Phosphorylated EGFR Dimers Are Not Sufficient to Activate Ras

Abstract

Growth factor binding to EGFR drives conformational changes that promote homodimerization and transphosphorylation, followed by adaptor recruitment, oligomerization, and signaling through Ras. Whether specific receptor conformations and oligomerization states are necessary for efficient activation of Ras is unclear. We therefore evaluated the sufficiency of a phosphorylated EGFR dimer to activate Ras without growth factor by developing a chemical-genetic strategy to crosslink and "trap" full-length EGFR homodimers on cells. Trapped dimers become phosphorylated and recruit adaptor proteins at stoichiometry equivalent to that of EGF-stimulated receptors. Surprisingly, these phosphorylated dimers do not activate Ras, Erk, or Akt. In the absence of EGF, phosphorylated dimers do not further oligomerize or reorganize on cell membranes. These results suggest that a phosphorylated EGFR dimer loaded with core signaling adapters is not sufficient to activate Ras and that EGFR ligands contribute to conformational changes or receptor dynamics necessary for oligomerization and efficient signal propagation through the SOS-Ras-MAPK pathway.

Keywords: EGFR nanoclusters; EGFR signaling; Ras-MAPK signaling; chemical-genetic dimerization; spatial reorganization.

Figure 1. Decoupling EGFR Dimerization and Transphosphorylation from Other EGFInduced Conformational and Spatial Changes

(A) EGFR exists in a tethered monomer or an inactive dimer formation. Upon EGF binding, it adopts an extended dimer conformation and undergoes auto-transphosphorylation. Phosphorylated dimers recruit adaptor proteins to EGFR, resulting in activation of the Ras-MAPK pathway. (B) EGF binding to EGFR also results in rapid changes in spatial organization from monomers (i) to dimers (ii); to higher order multimers and nanoscale clusters (iii-iv); to micron scale clusters in clathrin-coated pits (v); and, finally, to endosomes (vi). (C) A chemical genetic system utilizing a SNAP-tag on the N terminus of full-length EGFR and BG-modified DNA dimers as crosslinkers. (D) Representative western blot of lysates from cells treated with 8 nM EGF or 2 μM (DNA-BG)₂. To maintain DNA hybridization, SDS-PAGE samples were not boiled. EGFR dimers (d) and monomers (m) are indicated with arrows.

Figure 2. Quantitative Comparison of Tyrosine Phosphorylation after Dimerization by EGF or (DNA-BG)₂

(A) DNA-dimerized receptors can be revealed by PAGE without boiling or can be boiled to reveal a monomer for direct comparison to EGFR monomers. d, dimer; m, monomer; RT, room temperature. (B) Representative western blot of boiled lysates from cells treated with serum-free media (nt, no treatment), 8 nM EGF or 2 μM (DNA-BG)₂ at various tyrosines. (C) Mean fold increase of total EGFR and phosphotyrosines upon EGF or (DNA-BG)₂ treatment compared to no treatment control (n = 3; error bars indicate SD).

Figure 3. Trapped EGFR Dimers Recruit Adaptors with Similar Stoichiometry to EGF-Stimulated Cells but Do Not Activate Ras

(A) Representative western blot showing lysates from cells treated with either 8 nM EGF, 2 μM (DNA-BG)₂, or serum-free media (mock) for 5 min. The same lysates were used in a RasGTP pull-down, and samples were blotted for total Ras. (B) Mean RasGTP levels in each treatment compared to negative control (n = 3; error bars indicate SD). (C) Representative blot of Grb2 co-immunoprecipitation (coIP) with EGFR on lysates from treated cells. (D) Representative blot of SOS coIP with EGFR. (E) Representative blot of Shc coIP with EGFR. (F) Quantification of adaptor coIP in treated cells compared to negative controls. Signals for each adaptor were normalized to total EGFR levels in the pull-down sample and plotted as mean fold increase over mock treatment (n = 3; error bars indicate SD).

Figure 4. Trapped EGFR Dimers Do Not Form Nanoscale Spatial Intermediates or Traffic to Clathrin-Coated Pits

(A) Representative images of HEK293-SNAP-EGFR-mEos cells incubated with serum-free media (mock), 8 nM EGF or 2 μM (DNA-BG)₂ for 10 min and then imaged by STORM. Scale bars, 10 μm. (B) Pairwise correlation analysis of STORM images graphed as median and standard error (n = 10 cells per condition). (C) Representative images of single-molecule IP of SNAP-EGFR-EGFP cells treated with biotin (bt)-EGF or bt-(DNA-BG)₂ for 5 min. Ligand-bound receptors from lysates were immobilized on neutravidin-coated slides and imaged. EGFR monomers (m) appear as blue spots, dimers (d) appear as pink spots, and clusters (c) appear as larger yellow spots. (D) Mean monomer, dimer, and cluster populations of EGFR graphed as a percentage of the sample (n = 3 independent experiments; error bars indicate SD). (E) Representative EGF-biotin-treated sample with counts of relative intensity per molecule. The blue shaded region represents the monomer portion, the green shaded region represents the dimer portion, and zoom represents the cluster portion. The average number of EGFR molecules per cluster was estimated by dividing the average intensity of the clusters by the intensity of a monomer. (F) TIRF images of HEK293 cells co-transfected with SNAP-EGFR-EGFP and clathrin-light chain-dsRed after treatment with 8 nM EGF, 2 μM DNA, or serum-free media at 15 min. Scale bar, 1 μm. (G) Enrichment of SNAP-EGFR-EGFP at clathrin-coated pits over time after treatment with 8 nM EGF, 2 μM (DNA-BG)₂, or serum-free media graphed as mean and SD (n = 10 cells per condition).