib/dasatinib combination (Fig. S9B). There was no significant body weight loss in any group (Fig. S9C).

Discussion

In this study, we sought to determine the most effective rational drug combination therapy to delay or prevent acquired resistance to osimertinib. Among the top hits in our kinome-wide siRNA screen were components of the MAPK and AKT pathways. Our data reveal that signaling through these pathways is sustained or reactivated as early as 24-48 hours after addition of osimertinib, thereby attenuating the growth inhibitory effects of this agent. Recent studies have also demonstrated that MAPK signaling induces acquired resistance to third-generation EGFR TKIs, including WZ4002 (32) and osimertinib (33). Consistent with our work herein, these studies demonstrated that the combination of WZ4002 with the MEK inhibitor, trametinib, showed more profound inhibition of tumor cell growth (31). Additionally, we found that AKT pathway signaling is sustained in the presence of persistent EGFR inhibition, and that concomitant inhibition of EGFR and AKT pathway enhanced the effects of osimertinib. Interestingly, the efficacy was equivalent or better than that of EGFR plus MEK inhibition (Figs. 4A-C, S8A). In the previous studies, AKT pathway reactivation was reported to be a mechanism of acquired resistance to WZ4002 and trametinib combination therapy (31), suggesting that this pathway may play an important role in cancer cell survival under continuous EGFR inhibition.

Although the effects of simultaneous EGFR plus MAPK or AKT pathway inhibition are promising, it is unknown how these pathways are sustained or reactivated in the presence of continuous EGFR inhibition. A detailed mechanistic understanding of how and why EGFR-mutant cells have early downstream signaling reactivation is expected to result in improved therapeutic 10

strategies. To this end, we found that phosphorylation of SFKs and FAK increased after osimertinib treatment.

SFKs and FAK are known to interact with each other and mediate AKT and MAPK pathway signaling (21).Targeting SFKs with dasatinib enhanced the effects of osimertinib, resulting in both a more pronounced attenuation of effector signaling and a greater inhibition of cell growth. SFKs and FAK were also found to be activated through integrin signaling. Interestingly, integrin αvβ3 has been associated with resistance to EGFR TKI treatment through forming integrin-KRAS complex (34), and integrin-targeted therapies are currently being tested in the clinic (35, 36).

Others have previously reported on the potential utility of SFK-targeting drugs in oncogene-driven NSCLC. Crystal et al. identified Src as a common mediator of ALK TKI resistance in cell cultures established from patients who had progressed on ALK TKI therapy (37). Likewise, Yoshida et al. reported that dasatinib could enhance the anti-tumor activity of irreversible EGFR TKIs, including the mutant-selective WZ4002, in cell lines with T790M-mediated resistance to erlotinib and gefitinib (38). In addition, previous work using models of other solid tumors, such as breast and thyroid cancer, have shown that dasatinib enhances the effects of targeted therapies via inhibition of SFK/FAK signaling (39, 40). Taken together, these studies support the conclusion that SFKs and FAK are common mediators of resistance to oncogene-targeting therapies across multiple tumor types.

The combination of first-generation EGFR TKI plus dasatinib was evaluated in a phase II clinical trial in patients with EGFR-mutant lung cancer and acquired resistance to erlotinib or gefitinib (41). No radiographic responses were observed in the 12 patients who received combination therapy, two-thirds of whom were positive for T790M. However, we would not expect erlotinib to be sufficient to inhibit EGFR signaling in either T790M-induced acquired resistance or in non-T790M bypass-induced acquired resistance. Considering that we observed no increase in apoptosis with dasatinib monotherapy (Fig. 4B), dasatinib appears to be most efficacious when coupled with potent EGFR inhibition. Interestingly, a phase I/II clinical trial of osimertinib plus dasatinib in EGFR TKI-naïve patients recently opened in the United States (NCT02954523).

