Discovery of HCD3514 as a potent EGFR inhibitor against C797S mutation in vitro and in vivo

Osimertinib (AZD9291), a third-generation epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKI), has significantly improved the survival of non-small cell lung cancer (NSCLC) patients with EGFRT790M mutation, the major mechanism of acquired resistance to first-generation EGFR TKI. However, resistance to AZD9291 arises eventually and EGFRC797S mutation was reported to be a major resistance mechanism. Thus, it is highly valuable to develop novel EGFR fourth-generation inhibitors targeting C797S mutation to override the acquired resistance. In this study, we identified HCD3514 as a novel EGFR fourth-generation inhibitors targeting C797S triple mutation. It strongly inhibited EGFRL858R/T790M/C797S and EGFR19del/T790M/C797S mutations with IC50 values of 1.0 and 2.0 nM, respectively. HCD3514 dose-dependently inhibited the activation of EGFR in both engineered BaF3 cells and tumor cells harboring EGFRC797S triple mutant and thus effectively suppressed the proliferation of the cells. Moreover, HCD3514 caused a dose-dependent increase of apoptosis in C797S triple mutant cells accompanied by increased levels of cleaved caspase-3 and cleaved PARP. Furthermore, HCD3514 induced tumor growth inhibition in EGFR19del/T790M/C797S xenograft model as a single oral agent by decreasing the activation of EGFR. In addition to EGFRC797S triple mutations, HCD3514 also potently and selectively inhibited EGFRT790M double mutations (L858R/T790M and 19del/T790M). Collectively, HCD3514 is a highly selective and potent EGFR inhibitor against EGFRC797S triple mutations as well as EGFRT790M double mutations and is confirmed potently anti-tumor activity in preclinical models.


Introduction
Lung cancer is a malignant tumor with high incidence and mortality throughout the world, being responsible for almost one-quarter of all cancerrelated deaths [1]. The epidermal growth factor receptor (EGFR) is one of the most important driver oncogenes in lung cancer, especially non-small cell lung cancer (NSCLC), which accounts for about 85% of lung cancer [2,3]. Diverse activating mutations in EGFR occur in 10% to 50% of patients with NSCLC [4,5]. The two most common EGFR activating mutations are the small in-frame deletions in exon 19 (19del) and L858R point mutation in exon 21, occupying more Ivyspring International Publisher than 90% of all EGFR mutations [6,7]. NSCLC patients harboring activating mutations respond well to the first-generation EGFR tyrosine kinase inhibitors (TKIs), including gefitinib, erlotinib and icotinib [8][9][10]. But acquired resistance arises within 9-15 months after the treatment of first-generation EGFR TKIs [11]. The secondary T790M mutation in EGFR exon 20 is the primary resistance mechanism, contributing to approximately 60% of resistance cases [12,13]. Hence, the second-generation EGFR TKI such as afatinib has been developed, but it still fails to overcome resistance due to the poor selectivity over wild-type EGFR (EGFR WT ), resulting in severe side effects [14,15].
Consequently, the urgent need for highly selective third-generation EGFR TKIs that target both activating mutations and T790M resistance mutation leads to the development of osimertinib (AZD9291). AZD9291 exhibits remarkable efficacy in treating patients of T790M mutation-positive NSCLC [16,17]. More recently, third-generation EGFR TKIs, including almonertinib and furmonertinib have been approved to target mutant EGFR bearing T790M [18,19]. However, resistance inevitably emerges in most patients. The acquired resistance mechanisms to AZD9291 are complicated, among which C797S mutation is the most common on-target resistance mechanism [20]. C797S mutation occurs in EGFR exon 20 and accounts for 10-26% of resistant cases to AZD9291 as second-line therapy [21][22][23]. Therefore, the discovery of fourth-generation EGFR inhibitors that can effectively inhibit EGFR T790M/C797S resistance mutation has gained much attention.
Currently, encouraging achievements have been made in the development of the fourth-generation EGFR TKIs, which could be summarized into two categories based on their discovery strategies. 1) Exploiting allosteric binding pocket that is remote from the location of C797S mutation. For example, EAI045, an allosteric inhibitor, is reported as the first fourth-generation EGFR inhibitor overcoming T790M and C797S resistance mutation [24]. However, EAI045 is not effective in the absence of combination with anti-EGFR antibody such as cetuximab. Subsequently, its derivative JBJ-04-125-02 exhibits enhanced efficacy as a single agent but it is not effective to overcome 19del/T790M/C797S mutation [25]. 2) Developing ATP-competitive inhibitors through structural modification of ALK/EGFR dual inhibitor brigatinib or the third-generation EGFR TKIs. Brigatinib is reported to be potent against EGFR triple mutation in vitro and in vivo, but quite similar to that with EAI045 treatment, its antitumor efficacy is limited as a single agent [26]. Further potency improvement-oriented modification of brigatinib leads to the discovery of a series of EGFR T790M/C797S inhibitors showing potential antitumor activity of AZD9291-resistant triple mutant EGFR in vitro and in vivo, including TQB3804, LS-106 [27] and compound 12 [28]. Besides, modification based on scaffolds of third-generation EGFR TKIs is proposed to maintain inhibitory potency and mutant selectivity against EGFR T790M . Typical examples include BI-4020 [29] (from EGF-816 [30]), CH7233163 [31] (from compound 42 [32]) and JND3229 [33]/compound 8r-B [34] (from compound 2v [35]). We also reported a series of macrocyclic compounds based on the structure of AZD9291 [36]. Nonetheless, up to the present, there are still no fourth-generation EGFR inhibitor approved for C797S-resistant patients with disease progression following treatment with AZD9291, it is necessary to develop novel fourth-generation EGFR TKIs.
In this study, through structural hybridization of brigatinib with EGFR T790M inhibitor AZD9291, we identified a potent fourth-generation EGFR inhibitor, HCD3514, which is effective as a single agent to overcome EGFR triple mutation (EGFR L858R/T790M/C797S and EGFR 19del/T790M/C797S ) and shows promising antitumor activity both in vitro and in vivo.

