Narciclasine as a potential therapeutic agent to overcome EGFR-TKI resistance in non-small cell lung cancer

Abstract

Background

Non-small cell lung cancer (NSCLC) is associated with abnormal activation of the epidermal growth factor receptor (EGFR) due to overexpression or mutations. While EGFR tyrosine kinase inhibitors (TKIs), such as gefitinib, are used to treat NSCLC, resistance often develops, due to additional EGFR mutations or activation of alternative signaling pathways. Therefore, novel drugs to overcome EGFR-TKI resistance are needed for effective treatment of NSCLC. Narciclasin (Ncs) is a cytotoxic alkaloid from Narcissus species and exhibit antitumor and anti-inflammatory activities.

Methods

Cell viability assay was assessed using trypan blue staining and the Live/Dead viability assay. The growth inhibitory effects of Ncs were evaluated by WST-1 assay and cell cycle analysis across multiple NSCLC cell lines, including expressing A549 and H1299 (wt-EGFR), gefitinib-resistant H1975 (L858R/T790M-EGFR), gefitinib-sensitive PC-9 (exon 19 deleted-EGFR), and gefitinib-resistant PC-9 derivative, PC-9-GR. Ncs binding to wt-EGFR and mutant EGFRs was simulated with molecular docking models. Ncs effects on EGFR kinase activity was evaluated in vitro kinase assay using wt-EGFR and L858R/T790M-EGFR. Anti-tumor effects of Ncs in vivo were assessed using C. elegans tumor model expressing L858R/T790M-EGFR and mouse model xenografted with A549 and H1975. Histological analysis was conducted to measure EGFR, p-EGFR, and p-STAT3 levels in tumor tissues.

Results

Ncs exhibited significant growth inhibitory effects on various NSCLC cell lines, including, A549, H1299 and PC-9 cells (gefitinib-sensitive), and H1975 and PC-9-GR cells (gefitinib-resistant). Notably, Ncs dramatically reduced cell growth with IC50 of 22 nM in H1975 cells expressing gefitinib -resistant EGFR mutant, much lower than any other cell lines. Ncs dramatically induced G2/M arrest in H1975 cells. Ncs binds to both wt-EGFR and mutant EGFRs in molecular docking models and preferentially inhibited the kinase activity of L858R/T790M-EGFR compared to wt-EGFR. In a C. elegans tumor model, Ncs reduced the tumor-mimicking multivulva phenotype. Ncs treatment resulted in decreased tumor growth in mice xenografted with A549 and H1975 cells and lowered levels of EGFR, p-EGFR, and p-STAT3 in tumor tissues.

Conclusions

Our results suggest that Ncs exerts anti-tumor activity by inhibiting EGFR activity and downstream signaling. This effect is particularly evident in cases with EGFR mutations that confer resistance to TKIs, including gefitinib, supporting the potential of Ncs as a therapeutic agent for TKI-resistant NSCLC.

Supplementary information

The online version contains supplementary material available at 10.1186/s12967-025-07368-4.

Background

The cancer with the highest mortality rate in men and women is lung cancer [1]. The most relevant genes in lung cancer are epidermal growth factor receptor (EGFR), KRAS, MET, LKB1, BRAF, PIK3CA, ALK, RET, and ROS1 [2]. EGFR is a tyrosine kinase and its hyperactivity due to its overexpression or mutations are closely linked to many cancers, including lung cancer and breast cancer [3]. EGFR mutations are significant drivers of non-small cell lung cancer (NSCLC) [4]. Currently, 85% of the identified EGFR kinase mutations can be attributed to a single missense point mutation, L858R in exon 21, and to short in-frame deletion variants in exon 19 [5]. The L858R mutation and deletions in exon 19 result in increased and sustained phosphorylation of EGFR, and these mutations are constitutively active even in the absence of ligand binding and dimerization [6]. It has been reported that these mutant EGFRs activate signaling pathways, such as AKT and STAT pathway, which promote cell survival and cell proliferation [7]. To overcome mutation-induced EGFR activation in NSCLC, small molecules that bind to the ATP-binding pocket of the kinase domain of EGFR were generated to inhibit EGFR activity [8]. Tyrosine kinase inhibitors (TKIs) mimic ATP and potently inhibit constitutively activated EGFR [9]. TKIs improve progression-free survival (PFS) compared to standard chemotherapy in chemonaïve patients with advanced NSCLC due to EGFR mutations. Erlotinib and gefitinib, first-generation EGFR-specific TKIs, were found to be superior to chemotherapy in terms of PFS [10, 11]. However, despite good responses to first-generation TKIs initially, patients inevitably became resistant to these drugs within 10–14 months. Various mechanisms of resistance to erlotinib and gefitinib have been identified, and understanding these mechanisms is critical to developing treatment strategies for acquired resistance. The most frequently reported mechanism of acquired resistance is the EGFR T790M point mutation within exon 20 [12]. In addition to conferring resistance to first-generation TKIs, the T790M mutation increases EGFR kinase activity and activates downstream signaling, thereby increasing cancer cell proliferation [13]. Osimertinib, third-generation EGFR-TKI is designed to target specific mutations in the EGFR gene, particularly the T790M mutation compared to normal EGFR [14]. Although osimertinib is highly effective against NSCLC with the L858R/T790M mutation, resistance mechanisms have emerged [15]. Therefore, it remains to develop alternative treatment options for patients with L858/T790M mutation and/or who have developed resistance to EGFR TKIs. Despite the remarkable advancements achieved with osimertinib in managing L858/T790M mutation, therapeutic resistance persists as a formidable obstacle. Emerging strategies, including combinatorial therapeutic regimens, next-generation TKIs, and precision oncology approaches, offer substantial potential to surmount these limitations [15].

