Abstract
Background/Aim: Acquired resistance to epidermal growth factor receptor (EGFR)-tyrosine kinase inhibitors (TKIs) remains a substantial clinical obstacle in treating lung adenocarcinoma (LUAD). Identifying pro-survival pathways that allow tumor cells to evade TKI-induced apoptosis is critical for overcoming this resistance. The transcriptional repressor transducin-like enhancer of split 1 (TLE1) was previously identified as a crucial oncogenic factor that promotes survival in resistant cells. This study investigates the mitochondrial protein Bcl-2 inhibitor of transcription 1 (Bit1) as a key pro-apoptotic signal that overrides the TLE1-mediated survival program.
Materials and Methods: EGFR-TKI sensitive, resistant, and drug-tolerant persister LUAD models were utilized. Bit1 release and TLE1 translocation were assessed via subcellular fractionation. Cell fate following genetic manipulation was determined through viability and apoptosis assays, while RNA-sequencing identified TLE1-regulated transcriptional changes.
Results: In sensitive EGFR-mutant non-small cell lung cancer (NSCLC) cells, TKI treatment triggers the rapid cytosolic release of the mitochondrial outer membrane permeabilization (MOMP)-tethered protein Bit1. This early mobilization precedes cytochrome C release and occurs independently of full MOMP, identifying the Bit1 pathway as a novel, early-response death signal in lung cancer. While Bit1 downregulation attenuated EGFR-TKI-induced apoptosis in EGFR-mutant lung adenocarcinoma cells, ectopic mitochondrial Bit1 expression restored TKI sensitivity in resistant cells. Mechanistically, TKI exposure triggered cytosolic Bit1-AES complex formation, resulting in the nuclear exclusion and sequestration of TLE1. This axis is also pivotal in adaptive resistance; TLE1 was upregulated in drug-tolerant persister (DTP) cells and required for their survival, while Bit1 activation attenuated DTP formation. Transcriptomic analysis revealed that TLE1 coordinates a resistance program enriched for EMT and Notch signaling.
Conclusion: The Bit1/TLE1 axis is a pivotal regulatory switch dictating apoptotic outcomes following EGFR-TKI treatment. This mechanism identifies a therapeutic vulnerability, suggesting that pharmacological Bit1 pathway activation could be an effective strategy to overcome acquired resistance in LUAD.
Introduction
Lung adenocarcinoma (LUAD) continues to be a leading cause of cancer-related mortality globally, necessitating continuous advancements in therapeutic strategies. While the emergence of targeted therapies, particularly epidermal growth factor receptor (EGFR) - tyrosine kinase inhibitors (TKIs), has significantly improved outcomes for patients with EGFR-mutated lung adenocarcinoma, the development of acquired resistance remains a substantial clinical challenge (1, 2). This resistance limits the long-term efficacy of these otherwise potent drugs, underscoring the critical need to decipher its underlying mechanisms.
Among the transcriptional regulators implicated in these adaptive processes, transducin-like enhancer of split 1 (TLE1) has recently emerged as a critical oncogenic co-repressor in non-small cell lung cancer (NSCLC). TLE1 is a member of the Groucho (Gro)/TLE family of transcriptional co-repressors that regulate the transcriptional activity of a wide range of genes by interacting with numerous transcription factors, thereby influencing diverse cellular pathways (3, 4). It has been previously demonstrated that TLE1 expression is markedly elevated in EGFR-TKI–resistant NSCLC cells, where it contributes to EMT-mediated drug tolerance and reduced apoptotic response (5, 6). These findings underscore TLE1 as a promoter of a survival program that actively counteracts the apoptotic signals initiated by EGFR-TKI treatment.
Meanwhile, Bcl-2 inhibitor of transcription 1 (Bit1) protein has been previously identified as a potent antitumor effector in NSCLC through its unique, caspase-independent cell death mechanism. Under normal conditions, Bit1 localizes to mitochondria, but upon apoptotic stimuli, it is released into the cytoplasm where it complexes with the amino-terminal enhancer of split (AES) to antagonize the pro-survival TLE1 transcriptional program (7-10). Tethered to the mitochondrial outer membrane facing the cytosol, Bit1 cytosolic release mechanism may only require a release or cleavage event and is less reliant on complete mitochondrial outer membrane permeabilization (MOMP) (11). This ability of Bit1 to induce cell death independently of caspase activation positions it as a promising target for overcoming apoptosis resistance which is a hallmark of advanced and drug-refractory tumors.
