Registered report: Melanoma exosomes educate bone marrow progenitor cells toward a pro-metastatic phenotype through MET

The Reproducibility Project: Cancer Biology seeks to address growing concerns about reproducibility in scientific research by conducting replications of selected experiments from a number of high-profile papers in the field of cancer biology. The papers, which were published between 2010 and 2012, were selected on the basis of citations and Altmetric scores (Errington et al., 2014). This Registered Report describes the proposed replication plan of key experiments from “Melanoma exosomes educate bone marrow progenitor cells toward a pro-metastatic phenotype through MET” by Peinado and colleagues, published in Nature Medicine in 2012 (Peinado et al., 2012). The key experiments being replicated are from Figures 4E, as well as Supplementary Figures 1C and 5A. In these experiments, Peinado and colleagues show tumor exosomes enhance metastasis to bones and lungs, which is diminished by reducing Met expression in exosomes (Peinado 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 in eLife. DOI: http://dx.doi.org/10.7554/eLife.07383.001


Introduction
Exosomes are nanovesicles up to 100 nm in size that are derived from endosomal membranes and secreted by cells as a means of intercellular communication (Mathivanan et al., 2010). They contain a wide array of cargo including proteins, cytokines, and nucleic acids (Kharaziha et al., 2012). Recently, exosomes have been shown to play multiple roles in promoting carcinogenesis, including the regulation of metastatic niche formation, regulation of tumor immune response, and chemotherapeutic resistance (Tickner et al., 2014). Peinado and colleagues reported that exosomes derived from melanoma cells promoted metastasis through education of bone marrow-derived cells in order to prime the pre-metastatic niche and increase vascularization. They further showed that exosomemediated metastasis was dependent on expression of MET in exosomes, and that MET protein was increased in exosomes found in patients with advanced melanoma (Peinado et al., 2012). MET is an oncogenic receptor tyrosine kinase that promotes proliferation, motility, and migration, and is often aberrantly activated in tumors (Gherardi et al., 2012;Trusolino et al., 2010). These findings indicate that exosomal MET may be a potential therapeutic target or biomarker for metastatic disease.
Supplementary Figure 1C characterizes exosomes isolated from B16-F10 melanoma cells using electron microscopy imaging and Western blotting for exosome protein markers. Supplementary Figure 5A further characterizes these exosomes by assessing the levels of MET and pMET after shRNA-mediated depletion of MET in B16-F10 cells. These figures are essential to reproduce as they validate the expression of key proteins in exosomes that will subsequently be used to replicate Figure 4E. These experiments will be replicated in Protocols 1 and 2.
In Figure 4E, Peinado et al. (2012) reported that reduction of MET protein in melanoma exosomes reduced metastasis of B16-F10 melanoma cells to lung and bone (Peinado et al., 2012). Mice were pre-treated with exosomes isolated from B16-F10 melanoma cells expressing shRNAs directed against Met or control shRNAs, and then B16-F10 cells were injected subcutaneously. Primary tumor growth and metastasis to lungs and bone were assessed by luciferase imaging. This key experiment tests one of the central findings of the paper, namely that MET is necessary for melanoma exosome-mediated promotion of metastasis. This experiment will be replicated in Protocol 3.

Materials and methods
Protocol 1: Lentiviral knockdown of Met or non-silencing control in B16-F10 cells This protocol utilizes shRNA to knock down Met in B16-F10 cells. This experiment generates key reagents (B16-F10 shMet and B16-F10 shScramble) that will subsequently be used in Protocols 2 and 3.
Sampling . Experiment will be conducted once.
Note: Information that these are stable transfectants was communicated by authors. Deliverables . Data to be collected:

Materials and reagents
. Detailed notes on cell culturing of both stable cell lines generated. . Transduction efficiency as a percentage of GFP-positive cells.
. Fluorescent microscopy images of GFP-positive cells. . Sample delivered for further analysis: . B16-F10 shMet and B16-F10 shScramble cells for exosome purification and Western blot analysis of METexpression for Protocol 2.

