The EGF receptor and HER2 participate in TNF-α-dependent MAPK activation and IL-8 secretion in intestinal epithelial cells

Abstract

TNF-α activates multiple mitogen-activated protein kinase (MAPK) cascades in intestinal epithelial cells (IECs) leading to the secretion of interleukin 8 (IL-8), a neutrophil chemoattractant and an angiogenic factor with tumor promoting properties. As the epidermal growth factor receptor (EGFR) is a known transducer of proliferative signals and a potent activator of MAPKs, we hypothesized that the EGFR participates in TNF-dependent MAPK activation and IL-8 secretion by intestinal epithelial cells (IECs). We show that the EGFR is tyrosine-phosphorylated following treatment of IECs (HT-29 and IEC-6) with TNF-α. This requires EGFR autophosphorylation as it was blocked by the EGFR kinase inhibitor AG1478. Autophosphorylation was also inhibited by both a Src-kinase inhibitor and the metalloproteinase inhibitor batimastat. TNF treatment of IECs resulted in the accumulation of soluble TGF-α; treatment of IECs with batimastat suppressed TGF-α release and immunoneutralization of TGF-α resulted in decreased EGFR and ERK phosphorylations. TNF-α treatment of IECs resulted in an association between EGFR and HER2 and inhibition of HER2 using a specific inhibitor AG879 in combination with AG1478-suppressed TNF-α-dependent ERK phosphorylation and IL-8 release. Downregulation of HER2 via siRNA resulted in a significant decrease in ERK phosphorylation and a 50% reduction in IL-8 secretion.

Figure 1

EGF rapidly stimulates the ERK pathway in HT-29 cells. HT-29 cells were cultured in serum-free media overnight and stimulated with 50 ng/mL EGF or 10 ng/mL TNF (a). ERK activation in response to TNF is relatively delayed (apparent by 15 mins) relative to EGF (apparent by 5 mins). (b) shows an antiphosphotyrosine blot of immunoprecipitated EGFR following stimulation of serum-starved HT-29 cells with TNF-α. TNF-α treatment results in the time-dependent tyrosine phosphorylation of the EGF receptor. (c) shows the effect of EGF receptor tyrosine kinase inhibition using the EGFR tyrosine kinase inhibitor AG1478. Cells were treated for 15 mins with AG1478 (0–10 μM) and stimulated with 10 ng/mL TNF-α for 15 mins. AG1478 dose-dependently inhibits EGFR phosphorylation on tyrosine. Data are representative of at least three separate experiments.

Figure 2

EGFR tyrosine phosphorylation is inhibited by the EGF receptor tyrosine kinase inhibitor AG1478 (10 μM) but not the PDGF receptor inhibitor AG1298 (5 μM) (a). In contrast, only a modest decrease in Erk1/2 phosphorylation was noted in response to pretreatment of HT-29 cells with either of these agents (b). Serum-starved cells were pretreated for 30 mins prior to stimulation with TNF-α for 15 mins, and EGFR tyrosine phosphorylation and ERK phosphorylation were assessed as described in Materials and Methods. Transactivation of the EGFR in response to TNF-α was also observed in the rat intestinal epithelial cell line IEC-6 (c). IEC-6 cells were treated with 10 ng/mL TNF-α for 15 mins in the presence or absence of 1 μM AG1478. Data are representative of at least three experiments.