Finally, we identified amplification of the SFK, YES1, as a potential mechanism of osimertinib resistance in vitro. Previous reports demonstrated that SFK signaling activation causes intrinsic or acquired resistance to EGFR TKIs in an in vitro model (42, 43). Furthermore, Src was reported to be a mechanism of intrinsic resistance to EGFR TKIs in EGFR-mutant NSCLC (43). Notably, YES1 amplification has been detected in patients with EGFR-mutant NSCLC at the time of EGFR TKI resistance (Fan PD, Ladanyi M, et al. manuscript in preparation). These data suggest that SFK signaling can attenuate the effects of EGFR TKIs in both EGFR TKI-sensitive and TKI-resistant tumor cells. Furthermore, these data suggest that multiple mechanisms may converge on SFK mediated signaling and resistance e.g. through upstream activation via integrins or genetic amplification events. This is analogous to MAPK pathway resistance in

which studies suggest multiple different events e.g. KRAS amplification, MAPK1 amplification, BRAF mutation can all converge of activation of same pathway (33).

In conclusion, we found that inhibiting SFK/FAK, AKT or MAPK signaling enhanced the effects of osimertinib. Furthermore, osimertinib plus dasatinib was the most efficacious combination in several models of EGFR-mutant lung cancer, suggesting that this combination may produce more durable effects than osimertinib alone in a clinical setting. Overall, our studies provide mechanistic insights into early downstream signaling reactivation in the setting of continuous EGFR inhibition and suggest rational combination therapies for clinical use. Our results should prompt further investigation of third-generation EGFR TKIs and SFK/FAK inhibitors as combination therapy in patients whose tumors harbor EGFR mutations.

Acknowledgements We are grateful to Miika Ahdesmaki and Henry Brown for their assistance in obtaining NGS data and to Katie Hutchinson and Holli Loomans for critical review of the manuscript.

References

1. Lynch TJ, Bell DW, Sordella R, Gurubhagavatula S, Okimoto RA, Brannigan BW, et al. Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to gefitinib. N Engl J Med. 2004;350(21):2129-39. 2. Paez JG, Janne PA, Lee JC, Tracy S, Greulich H, Gabriel S, et al. EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science. 2004;304(5676):1497-500. 3. Pao W, Miller V, Zakowski M, Doherty J, Politi K, Sarkaria I, et al. EGF receptor gene mutations are common in lung cancers from "never smokers" and are associated with sensitivity of tumors to gefitinib and erlotinib. Proc Natl Acad Sci U S A. 2004;101(36):13306-11. 4. Maemondo M, Inoue A, Kobayashi K, Sugawara S, Oizumi S, Isobe H, et al. Gefitinib or chemotherapy for non-small-cell lung cancer with mutated EGFR. N Engl J Med. 2010;362(25):2380-8. 5. Mitsudomi T, Morita S, Yatabe Y, Negoro S, Okamoto I, Tsurutani J, et al. Gefitinib versus cisplatin plus docetaxel in patients with non-small-cell lung cancer harbouring mutations of the epidermal growth factor receptor (WJTOG3405): an open label, randomised phase 3 trial. Lancet Oncol. 2010;11(2):121-8. 6. Rosell R, Carcereny E, Gervais R, Vergnenegre A, Massuti B, Felip E, et al. Erlotinib versus standard chemotherapy as first-line treatment for European patients with advanced EGFR mutation-positive non-small-cell lung cancer (EURTAC): a multicentre, open-label, randomised phase 3 trial. Lancet Oncol. 2012;13(3):239-46. 7. Yang JC, Wu YL, Schuler M, Sebastian M, Popat S, Yamamoto N, et al. Afatinib versus cisplatin-based chemotherapy for EGFR mutation-positive lung adenocarcinoma (LUX-Lung 3 and LUX-Lung 6): analysis of overall survival data from two randomised, phase 3 trials. Lancet Oncol. 2015;16(2):141-51. 8. Ohashi K, Maruvka YE, Michor F, Pao W. Epidermal growth factor receptor tyrosine kinase inhibitor-resistant disease. J Clin Oncol. 2013;31(8):1070-80. 9. Sequist LV, Waltman BA, Dias-Santagata D, Digumarthy S, Turke AB, Fidias P, et al. Genotypic and histological evolution of lung cancers acquiring resistance to EGFR inhibitors. Sci Transl Med. 2011;3(75):75ra26. 10. Yu HA, Riely GJ, Lovly CM. Therapeutic strategies utilized in the setting of acquired resistance to EGFR tyrosine kinase inhibitors. Clin Cancer Res. 2014;20(23):5898-907. 11. Cross DA, Ashton SE, Ghiorghiu S, Eberlein C, Nebhan CA, Spitzler PJ, et al. AZD9291, an irreversible EGFR TKI, overcomes T790M-mediated resistance to EGFR inhibitors in lung cancer. Cancer Discov. 2014;4(9):1046-61. 12. Tjin Tham Sjin R, Lee K, Walter AO, Dubrovskiy A, Sheets M, Martin TS, et al. In vitro and in vivo characterization of irreversible mutant-selective EGFR inhibitors that are wild-type sparing. Mol Cancer Ther. 2014;13(6):1468-79. 13. Walter AO, Sjin RT, Haringsma HJ, Ohashi K, Sun J, Lee K, et al. Discovery of a mutant-selective covalent inhibitor of EGFR that overcomes T790M-mediated resistance in NSCLC. Cancer Discov. 2013;3(12):1404-15. 14. Zhou W, Ercan D, Chen L, Yun CH, Li D, Capelletti M, et al. Novel mutant-selective EGFR kinase inhibitors against EGFR T790M. Nature. 2009;462(7276):1070-4. 13

