1. Cancer Biology

Registered report: Widespread potential for growth factor-driven resistance to anticancer kinase inhibitors

  1. Edward Greenfield
  2. Erin Griner
  3. Reproducibility Project: Cancer Biology  Is a corresponding author
  1. Dana-Farber Cancer Institute, United States
  2. University of Virginia, United States
https://doi.org/10.7554/eLife.04037
Share this article
Cite this article
  1. Edward Greenfield
  2. Erin Griner
  3. Reproducibility Project: Cancer Biology
(2014)
Registered report: Widespread potential for growth factor-driven resistance to anticancer kinase inhibitors
eLife 3:e04037.
https://doi.org/10.7554/eLife.04037
  2. Download BibTeX
  3. Download .RIS

Abstract

The Reproducibility Project: Cancer Biology seeks to address growing concerns about reproducibility in scientific research by conducting replications of 50 papers in the field of cancer biology published between 2010 and 2012. This Registered Report describes the proposed replication plan of key experiments from ‘Widespread potential for growth-factor-driven resistance to anticancer kinase inhibitors’ by Wilson and colleagues, published in Nature in 2012 (Wilson et al., 2012). The experiments that will be replicated are those reported in Figure 2B and C. In these experiments, Wilson and colleagues show that sensitivity to receptor tyrosine kinase (RTK) inhibitors can be bypassed by various ligands through reactivation of downstream signaling pathways (Figure 2A; Wilson et al., 2012), and that blocking the receptors for these bypassing ligands abrogates their ability to block sensitivity to the original RTK inhibitor (Figure 2C; Wilson et al., 2012). The Reproducibility Project: Cancer Biology is a collaboration between the Center for Open Science and Science Exchange, and the results of the replications will be published by eLife.

https://doi.org/10.7554/eLife.04037.001

Introduction

A recurring theme in treatment of cancer is the acquisition of drug resistance. The effectiveness of therapies targeting specific mutations in receptor tyrosine kinases (RTKs) is limited by the acquisition of resistance to the drugs over the course of treatment (Mok et al., 2009; Camidge et al., 2014). Resistance can be acquired through new mutations that block the action of the RTK inhibitors or their uptake and/or genetic amplification of downstream target genes of the RTK (Chen and Fu, 2011; Garrett and Arteaga, 2011; Sequist et al., 2011; Gainor and Shaw, 2013; Yang, 2013). Several studies, including this work by Wilson and colleagues, elucidated another mechanism for this acquisition of resistance: the engagement of parallel RTK signaling pathways that converge on common downstream survival signals via signals from the tumor microenvironment. In this study, Wilson and colleagues examined several cancer cell lines for ligand-mediated drug resistance (Wilson et al., 2012).

In Figure 2B/C, Wilson and colleagues demonstrated that resistance to primary kinase inhibitor treatment can be induced by the addition of rescuing ligands that activate the PI(3)K–AKT and MAPK pro-survival signaling pathways. This resistance can be overcome with the addition of an appropriate secondary kinase inhibitor. Three different cancer cell line models were used to demonstrate this phenomenon. Treatment of A204 (a PDGFR amplified rhabdomyosarcoma cell line) with the ligand FGF activated pFRS2 and pERK, inducing resistance to sunitinib. The addition of a secondary kinase inhibitor, PD173074, blocked FGF-induced pFRS2 and pERK activation, restoring sensitivity to sunitinib. The treatment of M14 (a BRAF-mutated melanoma cell line) with the ligand NRG1 activated pHER3 and pAKT, inducing partial resistance to PLX4032. The addition of a secondary kinase inhibitor, lapatinib, blocked NRG1-induced pHER3 and pAKT activation, restoring sensitivity to PLX4032. Treatment of KHM-3S (an EGFR-mutated small cell lung cancer cell line) with the ligand HGF activated pMET and pERK, inducing resistance to Erlotinib. The addition of a secondary kinase inhibitor, crizotinib, blocked HGF-induced pMET and pERK activation, restoring sensitivity to erlotinib.

The cell viability assays examining drug sensitivity and the Western blots examining levels of phosphorylated kinases in Figures 2B and 2C, respectively, are the key experiments that demonstrate that growth factor ligands can reactivate downstream signaling components important for cancer cell survival, causing resistance to anticancer kinase inhibitors (Wilson et al., 2012). These experiments are replicated in Protocols 1 and 2.

Two studies published around the same time as the work of Wilson and colleagues also support the proposed mechanism of acquired resistance to RTK inhibition by signaling from the tumor microenvironment. Straussman and colleagues demonstrated that HGF signaling derived from the tumor microenvironment could bypass EGFR inhibition by activation of MET signaling (Straussman et al., 2012, also included for replication in the Reproducibility Project: Cancer Biology), and Harbinski and colleagues, in an approach similar to Wilson and colleagues, showed that multiple growth factor ligands could ‘bypass’ inhibitor-targeted RTKs (Harbinski et al., 2012).

