Fig 1.
In vitro validation of the H1975 EGFR mutant cell lines.
(A) Relative mutant allele frequency was compared in cDNA from each cell line by Digital droplet PCR. (B) WB of total EGFR levels in the H1975L858R/T790M cell line before and after lentivirus infection with a shEGFR and in the H1975 cell lines. hsc70 levels are shown as loading control. Values beneath blots are relative levels of T-EGFR compared to the H1975L858R/T790M cell line from 2 independent experiments (C) WB of phospho and total EGFR in H1975 derivative cell lines untreated or treated with EGF (100ng/mL) for 15 min, Erlotinib (1μM, 5μM or 10 μM) for 1 hour or both. hsc70 levels were used as loading control. Quantification of the WB is shown above the figure from 3 independent experiments. Error bars is SD (*p<0.05 and **p<0.001).
Fig 2.
EGFR-MET interaction is modulated by EGFR mutations.
(A) Western blot (WB) of total MET levels in H1975 derived cells. Tubulin levels are shown as loading control. Values beneath blots are relative levels of MET compared to the total levels in the H1975L858R/T790M cell line from 2 independent experiments +/−SD (B) MET (7q31) (red signal) copy number analysis by FISH in the three H1975 cell lines using the Leica Kreatech C-MET (7q31)/SE7 FISH probe (KBI-10719). The green signal indicates the chromosome 7 centromere control probe. Scale bar 10 mm. Average copy number and ratio between MET and chromosome 7 centromere probe are also indicated (n = 30 cells). (C) Immunofluorescence of total EGFR (Alexa546 –red in the image) and MET (Cyanine 5 –green in the image) in H1975 derived cells. Hoescht dye was used to stain the nuclei of the cells. Merge panels are also shown. Bars, 20 μm. (D) Co-immunoprecipitation (IP) of EGFR in H1975 derived cell lines. The EGFR antibody was used to immunoprecipitate. EGFR and MET levels are shown in both bound and input fractions. The gels shown in the figure were run separately for the bound and input fractions, as indicated by the dotted line, under the same experimental conditions. (E) Fluorescence lifetime imaging was performed on cells plated to sub-confluence on cover-slips and time-resolved analysis in Tri2. Quantification of average FRET efficiency (*** p < 0.0005) is shown, as well as representative pseudocolour lifetime images showing FRET efficiency and corresponding grayscale donor (EGFR-Alexa 546) and acceptor (MET-Cyanine 5) intensity images. Scale bar: 50μm.
Fig 3.
Inhibition of MET changes the EGFR-MET interaction both in vitro and in vivo.
(A) Fluorescence lifetime imaging performed in the three H1975 cell lines plated on coverslips with or without treatment with SGX523 (5 μM) for 24 hours. Representative pseudocolour lifetime images showing FRET efficiency accompanied by corresponding grayscale donor (EGFR-Alexa 546) and acceptor (MET-Cyanine 5) intensity images are shown. Scale bar: 50μm. (B) Bar graph showing quantification for average FRET efficiency in the three cell lines with or without SGX523 treatment (*** p < 0.0005, ** p = 0.001). (C) Quantification of average FRET of EGFR-MET interaction performed in xenograft tumors from each H1975 cell line in mice receiving mock or SGX523 treatment (60 mg/kg) for 12 days (*** p<0.001, * p<0.05). (D) Representative lifetime images for EGFR:MET FRET in xenograft tumors accompanied by corresponding grayscale donor (EGFR-Alexa 546) and acceptor (MET-Cyanine 5) intensity images. Scale bar, 50μm.
Fig 4.
In vitro and in vivo effects of MET inhibition on cell proliferation.
(A) Representative images of the three cell lines grown at 60% confluence on coverslips and treated or not with SGX523 (5 μM) for 24 hours. Scale bar, 20μm. The BrdU positive nuclei (red) show the cycling cells, and the Hoechst dye was used to stain all the nuclei of the cells (in blue). (B) Quantification of the proliferation rates as described in (A) (*p <0.04, **p<0.005, ***p<0.0001). (C) Soft agar colony formation in the H1975 derivate cell lines. The graph shows the number of colonies after 3 weeks of growing in the presence or absence of SGX523 (5 μM) (*p <0.001). Representative images of untreated and SGX523 treated colonies are also shown. (D) Graphs showing the fold change in volumes in 16 xenografts tumors (8 coming from mice subjected to vehicle and 8 from SGX523 treatment) coming from the three H1975 cell lines during the 12 days of treatment. Tumor volumes were measured at the indicated times using a calliper and calculated in based of the equation 0.4xAxB^2 (A, the long axis and B, the short axis of the tumor). (E) Quantification of phospho-histone H3 (P-H3) staining in the xenograft derived tissue in the presence or absence of SGX523 (**p<0.05) Representative images are also shown. Bar, 250 nm.
Fig 5.
In vitro and in vivo effects of MET inhibition on stroma remodeling.
Quantification and representative images of (A) collagen, (B) α-SMA and (C) cd31 staining of xenografts tumors (FFPE) grown from each H1975 derivate cell line coming from mice treated or not (mock) with SGX523 (60 mg/kg). Bar 250 nm. Quantification of the staining is shown above the images and was performed using Image J. (*p<0.05, **p<0.01, ***p<0.001). EGF (D) and HGF (E) quantification assessed by ELISA in each H1975 derivate cell line supernatants.
Fig 6.
In vitro and in vivo effects of MET inhibition on ERK, AKT and FAK phosphorylation.
(A) WB of phospho- and total EGFR, MET, AKT, ERK and FAK from cell lysates of the different H1975 cell lines treated or not with SGX523 (5 μM) for 24 hours. Gels shown are representative of three experiments, and were run separately for each cell line, as indicated by the dotted line, under the same experimental conditions. Hsc70 levels are also shown as loading control. Values beneath blots are relative levels of each phospho vs total protein comparing the untreated (UT) and the SGX523 treated conditions for each cell line for 3 independent experiments and for 2 independent experiments in the case of the P-FAK/FAK +/−SD (*p<0.05, **p<0.001) (B) WB of phospho- and total-ERK in the extracts coming from the H1975L858R/T790M-derived xenografts tumors in vehicle (mock) and SGX523 treated mice. hsc70 is also shown as loading control. (C) Quantification of the P-ERK vs T-ERK WBs shown in (A) and in (B) (*p<0.05). (D) Schematic model demonstrating effect of SGX523 on EGFR-MET dimerization pattern and consequences for tumor characteristics following treatment.