15. Janne PA, Yang JC, Kim DW, Planchard D, Ohe Y, Ramalingam SS, et al. AZD9291 in EGFR inhibitor-resistant non-small-cell lung cancer. N Engl J Med. 2015;372(18):1689-99. 16. Bauer JA, Ye F, Marshall CB, Lehmann BD, Pendleton CS, Shyr Y, et al. RNA interference (RNAi) screening approach identifies agents that enhance paclitaxel activity in breast cancer cells. Breast Cancer Res. 2010;12(3):R41. 17. Gong Y, Yao E, Shen R, Goel A, Arcila M, Teruya-Feldstein J, et al. High expression levels of total IGF-1R and sensitivity of NSCLC cells in vitro to an anti-IGF-1R antibody (R1507). PloS one. 2009;4(10):e7273. 18. Chmielecki J, Foo J, Oxnard GR, Hutchinson K, Ohashi K, Somwar R, et al. Optimization of dosing for EGFR-mutant non-small cell lung cancer with evolutionary cancer modeling. Sci Transl Med. 2011;3(90):90ra59. 19. Ohashi K, Sequist LV, Arcila ME, Moran T, Chmielecki J, Lin YL, et al. Lung cancers with acquired resistance to EGFR inhibitors occasionally harbor BRAF gene mutations but lack mutations in KRAS, NRAS, or MEK1. Proc Natl Acad Sci U S A. 2012;109(31):E2127-33. 20. Takezawa K, Pirazzoli V, Arcila ME, Nebhan CA, Song X, de Stanchina E, et al. HER2 amplification: a potential mechanism of acquired resistance to EGFR inhibition in EGFR-mutant lung cancers that lack the second-site EGFRT790M mutation. Cancer discovery. 2012;2(10):922-33. 21. Kim LC, Song L, Haura EB. Src kinases as therapeutic targets for cancer. Nat Rev Clin Oncol. 2009;6(10):587-95. 22. Hanke JH, Gardner JP, Dow RL, Changelian PS, Brissette WH, Weringer EJ, et al. Discovery of a novel, potent, and Src family-selective tyrosine kinase inhibitor. Study of Lck- and FynT-dependent T cell activation. J Biol Chem. 1996;271(2):695-701. 23. Golas JM, Arndt K, Etienne C, Lucas J, Nardin D, Gibbons J, et al. SKI-606, a 4-anilino-3-quinolinecarbonitrile dual inhibitor of Src and Abl kinases, is a potent antiproliferative agent against chronic myelogenous leukemia cells in culture and causes regression of K562 xenografts in nude mice. Cancer research. 2003;63(2):375-81. 24. Tabernero J, Cervantes A, Hoekman K, Hurwitz H, Jodrell D, Hamberg P, et al. Phase I study of AZD0530, an oral potent inhibitor of Src kinase: first demonstration of inhibition of Src activity in human cancers. Journal of Clinical Oncology. 2007;25(90180):3520-. 25. Davis MI, Hunt JP, Herrgard S, Ciceri P, Wodicka LM, Pallares G, et al. Comprehensive analysis of kinase inhibitor selectivity. Nat Biotechnol. 2011;29(11):1046-51. 26. Slack-Davis JK, Martin KH, Tilghman RW, Iwanicki M, Ung EJ, Autry C, et al. Cellular characterization of a novel focal adhesion kinase inhibitor. J Biol Chem. 2007;282(20):14845-52. 27. Chang AY, Wang M. Molecular mechanisms of action and potential biomarkers of growth inhibition of dasatinib (BMS-354825) on hepatocellular carcinoma cells. BMC Cancer. 2013;13:267. 28. Guo W, Giancotti FG. Integrin signalling during tumour progression. Nat Rev Mol Cell Biol. 2004;5(10):816-26. 29. Yang G-Y, Guo S, Dong C-Y, Wang X-Q, Hu B-Y, Liu Y-F, et al. Integrin αvβ6 sustains and promotes tumor invasive growth in colon cancer progression. World Journal of Gastroenterology: WJG. 2015;21(24):7457.