While TLE1 has been critically implicated in EGFR-TKI resistance and Bit1 is known to antagonize TLE1’s pro-survival effects, the mechanistic link between mitochondrial Bit1 release and the regulation of TLE1’s anti-apoptotic program in the context of EGFR-TKI-induced apoptosis in LUAD remains unexplored. In this study, we propose that the Bit1/TLE1 axis represents a previously unrecognized regulatory mechanism governing EGFR-TKI-induced apoptosis and tolerance. Utilizing a multi-disciplinary approach that incorporates biochemical fractionation, transcriptomic profiling of drug tolerant persisters, and translational peptide therapy, we map this signaling circuit and evaluate its therapeutic potential in lung adenocarcinoma (Figure 1). Specifically, we show here that Bit1 acts as an MOMP-caspase-independent effector of EGFR-TKI–mediated apoptosis by suppressing the anti-apoptotic transcriptional activity of TLE1, and that Bit1 activation may enhance therapeutic efficacy and overcome resistance to EGFR-TKIs in lung adenocarcinoma.
Experimental workflow and study design. Figure was generated using Biorender.
Materials and Methods
Cell culture. Human EGFR-mutant lung adenocarcinoma cell lines PC9 (cat. no.: CB_90071810, Sigma-Aldrich, St. Louis, MO, USA), HCC827 (cat. no.: CR-2868, ATCC, Manassas, VA, USA), and H1975 (cat. no.: CRL-5908) cell lines were cultured in RPMI-1640 with L-glutamine medium (cat. no.: 10-040-CM, Corning Inc, Corning, NY, USA), supplemented with 10% fetal bovine serum (FBS) (cat. no.: 900-108, Gemini Bioproducts, West Sacramento, CA, USA) and 1% penicillin-streptomycin (cat. no.: 15140-122, Gibco, Carlsbad, CA, USA).
Gefitinib-resistant lung cancer cell line (PC9GR, HCC827GR, and H1975OR) were established from their parental sensitive line (PC9, HCC827, and H1975) through a total of 6-month exposure to gefitinib (cat. no.: HY-50895, Medchem Express, Monmouth Junction, NJ, ISA) or osimertinib (cat. no.: HY-15772, Medchem Express). All cells were maintained in a humidified incubator at 37°C with 5% CO2.
Drug-tolerant persister (DTP) cells were established by treating the parental cells with high-dose EGFR-TKI for 9 days, with the remaining viable cells considered as the DTP. PC9 DTP cells were obtained by treating PC9 cells with 2 μM gefitinib for 9 days, HCC827 DTP by treating HCC827 with 0.5 μM gefitinib for 9 days, and H1975 DTP by treating H1975 with 3 μM osimertinib for 9 days.
Cell viability and apoptosis assays. Response to EGFR-TKIs were determined by treating cells with various concentrations of gefitinib (cat. no.: HY-50895, Medchem Express) or osimertinib (cat. no.: HY-15772, Medchem Express) for 48 h followed by assessment of cell viability or cell death.
Cell viability was determined via the metabolic activity based assay. Cells were treated with different concentrations Gefitinib or Osimertinib, then incubated for 48 h. The number of metabolically active cells was measured using PrestoBlue Cell Viability Reagent (cat. no.: A13262, Invitrogen, Carlsbad, CA, USA) and fluorescence reading at 485 nm excitation wavelength and 520 nm emission wavelength with a microplate plate reader. Cell death was assessed by quantifying the amount of cytoplasmic histone-associated DNA fragments using the Cell Death Detection ELISA PLUS (cat. no. 11774425001; Roche Molecular Diagnostics, Pleastanton, CA, USA) according to the manufacturer’s instructions.
Cell quantification and visualization. Cell numbers were determined using manual hemacytometer counting with trypan blue exclusion to ensure only viable cells were included in the quantification. Each experiment was performed in triplicate, and the results were expressed as the total number of cells per well. In situ visualization via crystal violet staining was done in parallel. The adherent cells were washed twice with PBS and fixed with 10% formaldehyde at room temperature for 20 min. Following fixation and two additional PBS washes, cells were stained with 0.1% crystal violet. The plates were subsequently rinsed with PBS to remove excess dye, air-dried, and imaged.
Lentiviral transduction. To express exogenous TLE1, parental cells were transduced with lentiviral vectors encoding empty control GFP (cat. no.: OHS5833, Horizon Discovery, Lafayette, CO, USA) or GFP-TLE1 (cat. no.: PLOHS_100005903, Horizon Discovery). Lentiviral transduction of parental cell lines was performed in 24-well plates using a multiplicity of infection of 0.5. After 48 h, the cells were cultured in medium containing 5 μg/ml Blasticidin S (cat. no.: ant-bl-l, Invitrogen) to select for selection. Control GFP clones and GFP-TLE1 clones were pooled separately to create the control GFP and GFP-TLE1 pools, respectively.
