Mutated EGFR or TGFα autocrine production impair the biological effects of MET targeting. A-B) Cells, bearing the doxycycline-inducible MET shRNAs, were transduced with the mutated form of EGFR (L858R) (A) or with the Hys-tagged TGFα (B). Top, WB showing the expression of transduced molecules, compared to cells transduced with the empty vector; Upper graph, growth curves of GTL16, GTL16-L858R (A) or GTL16-TGFα (B). The effect of MET silencing (DOXY) or inhibition (PHA) was compared with untreated cells. After six days, only cells bearing EGFR-L858R or TGFα were able to grow when MET was inactivated (A ** P < 0,01; B *** P < 0,001). Lower graph, anchorage-independent growth assays. Cells having an activated EGFR (GTL16-L858R, A or GTL16-TGFα, B) formed colonies in soft-agar in the absence of MET signaling, while wt cells did not (A-B ** P < 0,01). Stimulation of GTL16-L858R cells with a physiological concentration of EGF (1 ng/ml) resulted in further increase of anchorage-independent growth, only in the absence of MET signaling. C) The cells used for biological assays were injected subcutaneously in CD1 nude mice. MET silencing was maintained in vivo by adding doxycycline (1 mg/ml) to mice drinking water. The graph shows the Kaplan-Meier-like analysis of tumor latency. After 40 days, 70% of mice injected with GTL16 cells and treated with doxycycline were tumor-free. On the contrary, only 30% of mice injected with GTL16-L858R and GTL16-TGFα were tumor-free, despite MET silencing. EGFR-L858R and TGFα had no significative effect in promoting tumor growth in untreated GTL16.