30. Playford MP, Vadali K, Cai X, Burridge K, Schaller MD. Focal Adhesion Kinase regulates cell–cell contact formation in epithelial cells via modulation of Rho. Experimental cell research. 2008;314(17):3187-97. 31. Tricker EM, Xu C, Uddin S, Capelletti M, Ercan D, Ogino A, et al. Combined EGFR/MEK Inhibition Prevents the Emergence of Resistance in EGFR-Mutant Lung Cancer. Cancer Discov. 2015;5(9):960-71. 32. Ercan D, Xu C, Yanagita M, Monast CS, Pratilas CA, Montero J, et al. Reactivation of ERK signaling causes resistance to EGFR kinase inhibitors. Cancer Discov. 2012;2(10):934-47. 33. Eberlein CA, Stetson D, Markovets AA, Al-Kadhimi KJ, Lai Z, Fisher PR, et al. Acquired Resistance to the Mutant-Selective EGFR Inhibitor AZD9291 Is Associated with Increased Dependence on RAS Signaling in Preclinical Models. Cancer Res. 2015;75(12):2489-500. 34. Seguin L, Kato S, Franovic A, Camargo MF, Lesperance J, Elliott KC, et al. An integrin beta(3)-KRAS-RalB complex drives tumour stemness and resistance to EGFR inhibition. Nat Cell Biol. 2014;16(5):457-68. 35. Desgrosellier JS, Cheresh DA. Integrins in cancer: biological implications and therapeutic opportunities. Nature reviews Cancer. 2010;10(1):9-22. 36. Goodman SL, Picard M. Integrins as therapeutic targets. Trends in pharmacological sciences. 2012;33(7):405-12. 37. Crystal AS, Shaw AT, Sequist LV, Friboulet L, Niederst MJ, Lockerman EL, et al. Patient-derived models of acquired resistance can identify effective drug combinations for cancer. Science. 2014;346(6216):1480-6. 38. Yoshida T, Zhang G, Smith MA, Lopez AS, Bai Y, Li J, et al. Tyrosine phosphoproteomics identifies both codrivers and cotargeting strategies for T790M-related EGFR-TKI resistance in non–small cell lung cancer. Clinical Cancer Research. 2014;20(15):4059-74. 39. Caccia D, Micciche F, Cassinelli G, Mondellini P, Casalini P, Bongarzone I. Dasatinib reduces FAK phosphorylation increasing the effects of RPI-1 inhibition in a RET/PTC1-expressing cell line. Mol Cancer. 2010;9:278. 40. Seoane S, Montero JC, Ocana A, Pandiella A. Effect of multikinase inhibitors on caspase-independent cell death and DNA damage in HER2-overexpressing breast cancer cells. J Natl Cancer Inst. 2010;102(18):1432-46. 41. Johnson ML, Riely GJ, Rizvi NA, Azzoli CG, Kris MG, Sima CS, et al. Phase II trial of dasatinib for patients with acquired resistance to treatment with the epidermal growth factor receptor tyrosine kinase inhibitors erlotinib or gefitinib. Journal of thoracic oncology : official publication of the International Association for the Study of Lung Cancer. 2011;6(6):1128-31. 42. Ochi N, Takigawa N, Harada D, Yasugi M, Ichihara E, Hotta K, et al. Src mediates ERK reactivation in gefitinib resistance in non-small cell lung cancer. Exp Cell Res. 2014;322(1):168-77. 43. Park KS, Raffeld M, Moon YW, Xi L, Bianco C, Pham T, et al. CRIPTO1 expression in EGFR-mutant NSCLC elicits intrinsic EGFR-inhibitor resistance. J Clin Invest. 2014;124(7):3003-15.

