Natalia        Stamp

Mechanism for Activation of the EGF Receptor Catalytic Domain by the Juxtamembrane Segment

Natalia Jura, Nicholas F. Endres, Kate Engel, Sebastian Deindl, Rahul Das, Meindert H. Lamers, David E. Wemmer, Xuewu Zhang, and John Kuriyan

Cell, Volume 137, Issue 7, 1293-1307, 26 June 2009 doi:10.1016/j.cell.2009.04.025 (local copy)

Abstract / Figures from the paper / Supplemental Materials / Coordinates



Signaling by the epidermal growth factor receptor requires an allosteric interaction between the kinase domains of two receptors, whereby one activates the other. We show that the intracellular juxtamembrane segment of the receptor, known to potentiate kinase activity, is able to dimerize the kinase domains. The C-terminal half of the juxtamembrane segment latches the activated kinase domain to the activator, and the N-terminal half of this segment further potentiates dimerization, most likely by forming an antiparallel helical dimer that engages the transmembrane helices of the activated receptor. Our data are consistent with a mechanism in which the extracellular domains block the intrinsic ability of the transmembrane and cytoplasmic domains to dimerize and activate, with ligand binding releasing this block. The formation of the activating juxtamembrane latch is prevented by the C-terminal tails in a new structure of an inactive kinase domain dimer, suggesting how alternative dimers can prevent ligand-independent activation. 

Figures from the paper.

Click on the small image to get a bigger one.

Figure 1. Schematic Diagrams of EGFR (A) Activation of EGFR by EGF results in the formation of an asymmetric kinase domain dimer. (B) Domains of EGFR. Residue numbering corresponds to human EGFR, excluding the signal sequence.

Figure 2. The Effect of the Juxtamembrane Segment on Activity (A) Catalytic efficiency (kcat/KM) of the kinase core (residues 672-998) in solution (yellow) and on vesicles (blue), compared to the catalytic efficiency in solution of constructs that include the full juxtamembrane segment (JM-A and JM-B, residues 645-998) or only JM-B (residues 658-998). The values of kcat/KM were obtained from the linear dependence of reaction velocity on substrate concentration at low substrate concentration, and the error bars are derived from the linear fit (Zhang et al., 2006). (B) The activity of constructs that are either receiver-impaired (restricted to serve as activators, with the I682Q mutation) or activator-impaired (restricted to serve as receivers, with the V924R mutation). (C) Concentration-dependent change in specific activity, in solution, for the JM-kinase construct (containing both JM-A and JM-B, residues 645-998) and a construct containing JM-B but lacking JM-A (residues 658-998). Data shown are mean values from two independent experiments ± STD. (D) EGFR constructs were immunoprecipitated from cell lysates using an anti-FLAG antibody and EGFR autophosphorylation was examined by immunobloting using an anti-phosphotyrosine antibody (anti-pTyr).

Figure 3. Role of the Juxtamembrane Latch in Activation of EGFR (A) Comparison of the structures of asymmetric dimers of kinase domains for EGFR (PDB ID: 2GS6), (Zhang et al., 2006) and Her4 (PDB ID: 2R4B), (Wood et al., 2008). Residues are identified using EGFR numbering. (B) Sequence conservation in the juxtamembrane latch/C-lobe interface. Residues interacting with the juxtamembrane latch are indicated by asterisks. (C) Detailed view of the structure of the juxtamembrane latch in the Her4 structure (PDB ID: 2R4B), with residues identified by EGFR numbering. (D) Effect of mutating residues involved in formation of the juxtamembrane latch. The level of EGF-stimulated phosphorylation on Tyr 1173 relative to wild type, after normalizing for EGFR levels, is shown below each lane. (E) Comparison of the juxtamembrane latch with the docking of the EGFR inhibitor, Mig6 (PDB ID: 2RFE). (F) The effect of a mutation that prevents docking of the juxtamembrane latch (R953A). The results of co-transfection experiments using full length EGFR receptor variants that are receiver-impaired (I682Q) or activator-impaired (V924R) are shown. The level of EGF-stimulated phosphorylation relative to I682Q and V924R co-transfection in the wild type background, after normalizing for EGFR levels, is shown below each lane.

Figure 4. A Helical Dimer in the JM-A Segment (A) Alignment of the sequences of the juxtamembrane segments of EGFR family members. (B) Comparison of the effects of alanine and glycine substitutions in the JM-A segment. The first three panels compare activity of the full length wild type receptor with that for variants in which Arg 656 and Arg 657 are replaced by alanine and glycine. The next six panels shows results of co-transfection experiments using activator-impaired and receiver-impaired variants of the receptor, and compare the results of alanine and glycine substitutions in each variant. (C) The modeled antiparallel JM-A helical dimer, with Leu655 at the d position of a heptad motif in one helix and Leu658 and Leu659 at the g and a positions in the second helix. (D) Models for antiparallel (left and middle, with Leu 655 at the d and a positions, respectively) and parallel (right, with Leu 655 at the d position). The dotted lines indicate interatomic contacts that are either consistent or inconsistent with NMR data for a peptide containing two tandem repeats of the JM-A segment (see Supplemental Data).