Narcissus species have been used as folk medicines for cancer in Asia since the eighteenth century [16]. Narciclasine (Ncs), also known as lycoricidinol, is first isolated from bulbs of several Narcissus species, Amaryllidaceae genera in 1967 during a search for antigrowth factors [17]. Ncs is a cytotoxic alkaloid and it has been shown to exhibit anti-tumor and anti-inflammatory activities [18]. Specifically, Ncs is cytotoxic to lymphoma cells by arresting the cell cycle [19]. Ncs inhibits cell proliferation and induces autophagy-dependent apoptosis by activating the AMPK/ULK1 signaling pathway in triple-negative breast cancer (TNBC). Ncs inhibits glioblastoma multiforme (GBM) proliferation and migration in vitro by inducing actin polymerization, leading to increased cofilin inactivation and disruption of cell polarity [20]. More interestingly, lycorine, a pyrrolo [de] phenanthridine ring-type alkaloid extracted from Amaryllidaceae genera, is known to interact with EGFR in docking model and suppress EGFR activity, and thus inhibiting GBM [21]. It is probable that Ncs represents a compelling candidate as a novel TKI with therapeutic efficacy against NSCLC harboring EGFR mutations.

Here, we identified Ncs as a candidate to inhibit EGFR-driven NSCLC growth. In this study, we demonstrated the anti-cancer effects of Ncs on NSCLC cell lines. We found that Ncs inhibited cell proliferation and induces cell death in NSCLC cells which express either wild type (wt)-EGFR or L858R/T790M-EGFR. Interestingly, Ncs inhibited cell growth much more sensitively in the H1975 cells that express L858R/T790M-EGFR than in the cells with wt-EGFR cell lines. Computer modeling exhibited that Ncs appears to bind to L858R/T790M-EGFR. We also showed that Ncs inhibits kinase activity of L858R/T790M-EGFR as determined by in vitro kinase assay. Intriguingly, in the C. elegans tumor model, Ncs treatment significantly reduced tumor formation established by the expression of L858R/T790M-EGFR. Anti-tumor effects of Ncs in vivo were also observed with xenografted mice, suggesting Ncs as a potential therapeutic agent to control lung cancer.

Methods

Materials

Primary antibodies against total EGFR (#4267, Cell Signaling Technology(CST)), CDK1 (#9116, CST), phospho-EGFR Y1068 (#44-788 G, Invitrogen), total STAT3 (#4904, CST), phospho-STAT3 Y705 (#9145, CST), phospho-Src (#6943,CST), total-Src (#2109, CST), phospho-CDK1 Y15 (#4539, CST), phospho-Wee1 (#4910, CST), Myt1 (#4282, CST), p21 (#2947, CST), Cdc25B (#9525, CST), GAPDH (#5174, CST), β-actin (#A700-057, Bethyl Laboratory), phospho-ERK1/2 (#9101, CST), Bcl-2 (#sc-7382, Santa Cruz), and c-Myc (#sc-789, Santa Cruz) were used for immunoblotting. Horseradish peroxidase (HRP)-conjugated rabbit IgG and HRP-conjugated mouse IgG were purchased from Enzo Life Sciences (Farmingdale, NY). Gefitinib, osimertinib, cycloheximide (CHX), MG132 and propidium iodide (PI) were obtained from Sigma-Aldrich Corporation (St. Louis, MO).

Cell culture

Normal Human Bronchial/Tracheal Epithelial Cells (NHBE) were obtained from Lonza (Basel, Switzerland). Human lung cancer cell lines A549, H1299, PC-9, and H1975, which are non-small cell lung cancer (NSCLC) cell lines, were acquired from the American Type Culture Collection (ATCC, Rockville, MD). NHBE cells were cultured in BEGM medium supplemented with growth factors (Lonza). A549, H1299, PC-9, and H1975 cells were maintained in RPMI medium (Hyclone, Logan, UT) supplemented with 10% FBS (Hyclone, Logan, UT), 100 units/mL penicillin, and 100 μg/mL streptomycin (Thermo Scientific, Waltham, MA). PC-9-Gefitinib-Resistant (GR) cells were generated by continuously exposing PC-9 cells to increasing concentrations of gefitinib, as described in the literature [22] PC-9-GR cells were maintained in the presence of gefitinib (1 μM).

Growth curve determination

Cells were plated in a 96-well plate at a density of 4 × 10³ cells/well and treated with Ncs at the indicated concentrations for 48 hours at 37 °C in a humidified atmosphere containing 5% CO₂. Cell growth was monitored using the LionHeart FX automated microscope (Agilent Technologies, Santa Clara, CA). The growth curve and IC50 values were determined using the logistic growth equation and doubling time (DT) formula: DT = (ln2/K) (where K is a constant calculated by the software), using GraphPad software (New York, NY).

Cell viability assay

Cells were plated in culture plates and treated with Ncs for 24 or 48 hours. Cell viability was assessed using trypan blue staining and the Live/Dead viability assay. For trypan blue staining, cells were trypsinized, stained with 0.2% trypan blue solution, and counted for viable and non-viable cells [23]. For the Live/Dead viability assay, cells were treated with 1 μM PI and 1 μM Calcein AM (Invitrogen) for 30 minutes at 37 °C, followed by fluorescence microscopy using an Axio Observer Z1 (Carl Zeiss Microimaging, Thornwood, NY). For WST-1 assay [24], cells were seeded in 96-well plates (5 × 10³ cells per well) and allowed to adhere for 24 hours before treatment with Ncs at the indicated concentrations and times. Following treatment, 10 μL of WST-1 reagent (Sigma-Aldrich) was added to each well, and cells were incubated for 2 hours at 37 °C. Absorbance was measured using a Biotek microplate reader (Winooski, VT).

Apoptosis assay

Treated cells were harvested, stained with 5 μg/mL PI and 2.5 μg/mL fluorescein isothiocyanate (FITC)-conjugated annexin V reagent in binding buffer, and analyzed by flow cytometry. Data were analyzed using Cell Quest Software (BD Biosciences, Franklin Lakes, NJ).

Cell cycle analysis

Cells were treated with Ncs for the indicated times, fixed with cold 70% ethanol at 4 °C overnight, and stained with propidium iodide (PI) for 30 minutes at room temperature. The cell cycle distribution was analyzed using a FACSCalibur flow cytometer (BD Biosciences, Franklin Lakes, NJ).

