Search

Figure S1. Fluorescence in situ hybridization (FISH) analysis of FGFR1 gene copy number status in mesothelioma cell lines. Figure S2. Bioluminescence imaging of orthotopically implanted H226 tumors treated with or without ponatinib. Supplementary Tables S1 through S3.

Purpose: FGFR1 copy-number gain (CNG) occurs in head and neck squamous cell cancers (HNSCC) and is used for patient selection in FGFR-specific inhibitor clinical trials. This study explores FGFR1 mRNA and protein levels in HNSCC cell lines, primary tumors, and patient-derived xenografts (PDX) as predictors of sensitivity to the FGFR inhibitor, NVP-BGJ398.Experimental Design: FGFR1 status, expression levels, and BGJ398 sensitive growth were measured in 12 HNSCC cell lines. Primary HNSCCs (n = 353) were assessed for FGFR1 CNG and mRNA levels, and HNSCC TCGA data were interrogated as an independent sample set. HNSCC PDXs (n = 39) were submitted to FGFR1 copy-number detection and mRNA assays to identify putative FGFR1-dependent tumors.Results: Cell line sensitivity to BGJ398 is associated with FGFR1 mRNA and protein levels, not FGFR1 CNG. Thirty-one percent of primary HNSCC tumors expressed FGFR1 mRNA, 18% exhibited FGFR1 CNG, 35% of amplified tumors were also positive for FGFR1 mRNA. This relationship was confirmed with the TCGA dataset. Using high FGFR1 mRNA for selection, 2 HNSCC PDXs were identified, one of which also exhibited FGFR1 CNG. The nonamplified tumor with high mRNA levels exhibited in vivo sensitivity to BGJ398.Conclusions: FGFR1 expression associates with BGJ398 sensitivity in HNSCC cell lines and predicts tyrosine kinase inhibitor sensitivity in PDXs. Our results support FGFR1 mRNA or protein expression, rather than FGFR1 CNG as a predictive biomarker for the response to FGFR inhibitors in a subset of patients suffering from HNSCC. Clin Cancer Res; 21(19); 4356–64. ©2015 AACR.

Figure S6. Correlation of HPV-expression and FGFR1 mRNA expression.

XLSX file - 66KB, Table S1. Compiled and tabulated ponatinib IC50 values, FGF and FGFR expression, FGFR1 gene copy number and characteristics of the 58 lung cancer cell lines.

Table 1: Cell lines, fusions, tyrosine kinase inhibitors and targets. Table 2: IC50 values for Figure 1C and relative fold change.

Figure S5. In situ hybridization (ISH) assay for FGFR1 mRNA in primary HNSCC tissue.

Table S1. The panel of HNSCC PDXs was submitted to FGFR1 gene copy number (GCN) by CISH and mRNA levels by in situ hybridization (ISH) assay.

Oncogenic kinase fusions of ALK, ROS1, RET, and NTRK1 act as drivers in human lung and other cancers. Residual tumor burden following treatment of ALK or ROS1+ lung cancer patients with oncogene-targeted therapy ultimately enables the emergence of drug-resistant clones, limiting the long-term effectiveness of these therapies. To determine the signaling mechanisms underlying incomplete tumor cell killing in oncogene-addicted cancer cells, we investigated the role of EGFR signaling in drug-naïve cancer cells harboring these oncogene fusions. We defined three distinct roles for EGFR in the response to oncogene-specific therapies. First, EGF-mediated activation of EGFR blunted fusion kinase inhibitor binding and restored fusion kinase signaling complexes. Second, fusion kinase inhibition shifted adaptor protein binding from the fusion oncoprotein to EGFR. Third, EGFR enabled bypass signaling to critical downstream pathways such as MAPK. While evidence of EGFR-mediated bypass signaling has been reported after ALK and ROS1 blockade, our results extended this effect to RET and NTRK1 blockade and uncovered the other additional mechanisms in gene fusion–positive lung cancer cells, mouse models, and human clinical specimens before the onset of acquired drug resistance. Collectively, our findings show how EGFR signaling can provide a critical adaptive survival mechanism that allows cancer cells to evade oncogene-specific inhibitors, providing a rationale to cotarget EGFR to reduce the risks of developing drug resistance. Cancer Res; 77(13); 3551–63. ©2017 AACR.

