Abstract
Activating mutations in the EGFR gene are important targets in cancer therapy because they are key drivers of non–small cell lung cancer (NSCLC). Although almost all common EGFR mutations, such as exon 19 deletions and the L858R point mutation in exon 21, are sensitive to EGFR-tyrosine kinase inhibitor (TKI) therapies, NSCLC driven by EGFR exon 20 insertion mutations is associated with poor clinical outcomes due to dose-limiting toxicity, demonstrating the need for a novel therapy. TAS6417 is a novel EGFR inhibitor that targets EGFR exon 20 insertion mutations while sparing wild-type (WT) EGFR. In cell viability assays using Ba/F3 cells engineered to express human EGFR, TAS6417 inhibited EGFR with various exon 20 insertion mutations more potently than it inhibited the WT. Western blot analysis revealed that TAS6417 inhibited EGFR phosphorylation and downstream molecules in NSCLC cell lines expressing EGFR exon 20 insertions, resulting in caspase activation. These characteristics led to marked tumor regression in vivo in both a genetically engineered model and in a patient-derived xenograft model. Furthermore, TAS6417 provided a survival benefit with good tolerability in a lung orthotopic implantation mouse model. These findings support the clinical evaluation of TAS6417 as an efficacious drug candidate for patients with NSCLC harboring EGFR exon 20 insertion mutations. Mol Cancer Ther; 17(8); 1648–58. ©2018 AACR.
This article is featured in Highlights of This Issue, p. 1623
Introduction
Somatic mutation of the EGFR is a major oncogenic driver (1) and is present in approximately 30% to 50% and 10% to 20% of non–small cell lung cancer (NSCLC) in Asians and in Americans and Western Europeans, respectively (2–5). Activating mutations in the EGFR kinase domain induce ligand-independent constitutive activation and subsequent downstream molecule phosphorylation, leading to cancer cell growth and survival (6, 7). Because of its important role in cancer, mutant EGFR is an important therapeutic target for lung cancer. Among a wide variety of somatic mutations in EGFR, exon 19 deletion mutations and L858R substitution mutation in exon 21 are most common, accounting for over 80% of mutations (8, 9). These common mutations are sensitive to ATP-mimetic EGFR tyrosine kinase inhibitors (TKI) in preclinical models (10–12), and patients with NSCLC harboring these drug-sensitive mutations show significant responses to EGFR-TKI therapies, including gefitinib, erlotinib, and afatinib, compared with standard chemotherapies (13–18).
The next largest proportion of EGFR mutations is a family of exon 20 insertions, which accounts for roughly 4% to 13% of all EGFR mutations in patients with NSCLC (9, 19–22); these mutations consist of in-frame insertions of 3 to 21 base pairs predominantly within the range of codons 762 to 774. These insertions, as well as exon 19 deletions and L858R mutation, have characteristics of oncogenic driver mutations, which induce constitutive activation and cell transformation in preclinical studies. However, in contrast to exon 19 deletions and L858R mutation, exon 20 insertions do not sensitize the kinase domain to EGFR-TKIs, behaving as intrinsic resistant mutations (6, 23–25). Furthermore, patients with NSCLC harboring these mutations exhibit poor clinical responses to monotherapy with afatinib (26), gefitinib (25, 27), and erlotinib (21, 25, 28) because plasma concentrations of these drugs in clinical settings are kept low by dose-limiting toxicity caused by wild-type (WT) EGFR inhibition (22, 29, 30), indicating the importance of development of an EGFR inhibitor targeting exon 20 insertions.
Therefore, clinical trials for EGFR-TKIs designed to overcome exon 20 insertion–mediated resistance in patients are now in progress. AP32788, a TKI designed to inhibit exon 20 insertions in the EGFR or HER2 genes, is currently in a phase I/II clinical trial (NCT02716116) to determine a recommended phase II dose in the dose escalation cohort and the overall response rate in the expansion cohorts. In addition, interventional studies against EGFR exon 20 insertions have been initiated for poziotinib (NCT03066206 and NCT03318939), which is a pan-ErbB inhibitor, and are planned for osimertinib (NCT03191149), which has received approval for patients with NSCLC harboring T790M acquired resistance mutations in the exon 20 region.
In this study, we report the preclinical characterization of TAS6417, a novel EGFR-TKI with a unique scaffold fitting into the ATP-binding site of the EGFR hinge region. The structure differs from that of all previously known small-molecule EGFR-TKIs, and our results demonstrate the selective potent inhibitory effect of TAS6417 on EGFR exon 20 insertions compared with WT EGFR, leading to an efficacious anticancer effect in several in vivo models.
