Differential sensitivity of ERBB2 kinase domain mutations towards lapatinib

Abstract

Background: Overexpression of the ERBB2 kinase is observed in about one-third of breast cancer patients and the dual ERBB1/ERBB2 kinase inhibitor lapatinib was recently approved for the treatment of advanced ERBB2-positive breast cancer. Mutations in the ERBB2 receptor have recently been reported in breast cancer at diagnosis and also in gastric, colorectal and lung cancer. These mutations may have an impact on the clinical responses achieved with lapatinib in breast cancer and may also have a potential impact on the use of lapatinib in other solid cancers. However, the sensitivity of lapatinib towards clinically observed ERBB2 mutations is not known.

Methodology/principal findings: We cloned a panel of 8 clinically observed ERBB2 mutations, established stable cell lines and characterized their sensitivity towards lapatinib and alternative ERBB2 inhibitors. Both lapatinib-sensitive and lapatinib-resistant ERBB2 mutations were observed. Interestingly, we were able to generate lapatinib resistance mutations in wt-ERBB2 cells incubated with lapatinib for prolonged periods of time. This indicates that these resistance mutations may also cause secondary resistance in lapatinib-treated patients. Lapatinib-resistant ERBB2 mutations were found to be highly resistant towards AEE788 treatment but remained sensitive towards the dual irreversible inhibitors CL-387785 and WZ-4002.

Conclusions/significance: Patients harbouring certain ERBB2 kinase domain mutations at diagnosis may not benefit from lapatinib treatment. Moreover, secondary lapatinib resistance may develop due to kinase domain mutations. Irreversible ERBB2 inhibitors may offer alternative treatment options for breast cancer and other solid tumor patients harbouring lapatinib resistance mutations. In addition, these inhibitors may be of interest in the scenario of secondary lapatinib resistance.

Figure 1. Schematic representation of ERBB2 mutations analyzed.

(A and B) The side chains of mutants considered in this study are plotted (red sticks) together with a schematic representation of the protein fold using the crystal structure of EGFR kinase in complex with erlotinib (green sticks). B) is a view roughly orthogonal to A) and shows additional inhibitors gefitinib (yellow sticks) and lapatinib (blue sticks) superimposed at the ATP binding site.

Figure 2. Biochemical analysis of ERBB2 mutants.

(A) HEK293 cells were transfected with either wild type (WT) or mutant ERBB2 for 36 hours and analyzed for autophosphorylation and activation of downstream signaling molecules. Untransfected (UT) cells were taken as control for ERBB2 expression. (B) HEK293 cells were transfected with a combination of ERBB2 (WT or mutant) and EGFR (left two panels) or ERBB3 constructs (right two panels) for 36 hours followed by serum starvation for 12 hours. Cells were then stimulated with either EGF (left two panels) or heregulin (right two panlels) for 5 minutes and analyzed for the activation of ERBB2 as well as downstream signaling pathways by western blotting. Untransfected (UT) cells were used as control.

Figure 3. Anchorage-independent growth of ERBB2 mutants.

Figure 4. Analysis of ERBB2 kinase domain mutants identifies lapatinib-resistant mutations.

Ba/F3 cells stably expressing either wild type or mutant ERBB2 were treated with indicated concentrations of either lapatinib (A) or AEE 788 (B) for 48 hours and analyzed for cell proliferation inhibition.

Figure 5. Structural analysis of lapatinib resistant ERBB2 kinase domain mutants.

(A) L755 packs against helix C, closest to residues Ala763 and Ile767, and makes no contacts with the inhibitors (structure 1M17 with inhibitor erlotinib is depicted lower left). (B) Comparing the active structure of 1M17 (green) to an inactive representative 1XKK bound to lapatinib shows the loss of L755 interactions (cyan). (C) Overlay of AEE788 bound structures of EGFR (2J6M, active, blue) and EGFR T790M (2JIU, inactive, yellow). The existence of the salt bridge linking the active site lysine K753 with the helix C E770 is a marker for the active state. The T798M (ERBB2 numbering) mutation does not significantly alter binding, although a rotation of the inhibitor aromat is apparent. (D) Superposition of two binding modes of lapatinib onto the overlay of figure 2C and display of the T798M atoms as Van der Waals spheres shows how the binding mode seen in 1XKK (cyan) obviously clashes with the mutation, but the binding mode of 3BBT (pale blue, ERBB4, which also has threonine as gatekeeper) does not.

Figure 6. Lapatinib-resistant ERBB2 mutants show increased transformation potential of Ba/F3 cells to cytokine independence.

(A) Ba/F3 cells transformed by ERBB2 mutants were analysed by western blotting for the activation of ERBB2 and downstream ERK phosphorylation. (B) To test the activating nature of lapatinib-resistant mutations, Ba/F3 cells were transduced with wild type or mutant MSCV-eGFP-ERBB2 and outgrowth of ERBB2-positive (green) cells with respect to parental (non-green) Ba/F3 cells was measured by FACS analysis at indicated time points.

Figure 7. Irreversible inhibitors overcome lapatinib resistance due to ERBB2 kinase domain mutations.

Stable Ba/F3 cell lines expressing either wild type or mutant ERBB2 were treated with indicated concentrations of either CL-387785 (A) or WZ-4002 (B) for 48 hours and analyzed for inhibibtion of cell proliferation. Indicated Ba/F3-ERBB2 cell lines were treated with increasing concentrations (50 nM, 100 nM, 250 nM, 500 nM or 1000 nM) of either CL-387785 (C) or WZ-4002 (D) for 30 minutes and analyzed by western blotting with indicated antibodies.