Since the publication of Wilson and colleagues' work, several publications have reported similar results to those being replicated in Protocols 1 and 2. Similar to the experiments with A204 cells above, Welti and colleagues demonstrated that FGF ligands could induce resistance to sunitinib, which could be reversed by the addition of PD173074 (Welti et al., 2011). These experiments were performed in HUVEC cells, whereas A204 cells were used in the study being replicated. Similar to the experiments on M14 cells above, Montero-Conde and colleagues showed that NRG1 ligand could activate pHER3 and pAKT in the presence of PLX4032, and this activation could be reversed by the addition of lapatinib (Montero-Conde et al., 2013). These experiments were performed in 8505C cells, whereas M14 cells were used in the study being replicated. Similar to the experiments performed on KHM-S3 cells above, several groups have demonstrated that HGF ligand can induce resistance to erlotinib and that this resistance can be reversed by the addition of crizotinib (Nakagawa et al., 2012; Nakade et al., 2014). These experiments were performed in PC-9 and HCC827 cells, whereas KHM-3S cells were used in the study being replicated.

Materials and methods

Unless otherwise noted, all protocol information was derived from the original paper, references from the original paper, or information obtained directly from the authors. An asterisk (*) indicates data or information provided by the Reproducibility Project: Cancer Biology core team. A hashtag (#) indicates information provided by the replicating lab.

Protocol 1: Cell viability assays

This protocol describes cell viability assays to determine the IC50 values of three cancer cell lines treated with primary kinase inhibitor alone, primary kinase inhibitor in combination with rescuing ligand, and primary kinase inhibitor in triple combination with rescuing ligand and a drug targeting the rescuing ligand's receptor tyrosine kinase (RTK) (termed the secondary kinase inhibitor) (Figure 2B).

Sampling

Request a detailed protocol

  • The original data presented is qualitative, and the authors were unable to share the raw data values with the RP:CB core team. This prevents power calculations being performed a priori to determine the sample size (number of biological replicates). In order to determine an appropriate number of replicates to perform initially, we have estimated the sample sizes required based on a range of potential variance. We will also determine the sample size post hoc as described in Power Calculations.

    1. Please see Power Calculations for details.

  • Each experiment has three cohorts. In each cohort, a dilution series of the primary kinase inhibitor (10−4, 10−3, 10−2, 10−1, 100, and 101 µM) is run three times; once alone, once with the rescuing ligand, and once with both the rescuing ligand and the secondary kinase inhibitor. The effect of the secondary kinase inhibitor alone will also be assessed. Each condition will be run in triplicate.

    1. Cohort 1: A204 cell line.

      • Media only [additional].

      • Vehicle control.

      • 0.001 µM–10 µM sunitinib + no ligand.

      • 0.001 µM–10 µM sunitinib + 50 ng/ml FGF.

      • 0.001 µM–10 µM sunitinib + 50 ng/ml FGF + 0.5 µM PD173074.

      • 0.5 µM PD173074 + no ligand [additional].

    2. Cohort 2: M14 cell line.

      • Media only [additional].

      • Vehicle control.

      • 0.001 µM–10 µM PLX4032 + no ligand.

      • 0.001 µM–10 µM PLX4032 + 50 ng/ml NRG1.

      • 0.001 µM–10 µM PLX4032 + 50 ng/ml NRG1 + 0.5 µM lapatinib.

      • 0.5 µM lapatinib + no ligand [additional].

    3. Cohort 3: KHM-3S cell line.

      • Media only [additional].

      • Vehicle control.

      • 0.001 µM–10 µM erlotinib + no ligand.

      • 0.001 µM–10 µM erlotinib + 50 ng/ml HGF.

      • 0.001 µM–10 µM erlotinib + 50 ng/ml HGF + 0.5 µM crizotinib.

      • 0.5 µM crizotinib + no ligand [additional].

Materials and reagents

Request a detailed protocol
ReagentTypeManufacturerCatalog #Comments
96-well tissue culture platesMaterialsCorning (Sigma-Aldrich)CLS3516Original unspecified
KHM-3S cellsCellsJCRB Cell BankJCRB0138Original source of the cells unspecified
A204CellsATCCHTB-82Original source of the cells unspecified
M14CellsATCCHTB-129*Original source of the cells unspecified
LapatinibDrugLC LaboratoriesL-4804Original formulation unspecified
CrizotinibDrugSigma-AldrichPZ0191Originally from Selleck Chemicals
PD173074DrugSigma-AldrichP2499Originally from Tocris Bioscience
PLX4032DrugActive BiochemA-1130
SunitinibDrugSigma-AldrichPZ0012Originally from Selleck Chemicals, formulation unspecified
ErlotinibDrugLC LaboratoriesE-4007
HGFLigandSigma-AldrichH5791Originally obtained from Peprotech
FGF-basicLigandSigma-AldrichF0291Originally obtained from Peprotech
NRG1-β1LigandNovus BiologicalsH00003084-P01Originally obtained from R&D Systems
RPMI 1640MediaSigma-AldrichR8758Originally from Gibco, formulation unspecified
FBSReagentSigma-AldrichF4135Originally from Gibco
PenicillinAntibioticSigma-AldrichP4458Original unspecified
StreptomycinAntifungalOriginal unspecified
ParaformaldehydeReagentSigma-Aldrich158127Original unspecified
Syto 60ReagentLife TechnologiesS11342Original unspecified
Odyssey scannerEquipmentLiCOR
Odyssey application softwareSoftwareLiCOR
  1. *

    The breast cancer cell line MDA-MB-435 has been shown to be mislabeled; it is in fact identical to the M14 melanoma cell line (Rae et al., 2007; Chambers, 2009; Holliday and Speirs, 2011).

Procedure

Request a detailed protocol
Notes
Request a detailed protocol

  • All cells will be sent for mycoplasma testing and STR profiling.