Materials and Methods
Materials
TAS6417 was synthesized at Taiho Pharmaceutical Co., Ltd. The structure was confirmed by ESI-MS and NMR, and its purity exceeded 99% as measured by HPLC. Afatinib was purchased from Selleck Chemicals, and erlotinib from LC Laboratories.
Mass spectrometry analysis
The human recombinant EGFR D770_N771insNPG was purchased from Carna Biosciences. The recombinant enzyme at 1 μmol/L was incubated with 50 μmol/L TAS6417 at 37°C for 2 hours. After reduction and alkylation with DTT (Thermo Fisher Scientific) and iodoacetamide (Thermo Fisher Scientific), the protein was digested with trypsin/Lys-C (Promega) followed by Asp-N (Thermo Fisher Scientific). The peptide–compound complex was detected with an LC/MS system consisting of an Acquity UPLC I-Class and Xevo G2-S QT of mass spectrometer equipped with an ESI ion source (Waters Corporation) using an Aeris WIDEPORE 3.6μ XB-C18 150 × 2.1 mm column (Phenomenex) at a flow rate of 0.3 mL/minute and a column temperature of 50°C. The gradient of mobile phase A [water, 0.1% formic acid (v/v)] and B [acetonitrile, 0.1% formic acid (v/v)] was performed as follows: 2% B for 3 minutes, 2% to 50% B for 37 minutes, 50% to 90% B for 10 minutes. The data were acquired in MSE mode with collision energy of 4 eV for low-energy scans and ramped from 15 to 40 eV for high-energy scans, and further data analysis was performed using BioPharmaLynx1.3.3 software (Waters Corporation).
Cell culture
The NCI-H1975, HCC827, NCI-H23, NCI-H460, and NIH/3T3 cell lines were obtained from ATCC, and the Ba/F3 and PC-9 cell lines were from RIKEN BioResource Center. The LXF 2478L cell line was provided by Charles River Discovery Research Services, GmbH, who generated the original cell line from its proprietary patient-derived xenograft (PDX) model. The NHEK-Neo cell line was obtained from Lonza. Ba/F3 cells were cultured in RPMI1640 supplemented with 10% FBS (MP Biomedicals, LLC) and 1 ng/mL mIL-3 (Cell Signaling Technology), NIH/3T3 cells were cultured in D-MEM containing 10% NBCS (Thermo Fisher Scientific), NHEK-Neo cells were cultured in Keratinocyte Basal Medium (KBM) containing KGM-Gold SingleQuots (Lonza), and other human lung cancer cells were cultured in RPMI1640 containing 10% FBS (MP Biomedicals, LLC); all cells were cultured at 37°C with 5% CO2. All cell lines were authenticated by short tandem repeat profiling before purchase or use in this study.
Genetically engineered cell lines
cDNA constructs of WT and mutant EGFR were transfected into Ba/F3 and NIH/3T3 cells using Nucleofector (Lonza), followed by clone selection using puromycin. The mutant EGFRs tested in this study were A763_Y764insFQEA, V769_D770insASV, D770_N771insSVD, D770_N771insG, H773_V774insNPH, H773_V774insPH, delE746_A750 (exon 19 deletion), and delE746_A750 + T790M (exon 19 deletion + T790M). For NCI-H1975 EGFR D770_N771insSVD cells, NCI-H1975 cells expressing D770_N771insSVD mutant EGFR were established as described above, and then endogenous L858R/T790M mutant EGFR was knocked out by XTN TALEN-mediated mutagenesis (Transposagen Biopharmaceuticals). NCI-H1975 EGFR D770_N771insSVD_Luc cells were established by transfection of luciferase expression vector into NCI-H1975 EGFR D770_N771insSVD cells, followed by clone selection using hygromycin B. After establishment, all cell lines were authenticated by short tandem repeat profiling, and sequencing analysis was performed to confirm the integration of WT and mutant EGFR. As for NCI-H1975 EGFR D770_N771insSVD cells, the presence of knockout mutation by mutagenesis was also validated with both sequencing and immunoblotting analysis as described in Supplementary Methods.
Cell proliferation assay
Cell proliferation was assessed using the CellTiter-Glo luminescent cell viability assay (Promega). For a panel of Ba/F3 cell lines expressing human EGFRs, the cells were plated at an optimal density onto 96-well plates, followed by 72 hours of exposure to the compound in mIL-3-free conditions. Ba/F3 parent cells and Ba/F3 EGFR WT cells were stimulated with 1 ng/mL mIL-3 and 50 ng/mL EGF, respectively, during compound treatment. For a panel of human lung cancer cell lines and NHEK-Neo cells, the cells were plated as described above and incubated for 24 hours for attachment, followed by compound treatment for 72 hours. After compound exposure, a volume of CellTiter-Glo reagent equal to the volume of medium in each well was added, followed by measurement of luminescence in each well using an EnSpire Multimode Plate Reader (Perkin Elmer). IC50 and GI50 values were determined using the SAS software package in EXSUS (CAC Croit).