3D culture assay

A549 and H1975 cells were seeded in 8-well chamber slides (Nunc™ Lab-Tek™, Thermo Fisher Scientific, Waltham, MA) at a density of 2 × 10³ cells in 300 μL of 5% Matrigel on top of a 100 μL base layer of 100% Matrigel. Cells were incubated in a 5% CO₂ atmosphere at 37 °C for 6 days, and images were captured using an Axio Observer Z1 fluorescence microscope (Carl Zeiss Microimaging, Thornwood, NY). Spheroid areas were quantified using Gen3.1 program (Zeiss).

Immunofluorescence microscopy imaging

Cells were plated on coverslips, fixed with 4% paraformaldehyde for 15 minutes at room temperature, permeabilized with Triton X-100, and blocked with 5% BSA. After overnight incubation with primary antibodies at 4 °C, cells were washed and incubated with Alexa 488- and 568-conjugated secondary antibodies. Nuclei were stained with DAPI (0.5 μg/mL) and visualized using a Carl Zeiss fluorescence microscope (Gottingen, Germany).

RNA isolation and quantitative real time PCR

Total RNA was extracted using GeneAll (Seoul, Korea), and 1 μg of RNA was converted into cDNA using the iScript cDNA Synthesis Kit (Bio-Rad, Hercules, CA). Quantitative PCR was performed using the Roche LightCycler® 96 with SYBR-GREEN qPCR (Roche, Basel, Switzerland). Gene expression was normalized to GAPDH using the comparative Ct method. The sequences of primers were following: human EGFR (F) 5’-ACTGCTGCCACAACCAGTG-3’, human EGFR (R) 5’-GGCTTCGTCTCGGAATTTG-3’, human GAPDH (F) 5’-TCTCTGCTCCTCCTGTTC-3’, and human GAPDH (R) 5’- CGCCCAATACGACCAAAT-3’.

Xenograft model

A549 and H1975 cells (5 × 10⁶) were injected subcutaneously into BALB/c nude mice. When tumors reached 30 mm³, mice received intraperitoneal injections of either control or Ncs (2 mg/kg) three times per week. Tumor volume was measured using calipers and calculated as 1/2(length × width²). The study was approved by the Institutional Animal Care and Use Committee (IACUC) of the National Cancer Center Research Institute (NCCRI). NCCRI is an Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC International) accredited facility and abides by the Institute of Laboratory Animal Resources (ILAR) guide.

Immunohistochemical staining

Tumor tissues were fixed in 10% neutral buffered formalin, paraffin-embedded, and sectioned at a thickness of 4 μm. Sections were dried at 56 °C for 1 hour before immunohistochemical staining, which was performed using the Discovery XT system (Ventana Medical Systems, Tucson, AZ, USA). Sections were deparaffinized, rehydrated using EZ Prep (Ventana Medical Systems), and washed with reaction buffer. Antigen retrieval was performed by heat treatment in Tris-ethylenediaminetetraacetic acid (Tris-EDTA) buffer (pH 8.0, CC1, Ventana Medical Systems) at 90 °C for 30 minutes. The sections were then incubated with primary antibodies, including anti-EGFR (1:100 dilution; CST), anti-phospho-EGFR (Y1068), and anti-phospho-STAT3 (1:200 dilution; CST). Parallel sections incubated with normal IgG instead of primary antibodies served as negative controls. Stained sections were examined using a light microscope (Leica, Wetzlar, Germany), and images were captured using a LAS Image Analyzer.

C. elegans experiments

The wild type and jgIs25 strains were cultured on the nematode growth medium (NGM) plate at 20 °C. The jgIs25 strain was cultured on NGM plates containing 0.5% DMSO, 2 μM gefitinib, and 2 μM NCS, respectively, and observed at the adult stage. C. elegans was observed at high magnification by dropping M9 buffer containing 2 mM levamisole (Sigma, L9756) on a 2% agarose pad, mounting C. elegans on them, and observing them under an interference contrast microscope (Axio A1 microscope, Zeiss, Yena, Germany). C. elegans images were captured using Axiocam 705 mono (Zeiss). C. elegans with polyps were counted under a dissecting microscope in 3 plates each. This experiment was performed more than three times and showed similar results.

Molecular docking

Molecular docking studies performed using the program AMDock 1.5.2 and visualized with Pymol2.3.2 (Schrödinger, USA). We utilized several Protein Data Bank (PDB) models for docking simulations (1XKK, 6JRK, 6JXT, 7B85, and 7U99 for wt-EGFR; 3W3O, 3W2P, 3W2Q, 3W2R, 4RJ4, 4RJ5, 4RJ6, 4RJ7, 4RJ8, 5EDQ, 5EDR, 5y25, and 7OXB for L858R/T790M-EGFR; Alphafold2 generated Exon19 deletion-EGFR).

EGFR kinase assay

EGFR kinase activity was measured using the ADP-Glo™ Kinase Assay (Promega, Madison, WI), which quantifies kinase activity by detecting ADP production. The assay was performed according to the manufacturer’s instructions. Briefly, reactions were carried out in kinase buffer containing 1 mM ATP and 20 ng of purified wt-EGFR or L858R/T790M-EGFR (BPS Bioscience, San Diego, CA) in a total volume of 20 µL. After incubation at 30 °C for 30 minutes, the reaction was stopped by adding the ADP-Glo™ reagent to deplete unconsumed ATP. Subsequently, the kinase detection reagent was added to convert ADP to ATP, which drives a luciferase/luciferin reaction to produce luminescence proportional to kinase activity. Luminescence was measured using a Centro LB 96 microplate luminometer (Promega) with a 1-second integration time.

Statistical analysis

Data were analyzed using Student’s t-test or two-way ANOVA followed by Bonferroni’s post-hoc test. Statistical significance was defined as p < 0.05 and p < 0.01. Data represent means ± standard deviations from three independent experiments. Data represent average values and standard deviations (error bars) obtained from three independent experiments.