Short hairpin RNA (shRNA) transfection. To create stable knockdown cells lines, parental cell lines were transfected with 0.5 μg of control shRNA (cat. no.: sc-108080, Santa Cruz Biotechnology, Dallas, TX, USA) or TLE1-specific shRNA (cat. no.: sc-38558-V, Santa Cruz Biotechnology). After 48 h, transfected cells were treated with 1 μg/ml puromycin (cat. no.: A11138-02, Gibco) to select for stable clones.
Small interfering RNA (siRNA) transfection. For transient knockdown studies, control non-targeting siRNA (cat. no.: sc-37007, Santa Cruz Biotechnology) or pool of siRNAs specifically targeting TLE1 (cat. no.: sc-38558, Santa Cruz Biotechnology) or Bit1-targeting siRNA (cat. no.: AM16708, Invitrogen), or HES1-targeting siRNAs (cat. no.: sc-37938, Santa Cruz Biotechnology) were transfected into the cells using Lipofectamine RNAiMAX transfection reagent, following manufacturer’s instructions (cat. no.: 13778100, Invitrogen) and incubated for at least 24 h followed by subsequent experimentation.
Transfection assays. Transient transfection assays were conducted using Lipofectamine 3000 (cat. no.: L3000015, Invitrogen) as prescribed by the manufacturer. The Bit1-myc tagged plasmid construct (Bit-mito) used was made as described previously (12). The iRGD-CDD sequence was gifted by a collaborating laboratory. Transfected cells were incubated for at least 24 h before being used for subsequent experiments.
Protein extraction. Cells were lysed using Mammalian Protein Extraction Buffer (cat. no.: 78501, Thermo Fisher Scientific, Waltham, MA, USA) with 10% protease inhibitor (cat. no.: P8340-5ML, Sigma, Houston, TX, USA) for 30 min followed by centrifugation to remove cellular debris. Protein concentration was determined using DC Protein Assay Kit (cat. no.: 5000113, Bio-Rad, Hercules, CA, USA) with bovine serum albumin (cat. no. 23208, Thermo Fisher) as the standard.
Cytoplasmic and mitochondrial fractionation. Preparation of the mitochondrial and cytosolic containing fractions was performed using the Cell Fractionation Kit (cat. no.: ab109719, Abcam, Cambridge, MA, USA). The protein concentration in different fractions was measured using the Bio-Rad protein assay kit as the standard.
Western blot. For western blot analysis, 35 μg of protein lysate was resolved on 4-20% gradient Tris–glycine gels (cat. no.: XP04200BOX, Thermo Fisher Scientific) and electrophoretically transferred to nitrocellulose membrane. The membranes were incubated with primary antibodies overnight at 4°C: anti-Bit1 (1:200, sc-518195, Santa Cruz Biotechnology), anti-TLE1 (1:200, sc-137098, Santa Cruz Biotechnology), anti-HES1 (1:200, sc-166510, Santa Cruz Biotechnology), anti-GAPDH (1:2,000, sc-47724, Santa Cruz Biotechnology) and anti-β-actin (1:1,000, sc-81178, Santa Cruz Biotechnology). Afterwards, the membranes were incubated for 1 h with the appropriate secondary antibody: Amersham ECL Mouse IgG, HRP-linked whole Ab (1:25,000; cat. no.: NA931V; Cytiva, Marlborough, MA USA). Visualization of protein bands on the membranes was achieved using the ECL detection system (cat. no.: RPN2232, Cytiva).
Total RNA extraction and reverse transcription - quantitative PCR (RT-qPCR). Total RNA was extracted from cells using the RNeasy Mini Kit (cat. no.: 74104, Qiagen, Germantown, MD, USA) according to the manufacturer’s instructions. RNA quality and quantity were assessed spectrophotometrically. For RT-qPCR, the expression of TLE1 was quantified via the iTaq Universal SYBR Green One-Step Kit (cat. no.: 1725150, Bio-Rad) using human TLE1 forward (5′-CCTCCTACACAGCAGCAGTT-3′) and reverse (5′-TCTGCATCGTGGTGCTTCTT-3′) primers. Internal normalization was done using GAPDH human GAPDH forward (5′-CCC ACT CCT CCA CCT TTG AC-3′) and reverse (5′-TTG CTG TAG CCA AAT TCG TTG T-3′) primers. The thermocycling conditions included a reverse transcription step at 50°C for 10 min, an initial denaturation step at 95°C for 1 min, followed by 40 cycles of denaturation at 95°C for 10 s and annealing/extension at 60°C for 30 s. Relative gene expression levels were calculated using the ΔΔCt method.