  • Medium for all cell lines: RPMI 1640 supplemented with 10% FBS, 50 U/ml penicillin, and 50 µg/ml streptomycin.

  • Cells maintained at 37°C in a humidified atmosphere at 5% CO2.

    1. Seed 3000–5000 cells per well into 96-well plates. For each condition replicate seed 1 well as the media control, 1 well as the vehicle control, 1 well for treatment with the secondary kinase inhibitor alone, and 6 wells per concentration curve (10−4, 10−3, 10−2, 10−1, 100, and 101 µM), of which there are three.

      • a. 6 wells per concentration curve × 3 concentration curves = 18 wells + 3 wells = 21 wells per cohort.

    2. 18–24 hr after seeding treat 3 wells per condition with appropriate treatment (see Sampling).

      • a. Lab will record the vehicle used to solubilize the drugs.

    3. 72 hr after treatment, fix cells in 4% paraformaldehyde (PFA).

      • a. Lab will record the PFA incubation time.

    4. Stain with Syto 60 according to the manufacturer's recommendations and assay cell number using an Odyssey with Odyssey Application Software.

      • a. Include empty wells and media only wells.

    5. Calculate cell viability by dividing the fluorescence from the drug-treated cells by the fluorescence from the control (vehicle) treated cells. Fit normalized data to a sigmoidal dose–response curve.

      • a. Also calculate the effect of vehicle by dividing the fluorescence from the control vehicle cells by the fluorescence from the media only treated cells [additional control].

      • b. Determine the IC50 values for each curve.

      • c. Lab will document the software used to fit the data to a sigmoidal dose–response curve and calculate the IC50 values.

    6. Repeat independently two additional times.

Deliverables

Request a detailed protocol

  • Data to be collected:

    1. Raw fluorescence data and calculated cell viability.

    2. Semi-logarithmic graph for each condition of primary kinase inhibitor (log) vs normalized cell viability (linear) for each cell line [comparable to Figure 2B].

    3. Calculated IC50 for each condition.

Confirmatory analysis plan

Request a detailed protocol

  • Statistical analysis of the Replication Data:

    1. For each cell line compare the IC50 of primary kinase inhibitor alone, primary kinase inhibitor + ligand, and primary kinase inhibitor + ligand + secondary kinase inhibitor.

      • • ANOVA.

  • Meta-analysis of original and replication attempt effect sizes:

    1. We will plot the replication data (mean and 95% confidence interval) and will include the original data point, calculated directly from the representative image in Figure 2B, as a single point on the same plot for comparison.

Known differences from the original study

Request a detailed protocol

  • We are including two additional control conditions;

    1. Media alone.

      • a. To provide a baseline.

    2. Treatment of the cells with the secondary kinase inhibitor alone.

      • a. To assess any effects, the secondary kinase inhibitor may be independent of the ligand and primary kinase inhibitor.

Provisions for quality control

Request a detailed protocol

  • All data obtained from the experiment—raw data, data analysis, control data, and quality control data—will be made publicly available, either in the published manuscript or as an open access dataset available on the Open Science Framework (https://osf.io/h0pnz/).

  • Cell lines will be validated by STR profiling and screened for mycoplasma contamination.

  • A lab from the Science Exchange network with extensive experience in conducting cell viability assays will perform these experiments.

Protocol 2: Western blot assays

This protocol describes Western blot assays to determine the levels of activated phosphorylated signaling pathways in three cancer cell lines treated with primary kinase inhibitor alone, primary kinase inhibitor in combination with rescuing ligand, and primary kinase inhibitor in triple combination with rescuing ligand and a drug targeting the rescuing ligand's receptor tyrosine kinase (RTK) (termed the secondary kinase inhibitor) (Figure 2C).

Sampling

Request a detailed protocol

  • The original data presented is qualitative. This prevents power calculations being performed a priori to determine the sample size (number of biological replicates). In order to determine an appropriate number of replicates to perform initially, we have estimated the sample sizes required based on a range of potential variance. We will also determine the sample size post hoc as described in Power Calculations.

    1. Please see Power Calculations for details.

  • Each experiment has three cohorts. Each cohort will consist of cells treated with media alone, with vehicle alone, with the primary kinase inhibitor, with primary kinase inhibitor and the rescuing ligand and with the primary kinase inhibitor, the rescuing ligand and the secondary kinase inhibitor. The effect of the secondary kinase inhibitor alone will also be assessed. Each condition will be run once (i.e., no technical replicates will be performed).

    1. Cohort 1: A204 cell line.

      • Media only [additional].

      • Vehicle control.

      • 1 µM sunitinib + no ligand.

      • 1 µM sunitinib + 50 ng/ml FGF.

      • 1 µM sunitinib + 50 ng/ml FGF + 0.5 µM PD173074.

      • 1 µM PD173074 + no ligand [additional].

    2. Cohort 2: M14 cell line.

      • Media only [additional].

      • Vehicle control.

      • 1 µM PLX4032 + no ligand.

      • 1 µM PLX4032 + 50 ng/ml NRG1.

      • 1 µM PLX4032 + 50 ng/ml NRG1 + 0.5 µM lapatinib.

      • 1 µM lapatinib + no ligand [additional].

    3. Cohort 3: KHM-3S cell line.

      • Media only [additional].

      • Vehicle control.

      • 1 µM erlotinib + no ligand.

      • 1 µM erlotinib + 50 ng/ml HGF.