Figure 3

TNF-dependent EGFR transactivation requires metalloproteinase activity and results in TGF-α release. Serum-starved cells were treated for 30 min with the metalloproteinase inhibitor BB94 (batimastat, 10 μM) (a), or increasing concentrations of the TNF-converting enzyme (TACE) inhibitor TAPI-1 (b), and stimulated with 10 ng/mL TNF-α for 15 mins. EGFR tyrosine phosphorylation was assessed as described in Section 2. EGFR tyrosine phosphorylation is significantly reduced in the presence of BB94 and to a lesser extent by TAPI-1. (c) shows the effect of BB94 and TAPI-1 pretreatments on TNF-stimulated TGF-α release. Serum-starved cells were pretreated for 30 mins with BB94 or TAPI-1, stimulated with TNF-α for 3 hrs, and TGF-α measured via ELISA. (d) shows total TACE activity as measured using either recombinant TACE or membrane preparations from vehicle and TNF-treated HT-29 cells using a fluorescent substrate. Cells were pretreated with vehicle or TAPI-1 for 30 mins prior to stimulation with TNF-α (10 ng/mL) for 15 mins (see Section 2). Data are representative of at least three experiments.

Figure 4

Tyrosine kinase inhibitors inhibit EGFR phosphorylation in response to TNF-α. HT-29 cells were treated with 2 μM PP2 (Src-kinase inhibitor), 100 μM genistein (tyrosine kinase inhibitor), or 10 μM AG1478 for 15 mins prior to 15 mins of TNF-α treatment. The EGF receptor was immunoprecipitated and tyrosine phosphorylation assessed (a). Both PP2 and genistein abrogated TNF-α-dependent EGFR tyrosine phosphorylation. In contrast, neither PP2 nor genistein had an appreciable effect upon TNF-stimulated ERK1/2 phosphorylation (b) and (c). Data are representative of at least three experiments.

Figure 5

TGF-α release is required for TNF-α-stimulated EGFR transactivation and Erk1/2 phosphorylation. Serum-starved HT-29 cells were pretreated with increasing concentrations of TGF-α-neutralizing serum or Ig control and stimulated with TNF-α for 15 mins. EGFR was immunoprecipitated and phosphotyrosine content determined by western blotting (a). (b) shows the effect of TGF-α-neutralizing serum on TNF-stimulated Erk1/2 phosphorylation. Data are representative of at least 3 separate experiments.

Figure 6

TNF treatment stimulates EGFR-HER2 heterodimerization, and HER2 tyrosine phosphorylation. HT-29 cells were serum-starved for 24 hrs prior to stimulation with 10 ng/mL TNF-α. EGFR was immunoprecipitated, and coprecipitating HER2 was measured via western blotting (a). HT-29 cells were serum-starved for 24 hrs, stimulated with 10 ng/mL TNF-α for 15 mins, and phosphotyrosine-containing proteins immunoprecipitated. Samples were separated by SDS-PAGE and HER2 content quantified by western blotting (b). Figures are representative of at least 3 separate experiments.

Figure 7

Inhibition of HER2 tyrosine kinase activity (2.5 μM AG879) in addition to that of EGFR results in greater inhibition of ERK1/2 phosphorylation as compared to EGFR inhibition alone (10 μM) (a) in HT-29 cells. (b) is a loading control, 25 μM PD98058. Figure is representative of three separate experiments.

Figure 8

Effect of EGFR and HER2 inhibition on TNF-α-stimulated IL-8 secretion. HT-29 cells were treated with increasing doses of AG1478 for 30 mins prior to 6 hr stimulation with 10 ng/mL TNF-α (a). HT-29 cells were treated with increasing doses of the HER2 inhibitor AG879 or a combination of AG1478 and AG879 (b). Secreted IL-8 was measured via ELISA. Results are representative of three separate experiments. *P < 0.01, ^δ P < 0.001.

Figure 9

HER2 knockdown using siRNA attenuates TNF-dependant ERK activation and significantly inhibits IL-8 secretion. HT-29 cells were transfected with control or HER2-specific siRNA and incubated in serum-free media for 48 hrs. Cell lysates were prepared for western blotting as described in Section 2. HER2-specific siRNA reduced HER-2 protein expression but did not alter EGFR expression (a). siRNA-transfected cells were then stimulated with 10 ng/mL TNF-α for 15 mins, or 6 hrs and ERK phosphorylation, and IL-8 secretion measured as described previously ((b) and (c) resp.). Figures are representative of at least 3 separate experiments. *P < 0.001.