Results

Effects of Ncs on NSCLC cell growth

Ncs (C₁₄H₁₃NO₇), also known as lycoricidinol, is a toxic alkaloid compound purified from Narcissus plant (Fig. 1A) and it is known to have anti-proliferative effects on various cancer cells [18]. We examined anti-proliferative effects of Ncs on human lung cancer cell lines, including A549, H1299, and H1975, as well as normal human bronchial epithelial (NHBE) cells by the label-free kinetic live monitoring of cell proliferation under Lionheart automated microscope. A549 and H1299 cells express wt-EGFR while H1975 cells harbor L858R/T790M-EGFR mutation, which confers resistance to gefitinib, an EGFR-TKI [25]. Ncs efficiently inhibited cell proliferation of NSCLC cell lines in a dose-dependent manner and had a little effect on the growth of normal NHBE cells (Fig. 1B). Notably, H1975 cells were significantly more sensitive to Ncs than A549 and H1299 cells, with IC50 values of 22 nM, 110 nM and 88 nM, respectively, indicating that H1975 cells are more sensitive to Ncs (Fig. 1C and Supplementary Fig. 1). Additionally, treatment with 25 nM Ncs significantly suppressed the growth of both A549 and H1975 cells in 3D culture, which better mimics the in vivo tumor environment. Notably, H1975 cell spheroids were smaller than those of A549 cells. (Fig. 1F and 1G). These data indicate that Ncs has anti-proliferative effects in NSCLC cells, particularly in H1975 cells haboring the gefitinib-resistant L858R/T790M-EGFR mutation.

Fig. 1.

Effects of Ncs on cell growth in NSCLC cells. (A) Chemical structure of narciclasine (Ncs). (B) NHBE, A549, H1299, and H1975 cells were treated with the indicated concentrations of Ncs (0–10 μM) for 48 h and cell growth rates were measured with LionHeart FX automated microscope according to the manufacturer’s instruction. (C) Cells were treated with the indicated concentrations of Ncs for indicated times. Cell growth rates were measured as described in (B) and IC50 values were calculated using Gen5 software from Biothech. (D and E) cells were plated in 6-well plate (3 × 10³ cells/well) and treated with the indicated concentrations of Ncs for 5 days, followed by staining with 0.5% crystal violet. Representative images were taken (D) and scanned to quantify three different areas and measured relative staining areas (E). (F and G) 2 × 10³ cells in 300 μl of 5% matrigel were plated in the matrigel-coated 8-well chamber with indicated concentration of Ncs, as described in the “Materials and Methods” section. After 7 days of incubation, images were taken (F), and spheroid areas were quantified using three independent images (G). Scale bar = 50 μm. Similar results were observed in three independent experiments. Error bars represent standard deviations of the mean of three measurements, *p < 0.05, **p < 0.01

Effects of Ncs on apoptosis and cell cycle arrest in NSCLC cells

Ncs-induced cell growth inhibition could be due to cell death and/or cell cycle inhibition. Live/dead assays showed that Ncs increased dead cells (red color) as well as decreased live cells (green color) in both A549 and H1975 cells (Fig. 2A). Flow cytometric analysis of apoptosis using annexin V and propidium iodide (PI) staining also showed that Ncs induced apoptosis of both A549 and H1975 cells, with little difference between the two cells (Fig. 2B). Interestingly, cell cycle analysis revealed that Ncs increased dramatically the population of the G2/M phase population, especially in the H1975 cells, but not in the A549 cells (Fig. 2C). Because H1975 cells contain the L858R/T790M-mutation in EGFR, we examined whether Ncs affects mutant EGFR levels or activity. Ncs decreased both the EGFR activity as determined by EGFR phosphorylation and total EGFR protein levels in H1975 cells, with minimal effects on A549 cells (Fig. 2D). Consequently, in H1975 cells rather in A549 cells Ncs decreased ERK1/2 activation (Fig. 2D), which is a recognized effector in the EGF signaling pathway that controls cell apoptosis, cell growth, and cell cycle progression [26, 27]. These suggest that Ncs decreases EGFR activity and its protein levels, and thus attenuates EGFR signaling pathway.

Fig. 2.

Effects of Ncs on apoptosis and cell cycle. (A) Cells were treated with 50 nM Ncs for 24 h and stained with calcein-AM/PI, followed by image capture using fluorescence microscopy (green: live cells, red; dead cells), scale bars = 50 μm. (B) cells were treated with 50 nM Ncs for 24 h and stained with annexin V and PI and followed by flow cytometric analysis of apoptosis. Apoptotic cells were quantified with early and late apoptotic cell populations. (C) For cell cycle arrays, 50 nM Ncs-treated cells were fixed with 70% ethanol and stained with PI, followed by flow cytometry analysis. (D and E) Cells were treated with indicated concentrations (0–100 nM) of Ncs for 24 h and cell lysates were subjected to immunoblotting analysis using the indicated antibodies. Expression levels of p-EGFR and p-ERK1/2 were quantified from immunoblot bands and normalized to total EGFR and total ERK1/2 expression levels, respectively (D). Data were presented as mean values from three independent experiments and error bars represent standard deviations. *p < 0.05, **p < 0.01

We further investigated whether Ncs affects cell cycle pathway involved in G2/M arrest, including CDK1, cyclin B1, phosphorylated Histone H3 (Fig. 2E). During interphase of cell cycle, CDK1 is kept inactive by phosphorylation of Tyrosine 15 via CDK inhibitory kinases, such as Wee1 and Myt1. In the late G2 phase, cyclin B1 accumulates and activates CDK1, leading to G2/M progression [28]. Phosphorylation of Histone H3 at Ser10 by aurora kinase or ERK1/2 is essential for initiation of chromosome remodeling during mitosis [29] and EGF is known to activate the RAS-MAPK-ERK1/2 pathway, leading to phosphorylation of Histone H3 for mitosis progression [30]. Ncs treatment decreased cyclin B1 levels and increased levels of Wee1 phosphorylation and Myt1, and thus maintained inhibitory CDK1 phosphorylation. Furthermore, Ncs treatment decreased phosphorylation of Histone H3 (Fig. 2E). All these data indicate that Ncs inhibits G2/M transition by disrupting key regulators of G2/M progression, leading to G2/M arrest. It is of note that these Ncs-induced molecular changes in G2/M arrest are more pronounced in H1975 cells than in A549 cells, which aligns with the observation that Ncs dramatically induced G2/M arrest in H1975 cells harboring L858R/T790M-EGFR, known to confer resistance to gefitinib [12].

