Phosphorylation of TNF-alpha converting enzyme by gastrin-releasing peptide induces amphiregulin release and EGF receptor activation

Abstract

G protein-coupled receptors induce EGF receptor (EGFR) signaling, leading to the proliferation and invasion of cancer cells. Elucidation of the mechanism of EGFR activation by G protein-coupled receptors may identify new signaling paradigms. A gastrin-releasing peptide (GRP)/GRP receptor-mediated autocrine pathway was previously described in squamous cell carcinoma of head and neck. In the present study, we demonstrate that TNF-alpha converting enzyme (TACE), a disintegrin and metalloproteinse-17, undergoes a Src-dependent phosphorylation that regulates release of the EGFR ligand amphiregulin upon GRP treatment. Further investigation reveals the phosphatidylinositol 3-kinase (PI3-K) as the intermediate of c-Src and TACE, contributing to their association and TACE phosphorylation. Phosphoinositide-dependent kinase 1 (PDK1), a downstream target of PI3-K, has been identified as the previously undescribed kinase to directly phosphorylate TACE upon GRP treatment. These findings suggest a signaling cascade of GRP-Src-PI3-K-PDK1-TACE-amphiregulin-EGFR with multiple points of interaction, translocation, and phosphorylation. Furthermore, knockdown of PDK1 augmented the antitumor effects of the EGFR inhibitor erlotinib, indicating PDK1 as a therapeutic target to improve the clinical response to EGFR inhibitors.

Fig. 1.

GRP induces TACE and c-Src association. (a and b) Representative SCCHN cells (PCI-37A or PCI-15B) were treated with GRP (400 nM) for 10 min, followed by immunoprecipitation (IP) with TACE and immunoblotting (IB) for c-Src. The membrane was stripped to immunoblot for TACE to ensure equivalent loading. (c) PCI-15B cells were plated on coverslips, followed by serum starvation for 2 days. Upon incubation with GRP (400 nM) for 5 min, cells were double-stained with anti-TACE (green) and anti-c-Src (red) antibody. Arrows indicate the redistribution of TACE and c-Src to the cell membrane upon GRP stimulation.

Fig. 2.

TACE mediates GRP-induced EGFR activation. (a) Representative SCCHN cells (PCI-37A) were plated on 6-well plates, followed by TACE siRNA transfection or transfection with negative control GFP duplex. After serum starvation for 2 days, cells were treated with GRP (400 nM) for 10 min, followed by immunoblotting (IB) and immunoprecipitation (IP) as indicated. (b) An amphiregulin ELISA was performed on cell culture media according to the manufacturer’s instructions. TACE siRNA GRP treatment was compared to GFP siRNA GRP treatment. Cumulative results are shown from six independent experiments (P = 0.0011).

Fig. 3.

Src mediates GRP-induced TACE phosphorylation in SCCHN. (a) PCI-37A cells were transfected with c-Src siRNA, followed by serum starvation for 2 days and treatment with GRP (400 nM) for 10 min. Cell lysates were subject to Western blotting for c-Src expression and TACE immunoprecipitation. IB, immunoblotting; IP, immunoprecipitation. (b) An amphiregulin ELISA was performed on cell culture media according to the manufacturer’s instructions. Cumulative results are shown from six independent experiments (P = 0.0011).

Fig. 4.

PI3-K acts as an intermediary and is required for GRP-induced TACE phosphorylation. (a) PCI-37A cells were serum-starved for 2 days. After pretreatment with the Src family kinase inhibitor PD0180970 (500 nM) for 2 h, cells were treated with GRP or EGF for 10 min. Cell lysates were subject to immunoblotting with phospho-Akt (Ser-473) and total Akt. (b) PCI-37A cells were transfected with p85α siRNA, followed by serum starvation for 2 days and treatment with GRP (400 nM) for 10 min. Cell lysates were subject to Western blotting for p85α expression and TACE immunoprecipitation. IB, immunoblotting; IP, immunoprecipitation. (c) An amphiregulin ELISA was performed on cell culture media according to the manufacturer’s instructions. Cumulative results are shown from four independent experiments (P = 0.014).

Fig. 5.

PDK1 kinase phosphorylates TACE upon GRP treatment in SCCHN. (a) SCCHN cells (PCI-37A) were serum-starved for 2 days, followed by TACE immunoprecipitation. PDK1 enzyme was used to incubate with the cell pellets, followed by in vitro kinase assay. The same membrane was used to probe for total TACE. IB, immunoblotting; IP, immunoprecipitation. (b) Active PDK1 kinase was incubated with recombinant GST or GST-TACEc, followed by in vitro kinase assay. The same membrane was subject to immunoblotting with anti-GST antibody to ensure equal loading. (c) SCCHN cells (PCI-37A) were transfected with PDK1 siRNA or negative control (GFP) siRNA, followed by immunoblotting with PDK1 antibody, or immunoprecipitation with TACE, followed by immunoblotting with phosphothreonine or serine antibody.

Fig. 6.

Enhanced antitumor effects by combined targeting of PDK1 and EGFR. SCCHN (PCI-37A) cells were plated on 24-well plates, followed by transfection with PDK1 siRNA or GFP duplex siRNA. Twenty-four hours after transfection, erlotinib (1 μM) was added to the well. (a) 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide assay was performed after 24 h of erlotinib treatment. Results are from six independent experiments (P = 0.0011). (b) Cells were plated in a matrigel invasion chamber in duplicate, followed by transfection with PDK1 siRNA or GFP duplex siRNA. Twenty-four hours after transfection, erlotinib (1 μM) was added for 24 h. Invading cells in 10 representative fields were counted by using light microscopy at ×400 magnification.