L858R emerges as a potential biomarker predicting response of lung cancer models to anti-EGFR antibodies: Comparison of osimertinib vs. cetuximab

Abstract

EGFR-specific tyrosine kinase inhibitors (TKIs), especially osimertinib, have changed lung cancer therapy, but secondary mutations confer drug resistance. Because other EGFR mutations promote dimerization-independent active conformations but L858R strictly depends on receptor dimerization, we herein evaluate the therapeutic potential of dimerization-inhibitory monoclonal antibodies (mAbs), including cetuximab. This mAb reduces viability of cells expressing L858R-EGFR and blocks the FOXM1-aurora survival pathway, but other mutants show no responses. Unlike TKI-treated patient-derived xenografts, which relapse post osimertinib treatment, cetuximab completely prevents relapses of L858R⁺ tumors. We report that osimertinib's inferiority associates with induction of mutagenic reactive oxygen species, whereas cetuximab's superiority is due to downregulation of adaptive survival pathways (e.g., HER2) and avoidance of mutation-prone mechanisms that engage AXL, RAD18, and the proliferating cell nuclear antigen. These results identify L858R as a predictive biomarker, which may pave the way for relapse-free mAb monotherapy relevant to a large fraction of patients with lung cancer.

Keywords: EGFR; NSCLC; adaptive mutability; cetuximab; kinase inhibitors; lung cancer; monoclonal antibody; osimertinib; receptor tyrosine kinases.

Graphical abstract

Figure 1

An anti-EGFR antibody decreases viability of lung cancer cell lines expressing L858R-EGFR but it does not affect cells expressing the E746_A750 deletion mutant (Del19) or T790M-EGFR (A) H3255 cells (2 × 10⁴; L858R), along with 11-18 (5 × 10³; L858R), PC9 (3 × 10³; Del19), PC9ER (3 × 10³; Del19 and T790M), and H1975 cells (5 × 10³; L858R and T790M), were seeded in 96-well plates and treated for 72 h with cetuximab, trastuzumab (anti-HER2), or the combination of antibodies (2XmAbs) at 5, 10, or 20 μg/mL. Cell viability was assessed using the MTT assay. Data are presented as means ± SEM of three biological replicates. (B) The indicated NSCLC cell lines (H3255, 2 × 10⁴ cells; 11-18, 5 × 10³; PC9 and PC9ER, each at 3 × 10³) were seeded in 96-well plates and later treated for 72 h with the indicated anti-EGFR antibodies (each at 10 μg/mL). Cell viability was assessed using the MTT assay. Results are presented as means + SEM of three biological replicates. Significance was assessed using one-way ANOVA followed by Dunnett’s multiple comparisons test. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001; ns, not significant. (C) H3255, 11-18 and PC9 cells were treated for 48 h with cetuximab (Cetux., 10 μg/mL), trastuzumab (Trast., 10 μg/mL), or 2XmAbs (cetuximab + trastuzumab, each at 5 μg/mL). Protein extracts were resolved, blotted and probed with antibodies specific to the indicated apoptosis and cell-cycle markers, including an antibody specific to the cleaved form of caspase-3 (Cl.Casp.3). Vinculin and GAPDH were used as gel loading controls. See also Figures S1 and S2.

Figure 2

Both cetuximab and osimertinib reduce proliferation and induce apoptosis of cells driven by L858R-EGFR, but only osimertinib inhibits cells expressing Del19-EGFR (A) 11-18 (L858R-EGFR) and PC9 (Del19-EGFR) cells were seeded on six-well plates and on the next day they were treated for 48 h with cetuximab (Cetux., 10 μg/mL) or osimertinib (Osim., 50 nM for PC9 or 500 nM for 11-18 cells). Thereafter cells were treated with trypsin and counted. Five thousand (11-18) or 150 (PC9) cells were seeded in six-well plates to allow colony formation. Media (without drugs) were refreshed once every 3 days. After 14 days, cells were fixed and stained with crystal violet. For image quantification, five different fields were quantified per sample using ImageJ. Signals were normalized to the control wells. Values represent mean + SEM of three biological replicates. Significance was assessed using one-way ANOVA followed by Dunnett’s multiple comparison test. ∗∗∗∗p < 0.0001; ns, not significant. (B) H3255 and 11-18 cells were seeded on coverslips and treated for 48 or 72 h, respectively, with cetuximab (Cetux., 10 μg/mL) or osimertinib (Osim., 50 nM for H3255 and 500 nM for 11-18 cells). Cells were fixed in paraformaldehyde (4%) and incubated with an anti-KI67 antibody, followed by an Alexa Fluor 555-conjugated secondary antibody. DAPI (blue) was used to stain nuclei. Images were captured using confocal microscopy (40× magnification). The number of KI67-positive cells was normalized to the total number of nuclei. The signals shown in the graph bars are relative to Control. Scale bars, 20 μm. Values represent mean + SEM of three biological replicates. Significance was assessed using one-way ANOVA followed by Dunnett’s multiple comparison test. ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. (C) H3255, 11-18 and PC9 cells were treated for 48 h with cetuximab (Cetux., 10 μg/mL), erlotinib or osimertinib (Erlot. or Osim., 500 nM for 11-18, or 50 nM for the other cell lines). Protein extracts were blotted and probed with specific antibodies. Vinculin was used as the gel loading control. See also Figure S3.