Confirmatory analysis plan
No analysis performed.

Known differences from the original study
Similar to the original study, the transduction efficiency, determined by GFP expression, and the knockdown efficiency, determined by Western blot, will be measured (Protocol 2). The original protocol for selecting GFP-positive cells included a step to perform fluorescence-activated cell sorting (FACS) of the population to achieve a 95-99% GFP-positive population prior to puromycin selection (information provided by authors). The replication attempt will not include the FACS sorting step and instead will select the stable cells for at least 28 d before further analysis and will maintain the cell lines when not used in experimental procedures under puromycin selection. All known differences are listed in the Materials and reagents section above with the originally used item listed in the Comments section. All differences have the same capabilities as the original and are not expected to alter the experimental design.

Provisions for quality control
The cell lines used in this experiment will undergo STR profiling to confirm their identity and will be sent for mycoplasma testing to ensure there is no contamination. Untransduced B16-F10 cells will be used to confirm the GFP-negative cell population and during puromycin selection to ensure efficient transduction occurs. Detailed cell culture notes will be recorded and made available to monitor growth rates of B16-F10 shScramble and B16-F10 shMet cells. All of the raw data will be uploaded to the project page on the Open Science Framework (OSF) (https://osf.io/ewqzf/) and made publically available.
Protocol 2: Exosome purification and Western blot analysis of MET and phospho-MET expression This protocol isolates exosomes from B16-F10 shScramble and B16-F10 shMet cells and then utilizes Western blot to characterize protein expression in cells generated in Protocol 1 and exosomes purified from this protocol. MET and pMET protein expression will be determined to verify MET knock down in exosomes, and exosome markers will also be assessed. i. Estimated yield of exosomes should be around 1-2 mg/1 x 10 6 cells. k. Characterize exosome pellet using standard Nanosight NTA analysis.
i. 10-20 mg of exosome protein is needed for analysis.
ii. Report on size distribution and concentration of exosomes. l. Prepare 30 mg of total protein per sample with 4X Laemmli buffer.
i. Store at -20˚C until analysis. 6. Load 30 mg of total protein with sample buffer in each lane on an SDS-PAGE gel with a protein molecular weight marker. 7. Run electrophoresis at constant voltage (100 V) for~1-2 hr in 1X electrophoresis buffer following manufacturer instructions. 8. Transfer gel to membrane using a wet transfer system at constant amperage (350 mA) for 45 min in 1X Tris-glycine buffer supplemented with 20% methanol. a. Pre-soak membrane with methanol and then 1X transfer buffer before assembly. 9. Block membranes in SuperBlock following manufacturer's instructions. 10. Probe membranes with the following primary antibodies overnight at 4˚C diluted in SuperBlock. a. mouse anti-HSC70; use at 1:500; 70 kDa b. mouse anti-TSG101; use at 1:500; 45 kDa c. anti-CD63; use at 1:1000; 53 kDa d. mouse anti-MET; use at 1:1000; 145 kDa e. rabbit anti-pMET Tyr1234/5; use at 1:1000; 145 kDa f. rabbit anti-GAPDH; use at 1:1000; 37 kDa 11. Wash membranes three times 10 min in Ticket Tax Box Service (TTBS).

Deliverables
. Data to be collected: . Exosome characterization data (including protein concentration using a BCA kit). . Images of probed membranes (full images with ladder) (compare to Figure S1C [right panel]).
. Quantified levels of total MET and Phospho-MET normalized to GAPDH for all conditions (compare to Figure S5A).

Confirmatory analysis plan
This experiment assesses if knockdown of Met alters total MET and phospho-MET expression in cells and exosomes. This replication attempt will perform the following statistical analysis: . Statistical analysis: . One-way multivariate analysis of variance (MANOVA) comparing the relative mean of photon signal for total MET and phospho-MET normalized to GAPDH from B16-F10 shScramble and B16-F10 shMet cells.
. Planned comparisons with the Bonferroni correction: 1. Total MET in B16-F10 shScramble exosomes compared to B16-F10 shMet exosomes 2. Phospho-MET in B16-F10 shScramble exosomes compared to B16-F10 shMet exosomes . Meta-analysis of effect sizes: . Compute the effect sizes of each comparison, compare them against the reported effect size in the original paper, and use a meta-analytic approach to combine the original and replication effects, which will be presented as a forest plot.