Caspase induction analysis
Caspase induction was assessed using the Caspase-Glo 3/7 assay (Promega). Cells were plated at optimal density onto 96-well plates and incubated for 24 hours for attachment, followed by compound treatment for 24 hours. After compound exposure, a volume of Caspase-Glo 3/7 reagent equal to the volume of medium in each well was added, followed by measurement of luminescence in each well as described above. The measurement was corrected with respect to cell viability determined as described above, and relative caspase-3/7 activation against DMSO control was determined.
Immunoblotting analysis
For cultured cell lines, cells were seeded at optimal density onto culture plates or dishes with medium including 10% serum. Following culture for 24 hours, the cells were exposed to the compound and harvested at the indicated time points. NIH/3T3 EGFR WT was stimulated with 50 ng/mL EGF during compound exposure. Xenograft tumors and mouse skin tissues were excised at each time point after a single dosing of test compound. The collected samples were lysed and homogenized with RIPA buffer supplemented with a protease and phosphatase inhibitor cocktail (Thermo Fisher Scientific) on ice. Following centrifugation, the supernatants were collected, and protein concentrations were determined using the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific).
For quantification of pEGFR, proteins at optimal concentrations were dispensed into Wes 12 to 230 kDa prefilled plates (ProteinSimple) with primary antibody, anti-rabbit secondary antibody, and streptavidin horseradish peroxidase (HRP) and its substrate, followed by separation with capillary electrophoresis and chemiluminescence detection using Simple Western (ProteinSimple). Peak areas (in the range of 162–198 kDa) were used to determine IC50 values using the SAS software package in EXSUS (CAC Croit).
For Western blotting analysis, proteins were separated by SDS-PAGE using 4% to 15% Criterion TGX precast gels (Bio-Rad Laboratories) and transferred to PVDF membranes using a Trans-Blot Turbo Transfer System (Bio-Rad). After probing with the primary antibodies, the membranes were incubated with HRP-conjugated secondary antibody. The antigen–antibody complexes were visualized using ECL prime (GE Healthcare), followed by detection with the Amersham Imager 600 QC (GE Healthcare).
Primary and secondary antibodies were obtained from Cell Signaling Technologies and GE Healthcare, respectively.
In vivo efficacy experiments
All animal protocols were reviewed and approved according to regional Institutional Animal Care and Use Committees. For the LXF 2478 and LU0387 models, animal experiments were performed by Charles River Discovery Research Services, GmbH, and Crown Bioscience Inc., respectively. Male BALB/cAJcl-nu/nu mice and female F344/NJcl-rnu/rnu rats were obtained from CLEA Japan, NMRI nu/nu mice from Charles River Laboratories, and BALB/c nude mice from Harlan Laboratories; animals ranged from 4 to 6 weeks of age.
In the subcutaneous implantation model, LXF 2478 and LU0387, human lung cancer PDXs, were transplanted into NMRI nu/nu mice and BALB/c nude mice, respectively. The genetically engineered cell lines were suspended in PBS and inoculated into BALB/cAJcl-nu/nu mice for NCI-H1975 EGFR D770_N771insSVD (0.8 × 107 cells/100 μL) and NIH/3T3 EGFR H773_V774insNPH (0.5 × 107 cells/100 μL), and into F344/NJcl-rnu/rnu rats for NCI-H1975 EGFR D770_N771insSVD (1.6 × 107 cells/200 μL). After reaching optimal tumor volume, animals were randomized into groups (six to eight animals per group) and orally administered the compound once daily for 10, 14, or 28 days. In both PDX models, the treatment period was followed by an observation period of 14 days. Antitumor activity was evaluated using tumor volume calculated according to the following formula: [tumor volume] = [major axis] × [minor axis]2 × 0.5. Statistical analysis was performed using repeated measures ANOVA followed by Dunnett test for tumor volume (P < 0.05).