Figure S4. FGFR1 dependency in 584-A2 and CCL30 cells assessed by RNAi-mediated FGFR1 silencing.

PDF file - 786KB, Figure S1. FGFR1 immunoblots of extracts from lung cancer cell lines representative of the 58 cell lines employed in the studies. Figure S2. Images and quantification of clonogenic growth assays in selected lung cancer cell lines following shRNA-mediated FGFR1 silencing. Figure S3. Dose response curve for inhibition of Colo699 cell growth by the FGF ligand trap, FP-1039. Figure S4. Venn diagram of the overlap of FGFR1, FGF2 and FGF9 expression in ponatinib-sensitive lung cancer cell lines and splicing status of FGFR2 in H1581 and H1048 cells. Figure S5. Graph of ROC analysis of FGFR1 GCN and mRNA expression in a TMA comprised of 21 lung cancer cell lines. Figure S6. Kaplan-Meier analysis of overall survival of lung cancer patients with increased FGFR1 GCN or high mRNA. Figure S7. Venn diagram showing overlap of FGF2 and FGF9 expression with FGFR1 mRNA positivity in primary lung squamous cell carcinomas and adenocarcinomas analyzed by the Cancer Genome Atlas.

Figure S1. Representative images of FGFR1 GCN assayed by FISH in HNSCC cell lines.

XLSX file - 106KB, Table S5. Tabulated FGFR1, FGF2 and FGF9 mRNA expression values, FGFR1 CNV values and mutation status of TCGA lung adenocarcinomas and squamous cell carcinomas.

Figure S3. Correlation of FGFR1 mRNA and protein expression with BGJ398 sensitivity.

Fig. S1: EGFR inhibition or stimulation affects normal and catalytically inactive fusion oncogene signaling. Fig. S2: EGFR knockdown reduces fusion kinase cancer cell proliferation. Fig. S3: Conserved re-phosphorylated activation loop tyrosines enabled signaling PLA design and representative image analyses of fusion kinase and EGFR signaling PLAs. Fig. S4: Optimization of the EGFR-GRB2 PLA and fusion kinase PLA antibody controls. Fig. S5: The GRB2 adaptor does not switch from ROS1 to MET with FKI treatment, but MET can signal through GRB2. FIg. S6: GRB2 and SHC1 signaling rewire in a RET+ cell line treated with a fusion kinase inhibitor (FKI). Fig. S7: TRK, RET, ROS1 and ALK fusion kinases interact specifically with EGFR. Fig. S8: EGFR is highly expressed, phosphorylated, and signaling in vivo in an NTRK1+ patient-derived xenograft (PDX) model. Fig. S9: Fusion kinase, and EGFR signaling complexes are present in RET, and ALK, fusion resistant patient samples.

PDF file - 90KB, Table S2. Tabulated expression values for selected FGFs in lung cancer cell lines from the Cancer Cell Line Encyclopedia. Table S3. Tabulated FGFR1 gene copy number values determined by SISH and associated patient data for the surgical lung cancer cohort. Table S4. Tabulated FGFR1 mRNA expression values determined by ISH and associated patient data for the surgical lung cancer cohort.

Figure S2. Effect of BGJ398 on Ki67 staining, cleaved caspase 3 levels and phospho-ERK levels in HNSCC cell lines.

The FGFR1 is a therapeutic target under investigation in multiple solid tumors and clinical trials of selective tyrosine kinase inhibitors (TKI) are underway. Treatment with a single TKI represents a logical step toward personalized cancer therapy, but intrinsic and acquired resistance mechanisms limit their long-term benefit. In this study, we deployed RNAi-based functional genomic screens to identify protein kinases controlling the intrinsic sensitivity of FGFR1-dependent lung cancer and head and neck squamous cell cancer (HNSCC) cells to ponatinib, a multikinase FGFR-active inhibitor. We identified and validated a synthetic lethal interaction between MTOR and ponatinib in non–small cell lung carcinoma cells. In addition, treatment with MTOR-targeting shRNAs and pharmacologic inhibitors revealed that MTOR is an essential protein kinase in other FGFR1-expressing cancer cells. The combination of FGFR inhibitors and MTOR or AKT inhibitors resulted in synergistic growth suppression in vitro. Notably, tumor xenografts generated from FGFR1-dependent lung cancer cells exhibited only modest sensitivity to monotherapy with the FGFR-specific TKI, AZD4547, but when combined with the MTOR inhibitor, AZD2014, significantly attenuated tumor growth and prolonged survival. Our findings support the existence of a signaling network wherein FGFR1-driven ERK and activated MTOR/AKT represent distinct arms required to induce full transformation. Furthermore, they suggest that clinical efficacy of treatments for FGFR1-driven lung cancers and HNSCC may be achieved by combining MTOR inhibitors and FGFR-specific TKIs. Cancer Res; 75(20); 4398–406. ©2015 AACR.