Co-immunoprecipitation assay. Cells were lysed using RIPA lysis buffer (cat. no.: sc-24948, Santa Cruz Biotechnology), followed by a 1.5-h incubation at 4°C with vortexing every 10 min. Cell debris was removed by centrifugation. The resulting lysate was immunoprecipitated with 15 μl of anti-Bit1 (cat. no.: sc-518195, Santa Cruz Biotechnology), anti-TLE1 (cat. no.: sc-137098, Santa Cruz Biotechnology), anti-HES1 (cat. no.: sc-166410, Santa Cruz Biotechnology), or non-specific IgG (cat. no.: sc-137097, Santa Cruz Biotechnology), along with 20 μl of Protein A/G Plus-Agarose (cat. no.: sc-2003, Santa Cruz Biotechnology) at 4°C overnight and thoroughly washed with lysis buffer. Bound proteins were resolved by SDS-PAGE, and immunoblotting was performed using anti-AES (cat. no.: PA5-95534, Fisher Scientific), anti-myc-tag (cat. no.: 2272, Cell Signaling, Danvers, MA, USA), anti-TLE1 (1:1,000, cat. no.: PA5-78200, Invitrogen) or anti-HES1 (cat. no.: PA5-28802, Invitrogen) primary antibodies and Amersham ECL Rabbit IgG, HRP-linked whole Ab (1:50,000; cat. no.: NA934V; Cytiva). Visualization of protein bands on the membranes was achieved using the ECL detection system (cat. no.: RPN2232, Cytiva).
RNA sequencing and transcriptomic analysis. Total RNA was extracted from control and TLE1-overexpressing PC9 cells. Afterwards, RNA sequencing and analysis was performed by the Translational Genomics Core at the Louisiana State University Health Science Center, New Orleans, Louisiana, USA. RNA concentration was determined using the Qubit RNA High Sensitivity Assay kit (cat. no.: 11732088, Invitrogen), and structural integrity was verified using the Agilent 2100 Bioanalyzer (cat. no.: Agilent 200, Agilent Technologies, Santa Clara, CA, USA).
Sequencing libraries were prepared using the Illumina Stranded Total RNA Prep with Ribo-Zero Plus kit (cat. no.: 400800, Agilent Technologies) according to the manufacturer’s protocols. Briefly, 100 ng of total RNA underwent ribosomal RNA depletion followed by purification and fragmentation. Subsequently, first- and second-strand cDNA were synthesized. The resulting cDNA was adenylated at the 3′ ends, followed by the ligation of sequencing anchors. The library was then enriched via 13 cycles of PCR amplification to incorporate unique dual indexes. Final library quantification was performed with the Qubit dsDNA High Sensitivity Assay Kit, and library size distribution was confirmed via the Agilent 2100 Bioanalyzer. Multiplexed libraries were sequenced on an Illumina NextSeq 2000 platform using a P1 XLEAP 300 cycles kit, generating 76 bp paired end reads.
Raw sequencing data in FASTQ format were processed using Partek Flow. Sequence contaminants, including rDNA, tRNA, and mtDNA, were filtered and removed using Bowtie2 (v2.2.5). The remaining high-quality reads were aligned to the human reference genome (hg38) using the STAR aligner (v2.7.3a). Gene quantification was performed against the Ensembl v99 database. To reduce noise, low-abundance features with a maximum read count ≤10 were excluded from the analysis. Read counts were normalized using the Trimmed Mean of M-values (TMM) method. Differential gene expression (DGE) analysis was conducted using DESeq2 to compare TLE1-overexpressing cells against control counterparts. Genes exhibiting a False Discovery Rate (FDR) less than 0.05 and an absolute fold change greater than or equal to 2.0 were defined as significantly differentially expressed genes (DEGs). Functional enrichment analysis was performed using the Enrichr platform (https://maayanlab.cloud/Enrichr/) (13). The DEG list was queried against both the MSigDB Hallmark 2020 (14) and WikiPathways 2024 Human (15) databases to ensure robust pathway identification. Significance was determined based on the p-value and combined score to identify top-ranked biological processes. Visualization of the transcriptomic distribution was performed via volcano plots, and selected pathway enrichments were illustrated using bar charts.
Statistical analysis. Data are presented as the mean±standard deviation (SD) of at least three independent experiments. All statistical analyses were performed using GraphPad Prism software (GraphPad Software, Inc., Boston, MA, USA). Statistical significance was determined using two-tailed Student’s t-tests for comparisons between two groups and one-way ANOVA with Tukey’s post hoc test for comparisons among multiple groups. A statistically significant difference was defined as p<0.05.