Figure 5. Structural Coupling Between the Extracellular and Intracellular Domains in Active EGFR (A) A model for the JM-A helical dimer in the context of the asymmetric dimer of kinase domains. In the exploded view, arginine sidechains that face the membrane are shown. (B) Structure of the transmembrane domain dimer of Her2 (PDB ID: 2JWA), (Bocharov et al., 2008) and the modeled JM-A dimer. Positively charged sidechains that face the membrane are in blue. (C) A model for the activated EGF receptor. Two liganded EGFR extracellular domains are shown in an active dimeric assembly (PDB ID: 1IVO) (Ogiso et al., 2002), with domains IV based on the structure of the inactive EGFR extracellular domain (PDB ID: 1NQL) (Ferguson et al., 2003). This arrangement is compatible with the transmembrane domain dimer and couples to the asymmetric kinase domain dimer via the dimeric JM-A helices and the juxtamembrane latch.

Figure 6. A Symmetric Inactive Dimer of the EGFR Kinase Domain (A) Overview of the crystal structure of the symmetric inactive dimer. (B) Detailed view of the hydrophobic packing between the C-terminal AP-2 helix of monomer B and the N-lobe of monomer A. (C) Exploded view of the electrostatic hook formed between the C-terminal tail (residues 979-990) of EGFR and the hinge region in the kinase domain. (D) Effect of mutations in the electrostatic hook on autophosphorylation of full length EGFR in COS7 cells. (E) Alignment of the sequences of EGFR family members in the C-terminal tail regions encompassing residues in the electrostatic hook and AP-2 helix.

Figure 7. Proposed Role of the Inactive Dimer in EGFR Autoinhibition (A) The surface electrostatic potential of the inactive dimer, with positively and negatively charged regions in blue and red, respectively. The exploded view shows the proposed docking at the plasma membrane. (B) Schematic diagram comparing the juxtamembrane latch in the active asymmetric dimer and the docking of the C-terminal tail in the inactive symmetric dimer. (C) A schematic representation of the activation mechanism of EGFR.

Supplemental Materials

Figure S1. The JM-kinase Construct Is Predominantly a Dimer in Solution (A) Dynamic light scattering of the His-tagged kinase core construct in solution. Molecular weight distribution of the single protein peak is plotted as squares against the elution volume. Reflective index and light scattering (at 90° to the incident beam) of the peak are shown as solid and dashed lines, respectively. Data analysis using the program ASTRA 4.90.04 yielded a molecular weight of 40,000 Da, which corresponds to the monomeric state of kinase core. (B) Dynamic light scattering of the HIS-tagged JM-kinase construct in solution. The data collection and analysis was done as described in (A). The molecular weight of the predominant peak corresponds to 79,000 Da, which corresponds to the dimeric form of the JM-kinase domain (JM-kinase monomer has a molecular weight of 42,000 Da). Higher order oligomerization of the JM-kinase domain was also observed, as evidenced by the second peak corresponding to a 380,000 Da species. The nature of the higher oligomeric species of JM-kinase is unclear at present. In the absence of organization at the membrane (which restricts how many JM-A helices can form in a chain) the juxtamembrane latch may allow the formation of a chain of molecules, as observed in the Her4 crystal structure (Wood et al., 2008).

Figure S2. NMR Structural Analysis of the JM-A Peptide (A) Cross section of the dNN region of the 1H-1H NOESY spectra of JM peptide (652-666) in 20 mM phosphate buffer pH 6.8, 10% D2O and 4% deuterated acetonitrile at 293K. (B) Summary of NOESY connectivities for the JM peptide. NOE connections shown in grey are ambiguous due to spectral overlap. (C) and (D) The plots of Hα and Cα chemical shift index against the residue number for the JMA-peptide are shown, respectively. (E) Plot of normalized intensity as function of concentration. The intensity for the methyl group of Leu 658, Leu 659, Leu 664, Val 665 and Leu 655 was normalized against T3Hγ intensity. The error was calculated from the signal/noise ration of the fit height. The 1H-1H TOCSY spectra were acquired with 0.2 mM (orange), 0.5 mM (red) and 2 mM (green) peptide under identical condition.

Figure S3. Transfection and Co-transfection Analysis of the Effects of Mutations in the LRRLL Motif on EGFR Autophosphorylation (A) Comparison of the effects of alanine and glycine substitutions in the JM-A segment. The first three panels compare activity of the full length wild type receptor with that for variants in which Arg 656 and Arg 657 are replaced by alanine and glycine. The next four panels shows results of co-transfection experiments using activator-impaired and receiver-impaired variants of the receptor, and compare the results of alanine and glycine substitutions in each variant. (B) Effects of mutations in the JM-A segment on Tyr 1173 phosphorylation. The level of EGF-stimulated phosphorylation relative to the wild type, after normalizing for EGFR levels, is shown below each lane.

