NoxO1 regulates EGFR signaling by its interaction with Erbin

Abstract

NADPH oxidase organizer 1 (NoxO1) is a scaffold cytoplasmic subunit of the reactive oxygen species (ROS) forming Nox1 complex and involved in angiogenesis, differentiation, and atherosclerosis. We found that overexpression of NoxO1 without simultaneous overexpression of any other component of the active Nox1 complex inhibited EGF-induced wound closure and signaling, while NoxO1 KO yielded the opposite effect. Accordingly, we hypothesize NoxO1 to exert Nox1 independent functions. Using the BioID technique, we identified ErbB2 interacting protein (Erbin) as novel interaction partner of NoxO1. Colocalization of NoxO1 with EGFR, as well as with Erbin validated this finding. EGF treatment interrupted colocalization of NoxO1 and EGFR. EGF mediated kinase activation was delayed in NoxO1 overexpressing cells, while knockout of NoxO1 had the opposite effect. In conclusion, Erbin was identified as a novel NoxO1 interacting protein. Through the subsequent interaction of NoxO1 and EGFR, NoxO1 interferes with EGF signaling. The results of this study suggest a potential role of NoxO1 as an adaptor protein with functions beyond the well-established enabling of Nox1 mediated ROS formation.

Keywords: EGF signaling; Erbin; NADPH oxidase; NoxO1.

Graphical abstract

Fig. 1

NoxO1 decelerates wound closure. Representative images and statistical analysis of Scratch-Wound assays with (A) Hek293 cells overexpressing NoxO1 (OE) or empty vector (ctl) and (B) MCF7 cells with NoxO1 KO (KO) or not (ctl); time for 50 % wound closure is indicated in the graph; n = 4; ∗p < 0.05 in Two-Way ANOVA + Tukey post hoc test; mean ± SEM.

Fig. 2

EGF induces acute ROS formation, mediated by a Nox1 centered NADPH oxidase. Superoxide production (L-012) measurement in (A) Hek293 cells overexpressing only NoxO1 or all components of the Nox1 centered NADPH oxidase (Nox1+NoxA1+NoxO1) treated with EGF (50 ng/ml) 15 min before the actual measure. (B) Hek293 with constitutive overexpression of Nox1 and NoxA1 and transient expression of NoxO1 with acute addition of EGF to the cells in the measure chamber; Representative measurement traces; (C) Immunofluorescence of endogenous EGFR and overexpressed myc-taged NoxO1 in Hek293 cells.

Fig. 3

NoxO1, if not together with the Nox1 centered NADPH oxidase interacts with ErbB2IP (Erbin). (A) immunofluorescence of NoxO1 constructs as indicated (B) Superoxide measurement (L-012) in Hek293 cells overexpressing the NoxO1 construct indicated; (C) Identified biotinylated proteins in Hek293 cells transfected with the NoxO1-BioID construct in the absence or presents of Nox1 and NoxA1.

Fig. 4

EGF interrupts Erbin mediated interaction of NoxO1 and EGFR. (A) Immunoprecipitation for streptavidin (biotin) from Hek293 cells overexpressing empty vector (ctl) or NoxO1-BioID2 as indicated with detection of Erbin and myc (NoxO1); (B) Proximity ligation assay for myc (NoxO1) and Erbin in Hek293 cells overexpressing NoxO1; (C) Proximity ligation assay in Hek293 cells overexpressing empty vector (ctl) or NoxO1, detection of Erbin and ErbB2; (D) Immunoprecipitation for streptavidin (biotin) from Hek293 cells overexpressing empty vector (ctl) or NoxO1-BioID2 as indicated with detection of EGFR; (E) Proximity ligation assay in Hek293 cells overexpressing empty vector (ctl) or NoxO1 and treated without or with EGF as indicated, detection of myc (NoxO1) and EGFR; n = 3; ∗p < 0.05 ctl vs. NoxO1, #p < 0.05 ctl vs. NoxO1 + EGF, & p < 0.05 NoxO1 vs. NoxO1 + EGF in Two-Way ANOVA + Tukey post hoc test; mean ± SEM.

Fig. 5

NoxO1 and Erbin overexpression reduce EGF induced signaltransduction. (A & B) Western blot for phosphorylated and total (A) Erk1/2 and Akt (B) in Hek293 cells overexpressing empty vector (ctl), NoxO1 or Erbin (C) MCF7 without (ctl) and with NoxO1 knock out (KO) treated without or with EGF as indicated; (D) Tube formation assay in HUVECs overexpressing empty vector (ctl), NoxO1 or Erbin treated without or with EGF over night; n = 3–5; ∗p < 0.05 ctl vs. NoxO1, #p < 0.05 ctl vs. Erbin in Two-Way ANOVA + Tukey post hoc test; mean ± SEM.