Supplementary Methods, Figures S1-S12, Tables S1-S2. Figure S1. Overview of workflow for the synthetic lethal and essential gene screens. Figure S2. Procedure for isolating and barcoding PCR-amplified shRNA sequences. Figure S3. Growth inhibition of H157 cells by ponatinib in combination with MTOR inhibitor, AZD8055. Figure S4. Relative sensitivity of cancer cell lines to MTOR inhibitor AZD8055 using a cell proliferation assay. Figure S5. Dose response curves for growth inhibition of cancer cell lines by rapamycin. Figure S6. Growth inhibition by FGFR inhibitors in combination with MTOR inhibitor, AZD8055. Figure S7. Synergistic growth inhibition of CCL30 HNSCC cells by AZD4547 and AZD8055. Figure S8. Growth inhibition of Colo699 and H1581 cells by AZD4547 and rapamycin. Figure S9. Signal pathway inhibition by FGFR TKIs and rapamycin alone, and in combination. Figure S10. Growth inhibition by AKT inhibitors alone and combined with ponatinib on Colo699 cells. Figure S11. Combination ponatinib and AKT inhibitor GSK690693 yields synergistic in vitro growth inhibition. Figure S12. Combination AZD4547 and AZD2014 provides superior growth inhibition of Colo699 xenografts. Table S1. Primers used for sample preparation for Illumina sequencing. Table S2. Ranking of Synthetic Lethal Hits with ponatinib Treatment.

Boundary layer stability was analyzed for the HIFiRE-1 flight vehicle geometry for ground tests conducted at the CUBRC LENS I hypersonic shock test facility and the Langley Research Center (LaRC) 20- inch Mach 6 Tunnel. Boundary layer stability results were compared to transition onset location obtained from discrete heat transfer measurements from thin film gauges during the CUBRC test and spatially continuous heat transfer measurements from thermal phosphor paint data during the LaRC test. The focus of this analysis was on conditions at non-zero angles of attack as stability analysis has already been performed at zero degrees angle of attack. Also, the transition onset data obtained during flight testing was at nonzero angles of attack, so this analysis could be expanded in the future to include the results of the flight test data. Stability analysis was performed using the 2D parabolized stability software suite STABL (Stability and Transition Analysis for Hypersonic Boundary Layers) developed at the University of Minnesota and the mean flow solutions were computed using the DPLR finite volume Navier-Stokes computational fluid dynamics (CFD) solver. A center line slice of the 3D mean flow solution was used for the stability analysis to incorporate the angle of attack effects while still taking advantage of the 2D STABL software suite. The N-factors at transition onset and the value of Re(sub theta)/M(sub e), commonly used to predict boundary layer transition onset, were compared for all conditions analyzed. Ground test data was analyzed at Mach 7.2 and Mach 6.0 and angles of attack of 1deg, 3deg and 5deg. At these conditions, the flow was found to be second mode dominant for the HIFiRE-1 slender cone geometry. On the leeward side of the vehicle, a strong trend of transition onset location with angle of attack was observed as the boundary layer on the leeward side of the vehicle developed inflection points at streamwise positions on the vehicle that correlated to angle of attack. Inflection points are a strong instability mechanism that lead to rapid breakdown and transition to turbulence. The transition onset location on the windward side of the vehicle displayed no trend with angle of attack or freestream Reynolds number and transition was observed farther down the vehicle than observed on the leeward side of the vehicle. In analysis of both windward and leeward sides of the vehicle, use of the N factor methodology to develop trends to predict boundary layer transition onset showed improvements over the Re(sub theta)/M(sub e) empirical correlation methodology. Stronger correlations and less scatter in the data were observed when using the N factor method for these cases.