Figure Legends

Figure 1. Sustained AKT and MAPK pathway signaling following osimertinib treatment in EGFR-mutant lung cancer cells. (A) PC-9/BRc1 cells were reverse transfected with siRNAs targeting 714 kinases and kinase-related genes (4 individual siRNAs per gene) for 24 hours. Transfected cells were then treated with DMSO or osimertinib (5 nM), and cell viability was measured following 72 hours of drug treatment. The average of sensitivity index (SI) of the 4 individual siRNAs was calculated for each gene using the formula described in the Materials and Methods. Dots represent the SI value of each siRNA. Positive SI value indicates a sensitizing effect. siRNAs were plotted in alphabetical order. (B) PC-9/BRc1 cells were treated with 100 nM of osimertinib and harvested at the indicated time points. Drug was refreshed every 24 hours. Cellular lysates were probed with the indicated antibodies. (C) PC-9, HCC827, and H1975 cells were treated with 100 nM of osimertinib and harvested at the indicated time points. Drug was refreshed every 24 hours. Cellular lysates were probed with the indicated antibodies. (D) PC-9/BRc1 cells were treated with 100 nM osimertinib, 1 μM BKM120, 3 μM GSK2334470, or with combinations of these agents. Drugs were refreshed every 24 hours. Cellular lysates were probed with the indicated antibodies. Osim; osimertinib, BKM; BKM120, GSK; GSK2334470 (E) PC-9/BRc1 cells were treated with 100 nM osimertinib, 1 μM BKM120, 3 μM GSK2334470, or with combinations of these agents for 7 days. Drugs were refreshed every 24-48 hours. Viable cells were counted. Data are represented as % cell number relative to DMSO control. Bars indicate standard deviation (SD). *p<0.05 (Student’s t-test). Osim; osimertinib, BKM; BKM120, GSK; GSK2334470. (F) PC-9/BRc1 cells were treated with 100 nM osimertinib, 3 μM TAK632, 1 μM selumetinib, or with combinations of these agents. Drugs were refreshed every 24 hours. Cellular lysates were probed with the indicated antibodies. Osim; osimertinib, TAK; TAK632, Sel; selumetinib (G) PC-9/BRc1 cells were treated with 100 nM osimertinib, 3 μM TAK632, 1 μM selumetinib, or with combinations of these agents for 7 days. Drugs were refreshed every 24-48 hours. Viable cells were counted. Data are represented as % cell number relative to DMSO only control. Bars indicate SD. *p<0.05 (Student’s t-test). Osim; osimertinib, TAK; TAK632, Sel; selumetinib.

Figure 2. SFKs reactivate the MAPK pathway under continuous EGFR inhibition. (A) PC-9/BRc1 cells were treated with the indicated doses of osimertinib. Drugs were refreshed every 24 hours. Cellular lysates were probed with the indicated antibodies. Osim: osimertinib. (B) PC-9/BRc1 cells were treated with increasing amounts of osimertinib for 72 hours. Cell titer blue assays were performed to assess cell viability. Each point represents six replicates. Data are presented as the mean percentage of viable cells compared to vehicle control ± SD. (C) PC-9/BRc1 cells were treated with 100 nM osimertinib alone or in combination with 3 μM of PP2 or 100 nM dasatinib for 72 hours. Drug was refreshed every 24 hours. Cellular lysates were probed with the indicated antibodies. p-SFK signal was quantified using ImageJ software.