Effects of Ncs on cell growth in PC-9 and PC-9-GR cells

To further evaluate whether Ncs effectively inhibits cell growth in gefitinib-resistant NSCLC cells, we employed PC-9 and PC-9-GR cells. PC-9 cells express an EGFR mutant with exon 19 deletion (E746–A750 deletion) and are sensitive to gefitinib [31]. PC-9-GR cells were established by culturing PC-9 cells with increasing concentrations of gefitinib, which show acquired resistance to gefitinib (Fig. 3A). Interestingly, Ncs treatment decreased the growth of both PC-9 and PC-9-GR cells in a dose-dependent manner as determined by WST-1 assay for mitochondrial activity (Fig. 3B), and thus reduced cell confluence as observed by microscope (Fig. 3C). While Ncs increased the dead cell population in both cell lines, the effect was more profound in PC-9-GR cells compared to PC-9 cells (Fig. 3D). Similarly, Ncs more effectively decreased the live cell population in PC-9-GR cells than in PC-9 cells (Supplementary Fig. 2). Cell cycle analysis revealed that Ncs induced arrest at G1 and G2/M phases in PC-9 cells, and S and G2/M phases in PC-9-GR cells (Fig. 3E and 3F). Ncs appears to induce G2/M arrest because higher doses of Ncs also increased G2/M arrest in A549 cells (Supplementary Fig. 3). Nonetheless, it is an interesting observation that Ncs-induced G2/M arrest is more evident in EGFR-TKI-resistant cells, such as H1975 and PC-9-GR cells (Figs. 2C and 3F).

Fig. 3.

Effects of Ncs on cell growth in PC-9 and PC-9-GR cells. (A) PC-9 and PC-9-GR cells (3 × 10³ cells/well) were plated in 96-well plates and treated with gefitinib in a dose-dependent manner for 48 h. (B) Cells were treated with Ncs at the indicated concentrations for 48 hours, and cell growth rates were measured using the WST-1 assay. (C) Cells were treated with Ncs for 48 hours, and representative images are shown. (D) Cells (1 × 10⁴ cells/well) were plated in 24-well plates and treated with Ncs at the indicated concentrations for 48 h. Live and dead cells were counted using the trypan blue staining assay. (E and F) Cells were treated with Ncs for 48 h, fixed with 70% ethanol, stained with PI, and analyzed by flow cytometry to assess cell cycle distribution. (G) PC-9 and PC-9-GR cells were treated with the indicated concentrations (0-50 nM) of Ncs for 24 h. Cell lysates were then subjected to immunoblotting analysis using the indicated antibodies. Expression levels of p-EGFR, p-Src, and p-STAT3 were quantified and normalized to total EGFR, total Src, and total STAT3 expression levels, respectively. Data were presented as mean values from three independent experiments and error bars represent standard deviations of the mean of three measurements. Magnification: x 100, n.s. >0.05, *P < 0.05, **P < 0.01

EGFR-TKI resistance occurs as a result of EGFR-dependent and EGFR-independent molecular pathways, such as additional EGFR mutations, and activation of alternative survival signaling pathways, including STAT3 and Src activation [32–34]. EGFR sequencing of PC-9-GR cells revealed no additional mutations of EGFR (Supplementary Fig. 4), indicating that gefinitib resistance is not EGFR-dependent in PC-9-GR cells. Because Ncs downregulated EGFR levels in H1975 cells (Fig. 2D), we examined whether Ncs affects EGFR levels and/or activity in PC-9 and PC-9-GR cells. While EGFR levels were higher in PC-9-GR than in PC-9 cells, EGFR activity as determined by its phosphorylation was higher in PC-9 cells than PC-9-GR cells (Fig. 3G). However, the levels of activated STAT3 (p-STAT3) and activated Src (p-Src) were elevated in PC-9-GR cells compared to PC-9 cells, consistent with studies showing STAT3 and Src activation in gefitinib-resistant cells [33, 34]. Interestingly, Ncs decreased both total EGFR and phosphorylated EGFR levels in PC-9 cells, but not in PC-9-GR cells. Instead, Ncs dramatically reduced p-STAT3 and p-Src levels in PC-9-GR cells (Fig. 3G), suggesting that Ncs inhibits STAT3 and Src signaling, which could contribute to cell growth inhibition in PC-9-GR cells. H1975 cells and PC-9-GR cells possess different EGFR mutations, yet both exhibit resistance to gefitinib. Ncs significantly inhibited the growth of both cell lines at as low as 10 nM, suggesting similar sensitivity to Ncs (Supplementary Fig. 5).

Effects of Ncs on EGFR stability and activity

Ncs decreased levels of both total EGFR and phosphorylated EGFR in H1975 cells in a dose-dependent (Fig. 2D) and in a time-dependent manner (Fig. 4A). Interestingly, Ncs treatment increased EGFR mRNA levels (Fig. 4B), indicating that the reduction in EGFR occurs at the protein level, rather than at the transcriptional level. To investigate whether Ncs affects EGFR protein stability, H1975 cells were treated with cycloheximide (CHX), a protein synthesis blocker, either alone or in combination with Ncs. CHX treatment decreased EGFR protein levels over time and this decrease was further enhanced when cells were treated with Ncs together (Fig. 4C), indicating that Ncs shortens EGFR half-life. MG132, a proteasome inhibitor, and BafA1, an autophagosome-lysosome fusion inhibitor, also attenuated Ncs-indued EGFR degradation (Fig. 4D), indicating that EGFR is degraded through both the proteasomal and lysosomal degradation pathways. Immunostaining for EGFR and LAMP1 (a lysosome marker) revealed that EGFR was distributed throughout the cell under control condition whereas EGFR and LAMP1 were colocalized upon Ncs treatment (Fig. 4E), suggesting that lysosomal degradation could be involved in Ncs-induced EGFR downregulation. We further examined whether Ncs affects EGF-stimulated EGFR activation and found that Ncs inhibits EGF-induced EGFR phosphorylation as well as its basal phosphorylation (Fig. 4F). These results suggest that Ncs reduces EGFR activation and expression by EGFR degradation through proteasomal and lysosomal degradations.

Fig. 4.