Known differences from the original study
The cell lines when not used in experimental procedures will be maintained under puromycin selection. The replication attempt will include an additional exosome marker, CD63, not included in the original study. All known differences are listed in the Materials and reagents section above with the originally used item listed in the Comments section. All differences have the same capabilities as the original and are not expected to alter the experimental design.

Provisions for quality control
The cell lines used in this experiment will undergo STR profiling to confirm their identity and will be sent for mycoplasma testing to ensure there is no contamination. This protocol analyzes the knockdown efficiency of c-Met in B16-F10 cells. Isolated exosomes are characterized by NanoSight and are analyzed for typical exosome markers by Western blot, including CD63, an additional marker not included in the original study. All of the raw data, including the image files and quantified bands from the Western blot, will be uploaded to the project page on the OSF (https://osf.io/ewqzf/) and made publically available.

Protocol 3: Exosome-dependent MET signaling on primary tumor growth and metastasis
This experiment tests the effect of exosome-derived MET on primary growth and metastasis of melanoma cells. This is a replication of Figure 4E, which assesses metastasis in lungs and bone using bioluminescent imaging.
Sampling & Experiment will use seven mice per cohort for a minimum power of 82%.
.  The original study indicated mice were 8-10 weeks old, however the authors clarified that the mice were 6 weeks old. i. An aliquot of the exosome preparation can be stored at -20˚C until Nanosight analysis. j. Measure protein concentration with a BCA kit following manufacturer's instructions.
i. An aliquot of each exosome preparation will be stored at -20˚C and analyzed all at once following the final preparation. ii. 10-20 mg of exosome protein is needed for analysis. iii. Report on size distribution and concentration of exosomes. iv. Report the number of exosomes per mg protein (as measured by BCA) for each sample. l. Prepare samples at a concentration of 50 ng/ml to achieve 5 mg of total protein diluted in 100 ml of PBS. 5. Following protein quantification of each preparation, inject intravenously, via retro-orbital injection, freshly isolated B16-F10 shScramble or B16-F10 shMet exosomes, or synthetic unilamellar 100 nm liposomes into 6-week-old C57BL/6 female mice three times a week for a total of 28 d. a. Sample injection schedule: i. Each cohort will be injected on Monday, Wednesday, and Friday with a fresh exosome preparation (>35 mg total) each week for a total of 4 weeks. ii.
Step 4 will be performed for each injection day. b. It is crucial to inject fresh exosomes every time, do not freeze down, and inject right after purification and quantification following steps 2-4 above. c. Volume of injection is 100 ml. d. Amount of synthetic liposomes injected, 1.25 mg of L-a-phosphatidylcholine (PC) will mimic 5 mg of exosome protein, which is based on a theoretical 4:1 protein:PC ratio (communicated by authors). i. Dilute 1.92 ml of a 1:20 dilution of the stock concentration of synthetic liposomes into 100 ml PBS for each injection. 6. After 28 d, inject 1 x 10 6 B16-F10-luciferase cells diluted in 100 ml of PBS subcutaneously in the flank of mice. 7. Measure primary tumor volume three times a week for a total of 21 d.
a. Use calipers to measure width and height with volume determined as (length x width 2 )/2. (additional recorded information) i. Note: tumor volume detection will be limited to <1000 mm 3 . b. Record latency. (additional recorded information) i. Note: Perform Steps 8 through 10 (luciferin injection, euthanasia, dissection, and imaging) from mice from different cohorts in parallel (i.e., one from each of the three cohorts) so variation during the procedure is equal across cohorts. (additional detail) 8. After 21 d anesthetize mice and inject 50 mg/kg of D-luciferin via retro-orbital injection in 100-200 ml PBS (volume depending on the body weight). a. Weigh mice. b. Use isoflurane and O 2 to anesthetize mice. 9. Five min later euthanize mice by cervical dislocation under anesthesia and dissect tissues (lungs and femurs). a. Dissect out primary tumors and record weight (additional parameter). 10. Image dissected primary lungs and bones (femurs) for luciferase expression in IVIS Imaging system a. Record the time from euthanasia to imaging for each mouse. b. Record photon flux. a. Take two-three exposure times for each sample. i. Use same exposure time for each tissue from all mice during analysis.
Deliverables & Data to be collected: . Mouse health records (injection schedule, time from tumor cell injection to detectable tumors [latency], weight of mice at end of experiment, mortality report) . Exosome characterization data (including protein concentration using a BCA kit). . Raw numbers and calculated tumor volume for all mice, and graph of tumor volume versus time for all conditions during course of treatment.
. Time of euthanasia to imaging for each mouse. . Images of lungs and bones for luciferase expression (compare to Figure 4E). . Raw photon flux values of all analyzed images of each tissue for all conditions using the same exposure time (compare to Figure 4E).
. Final weight of tumors. Compute the effect sizes of each comparison, compare them against the reported effect size in the original paper and use a meta-analytic approach to combine the original and replication effects, which will be presented as a forest plot.
Comparison of primary tumor growth rates.
. This is exploratory analysis. We will measure tumor growth rates across all mouse cohorts over the length of the study. These data were collected, but not reported in the original study, and found to not be different. We will plot growth curves and calculate area under the curve for each mouse. We will then perform a one-way analysis of variance (ANOVA) analysis, with the following planned comparisons with the Fisher's LSD test: 1. Synthetic unilamellar liposomes compared to B16-F10 shScramble exosomes 2. B16-F10 shMet exosomes compared to B16-F10 shScramble exosomes . Comparison of final primary tumor weights. . This is exploratory analysis. We will measure tumor weights across all mouse cohorts at the end of the study. These data were not reported in the original study. We will perform a one-way ANOVA analysis, with the following planned comparisons with the Fisher's LSD test: 1. Synthetic unilamellar liposomes compared to B16-F10 shScramble exosomes 2. B16-F10 shMet exosomes compared to B16-F10 shScramble exosomes