In the orthotropic implantation model, NCI-H1975 EGFR D770_N771insSVD_Luc cells were suspended in 50% PBS and 50% Matrigel (BD Biosciences) and directly inoculated into the lungs of BALB/cAJcl-nu/nu nude mice (2.4 × 106 cells/20 μL). Six days after inoculation, a luciferin solution was administered via the tail vein, and then photon values were measured by the IVIS Imaging System (Perkin Elmer) to confirm engraftment, followed by grouping using the values. Starting on the following day, animals were orally administered the compound once daily. Survival benefit was evaluated according to median survival time (MST), and increase in life span (ILS) was calculated using the following formula: ILS (%) = ([MST of the treated group]/[MST of control] – 1) × 100. Statistical analysis was performed using the Wilcoxon test for MST (P < 0.05).
TAS6417 was formulated in 0.1 N HCl with or without 0.5% hypromellose (HPMC) solution. Afatinib was formulated in a 0.5% methylcellulose (MC), 0.4% Tween80 suspension.
Pharmacokinetic analysis
The plasma concentrations of TAS6417 were determined in BALB/cAJcl-nu/nu mice and F344/NJcl-rnu/rnu rats following multiple oral administrations for 14 days at effective doses (three animals per group). The plasma samples were collected before administration on Day 14, as well as at 0.5-, 1-, 2-, 4-, 8-, and 24-hour post-administration for mice and 0.25-, 0.5-, 1-, 2-, 4-, 8-, and 24-hour post-administration for rats. The plasma concentrations were measured with the LC/MS-MS.
Results
TAS6417 is a novel inhibitor of EGFR exon 20 insertion
TAS6417, with 6-methyl-8,9-dihydropyrimido[5,4-b]indolizine as its core structure, is a novel small-molecule inhibitor designed to be fit to the ATP binding site of the EGFR hinge region and to exhibit inhibitory activity against exon 20 insertion mutations. The chemical structure is shown in Fig. 1A and its synthetic procedure is described in Supplementary Fig. S1. MS analysis demonstrated that TAS6417 covalently modified the cysteine residue at position 797 of recombinant EGFR, harboring an in-frame insertion mutation in the exon 20 region (Fig. 1B; Supplementary Table S1).
TAS6417, an EGFR-TKI targeting exon 20 insertion mutations. A, Chemical structure of TAS6417, which contains 6-methyl-8,9-dihydropyrimido[5,4-b]indolizine as its core unit. B, MSE spectrum confirmation for protease-digested peptide [LLGICLTSTVQLITQLMPFGCLL] with TAS6417. TAS6417-treated EGFR D770_N771insNPG was digested with protease and analyzed by LC-MSE peptide mapping. Identification of TAS6417 attached peptide was confirmed using the MSE fragmentation data. Fragment ions containing C797 (red) were shifted corresponding to the molecular weight of TAS6417, suggesting that TAS6417 covalently bound to EGFR D770_N771insNPG. C, Inhibitory activity of TAS6417, afatinib, and erlotinib on cell proliferation in a panel of Ba/F3 cells engineered to express human EGFRs. The cells were cultured with the compounds for 72 hours in mouse IL3-free conditions. Ba/F3 EGFR WT cells were stimulated with 50 ng/mL EGF during compound treatment. IC50 values were determined with the CellTiter-Glo assay. D, Exon 20 insertion mutation selectivity of TAS6417, afatinib, and erlotinib. Using the IC50 values shown in C, the WT/mut ratio was determined.
TAS6417 inhibited the in vitro phosphorylation activity of EGFR and its mutants including an exon 20 insertion mutation (D770_N771insNPG), with IC50 values ranging from 1.1 ± 0.1 to 8.0 ± 1.1 nmol/L (Supplementary Table S2A). TAS6417 was further evaluated according to its selectivity for EGFR over other kinases as described in Supplementary Methods. Of the 255 kinases tested, TAS6417 inhibited 25 kinases other than EGFR with IC50 values less than 1,000 nmol/L, which was 100 times higher than its IC50 value against WT EGFR. It inhibited only six kinases containing TXK, BMX, HER4, TEC, BTK, and JAK3, with IC50 values less than 10 times that against WT EGFR, as shown in Supplementary Table S2B.