Results
Mitochondrial Bit1 is released to the cytosol by EGFR-TKIs and its downregulation attenuates EGFR-TKI-induced apoptosis in EGFR-TKI sensitive LUAD cells. The mitochondrial Bit1 protein is anchored to the mitochondrial outer membrane via its N-terminus, with its pro-apoptotic domain constitutively exposed to the cytosol. We therefore hypothesize that, in contrast to soluble intermembrane space (IMS) proteins such as cytochrome c, the activation and cytosolic release of OMM-tethered Bit1 may occur independently of, or upstream to, full BAX/BAK-mediated MOMP following therapeutic stress. To test this, we first examined the subcellular localization of Bit1 in TKI-sensitive NSCLC cell lines treated with EGFR-TKIs gefitinib and osimertinib. Cytosolic fractionation and western blot analysis revealed that EGFR-TKI treatment induced a marked increase in cytosolic Bit1 at early time points, which preceded any detectable release of cytochrome c (Figure 2A). The distinct kinetics of Bit1 accumulation suggest its release is mechanistically independent of the complete MOMP required for cytochrome c release.
EGFR-TKIs induce mitochondrial release of Bit1 to promote apoptosis in sensitive LUAD cells. (A) Western blot analysis of subcellular cytosolic fraction in PC9, HCC827, and H1975 cells treated with 0.1 μM gefitinib or 0.1 μM osimertinib over a 36-h time course. GAPDH was used as loading control. (B) Validation of Bit1 knockdown by western blot in PC9, HCC827, and H1975 cells transfected with control siRNA or Bit1-specific siRNA. (C) Cell viability assessed by PrestoBlue assay 72 h post-treatment in sensitive cells following Bit1 knockdown (*p<0.05). (D) Apoptosis quantified by Cell Death ELISA in the indicated cell lines following Bit1 knockdown and TKI treatment (*p<0.05). EGFR-TKI: Epidermal growth factor – tyrosine kinase inhibitor; Bit1: Bcl-2 inhibitor of transcription 1; LUAD: lung adenocarcinoma.
We next sought to determine whether Bit1 is required functionally for EGFR-TKI-induced cell death. Knockdown of Bit1 via a pool of Bit1-specific siRNAs in PC9, HCC827, and H1975 cells (Figure 2B) significantly attenuated apoptosis induced by gefitinib in the sensitive PC9 and HCC827 lines, as well as osimertinib in H1975 cells, as measured by PrestoBlue Cell Viability and DNA-histone-complex-based cell death ELISA assays (Figure 2C, D). Collectively, these results demonstrate that cytosolic Bit1 release is a specific response to effective EGFR-TKI treatment and that Bit1 is a critical effector of the ensuing apoptotic cell death.
Ectopic Bit1 expression restores EGFR-TKI-induced apoptosis in EGFR-TKI resistant LUAD cells. To investigate whether reactivation of the Bit1 pathway could overcome acquired resistance, we utilized TKI-resistant cell models, gefitinib-resistant PC9 (PC9GR) and osimertinib-resistant H1975 (H1975OR). We transfected these cells with either an empty control vector or a mitochondrial-targeted Bit1 construct (Bit1-mito) to achieve exogenous expression of Bit1. Western blotting confirmed successful expression of the Bit1-mito construct (Figure 3A). Critically, the introduction of exogenous Bit1 was sufficient to sensitize resistant cells to EGFR-TKI treatment. Exposure of Bit1-mito-expressing resistant cells to gefitinib or osimertinib resulted in a significant increase in TKI sensitivity, characterized by decreased cell viability with concomitant increase in apoptosis (Figure 3B and C). Furthermore, cytoplasmic/mitochondrial fractionation of these Bit1-mito transfected cells following treatment demonstrated that the exogenous Bit1 was also successfully released to the cytosol upon EGFR-TKI exposure, mirroring the EGFR-TKI effect observed in the sensitive lines (Figure 3D).
Ectopic expression of mitochondrial Bit1 sensitizes resistant LUAD cells to EGFR-TKIs. (A) Western blot of mitochondrial fractions from PC9GR and H1975OR cells stably transfected with a control vector or Bit1-Mito. (B) Cell viability in resistant lines (PC9GR, H1975OR) expressing Bit1-mito following treatment with gefitinib or osimertinib (*p<0.05). (C) Quantification of apoptosis by Cell Death ELISA in control vs. Bit1-mito expressing resistant cells post-EGFR-TKI treatment (*p<0.05). (D) Subcellular fractionation and western blot analysis showing the release of exogenous Bit1-mito from the mitochondria to the cytosol in resistant cells following EGFR-TKI treatment. GAPDH and COX IV are used as fraction-specific loading controls. EGFR-TKI: Epidermal growth factor – tyrosine kinase inhibitor; Bit1: Bcl-2 inhibitor of transcription 1; LUAD: lung adenocarcinoma.
