A Constitutive EGFR Kinase Dimer to Study Inhibitor Pharmacology

Abstract

Lung cancer is commonly caused by activating mutations in the epidermal growth factor receptor (EGFR). Allosteric kinase inhibitors are unaffected by common ATP-site resistance mutations and represent a promising therapeutic strategy for targeting drug-resistant EGFR variants. However, allosteric inhibitors are antagonized by kinase dimerization, and understanding this phenomenon has been limited to cellular experiments. To facilitate the study of allosteric inhibitor pharmacology, we designed and purified a constitutive EGFR kinase dimer harboring the clinically relevant L858R/T790M mutations. Kinetic characterization revealed that the EGFR kinase dimer is more active than monomeric EGFR(L858R/T790M) kinase and has the same K_m,ATP Biochemical profiling of a large panel of ATP-competitive and allosteric EGFR inhibitors showed that allosteric inhibitor potency decreased by >500-fold in the kinase dimer compared with monomer, yielding IC₅₀ values that correlate well with Ba/F3 cellular potencies. Thus, this readily purifiable constitutive asymmetric EGFR kinase dimer represents an attractive tool for biochemical evaluation of EGFR inhibitor pharmacology, in particular for allosteric inhibitors. SIGNIFICANCE STATEMENT: Drugs targeting epidermal growth factor receptor (EGFR) kinase are commonly used to treat lung cancers but are affected by receptor dimerization. Here, we describe a locked kinase dimer that can be used to study EGFR inhibitor pharmacology.

Fig. 1.

Design and characterization of the constitutive EGFR dimer. (A) Design of the EGFR(L858R/T790M) dimer. (B) Initial rates from ATP titration used to produce Michaelis-Menten kinetic plots (representative experiment shown from n = 3). (C) Michaelis-Menten kinetic plot for the EGFR(L858RR/T790M) dimer (n = 3 in triplicate, plotted as biologic replicate mean ± S.D.).

Fig. 2.

Pharmacological characterization of inhibitors against EGFR. (A) Potencies of allosteric inhibitors against the EGFR(L858R/T790M) dimer (n = 3, mean ± S.D.). (B) Correlation between monomeric (blue) or dimeric (red) enzyme and cellular potencies. Each point represents an allosteric inhibitor tested in enzymatic and cellular EGFR(L858R/T790M) assays in the presence or absence of the dimerization-disrupting antibody cetuximab. Two-tailed Pearson correlation analysis showed significant correlation between cellular potency without cetuximab and dimeric kinase (*P < 0.05) and between cellular potency with cetuximab and monomeric kinase (***P < 0.001). (C) Allosteric inhibitor JBJ-09-063 bound to EGFR in the inactive conformation (PDB 7JXQ) superimposed with EGFR kinase in the active conformation induced by asymmetric kinase dimerization (PDB 2GS2). The inward position of the C-helix in the active conformation clashes with the inhibitor and explains why allosteric inhibitor potency is antagonized by kinase dimerization.

Fig. 3.

Inhibitor profiling with monomeric and dimeric EGFR(L858R/T790M) kinase. (A) IC₅₀ fold changes between dimer and monomer enzymes (n = 3 in triplicate, plotted as biologic replicate mean ± S.D.). A one-sample T-test comparing the fold potency change to a hypothetical value of 1 (no change) was used to determine if a significant change occurred upon kinase dimerization (*P < 0.05, **P < 0.01, and ***P < 0.001). (B) Neratinib (PDB 2JIV) superimposed with active conformation EGFR kinase (PDB 2GS2). Neratinib cannot bind to the active conformation due to steric clashes with the inward C-helix position highlighted by E762 and M766 side chains. This explains the reduced potency of neratinib against the active conformation dimer construct. (C) The osimertinib metabolite AZ5104 (PDB 7JXL) superimposed with active conformation EGFR kinase (PDB 2GS2). The C-helix in position of the active conformation does not sterically clash with AZ5104 and explains why its potency does not dramatically change when the kinase is in the dimeric form.