Effects of Ncs on EGFR activity and EGFR signaling. (A) H1975 cells were treated with 50 nM of Ncs for indicated times and cell lysates were subjected to immunoblotting analysis using the indicated antibodies. (B) Cells were treated with 50 nM of Ncs for the indicated durations, and total RNA was isolated. EGFR mRNA levels were determined by real-time PCR analysis and normalized to GAPDH expression. (C) Cells were treated with 20 μM of cycloheximide (CHX) for the indicated time points in the presence of either DMSO or 50 nM of Ncs for 24 h. Cell lysates were subjected to immunoblotting analysis using the indicated antibodies (upper panel). EGFR expression levels were quantified from immunoblot bands and normalized to GAPDH, comparing CHX treatment in the presence of DMSO or Ncs (lower panel). (D) Cells were treated with 50 nM of Ncs for 24 h in the presence of BafA1 (20 nM, 12 h) or MG132 (20 μM, 2 h). Cell lysates were subjected to immunoblotting analysis using indicated antibodies. (E) Cells were plated on coverslips and treated with DMSO (control) or 50 nM of Ncs for 24 h. Cells were then fixed with 4% paraformaldehyde (PFA) and subjected to immunofluorescence staining using DAPI, EGFR, and LAMP1 antibody. (F) Cells were treated with 50 nM of Ncs for 24 h, followed by EGF (10 nM, 15 min) stimulation. Cell lysates were subjected to immunoblotting analysis using the specified antibodies. Similar results were observed in three independent experiments. Error bars represent standard deviations of the mean of three measurements, *P < 0.05, **P < 0.01, ***P < 0.001

Effects of Ncs on wt-EGFR and T790M/L858R-EGFR activity

Lycorine, an isocarbostyril compound with a structure close to Ncs, has been reported to interact with wt-EGFR in docking model [21]. Considering that Ncs inactivated EGFR and inhibited the growth of NSCLC cells, including A549, H1975, and PC-9 cells bearing wt-EGFR, L858R/T790M-EGFR, and exon 19 deleted-EGFR, respectively (Figs. 1–3), we investigated whether Ncs also interacts with these EGFRs by conducting docking simulations using various PDB models. Although the binding modes of Ncs to each wt-EGFR and mutant-EGFRs are quite variable, the binding energies were in similar ranges (7.2~8.4 kcal/mol). Remarkably, hydrophobic and van der Waals interactions were likely to be the major interaction forces in both wt-EGFR and L858R/T790M-EGFR, compared to hydrogen bonding (Fig. 5A). In detail, Ncs was predicted to interact with 16 residues in wt-EGFR, 18 residues in L858R/T790M-EGFR, and 16 residues in exon 19 deleted-EGFR, although specific interacting residues varied. Because three-dimensional structures of exon 19 deleted-EGFR (E746–A750) is not known yet, the deletion model was generated using AlphaFold. While five hydrogen bonds were predicted in exon 19 deleted-EGFR with Ncs (involving Glu758, Arg836, and Tyr869, with Arg836 probably forming triple hydrogen bonds), single hydrogen bond was predicted for wt-EGFR (Ser720) and L858R/T790M-EGFR (Asn842). Taken all together, these results suggest that Ncs interacts with wt-EGFR and mutant EGFRs variants.

Fig. 5.

Docking model and kinase activity of Ncs on EGFR. (A) Docking simulation results of Ncs binding to three EGFR (wt, L858R/T790M, E746-A750 deletion). Residues involved in interaction with Ncs were colored in magenta and hydrogen bonding are drawn in dashed lines. (B) Effect of Ncs on EGFR kinase activity, as assessed by the ADP-Glo kinase assay. Ncs, gefitinib (a known EGFR inhibitor), and osimertinib (a third-generation EGFR inhibitor) were incubated with 20 ng of wt-EGFR and mutant EGFR for kinase activity assessment. Similar results were observed in three independent experiments. Error bars represent standard deviations of the mean of three measurements, *P < 0.05, **P < 0.01

To further investigate whether Ncs inhibits kinase activity, we performed an in vitro kinase assay using purified wt-EGFR and L858R/T790M-EGFR. Gefitinib suppressed wt-EGFR kinase activity but not L858R/T790M-EGFR kinase activity (Fig. 5B), consistent with the known gefitinib resistance of the L858R/T790M-EGFR [25]. Osimertinib, a third-generation EGFR TKI that targets specific mutations particularly the T790M mutation in EGFR [35], effectively suppressed the kinase activities of both wt-EGFR and L858R/T790M-EGFR. Interestingly, while Ncs exhibited a little inhibition of wt-EGFR kinase activity, it significantly suppressed the kinase activity of L858R/T790M-EGFR, albeit to a lesser extent than osimertinib. Altogether, our data indicate that Ncs binds and suppresses the EGFR kinase activity, with a stronger inhibitory effect on L858R/T790M-EGFR than on wt-EGFR in vitro.

Anti-cancer effects of Ncs in vivo

To further explore the anti-cancer effects of Ncs in vivo, we employed the C. elegans tumor model jgIs25 strain that expresses human L858R/T790M-EGFR [36]. Compared with the wild type worm, the jgIs25 strain forms polyps on the ventral side, which is known as a multivulva (Muv) phenotype (Fig. 6A) and this Muv phenotype serves as a tumor formation [36]. When the jgIs25 strain was treated with 2 μM gefitinib, the polyp formation remained unchanged, similar to control group (Fig. 6B). In contrast, when jgIs25 strain was treated with 2 μM Ncs, polyp formation significantly decreased and thus the number of individuals showing the Muv phenotype was reduced to half of the control group (Fig. 6B and 6C). In addition, BALB/c nude mice were injected subcutaneously with either A549 or H1975 cells, and tumor-bearing mice were received intraperitoneal injections of DMSO or Ncs three times a week for 27 and 17 days, respectively. Both A549 tumor-bearing and H1975 tumor-bearing mice injected with Ncs displayed slower tumor growth and lighter tumor weights than mice injected with DMSO (Fig. 6D, 6E, 6G, and 6H). Immunohistochemistry analysis of tumor tissues revealed that levels of EGFR, p-EGFR, and p-STAT3 were lower in the Ncs-treated group than in control group (Fig. 6F and 6I). Taken all together these data suggest that Ncs has a therapeutic potential to control lung cancer especially with gefitinib resistant mutations through mechanisms that are both EGFR-dependent and EGFR-independent (Fig. 7).