EGFR-TKI triggers cytosolic Bit1-AES complex formation and cytoplasmic translocation of nuclear TLE1. It has been previously established that cytosolic Bit1 executes its pro-apoptotic function by forming a complex with the Amino-terminal Enhancer of Split (AES) to inhibit the nuclear, anti-apoptotic activity of TLE1 (8, 10). To confirm this pathway is engaged by EGFR-TKIs, co-immunoprecipitation assays and subcellular localization studies were performed. In gefitinib-sensitive PC9 cells, EGFR-TKI treatment robustly increased the physical association of Bit1 with AES in the cytoplasmic fraction (Figure 4A). Concurrently, we observed a striking translocation of TLE1 from the nucleus to the cytoplasm following treatment (Figure 4B). This nuclear-to-cytoplasmic redistribution of TLE1 is consistent with its functional inhibition, as it is physically removed from its nuclear transcriptional targets and sequestered in the cytosol. Significantly, the same molecular phenomena were observed in the osimertinib-sensitive H1975 cell line, underscoring the correlation between an intact Bit1/AES/TLE1 axis and drug sensitivity (Figure 4A and B). These findings suggest that the TKI-induced release of Bit1 serves as a molecular switch that initiates the Bit1/AES complex formation and subsequent TLE1 exclusion from the nucleus.
EGFR-TKIs promote the assembly of the pro-apoptotic Bit1/AES complex and nuclear exclusion of TLE1. (A) Co-immunoprecipitation analysis of the Bit1-AES interaction. PC9 and H1975 cells were treated with gefitinib or osimertinib, and cytoplasmic extracts were immunoprecipitated with an anti-Bit1 antibody. The presence of AES in the complex was detected by western blot. (B) Western blot analysis of cytoplasmic fraction showing the translocation of TLE1 to the cytoplasm upon TKI treatment. GAPDH was used as cytoplasmic loading control. EGFR-TKI: Epidermal growth factor – tyrosine kinase inhibitor; Bit1: Bcl-2 inhibitor of transcription 1; AES: amino-terminal enhancer of split; TLE1: transducin-like enhancer of split 1.
The Bit1/TLE1 pathway regulates the formation and survival of DTP cells. To investigate the clinical relevance of TLE1 in acquired resistance, we examined its expression in DTPs, a subpopulation of cells that survive initial drug exposure through non-mutational mechanisms. Consistent with a role for TLE1 in driving resistance, western blot and RT-qPCR analyses revealed a significant and consistent upregulation of TLE1 expression in DTPs derived from PC9, HCC827, and H1975 cells compared to their naive parental counterparts (Figure 5A and B). This suggests that transcriptional induction of TLE1 is a conserved feature of the adaptive response to EGFR-TKI therapy.
TLE1 is upregulated in DTPs and is required for their continued survival. (A) Western blot analysis of TLE1 expression in parental sensitive cells and their corresponding DTPs. (B) RT-qPCR analysis of TLE1 mRNA levels in sensitive parental cells versus derived DTP populations (*p<0.05). (C) Western blot confirming siRNA-mediated knockdown of TLE1 in pre-formed DTPs. (D) Survival assay of DTPs following TLE1 knockdown and subsequent TKI re-challenge (*p<0.05). (E) Apoptosis assay via Cell Death ELISA in DTP populations following TLE1 depletion and TKI challenge (*p<0.05). TLE1: Transducin-like enhancer of split 1; DTP: drug tolerant persister; TKI: tyrosine kinase inhibitor.
To determine if this elevated TLE1 expression is essential for the continued viability of the persister population, we performed a loss-of-function studies on established DTPs. Following the generation of DTPs, which inherently display high TLE1 levels, we transiently knocked down TLE1 via a pool of TLE1-specific siRNAs (Figure 5C). Subsequent re-treatment with EGFR-TKIs demonstrated that these TLE1-depleted DTPs exhibited significantly reduced survival, appearing more susceptible to treatment than control DTPs (Figure 5D and E). This partial restoration of sensitivity upon TLE1 depletion confirms its crucial role in maintaining the survival signaling of the persister state.