There are multiple bands for SFKs because the SFK antibody recognizes all the SFK members. Osim: osimertinib (D) PC-9/BRc1 cells were treated 100 nM osimertinib, 1 μM PP2, 100 nM Dasatinib, or with combinations of these agents for 7 days. Drugs were refreshed every 24-48 hours. Viable cells were counted. Data are represented as % cell number relative to DMSO only control. Bars indicate SD. *p<0.05 (Student’s t-test). Osim: osimertinib, Da: dasatinib.

Figure 3. SFK/FAK sustains the AKT and MAPK pathways in the absence of EGFR signaling. (A) PC-9/BRc1 cells were treated with 100 nM of osimertinib. Drug was refreshed every 24 hours. Cellular lysates were probed with the indicated antibodies. (B) PC-9/BRc1 cells were treated with 100 nM osimertinib alone or in combination with 3μM PF573228, 3 μM PP2, or 100 nM dasatinib. Drug was refreshed every 24 hours. Cellular lysates were probed with the indicated antibodies. Osim: osimertinib, PF: PF573228, Da: dasatinib. (C) PC-9/BRc1 cells were treated with the indicated drugs for 7 days and viable cells were counted. Each drug was refreshed every 3 or 4 days. Bars indicate SD. *p<0.05 (Student’s t-test). Osim: osimertinib, 100 nM; PF: PF573228, 3 µM; PP2: PP2, 1 µM. p-SFKs were quantified using ImageJ software. (D) Athymic nude mice with PC-9/BRc1 tumors were treated with osimertinib (5 mg/kg) or osimertinib (5 mg/kg) plus dasatinib (15 mg/kg) for 6 weeks followed by treatment cessation. The number of tumor cures, defined as no detectable tumor, was determined after 5 weeks of treatment cessation. Bars represent tumor volumes after 5 weeks of treatment cessation. Osim: osimertinib, Da: dasatinib. Difference in tumor regrowth was assessed with Fisher’s exact test. (E) PC-9/BRc1 cells were treated with 100 nM osimertinib for the length of time indicated, and lysates were harvested for immunoblot with the indicated antibodies. Drug was refreshed every 24 hours. Osim: osimertinib. (F) PC-9/BRc1 cells were treated with DMSO, 1 μM SB273005, 100 nM osimertinib, 1 μM SB273005 plus osimertinib for 96 hours. Drug was refreshed every 24 hours. Cellular lysates were probed with the indicated antibodies. Osim: osimertinib, SB: SB273005.

Figure 4. SFK/FAK pathway inhibition most potently enhances the effects of osimertinib. (A) EGFR-mutant cell lines were treated with the indicated drugs for 7 days and viable cells were counted. Drug was refreshed every 48-72 hours. Bars indicate SD. *p<0.05 (Student’s t-test). Osim: osimertinib 100 nM, BKM: BKM120 1 µM, Sel: selumetinib 1 µM. (B) PC-9/BRc1 cells were treated with the indicated drugs for 72 hours. Drug was refreshed every 24 hours. Cellular lysates were probed with the indicated antibodies. Osim: osimertinib, 100 nM; BKM: BKM120 1 µM, Sel: selumetinib 1 µM, Da: dasatinib 100 nM. (C) PC-9/BRc1 cells were treated with the indicated drugs for 30 days and then stained with crystal violet. Drugs were refreshed twice a week. Osim: osimertinib 100 nM; BKM: BKM120, 1 µM; Sel: selumetinib, 1 µM; Da: dasatinib, 100 nM.

Figure 5. YES1 overexpression contributes to osimertinib resistance.