      • 1 µM erlotinib + 50 ng/ml HGF + 0.5 µM Crizotinib.

      • 1 µM crizotinib + no ligand [additional].

    4. Cohort 4: positive control cell lines.

      • For Cohort 1: HL60 cells treated with FGF [additional control].

      • For Cohort 2: MCF7 cells treated with NRG1 [additional control].

      • For Cohort 3: HEK293 cells treated with HGF [additional control].

        • a. Treatment of these cell lines with their cognate growth factor ligands will serve as a positive control for ligand activity.

Materials and reagents:

Request a detailed protocol
ReagentTypeManufacturerCatalog #Comments
96-well Tissue culture platesMaterialsCorning (Sigma-Aldrich)CLS3596Original unspecified
6-well tissue culture platesMaterialsCorning (Sigma-Aldrich)CLS3516Original unspecified
KHM-3S cellsCellsJCRB Cell BankJCRB0138Original source of the cells unspecified
A204 cellsCellsATCCHTB-82Original source of the cells unspecified
M14 cellsCellsATCCHTB-129Original source of the cells unspecified
HL60 cellsCellsATCCCCL-240
MCF7 cellsCellsATCCHTB-22
HEK293 cellsCellsATCCCRL-1573
LapatinibDrugLC LaboratoriesL-4804Original formulation unspecified
CrizotinibDrugSigma-AldrichPZ0191Originally from Selleck Chemicals
PD173074DrugSigma-AldrichP2499Originally from Tocris Bioscience
PLX4032DrugActive BiochemA-1130
SunitinibDrugSigma-AldrichPZ0012Originally from Selleck Chemicals, formulation unspecified
ErlotinibDrugLC LaboratoriesE-4007
HGFLigandSigma-AldrichH5791Originally obtained from Peprotech
FGF-basicLigandSigma-AldrichF0291Originally obtained from Peprotech
NRG1-β1LigandNovus BiologicalsP1426Originally obtained from R&D Systems
RPMI 1640MediaSigma-AldrichR8758Originally from Gibco, formulation unspecified
FBSReagentSigma-AldrichF4135Originally from Gibco
PenicillinAntibioticSigma-AldrichP4458Original unspecified
StreptomycinAntifungalOriginal unspecified
Halt protease and phosphatase cocktail inhibitorReagentThermo Scientific78440
Image JSoftwareNational Institutes of Health (NIH)N/A
p-PDGFRαAntibodySanta CruzSC-12911190 kDa
PDGFRαAntibodyCell Signaling5241190 kDa
p-AKT S473AntibodyInvitrogen44-621 G65 kDa
AKTAntibodyCell Signaling927265 kDa
p-ERK T202/Y204AntibodyCell Signaling910144,42 kDa
ERKAntibodyCell Signaling910244,42 kDa
pFRS2α Y196AntibodyCell Signaling386485 kDa
FRS2αAntibodySanta CruzSC-831885 kDa
β-tubulinAntibodyCell Signaling214655 kDa
pHER3 Y1289AntibodyCell Signaling4791185 kDa
HER3AntibodySanta CruzSC-285185 kDa
p-EGFR Y1068AntibodyAbcamab5644185 kDa
EGFRAntibodyBD Biosciences610017185 kDa
p-MET Y1234/5AntibodyCell Signaling3126145 kDa
METAntibodySanta CruzSC-10145 kDa
Anti-Mouse IgG-HRPAntibodyCell Signaling Technology7076P2Original unspecified
Anti-Rabbit IgG-HRPAntibodyCell Signaling Technology7074P2Original unspecified
Anti-Goat IgG-HRPAntibodySanta Cruz Biotechnologysc-2020Original unspecified
Trypsin-EDTA solution (1X)ReagentSigma-AldrichT3924Original unspecified
Dulbecco’s Phosphate Buffered SalineReagentSigma-AldrichD1408Original unspecified
Mini Protean TGX 4–15% Tris-Glycine gels; 15-well; 15 μlReagentBio-Rad456-1086Original unspecified
2X Laemmli sample bufferReagentSigma-AldrichS3401Original unspecified
ECL DualVue Western Markers (15 to 150 kDa)ReagentSigma-AldrichGERPN810Original unspecified
Nitrocellulose membrane; 0.45 μm, 20 × 20 cmReagentBio-Rad162-0113Original unspecified
Ponceau SReagentSigma-AldrichP7170Original unspecified
Tris Buffered Saline (TBS); 10X solutionReagentSigma-AldrichT5912Original unspecified
Tween 20ReagentSigma-AldrichP1379Original unspecified
Nonfat-Dried MilkReagentSigma-AldrichM7409Original unspecified
Super Signal West Pico SubstrateReagentThermo-Fisher (Pierce)34087

Procedure

Request a detailed protocol
Notes
Request a detailed protocol

  • All cells will be sent for mycoplasma testing and STR profiling.

  • Medium for cell lines: RPMI 1640 supplemented with 10% FBS, 50 U/ml penicillin, and 50 µg/ml streptomycin.

    1. MCF7 cells and HEK293 cells are maintained in DMEM + 10% FBS.

  • Cells maintained at 37°C in a humidified atmosphere at 5% CO2.

    1. Seed cells in plates.

      • a. Two control and four experimental wells (6 wells total) are needed for each cell line in Cohorts 1–3.

        • i. Lab will determine and record the number of cells seeded and well size used.

