Inhibition of mTOR pathway by everolimus cooperates with EGFR inhibitors in human tumours sensitive and resistant to anti-EGFR drugs

Abstract

Inhibition of a single transduction pathway is often inefficient due to activation of alternative signalling. The mammalian target of rapamycin (mTOR) is a key intracellular kinase integrating proliferation, survival and angiogenic pathways and has been implicated in the resistance to EGFR inhibitors. Thus, mTOR blockade is pursued to interfere at multiple levels with tumour growth. We used everolimus (RAD001) to inhibit mTOR, alone or in combination with anti-EGFR drugs gefitinib or cetuximab, on human cancer cell lines sensitive and resistant to EGFR inhibitors, both in vitro and in vivo. We demonstrated that everolimus is active against EGFR-resistant cancer cell lines and partially restores the ability of EGFR inhibitors to inhibit growth and survival. Everolimus reduces the expression of EGFR-related signalling effectors and VEGF production, inhibiting proliferation and capillary tube formation of endothelial cells, both alone and in combination with gefitinib. Finally, combination of everolimus and gefitinib inhibits growth of GEO and GEO-GR (gefitinib resistant) colon cancer xenografts, activation of signalling proteins and VEGF secretion. Targeting mTOR pathway with everolimus overcomes resistance to EGFR inhibitors and produces a cooperative effect with EGFR inhibitors, providing a valid therapeutic strategy to be tested in a clinical setting.

Figure 1

(A) Cells were treated with everolimus (0.01–20 μM). (B and C) Cells were treated with cetuximab (0.1–1 μg ml⁻¹), and gefitinib (0.1–1 μM), in presence of everolimus, 0.1 μM. (D) Western blotting on total cell lysates from cells treated with 1 μM gefitinib, 1 μg ml⁻¹ cetuximab or 0.1 μM everolimus. Lane 1: untreated control; lane 2: cetuximab; lane 3: gefitinib; lane 4: everolimus.

Figure 2

(A) GEO and GEO-GR cells were treated with everolimus (0.1 μM) or gefitinib (5 μM) alone or in combination. The results are statistically significant for everolimus plus gefitinib vs control, vs everolimus alone and vs gefitinib alone (two-sided P<0.0001). For GEO cells, the results are statistically significant for both gefitinib and everolimus vs control (two-sided P<0.0001), whereas for GEO-GR cells only for everolimus vs control (two-sided P<0.0001). (B) Western blotting on total cell lysates from GEO and GEO-GR cells treated with 1 μM gefitinib and/or 0.1 μM everolimus. Lane 1: untreated control; lane 2: gefitinib; lane 3: everolimus; lane 4: gefitinib plus everolimus.

Figure 3

ELISA assays for hVEGF on conditioned media from (A) cancer cells treated with everolimus (0.1 μM) and from (B) GEO and (C) GEO-GR cancer cells treated with everolimus (0.1 μM) and gefitinib (5 μM) alone or in combination. HUVEC cells were treated with everolimus from 1 nM to 1 μM (D) or with everolimus 0.01 μM and gefitinib 1 μM alone or the combination (E). HUVECs were incubated on solidified matrigel in presence of everolimus 0.1 μM, gefitinib 5 μM or the combination. The positive control was matrigel with VEGF 100 ng ml⁻¹. Photographies were performed at 0 and 24 h (F). All the results are statistically significant for each treatment vs control (two-sided P<0.0001).

Figure 4

(A and B) After 7 days following tumour injection, 10 mice were treated. Tumour sizes among different treatment groups at day 56 following GEO or GEO-GR cell injection resulted statistically significant for everolimus plus gefitinib vs control, vs everolimus alone and vs gefitinib alone (two-sided P<0.0001). Bars, s.d. (C and D) Mice survival resulted statistically significant for everolimus plus gefitinib vs control, vs everolimus alone and vs gefitinib alone (two-sided P<0.0001).

Figure 5

(A) Western blotting on total lysates from tumour specimens of two mice killed on day 25. Lane 1: untreated control; lane 2: gefitinib; lane 3: everolimus; lane 4: everolimus plus gefitinib. Bars, s.d. ELISA assays for hVEGF on total lysates from tumour specimens (B) and on mice serum (C). The results are statistically significant for everolimus plus gefitinib vs control and for everolimus vs control (two-sided P<0.0001), whereas they are not statistically significant for gefitinib vs control.