Fig. 6.

Anti-cancer effects of ncs in vivo. (A) Adult-stage C. elegans models: wild type (WT) and jgIs25 strains. The jgIs25 strain exhibits polyps on the ventral side (arrows in the lower panel). Scale bar = 20 μm. (B) Effects of gefitinib and ncs treatment on jgIs25. the arrows in the upper panel indicate polyps, and the rectangular area is magnified in the lower panel. Scale bar = 200 μm (upper), 20 μm (lower). (C) Quantification of polyp formation in jgIs25 following treatment with the indicated drugs. (D and G) In vivo tumor model using subcutaneous transplantation of A549 and H1975 cells into nude mice, followed by ncs treatment. Tumor volumes were measured at the indicated time points in control and ncs-treated mice. Tumor weights were measured after sacrifice. (E and H) Representative image of xenograft tumors after sacrifice. (F and I) Immunohistochemical (IHC) staining of xenograft tumors using anti-EGFR, anti-phospho-EGFR (Y1068), and anti-phospho-STAT3 (Y705) antibodies. Scale bar = 500 μm. Similar results were observed in three independent experiments. Error bars represent standard deviations of the mean of three measurements, *p < 0.05, **p < 0.01, *p < 0.001

Fig. 7.

Ncs-mediated cell death in NSCLC cells with different EGFR mutations. (A) In NSCLC cells expressing wild type-EGFR (wt-EGFR), ncs induces cell death through EGFR inhibition or alternative mechanisms (indicated by “?”). (B) In NSCLC cells expressing the L858R/T790M EGFR mutation, which are resistant to gefitinib, ncs induces cell death through EGFR inhibition and/or inhibition of alternative pathways. (C) Although NSCLC cells expressing exon 19-deleted EGFR are initially sensitive to gefitinib, they can develop gefitinib resistance. However, these resistant cells remain sensitive to ncs via the inhibition of STAT3 and Src

Discussion

In this study, we demonstrate that Ncs, an alkaloid from Narcissus plants, is effective at inhibiting growth of lung cancer cells with EGFR mutations, which are resistant to an EGFR TKI, gefitinib, indicating that Ncs is an alternative chemotherapeutic reagent for NSCLC bearing TKI-resistant EGFR mutants.

Both Ncs and lycorine share a crinine skeleton structure, which consists of a bicyclic system with a perhydroindole fused to a pyrrolidine ring [18]. Lycorine has been reported to have anti-tumor effects on glioblastoma and molecular docking model has shown that lycorine binds to the EGFR tyrosine kinase pocket domain via its hydrogen bond interacting to Asn842, Lys 745, and Thr854 [21]. Since Ncs has been shown to have more potent anti-proliferative effects than lycorine [19], and it inhibited cell growth more efficiently in H1975 cells (Fig. 1 and Supplementary Fig. 1), we examined whether Ncs also interacts with wt-EGFR and EGFR mutants using molecular docking models. Ncs was predicted to bind to both wt-EGFR and EGFR mutants primarily through hydrophobic and van der Waals interactions. While hydrogen bond formation was predicted in wt-EGFR and EGFR mutants, the specific residues involved in hydrogen bonding differed due to the presence of distinct mutations (Fig. 5A). Interestingly, in vitro kinase assay revealed that Ncs inhibited kinase activity of L858R/T790M-EGFR more effectively than that of wt-EGFR (Fig. 5B). Specifically, the kinase activity of wt-EGFR was slightly inhibited at 50 nM Ncs and significantly inhibited at 100 nM Ncs. However, L858R/T790M-EGFR kinase activity was significantly inhibited at 25 nM Ncs, with aptoximately 50% inhibition at 50 nM Ncs, reaching a plateau thereafter (Supplementary Fig.6). These data indicate that while Ncs inhibits the kinase activity of both wt-EGFR and L858R/T790M-EGFR, a relatively lower concentration of Ncs is required to inhibit L858R/T790M-EGFR in vitro. Consistent with these kinase assay results, Ncs dramatically inhibited EGFR activity in H1975 cells harboring L858R/T790M-EGFR compared with in A549 cells harboring wt-EGFR (Fig. 2D). In addition, Ncs inhibited cell growth more effectively in H1975 cells harboring than in A549 cells or H1299 cells with IC50 of 22 nM, 110 nM and 88 nM Ncs, respectively (Fig. 1B). To determine whether L858R/T790M-EGFR is critical for Ncs sensitivity, endogenous wt-EGFR and L858R/T790M-EGFR were knockdowned using si-EGFR RNA, and L858R/T790M-EGFR and wt-EGFR were overexpressed in A549 and in H1975 cells, respectively (Supplementary Fig. 7A). Notably, in A549 cells, ectopic expression of L858R/T790M-EGFR combined with silencing of wt-EGFR led to increased sensitivity to Ncs. In H1975 cells, the enhanced sensitivity to Ncs was abolished when wt-EGFR was overexpressed and L858R/T790M-EGFR was silenced (Supplementary Fig. 7B). These results demonstrate that Ncs sensitivity is dependent on the presence of the L858R/T790M-EGFR.

In A549 cells, treatment with 50 nM of gefitinib, erlotinib, or Ncs resulted in similar levels of cell growth inhibition, comparable to the effect of 10 nM osimertinib. However, in H1975 cells, both gefitinib and erlotinib had little effects on cell growth inhibition, whereas Ncs and osimertinib significantly inhibit cell growth, with osimertinib showing a somewhat stronger effect (Supplementary Fig. 8). Furthermore, 50 nM Ncs treatment significantly induced G2/M cell cycle arrest in H1975 cells, but not in A549 cells (Fig. 2C). However, higher doses of Ncs (50–200 nM) induced G2/M arrest in A549 cells (Supplementary Fig. 3). Interestingly, Ncs treatment significantly inhibited the growth of osimertinib-resistant H1975 cells in a dose-dependent manner (Supplementary Fig. 9), with an IC50 of 49.7 nM. These results suggest that Ncs could be a potential chemotherapeutic agent for patients with osimertinib-resistant lung cancer.