Cell culture
BaF3 cells and PC-9 cells were obtained from Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH (DSMZ) and the European Collection of Authenticated Cell Cultures (ECACC), respectively. MRC-9 cells and GSE-1 cells were purchased from the American Type Culture Collection (ATCC). All the cells were maintained in RPMI-1640 medium (Gibco) or EMEM medium (Gibco) supplemented with 10% FBS (Gibco) at 37 °C in 5% CO2.

Compounds
Compound HCD3514 was designed and synthesized by Ke Ding laboratory. 1 H NMR and 13 C NMR spectra was recorded on a Bruker AV-400 spectrometer at 400 MHz and Bruker AV-600 spectrometer at 151 MHz, respectively, in CDCl3. Coupling constants (J) are expressed in hertz (Hz). Chemical shifts (δ) of NMR are reported in parts per million (ppm) units relative to internal control (TMS). The first-order peak patterns are indicated as s (singlet), d (doublet), t (triplet), q (quadruplet). Complex non-first order signals are indicated as m (multiplet). The high-resolution ESI-MS results were recorded on an Applied Biosystems Q-STAR Elite ESI LC-MS/MS mass spectrometer. The purity of compound was determined by reverse-phase high-performance liquid chromatography (HPLC) analysis using an Agilent 1260 system (G1310B Iso pump and G1365D MWD VL detector) with an YMC Triart C18 reversed-phase column (250 mm × 4.6 mm, 5 μm) at 254 nm. Elution was MeOH in water, and flow rate was 1.0 mL/min. 1

Molecular docking
The molecular docking procedure was performed within Glide 7.9 software [38], and the results were prepared by PyMol 2.6 software. The protein-ligand complex crystal structure of AZD9291 with EGFR L858R/T790M/C797S was selected as the template to elucidate the binding mode of HCD3514. Protein structure was downloaded from Protein Data Bank (PDB 6LUD). The EGFR enzyme was defined as a receptor and the docking grid was centered on the ligand binding location. Docking simulations were performed in standard precision (SP) mode. Other parameters were set as default. After accomplishment of the molecular docking procedure, eight docking poses were scored and selected based on calculated energy.

Cell proliferation assay
BaF3 cells, PC-9-OR cells, MRC-9 cells and GSE-1 cells were plated in 96-well plates in the corresponding medium and incubated overnight, followed by exposing to medium containing serial dilutions of compounds (HCD3514, brigatinib and AZD9291). After 72 h of drug treatment, cell counting kit-8 (CCK8) or sulforhodamine B (SRB) was added to each well and then measured by SoftMax Pro software at the absorbance of 450 nm or 560 nm [39].

Cell apoptosis analysis
BaF3-EGFR L858R/T790M/C797S cells, BaF3-EGFR 19del/ T790M/C797S cells and PC-9-OR cells were seeded in 6-wells plates overnight, and exposed to indicated concentrations of compounds HCD3514 or brigatinib for 48 h. Then cells were collected and washed with PBS and measured by Annexin V-FITC Apoptosis detection kit (cat. no. A211-02, Vazyme Biotechnology). Signals were measured using FACS Calibur flow cytometer, and FlowJo software was used to analyze the data.

