Figure 1.
Design of transgenic constructs.
Tet–tetracycline; mp-1 pA–poly A tract from the mouse protamine 1 gene; C–CCSP-rtTA. Bitransgenic mice harboring both CCSP-rtTA and EGFRL858R+T790M transgenes were labeled “C/L858R+T790M”. Bitransgenic mice harboring the EGFRT790M and EGFRL858R transgenes were labeled “C/T790M” and “C/L858R”, respectively. The latter strain of mice were previously described [9].
Figure 2.
Inducible, lung-specific expression of the mutant EGFR transgene in C/L858R+T790M mice (line 51).
A, B. RT-PCR performed in the presence or absence of reverse transcriptase (RT) using transgene-specific primers on mRNA from a bitransgenic animal on dox for 17.5 weeks (A) and various mice (genotypes as indicated) on dox for 5.5 weeks (B). “+” and “−“ denote known positive and negative lung samples derived from bitransgenic and non-transgenic mice on dox, respectively. “C/L+T” denotes C/L858R+T790M animals. C. Immunoblotting with antibodies against EGFRL858R, EGFR Y1092, and actin was performed on lung lysates derived from various mice on and/or off dox for varying periods of time; W–weeks. Genotypes are as indicated. “C/L+T” denotes C/L858R+T790M animals.
Figure 3.
C/L858R+T790M mice fed doxycycline develop lung adenocarcinomas.
Hematoxylin and eosin (H&E)-stained sections (A–F) of lungs derived from three different bitransgenic progeny of line 51 displaying different adenocarcinoma histological subtypes, including solid (A), bronchioloalveolar (A, C), and papillary (B, D), and one of line 12 with solid and bronchioloalveolar subtypes (E), fed dox for 11 (A, C), 16 (D), 17.5 (B), and 32 weeks (E), respectively. Lungs from bitransgenic mice fed a normal diet are shown in F. Lung tumors are composed mainly of type II pneumocytes, as indicated by the lack of staining for the Clara cell marker, CC26 (G) and positive staining for the type II cell marker, surfactant protein-C (SP-C) (H). Bars, 100 microns.
Figure 4.
Lung tumors in C/L858R+T790M mice require expression of mutant EGFR for tumor maintenance.
A. Serial axial lung field MR images from a mouse administered dox for 21 weeks (W) and then fed a normal diet for the indicated amounts of time. Arrows indicate lung opacities; H, heart. B. Tumor volumes were quantified in the images depicted in A, using imaging software. See methods for details. C. H&E-stained sections from lungs of bitransgenic mice on dox for 17 weeks with tumor by MRI and then taken off dox for the indicated times. Left–degenerating tumor surrounding a scar; right–dying tumor cells surrounded by an inflammatory response; arrowed–island of viable tumor cells.
Figure 5.
Lung tumors in C/L858R+T790M mice are resistant to treatment with erlotinib.
A. Serial axial (left) and coronal (right) images from C/L858R or C/L858R+T790M mice, treated with erlotinib for 8 days. Arrow depicts tumor; H–heart. B. Histological sections derived from the lungs of C/L858R (upper panel) or C/L858R+T790M (lower panel) mice treated with erlotinib for 0 or 8 days. Tumors were stained for the proliferation marker, Ki-67, and phospho-EGFR (pEGFR, Y1016). Far right: H&E staining reveals significant treatment effect in a representative C/L858R tumor, but viable cells in the corresponding C/L858R+T790M tumor. C. Tumor volumes were quantified in the MR images obtained from the individual mice pre- and post-treatment with placebo or erlotinib. See Table S1 and methods for details. For comparison, tumor volumes from a drug-sensitive C/L858R mouse pre- and post-erlotinib treatment and from a C/L858R+T790M mouse on and off dox for the indicated times are also displayed. W–weeks; D–days. D. Immunoblotting with antibodies against EGFRL858R, EGFR Y1092, and actin was performed on lysates derived from lungs of tumor-bearing C/L858R+T790M mice, treated with placebo or erlotinib for the indicated times.
Figure 6.
Characteristics of C/T790M mice.
A. Axial MR image from a line 37-derived bitransgenic animal fed dox for 32 weeks, revealing tumor (T); H: heart. B. Gross histology shows corresponding lung lesions. C. RT-PCR was performed on lungs from this animal in the presence or absence of RT using transgene-specific primers on mRNA from various tissues; “+”–positive control; “−“– negative control. D. Images of H&E-stained lung sections from various T790M transgene bearing mice. Lungs from bitransgenic line 37 (i) and 8 animals (ii-iii) displayed features of papillary adenocarcinoma (i), solid adenocarcinoma surrounded by bronchioloalveolar carcinoma (ii), and solid/papillary adenocarcinoma surrounded by bronchioloalveolar carcinoma (iii). Tumors were negative for CC26 (iv), positive for SP-C (v), and positive for phospho-EGFR (Y1016) (vi). Bars, 100 microns. E. Serial MR images of a bitransgenic C/T790M animal (line 8) administered dox 18 weeks (left) and then fed a normal diet for the indicated times; W–weeks. F. H&E-stained sections from lungs of bitransgenic tumor-bearing mice withdrawn from dox for the indicated times. Left–degenerating tumor with scattered viable cells; Right–hemosiderin-laden macrophages engulfing a degenerating tumor. G. Tumor volumes were quantitated in the MR images obtained from the individual mice pre- and post-treatment with placebo or erlotinib. See Table S2 and methods for details. W–weeks, D–days.
Figure 7.
Comparison of phenotypes induced by various EGFR mutants.
Representative pictures of H&E-stained lung sections from C/T790M (line 8) and C/L858R+T790M (line 51) animals, fed dox for the indicated amounts of time. W–weeks.
Figure 8.
Effect of 17-AAG on lung tumors from C/L858R+T790M animals.
A. Immunoblotting analyses of lung lysates derived from bitransgenic animals (line 51) treated with placebo or 17-AAG and sacrificed 6 hours later. Lane numbers correspond with the animal study numbers in Table S3. Lysates were probed with the indicated antibodies; p-phospho. B. Histology of lung tumors from bitransgenic mice treated with either placebo or 17-AAG for indicated times; H–hours; D–days. Lungs from placebo treated mice had viable tumor (upper panels), while lungs from drug-treated animals displayed areas of tumor necrosis surrounded by rims of viable cells (two middle panels), and degenerating tumors with inflammatory response (lower panels). Bars, 100 microns. C. Tumor volumes were quantitated in the MR images obtained from the individual mice pre- and post-treatment with placebo or 17-AAG. See Table S3 and methods for details. D–days; W–weeks.