Figure 3

Unlike cells expressing other EGFR mutants, the FOXM1 pathway and compensatory RTKs are inhibited in L858R expressors following cetuximab treatment (A–C) H3255 (L858R-EGFR), 11-18 (L858R-EGFR), and PC9 cells (Del19-EGFR) were treated for 24 h with cetuximab (10 μg/mL), or erlotinib or osimertinib (500 nM for 11-18; 50 nM for the other cell lines). Protein extracts were resolved and probed with antibodies specific to the indicated receptors and downstream pathways. GAPDH was used as the loading control. (D–F) The indicated cell lines, including PC9ER (Del19 and T790M EGFR), were treated for 48 h with cetuximab (10 μg/mL), erlotinib, or osimertinib (each at 50 nM). Protein extracts were resolved, blotted, and probed with antibodies specific to components of the FOXM1 and other pathways. GAPDH and vinculin were used to ensure equal gel loading.

Figure 4

In contrast to treatments with TKIs, cetuximab induces no ROS production or emergence of persister cells (A) PC9 (Del19-EGFR), H3255 (L858R-EGFR), HCC827 (Del19-EGFR), and 11-18 (L858R-EGFR) cells were treated for 8 h with cetuximab (Cetux., 10 μg/mL) or osimertinib (Osim., 500 nM for 11-18; 50 nM for the other cell lines). NAC (N-acetyl-L-cysteine; 10 mM) was used as a ROS scavenger. DCFDA (2ʹ,7ʹ-dichlorofluorescin diacetate) was employed for determining the intracellular content of hydrogen peroxide. Representative images of epifluorescence microscopy (original magnification, ×100) are shown. Scale bar, 200 μm (B) Shown are quantifications of the hydrogen peroxide fluorescence signals from (A) (mean +SEM of three biological replicates). Significance was assessed using one-way ANOVA followed by Dunnett’s multiple comparison test. ∗p < 0.05, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. (C) H3255, PC9, and HCC827 cells were seeded on six-well plates at high confluence and on the next day they were treated for 9 days with either cetuximab (Cetux., 10 μg/mL) or osimertinib (Osim., 300 nM). Media and drugs were refreshed once every 3 days. After 9 days, cells were fixed and stained with crystal violet. Images corresponding to five different fields were quantified using ImageJ. The experiment was repeated thrice. Representative images and the respective histograms are shown. Signals were normalized to the control. Values represent mean + SEM of three biological replicates. Significance was assessed using one-way ANOVA followed by Dunnett’s multiple comparison test. Scale bars, 200 μm. ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001; ns, not significant. See also Figure S4.

Figure 5

Cetuximab inhibits relapses of animal models carrying either L858R-EGFR (11-18) or an overexpressed Del19-EGFR (HCC827), but a model expressing normal levels of Del19-EGFR (PC9) does not respond to cetuximab 11-18 (A, 1 × 10⁷ cells/mouse), PC9 (B, 3 × 10⁶/mouse), or HCC827 cells (C, 4 × 10⁶/mouse) were subcutaneously implanted in the flanks of CD1-nu/nu mice. When tumors became palpable, mice were randomized in groups of 5–7 animals (each) that were treated (gray areas) for 58 days (A) or 21 days (B and C) with cetuximab (0.2 mg/mouse/injection) once every 3 days, or daily with either osimertinib (10 mg/kg/day in A or 5 mg/kg in C) or erlotinib (50 mg/kg/day in B). Shown are tumor volumes of individual mice from each group. Animal numbers per group are indicated (N). See also Figures S5 and S6.

Figure 6

Cetuximab treatment prevents relapses of two PDX models expressing L858R-EGFR but a third model expressing Del19-EGFR does not respond to the antibody NSG mice were pre-implanted with tumor fragments derived from three different PDX models: TM00199 (PDX1, L858R-EGFR) (A), TM00253 (PDX2, L858R-EGFR) (B), and TM00193 (E746_A750 Del19-EGFR) (C). Once tumors reached approximately 500 mm³, mice were treated for 5 weeks (A and B) or 3 weeks (C), as indicated by the gray areas. Shown are tumor volumes corresponding to individual mice. The drugs were administered as follows: cetuximab (0.2 mg/intraperitoneal injection), twice a week, or TKIs, either erlotinib (50 mg/kg) or osimertinib (10 mg/kg), which were orally administered daily. Animal numbers per group are indicated (N). Note that each color represents one animal. See also Figure S7.

Figure 7

Immunoblot analyses of extracts derived from a PDX model (L858R-EGFR) uncover differences between the modes of action of cetuximab and osimertinib (A and B) NSG mice were pre-implanted with tumor fragments derived from the PDX model TM00199 (L858R-EGFR). Once tumors reached approximately 500 mm³, mice were randomized into groups of four animals each, which were untreated (Control), or treated for 1 week only with either cetuximab (0.2 mg/injection, twice a week), or osimertinib (10 mg/kg, daily). Thereafter, all tumors were removed and their volumes were determined (A). Also shown are growth curves of individual tumors (B). The timing of tumor excisions is marked by red arrows. (C–F) Cleared whole extracts of individual tumors were resolved and probed with the indicated antibodies, with the aim of evaluating signaling pathways (C), apoptosis markers (D), cell-cycle markers (E), and specific components of the DNA damage response (F). Vinculin and GAPDH were used as gel loading controls.