Animal study
BaF3-EGFR 19del/T790M/C797S cells (2 × 10 6 ) were injected subcutaneously into BALB/c nude mice (Shanghai institute of medicine). After animals were randomly assigned to groups, tumor-bearing mice were orally administered vehicle, test compound HCD3514 or AZD9291 at the indicated doses once a day and tumor size were measured once per week. Tumor volume was calculated as length × width 2 × 0.5 (mm 3 ) and the tumor growth inhibition (TGI) = [1-RTV (treated) / RTV (control)] × 100%. Mice were sacrificed and tumors were harvested for Western blot analysis. All animal study was approved by the Institutional Animal Care and Use Committee of the Shanghai Institute of Materia Medica and strictly performed according to the institutional ethical guidelines on animal care.

Statistical analysis
Data were presented as mean ± Standard Deviation (SD) and were analyzed by GraphPad Prism 8.0 software. Statistical analysis of the difference between vehicle and compound treated groups was compared by a two-tailed Student's t-test. Statistical significance was defined as *P < 0.05, **P < 0.01, ***P < 0.001.

Identification of HCD3514 as a selective inhibitor of EGFR C797S triple mutant
To obtain a novel and selective inhibitor targeting EGFR T790M/C797S , we incorporated important pharmacophore of brigatinib to the 4-indolyl-2phenylaminopyrimidine scaffold of EGFR T790M inhibitor AZD9291 and developed a novel series of potent and mutant selective EGFR T790M/C797S inhibitor. Following extensive medicinal chemistry optimization by adding H-bond receptor of the indole ring and hydrophobic moiety on the benzene ring, we identified HCD3514 as a selective EGFR T790M/C797S inhibitor ( Figure 1A).
To further elucidate the binding interaction of HCD3514 and EGFR L858R/T790M/C797S , a molecular docking experiment was performed. As illustrated in Fig 1C, HCD3514 bound to the ATP binding site of EGFR with a reversible "U-shaped" configuration similar to AZD9291. The aminopyrimidine core formed a bidentate hydrogen bond interaction with the "hinge" residue Met793 and the chloride atom was directed toward the mutated Met790. Notably, an additional hydrogen bond was observed between sulfonate oxygen and Lys745, which may explain the high potency against EGFR C797S mutants.
To further verify the mechanism of the antiproliferative activity, Western blot assay was employed to examine the ability of HCD3514 to inhibit cellular EGFR phosphorylation. The results revealed that HCD3514 potently and dosedependently suppressed the EGFR phosphorylation in both triple mutant BaF3 cells and PC-9-OR cells ( Figure 2B). Protein expression of phosphorylated EGFR was significantly reduced following brigatinib treatment, but AZD9291 at 1 μM concentration failed to suppress the level of EGFR phosphorylation in all three EGFR C797S triple mutant cells.
Moreover, to evaluate the cytotoxic effect of HCD3514, we also tested the antiproliferative activity of HCD3514 in normal human lung (MRC-9) and gastric mucosal (GSE-1) cell lines. As shown in Supplementary Figure S1, both the human normal cell lines (MRC-9 and GSE-1 cells) were less responsive and showed weak growth inhibition after HCD3514 treatment with IC50 of 1.03 μM and 1.04 μM, respectively, which was comparable to that in BaF3 parental cells with IC50 of 1.31 μM (Figure 2A), indicating the weak cytotoxic effect of HCD3514.
Thus, these results demonstrated the in vitro anti-tumor activity of HCD3514 in EGFR T790M/C797S triple mutant cells through suppression of EGFR phosphorylation, leading to the inhibition of cellular proliferation.

HCD3514 induced apoptosis in cells harboring EGFR C797S triple mutation
Given that NSCLC cells harboring EGFR mutations have been reported to undergo apoptosis [41,42], we assessed the effect of HCD3514 treatment on cell apoptosis. As shown in Figure 3, the proportion of apoptotic cells in HCD3514-treated cells increased in a dose-dependent manner compared to vehicle-treated cells. In BaF3-EGFR L858R/T790M/C797S cells, HCD3514 triggered significant apoptosis with apoptosis rates of 72.39% and 98.27% at the concentration of 1 μM and 3 μM, respectively ( Figure  3A), similar results were detected in BaF3-EGFR 19del/T790M/C797S cells ( Figure 3B) and PC-9-OR cells ( Figure 3C). In contrast, brigatinib only induced significant apoptosis at the concentration of 3 μM with apoptosis rates around 80% in all three EGFR C797S mutant cells, indicating that HCD3514 exhibited more potent apoptosis-inducing efficacy than brigatinib (Supplementary Figure S2).