EGFR endocytosis and downregulation decreases the availability of EGFR on the cell surface, thereby limiting EGFR intracellular signaling, which is crucial for cell growth and survival [37]. There are two major types of therapeutic strategies to target EGFR, such as anti-EGFR antibodies and TKIs. Anti-EGFR antibodies bind to the extracellular domain of the receptor, preventing ligand binding and/or triggering receptor internalization and downregulation [38]. For example, Okada et. al. has demonstrated that anti-EGFR antibody promotes EGFR downregulation through the endosomal/lysosomal pathway, a process strongly associated with its anti-tumor activity in colorectal cancer cells [39]. EGFR TKIs inhibit EGFR kinase activity [40] and in some cases, they also downregulate EGFR. For example, gefitinib has been reported to reduce protein levels of EGFRvIII mutant at high doses [41] and erlotinib treatment downregulates EGFR protein levels in head and neck cancer patients and cell lines [42]. In our study, we also observed that Ncs treatment significantly downregulated EGFR protein levels, despite an increase in EGFR mRNA levels (Fig. 4B). Ncs induced the colocalization of EGFR with LAMP1 (Fig. 4E), and Ncs-induced EGFR downregulation was reversed by BafA1, and MG132 (Fig. 4D). Moreover, Ncs not only reduced basal EGFR activity but also inhibited EGF-stimulated EGFR activation (Fig. 4F). These data indicate that Ncs inhibits EGFR kinase activity both by direct kinase inhibition and by promoting EGFR downregulation.

Gefitinib is known to sensitively inhibits kinase activity of exon 19 deleted-EGFR and PC-9 cells harboring this mutation are sensitive to gefitinib treatment [43]. We established gefitinib resistant PC-9-GR cell line (Fig. 3A). Since no additional mutations were found in the exon 19 deleted-EGFR in PC-9-GR cells (Supplementary Fig. 4), gefitinib resistance appears to be independent of EGFR and is likely dependent on alternative survival pathways. Indeed, despite elevated EGFR levels, p-EGFR levels were reduced in PC-9-GR cells compared to PC-9 cells (Fig. 3G). However, enhanced activation of Src and STAT3 was observed in PC-9-GR cells relative to PC-9 cells. Ncs treatment decreased p-EGFR levels in PC-9 cells but not in PC-9-GR cells. However, Ncs decreased levels of p-Src and p-STAT3 in PC-9-GR cells. It has been reported that Ncs exhibits anti-tumor effects through STAT3 inactivation, and Ncs directly binds to STAT3 in tamoxifen-resistant breast cancer cells [44]. Because Ncs induced cell growth inhibition in both PC-9 and PC-9-GR cells, it is likely that Ncs inhibits cell growth through both EGFR-dependent and EGFR-independent pathways, such as the Src and STAT3 signaling pathways.

We employed the C. elegans tumor model expressing L858R/T790M-EGFR, which resulted in multivulva formation (Fig. 6A). Gefitinib treatment did not attenuate multivulva formation, indicating resistance to gefitinib. Interestingly, Ncs treatment significantly decreased multivulva formation in the C. elegance tumor model (Fig. 6B and 6C). In addition, Ncs exhibited significant anti-tumor effects in xenografted mice with either A549 or H1975 cells (Fig. 6D and 6G). Furthermore, Ncs treatment decreased levels of EGFR, p-EGFR, and p-STAT3 in xenografted tumor tissues (Fig. 6I). These data indicate that Ncs exerts anti-tumor effects on NSCLC cells through both EGFR kinase inhibition and EGFR downregulation, as demonstrated in vitro and in vivo. Taken all together, Ncs could be developed as a potential therapeutic agent for targeting EGFR-driven lung cancer, particularly in case with gefitinib-resistant mutations.

Conclusions

NSCLC is driven by abnormal activation of EGFR, and resistance to EGFR-TKIs, such as gefitinib remains a major therapeutic challenge. In this study, we identify narciclasine (Ncs) as a promising inhibitor of NSCLC, particularly in the context of TKI-resistant EGFR mutations. Ncs effectively targets EGFR variants, induces cell cycle arrest, and suppresses tumor growth in preclinical models, demonstrating greater efficacy than gefitinib in resistant settings. These findings suggest that Ncs holds potential as a novel therapeutic agent, either as monotherapy or in combination with EGFR-TKIs, to overcome drug resistance in NSCLC.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

Ncs

Narciclasine

NSCLC

Non-small Cell Lung Cancer

EGFR

Epidermal Growth Factor Receptor

TKIs

Tyrosine Kinase Inhibitors

PFS

Progression-Free Survival

TNBC

Triple-Negative Breast Cancer

GBM

Glioblastoma Multiforme

STAT3

Signal Transducer and Activator of Transcription 3

CDK

Cyclin-Dependent Kinases

ERK

Extracellular Signal-Regulated Kinases

Author contributions

Sung, JY, performed the experiments, analyzed the data, and wrote the manuscript. Yoo, HE, Kim, S, Yu, S, Lim, GE were performed the experiments, analyzed the data. Lee, BI performed molecular docking simulations. Park, S, performed cloning and wrote the manuscript Yoon, K, Kim, H provided material and methodological support and conceptual advice. Shim, J performed experiments, wrote the manuscript and designed this study. Kim YN: Conceived this project, designed this study, and wrote the manuscript.

Funding

This study was supported by a research grant from the National Cancer Center, Korea (24H1120, 2210722, and 2110261). Animal study was in accordance with the Institutional Animal Care and Use Committee (IACUC) of National Cancer Center Research Institute (NCCRI) and IACUC approval number is NCC-21-606.

Data availability

Not applicable

Declaration

Ethics approval and consent to participate

This study was reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) of National Cancer Center Research Institute (NCCRI). NCCRI is an Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC International) accredited facility and abides by the Institute of Laboratory Animal Resources (ILAR) guide.

Consent for publication

Not applicable.

Competing interests

The authors state no conflicts of interest.