Acquisition of EGFR TKI resistance and EMT phenotype is linked with activation of IGF1R/NF-κB pathway in EGFR-mutant NSCLC

Abstract

Epithelial-mesenchymal transition (EMT) is clinically associated with acquired resistance to epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKI) in non-small cell lung cancers (NSCLC). However, the mechanisms promoting EMT in EGFR TKI-resistant NSCLC have not been fully elucidated. Previous studies have suggested that IGF1R signaling is involved in both acquired EGFR TKI resistance in NSCLC and induction of EMT in some types of tumor. In this study, we further explored the role of the IGF1R signaling in the acquisition of EMT phenotype associated with EGFR TKI resistance in mutant-EGFR NSCLC. Compared to gefitinib-sensitive parental cells, gefitinib-resistant (GR) cells displayed an EMT phenotype associated with increased migration and invasion abilities with the concomitant activation of IGF1R and NF-κB p65 signaling. Inhibition of IGF1R or p65 using pharmacological inhibitor or specific siRNA partially restored sensitivity to gefitinib with the concomitant reversal of EMT in GR cells. Conversely, exogenous IGF1 induced both gefitinib resistance and accompanying EMT in parental cells. We also demonstrated that IGF1R could phosphorylate downstream Akt and Erk to activate NF-κB p65. Taken together, our findings indicate that activation of IGF1R/Akt/Erk/NF-κB signaling is linked to the acquisition of EGFR TKI resistance and EMT phenotype in EGFR-mutant NSCLC and could be a novel therapeutic target for advanced NSCLC.

Keywords: EGFR TKI resistance; EMT; IGF1R; NF-κB; NSCLC.

Figure 1. Gefitinib-resistance (GR) cells acquired an epithelial-mesenchymal transition (EMT) phenotype and increased migration and invasion abilities

(A) The sensitivity to gefitinib of PC9GR, HCC827GR and their respective parental cells was examined by CCK8 assay. Cells were treated with the indicated concentration of gefitinib for 72h. IC50 values were calculated and marked for each curve. (B) The expression of p-Akt, Akt, p-Erk and Erk of PC9 and PC9GR cells treated with 20nM gefitinib for 2h was measured by Western blot. (C) Morphology of GR and parental cells. Photographs were taken at × 40 magnification. Scale bar: 100μm. The expression of E-cadherin and vimentin was analyzed by immunofluorescence staining. Nuclei were visualized with DAPI staining. (D) The expression of E-cadherin, vimentin, β-catenin, Snail, Slug and ZEB1 of GR and parental cells was examined by Western blotting. (E) Migration and invasion abilities of PC9 and PC9GR cells were examined by transwell assays. The cells were incubated for 8h (for migration) or 24h (for invasion). Those migrated cells remaining on the bottom surface were fixed, stained and photographed. Photographs were taken at ×40 magnification.

Figure 2. IGF1R/NF-κB p65 signaling was activated in GR cells and inhibition of IGF1R/NF-κB p65 increased the sensitivity of GR cells to gefitinib

(A) The expression of IGF1Rβ, p-IGF1Rβ, NF-κB p65, p-p65 of GR and parental cells was examined by Western blotting. (B) The expression of p-p65, p65 in cytoplasmic extracts or nuclear extracts of PC9GR cells was measured by Western blot. β-actin and H3 histone was used as loading control for cytoplasmic extracts and nuclear extracts, respectively. (C) and (D) The sensitivity to 5μM gefitinib of PC9GR cells pre-treated with 0.1μM PPP (C) or 1μM PTL (D) was examined by CCK8 assay. Cells were pre-treated with 0.1μM PPP (C) or 1μM PTL (D) for 2h and then subjected to 5μM gefitinib for 72h. The data are presented as the means ± SEM and normalized to control cells treated with solvent (DMSO). ^* P < 0.05; ^** P < 0.01; ^*** P < 0.001, compared with control.