      • b. *For Cohort 4 seed cells as needed into wells of a 6-well plate.

    2. 18–24 hr after seeding treat wells in Cohorts 1–3 with conditions as described in the Sampling section.

      • a. Lab will determine and record vehicle for preparation of drug solutions.

      • b. Harvest protein as in Step 5 after 2 hr of treatment.

    3. Simultaneously treat cells in Cohort 4 as follows:

      • a. HL60 cells. Note: This protocol is based on Krejci et al. (2003).

        • i. Serum starve HL60 cells for 24 hr prior to protein harvesting.

          • Serum starve = DMEM + 0% FBS.

        • ii. Treat cells for 10 min with 100 ng/ml FGF.

        • iii. Harvest cell lysates as noted in Step 5.

      • b. MCF7 cells. Note: This protocol is based on Sarup et al. (2008).

        • i. Serum starve cells for 48 hr prior to protein harvesting.

          • Serum starve = DMEM + 0.1% BSA.

        • ii. Treat cells with 1 nmol/l NRG1 for 10 min at 37°C.

        • iii. Harvest cell lysates as noted in Step 5.

      • c. HEK293 cells. Note: This protocol is based on Wright et al. (2012).

        • i. Serum starve HEK293 cells for 24 hr prior to protein harvesting.

          • Serum starve = DMEM + 0% FBS.

        • ii. Treat cells with 29 ng/ml HGF for 10 min at 37°C.

        • iii. Harvest cell lysates as noted in Step 5.

    4. #Preparation of cell lysate:

      • a. Note: from here on, the replicating lab will use their in-house Western blot protocol, as recommended by the original authors.

      • b. Harvest cells from the tissue culture plate using 1× trypsin–EDTA.

      • c. Wash cells with 1× cold PBS and spin at 1200 rpm for 5 min.

      • d. Decant the PBS and add lysis buffer to the cell pellet and resuspend well.

      • e. Incubate at room temperature for 5 min.

      • f. Spin solution at 13,000 rpm for 30 min at 4°C using a benchtop centrifuge.

      • g. Collect the lysate/protein sample and store at −20°C or −80°C for later use.

    5. #SDS-PAGE separation:

      • a. Prepare the lysate sample by adding SDS reducing loading dye to ∼25–30 µg of protein sample and boiling at 95°C–100°C for 5 min.

        • i. Lab will record exact amount of protein loaded and provide data from determining protein concentration.

      • b. Let samples cool on ice and quick-spin the tubes to collect any droplets on the cap of the tube.

      • c. Prepare the gel for sample loading—insert the gel in the gel box with 1× running buffer and ensure there is no leak.

        • i. Based on the expected MWs of the targets, lab will determine the optimal percentage gel to use.

      • d. Load 16 µl of sample (25–30 µg/lane) in each well of the Tris–glycine gel.

      • e. Run the sample at 175 V for 25 min.

      • f. Remove the gel from the cassette and rinse with water.

    6. #Transfer and blocking:

      • a. Transfer protein on the gel to a nitrocellulose membrane for 1 hr at 12 V using a semi-dry transfer apparatus, 1× transfer buffer, and blotting sheets.

      • b. Verify the efficiency of the transfer by Ponceau staining of the membrane.

        • i. Lab will record an image of the Ponceau-stained membrane.

      • c. Incubate the blots in 5% non-fat skim milk for 1 hr at room temperature.

    7. #Antibody probing:

      • a. Dilute the primary antibodies according to the manufacturer's recommendations, as suggested by the original authors.

        • i. If the manufacturer recommends a range of dilutions, lab will use a dilution in the middle of the recommended dilution range.

        • ii. A204:

          • p-PDGFRα.

          • PDGFRα.

          • p-AKT S473.

          • AKT.

          • p-ERK T202/Y204.

          • ERK.

          • pFRS2α Y196.

          • FRS2α.

          • β-tubulin [additional control].

            • A. Loading control.

        • iii. M14:

          • pHER3 Y1289.

          • HER3.

          • p-AKT S473.

          • AKT.

          • p-ERK T202/Y204.

          • ERK.

          • β-tubulin [additional control].

            • A. Loading control.

        • iv. KHM-3S:

          • p-EGFR Y1068.

          • EGFR.

          • p-AKT S473.

          • AKT.

          • p-ERK T202/Y204.

          • ERK.

          • p-MET Y1234/5.

          • MET.

          • β-tubulin [additional control].

            • A. Loading control.

        • v. HL60:

          • pERK T202/Y204.

          • ERK.

          • β-tubulin [additional control].

            • A. Loading control.

        • vi. MCF7:

          • pHER3.

          • HER3.

          • β-tubulin [additional control].

            • A. Loading control.

        • vii. HEK293:

          • pMET.

          • MET.

          • β-tubulin [additional control].

            • A. Loading control.

      • b. Add the antibody solutions to the membranes and incubate them for 12–16 hr at 4°C.

      • c. Wash the blots with Tris-buffered saline (TBS) and with 0.5% Tween-20 three times for 10 min each.

      • d. Dilute HRP-secondary antibody in 5% milk and add to the blots.

        • i. Lab will record the dilution factor of the secondary antibody.

      • e. Incubate at room temperature for 1 hr.

      • f. Wash the blots with TBS +0.5% Tween-20 four times for 15 min each.

    8. #Developing:

      • a. Remove as much wash buffer as possible.