(A) Lysates from PC-9/BRc1 and PC-9/BRc1/9291 cells were probed with the indicated antibodies. (B) PC-9/BRc1/9291 cells were treated with 100 nM osimertinib alone or in combination with 3 μM PP2 or 100 nM dasatinib for 72 hours. Drug was refreshed every 24 hours. Cellular lysates were probed with the indicated antibodies. Cleaved PARP was quantified using ImageJ software. Osim: osimertinib, Da: dasatinib. (C) PC-9/BRc1/9291 cells were treated with PC-9 increasing amounts of osimertinib in the presence or absence of 30 nM dasatinib for 72 hours. Cell titer blue assays were performed to assess cell viability. Each point represents six replicates. Data are presented as the mean percentage of viable cells compared to vehicle control ± SD. (D) PC-9/BRc1/9291 cells were transfected with non-targeting siRNA (siNT), siYES1, or siSRC. 24 hours post transfection, cells were plated in 96-well plates and drug treatment was started. Cell titer blue assays were performed at 72hours post drug addition to assess cell viability. Each point represents six replicates. Data are presented as the mean percentage of viable cells compared to vehicle control ± SD. (E) PC-9/BRc1 cells were transduced with YES1 or control virus and treated with 10 nM osimertinib for 72 hours. Cell viability was determined using the Cell Titer Blue assay. (F) Athymic mice were subcutaneously injected with PC-9/BRc1/9291 cells. 7 days post injection, mice were randomized to receive vehicle, osimertinib (5 mg/kg), dasatinib (15 mg/kg), or osimertinib (5 mg/kg) plus dasatinib (15 mg/kg) by oral gavage 5 days a week. Tumor volume was determined at the indicated times after the onset of treatment. Data represent the mean tumor volume ± SE. *p<0.05 (Student’s t-test).

A 0.5 B PC-9/BRc1 (ex19 del+T790M) BRAF PIK3C2A osimertinib Fig. 1 PDPK1 sensitize 0.3 Time (h) 0 8 24 48 72 96 MAPK1 PIK3CB p-EGFR 0.1 EGFR -0.1 p-PDK1 PDK1 -0.3 antagonize Sensitivity Index value p-S473-AKT -0.5 Kinase genes included in the library p-T308-AKT AKT C PC-9 HCC827 H1975 (ex19 del) (ex19 del) (L858R+T790M) p-S6 Time (h) 0 24 96 0 24 96 0 24 96 S6 p-S743-AKT p-BRAF AKT BRAF p-ERK p-MEK ERK MEK Actin p-ERK

24h 96h ERK

Actin

BKM PC-9/BRc1 Osim Osim DMSO Osim BKM GSK Osim

D GSK +

+ E 100 p-EGFR 90 EGFR 80 p-S473-AKT 70 * p-T308-AKT 30 * AKT 20

p-ERK 10 ERK 0 %cell number relative to DMSO Actin GSK BKM Osim Osim Osim PC-9/BRc1 +GSK+BKM

24h 96h PC-9/BRc1 90

G 80 TAK Sel Osim Osim DMSO Osim TAK Sel Osim + + p-EGFR 70 EGFR 60 40 p-S473-AKT 50 * AKT 30 * p-ERK 20 %cell number relative to DMSO ERK 10

Actin 0 DownloadedPC-9/BRc1 from cancerres.aacrjournals.org on SeptemberTAK Sel 28,Osim 2021.Osim © 2017Osim American Association for Cancer Research. +TAK +Sel Author Manuscript Published OnlineFirst on April 17, 2017; DOI: 10.1158/0008-5472.CAN-16-2300 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Osim Osim Osim Fig. 2 A 100nM 1μM 3μM B PC-9/BRc1 Time (h) 0 24 96 24 96 24 96 120 p-EGFR 100

EGFR 80 p-AKT 60 AKT 40 p-ERK %cell viability 20

ERK relative to DMSO control 0 Actin 0.0001 0.001 0.01 0.1 1 Osimertinib (μM) C Osim - + + + PP2 - - + - D PC-9/BRc1 Dasatinib - - - + 120 110

p-EGFR EGFR 100 90 p-SFKs 1.0 1.3 0.7 0.0 80 SFKs 70 60 p-ERK * 50 ERK 40 p-AKT 30 * AKT 20 %cell number relative to DMSO 10 PARP 0 DMSO PP2 Da Osim Osim Osim +PP2 +Da Actin Downloaded from cancerres.aacrjournals.org on September 28, 2021. © 2017 American Association for Cancer 1 2 3 4 Research. Author Manuscript Published OnlineFirst on April 17, 2017; DOI: 10.1158/0008-5472.CAN-16-2300 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Fig. 3