Figure 3. Inhibition of IGF1R reversed EMT and suppressed migration and invasion in PC9GR cells

(A) The expression of IGF1R, p-IGF1R, β-catenin, vimentin, Snail and Slug of PC9GR cells treated with indicated concentrations of PPP for 72h was examined by Western blotting. (B) The expression of IGF1R, vimentin, Snail and Zeb1 of PC9GR cells transfected with siRNA IGF1R for 72h was examined by Western blotting. (C) and (D) Migration and invasion abilities of PC9GR cells treated with 0.1μM PPP (C) or transfected with siRNA IGF1R (D) were examined by transwell assays. The cells were incubated for 8h (for migration) or 24h (for invasion). Those migrated cells remaining on the bottom surface were fixed, stained, photographed and counted. Photographs were taken at ×40 magnification. (E) Quantification of transwell migration and invasion assays (C and D). The number of cells was counted from at least four independent microscopic fields. (F) Quantification of wound healing assay of PC9GR cells treated with 0.1μM PPP or transfected with siRNA IGF1R. The open wound area was quantified by ImageJ from at least four independent microscopic fields and normalized to the area at time 0. The data are presented as the means ± SEM and normalized to control cells treated with solvent (DMSO) or transfected with siRNA NC. ^* P < 0.05; ^** P < 0.01; ^*** P < 0.001, compared with control.

Figure 4. Inhibition of NF-κB p65 reversed EMT and suppressed migration and invasion in PC9GR cells

(A) The expression of p65, p-p65, IkBa, p-IkBa, vimentin, Slug and Zeb1 of PC9GR cells treated with 1μM PTL for 72h was examined by Western blotting. (B) The expression of p65, p-p65, vimentin, Snail, Slug and Zeb1 of PC9GR cells transfected with siRNA p65 for 72h was examined by Western blotting. (C) and (D) Migration and invasion abilities of PC9GR cells treated with 1μM PTL (C) or transfected with siRNA p65 (D) were examined by transwell assays. The cells were incubated for 8h (for migration) or 24h (for invasion). Those migrated cells remaining on the bottom surface were fixed, stained, photographed and counted. Photographs were taken at ×40 magnification. (E) Quantification of transwell migration and invasion assays (C and D). The number of cells was counted from at least four independent microscopic fields. (F) Quantification of wound healing assay of PC9GR cells treated with 1μM PTL or transfected with siRNA p65. The open wound area was quantified by ImageJ from at least four independent microscopic fields and normalized to the area at time 0. The data are presented as the means ± SEM and normalized to control cells treated with solvent (DMSO) or transfected with siRNA NC. ^* P < 0.05; ^** P < 0.01; ^*** P < 0.001, compared with control.

Figure 5. IGF1R activation by exogenous IGF1 led to Akt/Erk/NF-κB p65 activation, decreased sensitivity to gefitinib and EMT induction

(A) The expression of IGF1R, p-IGF1R, E-cadherin, β-catenin, vimentin, Snail and Slug of PC9 cells treated with 100ng/ml IGF1 for 72h was examined by Western blotting. (B) Morphology of PC9 cells treated with 100ng/ml IGF1 for 72h. Migration and invasion abilities of PC9 cells treated with 100ng/ml IGF1 were examined by transwell assays. Photographs were taken at ×40 magnification. Scale bar: 100μm. The number of cells was counted from at least four independent microscopic fields. (C) The sensitivity to 20nM gefitinib of PC9 cells pre-treated with 100ng/ml IGF1 was examined by CCK8 assay. PC9 cells were pre-treated with 100ng/ml IGF1 for 2h and then subjected to 20nM gefitinib for 72h. (D) The expression of p65, p-p65, p-Akt, Akt, p-Erk and Erk of PC9 cells treated with 100ng/ml IGF1 for 72h or PC9GR cells treated with indicated concentrations of PPP for 72h was examined by Western blotting. (E) The expression of p65, p-p65, p-Akt, Akt, p-Erk and Erk of PC9 cells treated with 100ng/ml IGF1 for 72h was examined by Western blotting. PC9 cells were pre-treated with 25μM LY294002 and 5μM U0126 alone or in combination and then subjected to 100ng/ml IGF1 for 72h. (F) A proposed model for IGF1R/Akt/Erk/NF-κB p65 signaling pathways in the acquisition of EMT phenotype of gefitinib-resistant NSCLC. The data are presented as the means ± SEM and normalized to control cells treated with solvent (DMSO). ^* P < 0.05; ^** P < 0.01; ^*** P < 0.001, compared with control.