      • b. Mix Super Signal West Pico Chemiluminescent Substrate solutions in equal proportions and add it to the blot.

      • c. Incubate for ∼1 min.

      • d. Insert the blot in the developing cassette and develop the blot in the dark.

      • e. Expose the blot to the film at three time points, starting with 15 s. Determine the other two time points based on the strength of the signal in the 15 s exposure.

    9. #Scan film and quantify band intensity using densitometric analysis software.

    10. Repeat independently two additional times.

Deliverables

Request a detailed protocol

  • Data to be collected:

    1. Images of probed membranes (images of full films with molecular weight ladders).

    2. Scanned image of Ponceau-stained membranes after protein transfer.

    3. Quantified signal intensities and bar graphs of mean signal intensities normalized for β-tubulin loading and total pan-protein levels.

Confirmatory analysis plan

Request a detailed protocol

  • Statistical analysis of the Replication Data:

    1. For each cell line compare the following normalized phosphorylated kinase levels of primary kinase inhibitor alone, primary kinase inhibitor + ligand, and primary kinase inhibitor + ligand + secondary kinase inhibitor.

      • • One-way ANOVA.

      • • Note: at the time of analysis, we will generate a histogram of all the data to determine if it follows a Gaussian distribution or not. If it is skewed, we will perform the appropriate transformation in order to proceed with the proposed statistical analysis.

  • Meta-analysis of original and replication attempt effect sizes:

    1. We will plot the replication data (mean and 95% confidence interval) and will include the original data point, calculated directly from the representative image in Figure 2C, as a single point on the same plot for comparison.

Known differences from the original study

Request a detailed protocol

  • We are including three additional control conditions;

    1. Media alone.

      • i. To provide a baseline.

    2. Treatment of the cells with the secondary kinase inhibitor alone.

      • i. To assess any effects, the secondary kinase inhibitor may be independent of the ligand and primary kinase inhibitor.

    3. Treatment of a control cell line with the growth factor ligand alone.

      • i. To ensure the growth factor ligand is active.

        • FGF should cause phosphorylation of ERK1/2 in HL60 cells.

        • NRG1 should cause phosphorylation of HER3 in MCF7 cells.

        • HGF should cause phosphorylation of MET in HEK293 cells.

  • The original authors recommended that the replicating lab follows a standard Western blot protocol.

Provisions for quality control

Request a detailed protocol

  • All data obtained from the experiment—raw data, data analysis, control data and quality control data—will be made publicly available, either in the published manuscript or as an open access dataset available on the Open Science Framework (https://osf.io/h0pnz/).

  • Cell lines will be validated by STR profiling and screened for mycoplasma contamination.

  • A lab from the Science Exchange network with extensive experience in conducting Western blot assays for phosphorylated proteins will perform these experiments.

Power Calculations

Protocol 1

Request a detailed protocol

The original data presented is qualitative (images of survival curves) and the authors were unable to share the raw data values with the RP:CB core team. To estimate original effect sizes, we determined approximate IC50 concentrations from the original survival curve images.

Summary of the original data.

A204 cellsIC50
Sunitinib0.05 μM
Sunitinib + FGF2.5 μM
Sunitinib + FGF + PD1730740.025 μM

  1. • FGF induces resistance to Sunitinib.

  2. • PD173074 blocks FGF-induced resistance to Sunitinib, restoring sensitivity.

M14IC50
PLX40320.1 μM
PLX4032 + NRG10.2 μM
PLX4032 + NRG1 + Lapatinib0.1 μM

  1. • NRG1 induces partial resistance to PLX4032.

  2. • Lapatinib blocks NRG1-induced resistance to PLX4032, restoring sensitivity.

KHM-3SIC50
Erlotinib0.5 μM
Erlotinib + HGF>10 μM
Erlotinib + HGF + Crizotinib0.3 μM

  1. • HGF induces resistance to Erlotinib.

  2. • Crizotinib blocks HGF-induced resistance to Erlotinib, restoring sensitivity.

We have calculated the projected sample size based on a variety of different possible levels of variance using a one-way ANOVA test with an alpha error of 0.05.

A204
VarianceF (2, 6)ηP2Effect size fPowerTotal sample size across all groups
2%7273.61320.99958849.2563199.99%6
15%129.30870.9773266.56531699.99%6
28%37.11030.9252063.51710998.53%6
40%18.1840.8583842.46198185.32%6

For each percent variance, the relative standard deviation of the approximated IC50 was used to calculate the F statistic from a one-way ANOVA analysis, which was converted to ηP2 (the ratio of variance attributed to the effect and the effect plus its associate error variance from the ANOVA), and then used to determine the effect size (Cohen's f) and the needed sample size to obtain at least 80% power. The actual power obtained is listed.

M14
VarianceF (2, 6)ηP2Effect size fPowerTotal sample size across all groups
2%12500.99760620.413599.99%6
15%22.22220.8810572.72165290.90%6
28%6.37760.6800891.45803685.39%9
40%3.1250.5102041.02062188.33%15
KHM-S3
VarianceF (2, 6)ηP2Effect size fPowerTotal sample size across all groups
2%6890.82120.99956547.935999.99%6
15%122.50350.9760966.39014999.99%6
28%35.15730.9213783.42331598.12%6
40%17.22710.8516842.39632283.59%6

In order to produce quantitative replication data, we will run the experiment three times. Each time we will quantify the IC50. We will determine the standard deviation of the IC50 across the three biological replicates and combine this with the means from the original study to simulate an effect size. Using this simulated effect size, we will then determine the number of replicates necessary to reach a power of at least 80%. We will then perform additional replicates, if required, to ensure that the experiment has more than 80% power to detect the original effect.