Known differences from the original study
The cell lines when not used in experimental procedures will be maintained under puromycin selection. The number of exosomes injected (based on protein content) will be reported for each preparation from each cohort. The original data on primary tumor growth were not shown and final primary tumor weights were not recorded. The replication attempt will record and present these data as well as tumor latency. All known differences are listed in the Materials and reagents section above with the originally used item listed in the Comments section. All differences have the same capabilities as the original and are not expected to alter the experimental design.

Provisions for quality control
The cell lines used in this experiment will undergo STR profiling to confirm their identity and will be sent for mycoplasma testing to ensure there is no contamination. Isolated exosomes will be injected immediately after protein quantification and will be characterized by NanoSight following the final preparation to ensure the integrity of the samples. Exosomes and luciferin will be injected intravenously, via retro-orbital injection, similar to the original study. While it will be attempted to be the same for all animals, the time from euthanasia to imaging for each mouse will be recorded as an additional quality control parameter. All of the raw data, including the image files and photo flux values, will be uploaded to the project page on the OSF (https://osf.io/ewqzf/) and made publically available.

Power calculations
For additional details on power calculations, please see analysis scripts and associated files on the OSF: https://osf.io/nyb8d/

Protocol 1
Not applicable

Protocol 2
Summary of original data reported in Figure S5A (calculated from data shared by original authors): The replication will also include analysis of total MET and phospho-MET in exosomes from B16-F10 shScramble and B16-F10 shMet cells. We will use the same calculated sample size as determined for the cells with the assumption that the cells and exosomes have similar values, which Peinado and colleagues (2012) present in Figure 4A.
Test family . two-tailed t test, difference between two independent means, Bonferroni's correction: alpha error = 0.025 Power calculations performed with G*Power software, version 3.1.7 (Faul et al., 2007).