A B C 120 * Osim * Time (h) 0 24 96 100 DMSO Osim Osim+PF Osim+PP2 Osim+PF+PP2 Osim+Da

p-FAK p-EGFR 80 * alone FAK EGFR p-FAK 60 p-SFKs FAK 40 SFKs p-SFKs osimertinib

Cell number relative to 20 Actin SFKs p-AKT 0 Osim + + + + AKT PF - + - + D p = 0.077 p-ERK PP2 - - + + 500 ERK

)

3 PARP 400 SB Osim DMSO SB Osim + Actin p-EGFR 300 1 2 3 4 5 6 EGFR p-FAK 200 E Osim Time (h) Integrin: 0 24 96 FAK a4 p-SFKs Tumor volume (mm 100 a5 SFKs 0 av p-AKT 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 b1 AKT Osim Osim+Da b3 p-ERK Tumor cures: 0% (0/8) 43% (3/7) b4 ERK b5 Actin 1 2 3 4 Downloaded from cancerres.aacrjournals.orgActin on September 28, 2021. © 2017 American Association for Cancer Research. Author Manuscript Published OnlineFirst on April 17, 2017; DOI: 10.1158/0008-5472.CAN-16-2300 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

A 120 PC-9/BRc1 120 PC-9 120 H3255 Fig. 4

100 100 100 * 80 80 * 80 60 * 60 * 60 40 * 40 40 * 20 20 20

0 0 0 Osim Osim Osim Osim Osim Osim Osim Osim Osim Osim Osim Osim +Sel +BKM +Da +Sel +BKM +Da +Sel +BKM +Da 120 120 120 HCC827 11-18 H1975 100 100 100 * * 80 80 80 60 60 * 60 * * 40 40 * 40

%cell number relative to osimertinib monotherapy 20 20 20

0 0 0 Osim Osim Osim Osim Osim Osim Osim Osim Osim Osim Osim Osim +Sel +BKM +Da +Sel +BKM +Da +Sel +BKM +Da

B +DMSO +BKM +Sel +Da Osim - - - - + + + + C

DMSO DMSO Sel BKM Da DMSO Sel BKM Da

PARP Osim Actin PC-9/BRc1 Downloaded from cancerres.aacrjournals.org on SeptemberPC 28,-9/BRc1 2021. © 2017 American Association for Cancer Research. Author Manuscript Published OnlineFirst on April 17, 2017; DOI: 10.1158/0008-5472.CAN-16-2300 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Fig. 5 120 PC-9/BRc1/9291

A B C 100

80 DMSO PP2 Da Osim Osim+PP2 Osim+Da

p-EGFR 60

EGFR

%cell viability 40 -9/BRc1/9291 -9/BRc1 osimertinib PC PC p-SFKs relative to DMSO control 20 p-FAK YES osimertinib+30nM dasatinib FAK 0 p-ERK 1 10 100 1000 p-SFKs Osimertinib (nM) ERK D Yes PC-9/BRc1/9291 p-AKT 120 Actin AKT 100 p-S6 80 S6 60 PARP siNT %cell viability 40 1.0 2.9 5.7 siYES1 Actin relative to DMSO control 20 siSRC PC-9/BRc1/9291 0 1 10 100 1000 Osimertinib (nM)

E F 1000 PC-9/BRc1 Vehicle 70 Osimertinib

60 * ) 3 800 Dasatinib 50 Osim+Da

40 600 * * 30

%cell viability 20 400

10 Tumor volume (mm relative to DMSO control 0 200 mock YES1

0 0 5 10 15 20 25 30

Downloaded from cancerres.aacrjournals.org on SeptemberTime 28, (days) 2021. © 2017 American Association for Cancer Research. Author Manuscript Published OnlineFirst on April 17, 2017; DOI: 10.1158/0008-5472.CAN-16-2300 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

SFK/FAK signaling attenuates osimertinib efficacy in both drug-sensitive and drug-resistant models of EGFR-mutant lung cancer

Eiki Ichihara, David Westover, Catherine B Meador, et al.

Cancer Res Published OnlineFirst April 17, 2017.

Updated version Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-16-2300

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2017/04/15/0008-5472.CAN-16-2300.DC1

Author Author manuscripts have been peer reviewed and accepted for publication but have not yet been Manuscript edited.

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/early/2017/04/17/0008-5472.CAN-16-2300. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.