Protocol 2

Request a detailed protocol

The original data presented is qualitative (images of Western Blots). We used Image Studio Lite v. 4.0.21 (LICOR) to perform densitometric analysis of the presented bands to quantify the original effect size. Levels of phospho-protein were normalized to total protein and then normalized to the control.

Summary of original data.

A204 cellspPDGFRpAKTpERKpFRS2
Control1111
Sunitinib alone0.2640.08451.9521.473
Sunitinib + FGF0.3370.0925.3508.069
Sunitinib + FGF + PD1730740.3040.0710.3691.013

  1. • FGF activates pFRS2 and pERK in the presence of Sunitinib.

  2. • PD173074 blocks FGF-induced pFRS2 and pERK activation.

M14 cellspHER3pAKTpERK
Control111
PLX4032 alone0.36671.86450.0524
PLX4032 + NRG13.944711.2110.0539
PLX4032 + NRG1 + Lapatinib1.06661.78630.0571

  1. • NRG1 activates pHER3 and pAKT in the presence of PLX4032.

  2. • Lapatinib blocks NRG1-induced pHER3 and pAKT activation.

KHM-3S cellspEGFRpAKTpERKpMET
Control1111
Erlotinib alone0.0080.6090.181.098
Erlotinib + HGF0.0141.3810.97911.66
Erlotinib + HGF + Crizotinib0.0230.4170.0851.095

  1. • HGF activates pMET and pERK in the presence of Erlotinib.

  2. • Crizotinib blocks HGF-induced pMET and pERK activation.

We have calculated the projected sample size based on a variety of different possible levels of variance (Koller and Wätzig, 2005) using a one-way ANOVA test with an alpha error of 0.05.

  • These power calculations were performed with G*Power software, version 3.1.7 (Faul et al., 2007).

  • The F statistic was calculated at http://statpages.org/anova1sm.html.

  • The ηP2 was calculated using the formula on the spreadsheet accessed from Lakens and colleagues (Lakens, 2013).

A204 cells
2% VariancepPDGFRpAKTpERKpFRS2
F(3, 8)2884.51336189.00644400.83415183.0738
ηp²0.9990763770.9995693140.9993944210.999485769
Effect size f32.889148.1754840.6240344.08686
Power99.99%99.99%99.99%99.99%
Total sample size across all groups8888
15% variancepPDGFRpAKTpERKpFRS2
F(3, 8)51.28023644110.026780478.2370506792.14353422
ηp²0.9505686790.9763369860.9670390090.971873631
Effect size f4.3852126.4233985.4165395.87825
Power99.99%99.99%99.99%99.99%
Total sample size across all groups8888
28% variancepPDGFRpAKTpERKpFRS2
F(3, 8)14.7169045931.5765632722.453235226.44425408
ηp²0.8465984560.9221257260.8938424730.908396348
Effect size f2.3492213.4411062.9017173.149063
Power91.97%99.79%98.43%99.35%
Total sample size8888
40% variancepPDGFRpAKTpERKpFRS2
F(3, 8)7.2112832515.47251611.0020852512.9576845
ηp²0.730038450.8529885980.8049078160.829326246
Effect size f1.6444552.4087742.0312022.204344
Power96.95%93.12%83.18%88.55%
Total sample size across all groups12888

For each percent variance, the relative standard deviation of the approximated phospho-protein level was used to calculate the F statistic from a one-way ANOVA analysis, which was converted to ηP2 (the ratio of variance attributed to the effect and the effect plus its associated error variance from the ANOVA), and then used to determine the effect size (Cohen's f) and the needed sample size to obtain at least 80% power. The actual power obtained is listed.

M14 cells
2% VariancepHER3pAKTpERK
F(3, 8)4297.46015283.29946645.7378
ηp²0.9993798630.999495520.999598901
Effect size f40.1440844.5111149.92144
Power99.99%99.99%99.99%
Total sample size across all groups888
15% variancepHER3pAKTpERK
F(3, 8)76.3992906793.92532267118.1464498
ηp²0.9662728850.9723924660.977927341
Effect size f5.3525455.9348126.656194
Power99.99%99.99%99.99%
Total sample size across all groups888
28% variancepHER3pAKTpERK
F(3, 8)21.9258168426.9556091833.90682551
ηp²0.8915657840.9099776570.927087448
Effect size f2.8674353.1793643.565818
Power98.24%99.42%99.88%
Total sample size888
40% variancepHER3pAKTpERK
F(3, 8)10.7436502513.208248516.6143445
ηp²0.8011481250.83202010.861694667
Effect size f2.0072042.2255552.496073
Power82.32%89.11%94.57%
Total sample size across all groups888
KHM-S3 cells
2% VariancepEGFRpAKTpERKpMET
F(3, 8)7271.8941594.15613697.78226041.5258
ηp²0.9996334260.9983300170.9992793670.999558805
Effect size f52.2203224.4501237.23847.59802
Power99.99%99.99%99.99%99.99%
Total sample size across all groups8888
15% variancepEGFRpAKTpERKpMET
F(3, 8)129.278115628.3405528965.73835022107.4049031
ηp²0.9797895250.9139985230.9610165050.975773338
Effect size f6.9627073.2600164.9650666.346404
Power99.99%99.57%99.99%99.99%
Total sample size across all groups8888
28% variancepEGFRpAKTpERKpMET
F(3, 8)37.10158.1334494918.8662357130.82411122
ηp²0.9329446920.7530890750.8761585120.920376091
Effect size f3.7300221.7464372.6598573.399859
Power99.94%98.31%96.62%99.75%
Total sample size across all groups81288
40% variancepEGFRpAKTpERKpMET
F(3, 8)18.1797353.985390259.244455515.1038145
ηp²0.8720802420.599121490.7761196110.84993841
Effect size f2.6110151.2225061.86192.379901
Power96.09%94.83%99.19%92.58%
Total sample size across all groups816128