We next investigated whether the Bit1/TLE1 axis plays roles in the initial transition of sensitive cells into the DTP state, referred to as DTP formation. To test this, we modulated TLE1 levels in parental sensitive cell lines prior to drug exposure. We observed that silencing of endogenous TLE1 in PC9, HCC827, and H1975 cells (Figure 6A) significantly impeded the emergences of DTPs upon EGFR-TKI treatment, as reflected by the number of DTPs formed (Figure 6B). Conversely, overexpression of exogenous TLE1 in parental cells (Figure 6C) led to a notable increase in the number of DTPs formed against the EGFR-TKIs (Figure 6D). These results indicate that TLE1 expression levels influences a cell’s competence to enter the drug-tolerant state.
The Bit1/TLE1 axis dictates the competence of sensitive cells to transition into the DTP state. (A) Western blot confirming TLE1 knockdown in parental cells. (B) Quantification of DTP colonies formed after 9 days of TKI treatment following TLE1 silencing (*p<0.05). (C) Western blot of ectopic TLE1 expression in sensitive parental cell lines. (D) Quantification of resulting DTP colonies in TLE1-overexpressing versus control parental cells (*p<0.05). (E) Western blot confirmation of Bit1-mito expression in parental cells. (F) Quantification of DTP colony counts in parental cells expressing Bit1-mito compared to control (*p<0.05). Bit1: Bcl-2 inhibitor of transcription 1; TLE1: transducin-like enhancer of split 1; DTP: drug tolerant persister; TKI: tyrosine kinase inhibitor.
Finally, we explored whether activation of the Bit1 pathway could prevent this TLE1-driven formation of persisters. Activation of the Bit1 pathway via Bit1-mito expression (Figure 6E) significantly attenuated the initial formation of DTPs (Figure 6F). This establishes the Bit1/TLE1 axis as a critical regulator of the persister cell state, where the Bit1-mediated sequestration of TLE1 can effectively block the transition of sensitive cells into a drug-tolerant phenotype.
TLE1 interacts with HES1 to drive EGFR-TKI resistance. Given TLE1’s function as a transcriptional corepressor known to interact with various transcription factors, we investigated novel binding partners that might mediate its role in EGFR-TKI resistance. In light of numerous evidence indicating that the transcription factor HES1 acts as a determinant of EGFR-TKI resistance in NSCLC, we hypothesized that it might collaborate with TLE1 in LUAD cells. Our Co-IP experiments revealed a physical interaction between endogenous TLE1 and the basic helix-loop-helix transcription factor HES1 in lung adenocarcinoma cells (Figure 7A). This interaction suggests that HES1 acts as a crucial DNA-binding effector of TLE1’s pro-resistance activities.
TLE1 interacts with HES1 to promote resistance against EGFR-TKIs. (A) Co-immunoprecipitation analysis demonstrating the interaction between TLE1 and HES1 in PC9, HCC827, and H1975 cells. (B) Western blot confirmation of HES1 knockdown in TLE1-overexpressing cell lines. (C) Cell viability assay showing the effect of HES1 knockdown on TLE1-mediated resistance (*p<0.05). (D) Apoptosis quantification via Cell Death ELISA illustrating the restoration of EGFR-TKI-induced death in TLE1-overexpressing cells following HES1 silencing (*p<0.05). TLE1: Transducin-like enhancer of split 1; EGFR-TKI: epidermal growth factor – tyrosine kinase inhibitor.
To determine if this interaction has functional implications for EGFR-TKI resistance, we assessed the effect of HES1 knockdown in TLE1-overexpressing cells. Western blot analysis confirmed successful reduction of HES1 protein levels (Figure 7B). Importantly, subsequent EGFR-TKI sensitivity assays demonstrated that silencing HES1 expression significantly attenuated the resistance driven by TLE1. Specifically, cell viability assays (Figure 7C) showed a marked decrease in survival, and cell death ELISA (Figure 7D) indicated an increase in apoptosis in TLE1-overexpressing cells upon HES1 knockdown when treated with EGFR-TKIs. These results indicate that the interaction between TLE1 and HES1 is critical for TLE1-mediated EGFR-TKI resistance. These results indicate that the physical interaction between TLE1 and HES1 is critical for TLE1-mediated pro-survival and anti-apoptotic activities and suggest that the TLE1-HES1 axis serves as a driver of EGFR-TKI resistance.
Transcriptomic analysis identifies downstream genes regulated by TLE1. To globally map the transcriptional changes driven by TLE1 and identify the gene networks underpinning EGFR-TKI resistance, we performed an RNA-sequencing analysis comparing TLE1-overexpressing PC9 cells to control counterparts. Here we identified a total of 2,849 differentially expressed genes. Among these, 1,366 genes were significantly upregulated, while 1,483 genes were significantly downregulated (Figure 8A). Confirming the validity of our experimental model, TLE1 itself was significantly overexpressed in the treatment group (FC=3.14).