Figure 2. Cellular activity of HCD3514 against engineered-BaF3 cells and PC-9-OR tumor cells expressing EGFR C797S triple mutation. (A)
Antiproliferation activity of HCD3514, brigatinib and AZD9291 in BaF3 parental cells, BaF3 cells that overexpressed EGFR L858R/T790M/C797S or EGFR 19del /T790M/C797S triple mutation, as well as PC-9-OR cells. The IC50 values were shown as mean ± SD of three independent experiments. (B) The activities of the compounds on the activation of EGFR mutant in these cells. All cells were treated with HCD3514, brigatinib or AZD9291 at indicated concentration for 2 h, and the expression of phospho-EGFR and total EGFR was detected. The phospho-EGFR expression was quantified by ImageJ software and data shown as mean ± SD of three independent experiments. *P < 0.05, **P < 0.01.
To gain deeper insight into the potential mechanism of apoptosis induction in EGFR C797S triple mutant cells, the expressions of apoptosis-related genes were tested by Western blot analysis. Poly ADP-ribose polymerase (PARP) and cysteinyl aspartate specific proteinase-3 (caspase-3) play an important role in cellular processes, including apoptosis [43,44]. It was observed that the expression of caspase-3 was down-regulated after treatment with HCD3514 in triple mutant BaF3 and PC-9-OR cells. Correspondingly, the expressions of pro-apoptotic proteins, including cleaved caspase-3 and cleaved PARP were dose-dependently up-regulated ( Figure  3D). In conclusion, HCD3514 induced apoptosis through a caspase-related mechanism.

HCD3514 suppressed tumor growth in EGFR C797S triple mutant BaF3 xenograft model
Based on the high potency of HCD3514, we then evaluated the activity of HCD3514 in vivo in BaF3-EGFR 19del/T790M/C797S xenograft mouse model. BALB/c nude mice bearing established BaF3-EGFR 19del/T790M/ C797S mouse xenograft tumors were daily oral treatment with HCD3514 (50 and 75 mg/kg), AZD9291 (25 mg/kg) or vehicle control. Consequently, HCD3514 resulted in an obvious growth inhibition of BaF3-EGFR 19del/T790M/C797S mouse xenograft tumors in a dose-dependently manner with tumor growth inhibition (TGI) values of 22.53% and 37.73% at the dosage of 50 mg/kg and 75 mg/kg, respectively, while AZD9291 treatment did not show significant reduction of tumor growth in the xenograft model compared to the vehicle treated mice ( Figure  4A). Then tumor tissues were collected for follow-up detection after 2 h of the last treatment with HCD3514, AZD9291 or vehicle. Tumor weight was reduced with the increased concentration of HCD3514 ( Figure 4B). In addition, the treatment was welltolerated, with body weight maintained throughout the study, suggesting a favorable safety of HCD3514 ( Figure 4C).
The antitumor potency of HCD3514 in BaF3-EGFR 19del/T790M/C797S mouse xenograft model was further confirmed by Western blotting that showed decreased phosphorylation of EGFR and the downstream signaling phosphorylation of Erk level, indicating that the tumor growth inhibition of the resistant triple mutant model due to the inhibition of the activation of EGFR and EGFR-related molecule. However, as expected, AZD9291 treatment showed no obvious inhibitory activities of EGFR and Erk ( Figure 4D).
Taken together, we identified HCD3514 as a potent and orally active fourth-generation EGFR inhibitor and exhibited an excellent antitumor effect in vitro and in vivo.