In order to produce quantitative replication data, we will run the experiment three times. Each time we will quantify band intensity. We will determine the standard deviation of band intensity across the three biological replicates and combine this with the mean from the original study to simulate the original effect size. We will use this simulated effect size to determine the number of replicates necessary to reach a power of at least 80%. We will then perform additional replicates, if required, to ensure that the experiment has more than 80% power to detect the original effect.

References

Article and author information

Author details

  1. Edward Greenfield

    Monoclonal Antibody Core Facility, Dana-Farber Cancer Institute, Boston, United States
    Contribution
    EG, Drafting or revising the article
    Competing interests
    EG: The Monoclonal Antibody Core Facility is a Science Exchange associated laboratory.

  2. Erin Griner

    University of Virginia, Charlottesville, United States
    Contribution
    EG, Drafting or revising the article
    Competing interests
    No competing interests declared.

  3. Reproducibility Project: Cancer Biology

    Contribution
    RP:CB, Conception and design, Drafting or revising the article
    For correspondence
    fraser@scienceexchange.com
    Competing interests
    RP:CB: EI, FT and JL are employed and holds shares in Science Exchange Inc.
    1. Elizabeth Iorns, Science Exchange, Palo Alto, United States
    2. William Gunn, Mendeley, London, United Kingdom
    3. Fraser Tan, Science Exchange, Palo Alto, United States
    4. Joelle Lomax, Science Exchange, Palo Alto, United States
    5. Timothy Errington, Center for Open Science, Charlottesville, United States

Funding

Laura and John Arnold Foundation

  • Reproducibility Project: Cancer Biology

The Reproducibility Project: Cancer Biology is funded by the Laura and John Arnold Foundation, provided to the Center for Open Science in collaboration with Science Exchange. The funder had no role in study design or the decision to submit the work for publication.

Acknowledgements

The Reproducibility Project: Cancer Biology core team would like to thank the original authors, in particular Dr Jeff Settleman, for generously sharing critical information to ensure the fidelity and quality of this replication attempt. We would also like to thank the following companies for generously donating reagents to the Reproducibility Project: Cancer Biology: American Type Culture Collection (ATCC), BioLegend, Charles River Laboratories, Corning Incorporated, DDC Medical, EMD Millipore, Harlan Laboratories, LI-COR Biosciences, Mirus Bio, Novus Biologicals, and Sigma-Aldrich.

Copyright

© 2014, Greenfield et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 1,730
    views
  • 108
    downloads
  • 4
    citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Edward Greenfield
  2. Erin Griner
  3. Reproducibility Project: Cancer Biology
(2014)
Registered report: Widespread potential for growth factor-driven resistance to anticancer kinase inhibitors
eLife 3:e04037.
https://doi.org/10.7554/eLife.04037

Share this article

https://doi.org/10.7554/eLife.04037

Categories and tags

Research organism

Further reading

  1. Widespread potential for growth-factor-driven resistance to anticancer kinase inhibitors

    TR Wilson, J Fridlyand ... J Settleman
    Nature
    1. Cancer Biology

    Reproducibility Project: Cancer Biology

    Edited by Roger J Davis et al.
    Collection Updated

    Investigating reproducibility in preclinical cancer research.

    1. Cancer Biology
    2. Computational and Systems Biology

    Unveiling the influence of tumor and immune signatures on immune checkpoint therapy in advanced lung cancer

    Nayoung Kim, Sehhoon Park ... Myung-Ju Ahn
    Research Article

    This study investigates the variability among patients with non-small cell lung cancer (NSCLC) in their responses to immune checkpoint inhibitors (ICIs). Recognizing that patients with advanced-stage NSCLC rarely qualify for surgical interventions, it becomes crucial to identify biomarkers that influence responses to ICI therapy. We conducted an analysis of single-cell transcriptomes from 33 lung cancer biopsy samples, with a particular focus on 14 core samples taken before the initiation of palliative ICI treatment. Our objective was to link tumor and immune cell profiles with patient responses to ICI. We discovered that ICI non-responders exhibited a higher presence of CD4+ regulatory T cells, resident memory T cells, and TH17 cells. This contrasts with the diverse activated CD8+ T cells found in responders. Furthermore, tumor cells in non-responders frequently showed heightened transcriptional activity in the NF-kB and STAT3 pathways, suggesting a potential inherent resistance to ICI therapy. Through the integration of immune cell profiles and tumor molecular signatures, we achieved an discriminative power (area under the curve [AUC]) exceeding 95% in identifying patient responses to ICI treatment. These results underscore the crucial importance of the interplay between tumor and immune microenvironment, including within metastatic sites, in affecting the effectiveness of ICIs in NSCLC.