Figure S4. Models of the Helical Parallel Dimer Between the JM-A Segments of EGFR (A) The modeled parallel JM-A dimers present two different helical packing scenarios: Leu655 placed in a position or in d position. The rotated views show ion pair interactions involving Arg 662. (B) Effect of mutations of the residues involved in ion pairs in the models of the parallel JM-A dimer (Arg 662) and the antiparallel JM-A dimer (Lys 652 and Arg 656). Phosphorylation at Tyr1173 was examined by immunoblotting of whole cell lysates with an anti-pY1173 antibody and the level of EGF-stimulated phosphorylation was normalized relative to the wild type. (C) Schematic diagrams of all combinations of heterodimeric parallel arrangement of JM-A helices between Her4 and EGFR or Her2. The unfavorable juxtaposition of the Glu residues in the parallel helical dimers is marked by red stars.

Figure S5. Model of the Helical Antiparallel JM-A EGFR Homodimer With Leu 655 at Position a Schematic diagram of the EGFR JM-A antiparallel helical homodimer with Leu 655 placed at the a position. The electrostatic interactions between the charged residues are marked by the black dashed line. This arrangement does not provide a specific role for Arg 656 compared to the arrangement when Leu 655 is placed at the d position.

Figure S6. Models of Helical Antiparallel JM-A Dimers Between Her Family Members. Schematic diagrams of various combinations of antiparallel homo- and heterodimers of JM-A helices between Her family members. In all dimers, Leu 655 in EGFR and the corresponding residues in Her2, Her3 and Her4, were placed at the d position. All these arrangements lead to tight packing of the hydrophobic residues in the dimer interface and the favorable electrostatic interactions (dashed green lines) for each combination.

Figure S7. NMR Based Structural Analysis of 35 Residue JMA peptide (A) The JM-A peptide used for the NMR investigation is shown. The two JM-A segments are denoted as Segment-A and Segment-B, respectively. The residues highlighted with red stars (L655 and L659) in Segment-A are labeled with 15N, and E663 in Segment-A, which is highlighted with green star, is double labeled with 15N13C. (B) Aliphatic region of the 2D 15N-1H HSQC –NOESY spectra of the JM-A peptide in 50 mM deuterated Acetate buffer pH 5.1 10% D2O at 293K. The amino acid residues from Segment-A and Segment-B are denoted by suffixes A and B, respectively. (C) Upfield region of the 2D 13C-1H HSQC –NOESY spectra of the JMA peptide.

Figure S8. Molecular Dynamics of the JM-A Helical Dimer Docked on the Kinase Domain (A) Overview of the structure used for molecular dynamics. (B) The stability of the hydrophobic and electrostatic interactions in the JM-A helical dimer in molecular dynamics trajectories. The results of four independent trajectories, each extending for 10 ns, are shown. For each monitored ion pair, the distance between the penultimate carbon atoms of the two residues was calculated over the course of each trajectory, and ion pairs with distances smaller than a threshold value of 5Å were considered “intact”. In order to quantify the hydrophobic packing of leucine residues in the helical dimer, the distance of the Cβ-atom of the leucine residue at the a position in one helix to the midpoint between the Cβ-atoms of two adjacent leucine residues at the d and e positions of the other helices (see Figure 4B in the main text) was computed and recorded as a function of time. Hydrophobic packing at a given time point was considered “intact” when this distance did not exceed its mean value (over the course of the entire trajectory) by more than 0.5 Å. The close packed configuration of the six leucine sidechains is stable in each of the simulations. The intramolecular and intermolecular ion pairs are broken and reformed due to transient interactions with water and ions, but are maintained on average (Figure S3B). These results suggest that the formation of an antiparallel helical dimer is a plausible model for the JM-A segments. In the diagram intact ion pairs and intact hydrophobic packing of leucine residues in the helical dimer between the helix of the receiver kinase (Rec) and the helix of the activator kinase (Act) is marked as green and a break in the interaction is marked as red.

Figure S9. The Surface Electrostatic Potential of the Inactive Dimer. Two representative views of the inactive dimer and the corresponding calculated surface electrostatic potentials are shown. The surface electrostatic potential was calculated using GRASP2 (Petrey and Honig, 2003).

Figure S10. The Effect of Mutations in the AP-2 helix. The effect of deletion (A) or mutations (B) of residues in the AP-2 helix on EGFR autophosphorylation in COS7 cells was determined by immunoblotting of whole cell lysates with anti-pY1173 antibody. (C) Catalytic efficiency (kcat/KM) of the EGFR kinase core (672-998) and mutants of vesicles. Error bars correspond to the standard deviation of the linear fittings (Zhang et al., 2006).


Coordinates in the Protein Data Bank: 3GT8