The cellular potency of TAS6417 against EGFR exon 20 insertion mutations was evaluated using genetically engineered NIH/3T3 cells. After 6 hours of exposure, TAS6417 showed potent inhibition of EGFR phosphorylation in a panel of NIH/3T3 cells stably expressing exon 20 insertion mutant human EGFRs, with an IC50 value of <4.48 nmol/L for A762_Y763insFQEA, 78.2 ± 27.4 nmol/L for V769_D770insASV, 23.4 ± 2.5 nmol/L for D770_N771insSVD, 45.7 ± 19.8 nmol/L for D770_N771insG, 66.7 ± 19.0 nmol/L for H773_V774insNPH, and 117 ± 31 nmol/L for H773_V774insPH (Table 1A). In contrast, the IC50 value of TAS6417 for WT EGFR was 368 ± 181 nmol/L. Furthermore, TAS6417 inhibited proliferation of Ba/F3 cells driven by various EGFR exon 20 insertion mutations, with IC50 values ranging from 5.05 ± 1.33 nmol/L to 150 ± 53 nmol/L (Fig. 1C). Consistent with an intracellular target inhibition assay in NIH/3T3 cells, TAS6417 showed selectivity against WT EGFR (IC50 = 676 ± 304 nmol/L). In a cell proliferation assay in a Ba/F3 cell panel, the selectivity for exon 20 insertions over the WT, represented as the WT/mut ratio, was 134-fold for A763_Y764insFQEA, 6.37-fold for V769_D770insASV, 17.4-fold for D770_N771insSVD, 17.2-fold for D770_N771insG, 4.51-fold for H773_V774insNPH, and 4.55-fold for H773_V774insPH (Fig. 1D). In contrast, afatinib and erlotinib showed no selectivity against exon 20 insertions, with less than one-fold WT/mut ratios except for that of A763_Y764insFQEA. Although A763_Y764insFQEA is an insertion mutation at the exon 20 region in the EGFR gene, it has been reported as an EGFR-TKI-sensitive mutation, and patients with NSCLC harboring this mutation have shown clinical responses to both afatinib and erlotinib (9, 25). These data demonstrated the potency of TAS6417 against cells driven by EGFR exon 20 insertions with mutant selectivity, the spectrum of which was different from that of other EGFR-TKIs.
Cellular potency of TAS6417
TAS6417 inhibits in vitro cell growth and EGFR signaling in human cancer cell lines driven by EGFR exon 20 insertions
The effect of TAS6417 on tumor cell proliferation was investigated using a panel of seven human lung cancer cell lines expressing different EGFRs. The panel consisted of LXF 2478L cells established from a PDX expressing an EGFR exon 20 insertion mutation (V769_D770insASV), NCI-H1975 EGFR D770_N771insSVD cells in which the endogenous alleles with L858R and T790M substitutions were knocked out and exogenous EGFR with an exon 20 insertion mutation (D770_N771insSVD) was transfected (Supplementary Fig. S2), HCC827 and PC-9 cells expressing an exon 19 deletion mutation (delE746_A750) of the EGFR gene, NCI-H1975 cells with substitution mutations in exons 20 and 21 (T790M with L858R), and NCI-H23 and NCI-H460 cells expressing a KRAS mutation without EGFR mutation. TAS6417 suppressed the growth of cells expressing exon 20 insertion mutations of the EGFR gene, with a GI50 value of 86.5 ± 28.5 nmol/L for LXF 2478L cells and 45.4 ± 2.6 nmol/L for NCI-H1975 EGFR D770_N771insSVD cells (Table 1B). TAS6417 also potently inhibited proliferation in other cell lines harboring activating mutations or acquired resistance mutations, with mean GI50 values of 1.92 ± 0.21 nmol/L to 7.12 ± 0.60 nmol/L. In contrast, the effect of TAS6417 on cell proliferation in normal human epidermal keratinocytes (NHEK-Neo), of which WT EGFR is implicated in the growth and survival (31, 32), was moderate. Also, TAS6417 had no effect on EGFR-independent proliferation in NCI-H23 or NCI-H460 cells.
Furthermore, EGFR signal inhibition and apoptosis induction by TAS6417 were confirmed by assessing the expression of phospho-proteins and caspase activation markers (Fig. 2A and B). Culture with cells for 6 hours resulted in a concentration-dependent decrease of phospho-EGFR and its downstream molecules, including ATK and ERK, in both LXF 2478L cells and H1975 EGFR D770_N771insSVD cells, indicating inhibition of EGFR signal transduction. Accordingly, TAS6417 led to increased levels of Bim, which is a pro-apoptosis protein whose expression is regulated by AKT and ERK signaling (33), and cleaved PARP, which is cleaved by caspases and is a marker of cells undergoing apoptosis (34), in both cell lines after treatment for 24 hours. Consistent with Western blotting analysis, chemiluminescence analysis revealed that TAS6417 caused a concentration-dependent increase in relative caspase-3/7 activity in both cell lines compared with that in the control 24 hours after treatment, implying induction of apoptosis (Fig. 2C and D). In summary, these findings suggest that TAS6417 inhibits EGFR signal transduction, leading to cell growth inhibition and apoptosis induction in NSCLC cells driven by EGFR exon 20 insertion mutations.