HCD3514 inhibits not only EGFR T790M/C797S triple mutations, but also EGFR T790M double mutations
In addition, we checked whether HCD3514 inhibits EGFR-double mutations (EGFR L858R/T790M and EGFR 19del/T790M ), as suggested by the molecular docking experiment showed in Figure 1C. According to the docking results, the aminopyrimidine core of HCD3514 formed a bidentate hydrogen bond interaction with the "hinge" residue Met793 and the chloride atom was directed toward the mutated Met790. The hydrophobicity conferred by Met790 mutation likely contributes to the potency of HCD3514 against the T790M mutant. As shown in Figure 5, HCD3514 potently inhibited EGFR L858R/T790M and EGFR 19del/T790M mutations with IC50 values of 4.9 and 8.9 nM in biochemical, respectively. Moreover, HCD3514 showed antiproliferative activities in BaF3-EGFR L858R/T790M and BaF3-EGFR 19del/T790M cells with IC50 of 0.66 μM and 0.52 μM, respectively, although the IC50 values of HCD3514 were higher than that of AZD9291. Furthermore, Western blot analysis revealed that HCD3514 dose-dependently suppressed the EGFR phosphorylation in both BaF3-EGFR L858R/T790M cells and BaF3-EGFR 19del/T790M cells. Besides, we also evaluated the in vitro anti-tumor activity of HCD3514 in NCI-H1975 cells harboring EGFR L858R/T790M . HCD3514 also displayed an excellent antiproliferative potency in NCI-H1975 cells with IC50 of 0.49 μM and the expression of phospho-EGFR was down-regulated after treatment with HCD3514 in a dose-dependent manner. Thus, HCD3514 could target EGFR T790M double mutations as well as EGFR T790M/ C797S triple mutations.

Discussion
Oncogenic mutations in EGFR are prevalent in NSCLC. Though third-generation EGFR inhibitor AZD9291 has achieved a great clinical benefit in NSCLC patients, the EGFR C797S acquired mutation would compromise the therapeutic efficacy. Currently, no approved small molecule EGFR inhibitors are available for NSCLC patients with resistant triple mutant EGFR, and thus, discovery of a potent and high selective EGFR inhibitor to overcome the AZD9291 mediated resistance was urgent and meaningful. Current discovery strategies of the fourth-generation EGFR inhibitors include: 1) exploiting allosteric binding pocket by allosteric inhibitors. 2) developing ATP-competitive inhibitors by structural modification of ALK/EGFR inhibitor brigatinib or the third-generation EGFR TKIs.  The activities of the compounds on the activation of EGFR mutant in EGFR T790M double mutant cells. All cells were treated with HCD3514 or AZD9291 at indicated concentration for 2 h, and the expression of phospho-EGFR and total EGFR was detected. The phospho-EGFR expression was quantified by ImageJ software and data shown as mean ± SD of three independent experiments. *P < 0.05, **P < 0.01.
In this study, through structural hybridization of brigatinib with EGFR T790M inhibitor AZD9291, we identified a new fourth-generation EGFR triple mutant-selective inhibitor HCD3514, which possessed a promising antitumor activity both in vitro and in vivo. Compared to AZD9291, molecular docking of HCD3514 identified an additional hydrogen bond interaction towards Lys745, which was proposed to compensate for the loss of covalent bound due to the C797S point mutation and ultimately enhanced the inhibitory efficacy of HCD3514 on EGFR C797S mutations. HCD3514 displayed a single digitnanomolar kinase inhibitory activity against triple mutant EGFR, including EGFR L858R/T790M/C797S and EGFR 19del/T790M/C797S , which showed superior bioche-mical activity than brigatinib. Importantly, HCD3514 exhibited a sufficient EGFR WT sparing window, suggesting its low toxicity. In addition, HCD3514 simultaneously inhibited the activities of both mutant forms of EGFR at the cellular level. It showed that both engineered BaF3 cells harboring EGFR L858R/ T790M/C797S or EGFR 19del/T790M/C797S were sensitive to HCD3514, as well as the PC-9-OR cells (EGFR 19del/ T790M/C797S expressing PC-9 lung cancer cells), with enhanced inhibition the phosphorylation of EGFR. Besides, cleaved PARP and cleaved caspase-3 levels were showed to increase in three C797S mutant cells, which was consistent with apoptosis induction in a concentration-dependent manner. Furthermore, unlike EAI045 or brigatinib which showed limited antitumor effect without combined with anti-EGFR antibody [24,26], HCD3514 as a single agent exhibited potent antitumor efficacy in tumor xenograft model in a dose-dependent manner without toxicity suggested the potential to target C797S mutant tumors. HCD3514 treatment revealed a significant inhibition of phosphorylation of EGFR in tumor xenografts, supporting that the antitumor efficacy of HCD3514 was through the inhibition of AZD9291-resistant EGFR mutations. Moreover, HCD3514 potently inhibited not only EGFR C797S triple mutations but also EGFR T790M double mutations.
In summary, our preclinical data demonstrated that HCD3514 was effective in overcoming both triple mutant and double mutant forms of EGFR-dependent resistance mechanisms in vitro and in vivo and we also provided a novel chemical scaffold and a new lead compound that worthy of further investigation.