Transcriptomic profiling of TLE1-overexpressing cells reveals activation of EMT and Notch signaling programs. (A) Volcano plot showing differentially expressed genes in TLE1-overexpressing versus control PC9 cells. Genes that are less expressed in TLE1-overexpressing cells compared to control cells are colored red while gene more expressed in TLE1-overexpressing cells are in blue. (B) Pathway enrichment analysis using the MSigDB Hallmark 2020 library. (C) Pathway enrichment analysis using the WikiPathways 2024 library. (D) mRNA fold changes of the epithelial marker CDH1 and mesenchymal markers CDH2 and VIM. (E) mRNA fold changes of key Notch signaling components including ligands JAG1 and DLL3, and effectors DTX3 and HES1. TLE1: Transducin-like enhancer of split 1; EMT: epithelial-mesenchymal transition.
To identify the biological significance of these changes, the DEG list was subjected to pathway enrichment analysis using the Enrichr platform. Analysis against the MSigDB Hallmark 2020 and WikiPathways 2024 Human databases independently identified the “Epithelial Mesenchymal Transition” as the most significantly enriched biological program (Figure 8B and C). These findings further solidify our previous reports indicating that TLE1 promotes a mesenchymal state (16). Consistent with our earlier work showing that TLE1 suppresses CDH1 which encodes for E-cadherin, the transcriptomic data revealed a definitive “cadherin switch.” This included a profound loss of the epithelial marker E-cadherin (FC=−2.83) and significant induction of mesenchymal markers N-cadherin, encoded by CDH2 (FC=5.99), and Vimentin (FC=3.16) (Figure 8D).
Beyond the established ZEB1-mediated EMT axis, we sought to characterize the functional output of the newly identified TLE1/HES1 physical interaction. While the precise direct targets of this complex are still being elucidated, the transcriptomic profile confirms a focused activation of Notch signaling components. Notably, the “Notch Signaling” hallmark was also enriched in the MsigDB database (Figure 8B). This was specifically supported by a robust induction of the Notch ligand JAG1 (FC=8.07) and the effectors DTX3 (FC=41.97) and DLL3 (FC=33.05) (Figure 8E).
Taken together, these transcriptomic findings provide a functional bridge between our biochemical observations and the observed drug-tolerant phenotype. By confirming that TLE1 overexpression not only reinforces the ZEB1-driven EMT program but also triggers a specialized Notch signaling circuit through the TLE1/HES1 axis, we establish TLE1 as a master regulator of phenotypic plasticity in LUAD.
Pharmacological induction of the Bit1 pathway via iRGD-CDD bypasses TLE1-mediated resistance. Targeting the MOMP-independent and TLE1-mediated epigenetic inhibitory functions of the Bit1 apoptosis pathway offers a viable strategy to circumvent EGFR-TKI tolerance and resistance in lung cancer. Here, we employed a tumor-penetrating iRGD-fused Bit1 cell death domain (iRGD-CDD) peptide as a pharmacological means of Bit1 activation. This construct is designed to penetrate tumor cells and mimic the endogenous release of Bit1 to the cytosol, thereby triggering the AES-mediated sequestration of TLE1.
Using our previously established TLE1-overexpressing PC9 model and its corresponding control, we evaluated the impact of pharmacological Bit1 pathway activation on gefitinib sensitivity. As shown, co-treatment with iRGD-CDD significantly increased the susceptibility to gefitinib in both cell lines (Figure 9A). Notably, while the peptide successfully diminished TLE1-mediated resistance in the TLE1-expressing cells, the sensitizing effect was more pronounced in the control cells. To further confirm that the sensitizing effect of iRGD-CDD is specifically dependent on TLE1, we evaluated the peptide’s efficacy following TLE1 knockdown. While iRGD-CDD promoted gefitinib-induced apoptosis in control shRNA cells, this effect was significantly attenuated in TLE1-depleted cells (Figure 9B). The reduced potency of the Bit1-mimetic in the absence of its nuclear target suggests that the therapeutic activity of iRGD-CDD is, at least in part, mediated through the neutralization of TLE1.
Pharmacological activation of the Bit1 pathway via iRGD-CDD sensitizes LUAD cells to EGFR-TKIs. (A) Cell viability of control and TLE1-overexpressing PC9 cells treated with gefitinib in the presence or absence of the iRGD-CDD peptide (*p<0.05). (B) Cell viability of control and TLE1-shRNA knockdown cells treated with gefitinib in the presence or absence of the iRGD-CDD peptide (*p<0.05, ns: not significant). Bit1: Bcl-2 inhibitor of transcription 1; LUAD: lung adenocarcinoma; EGFR-TKI: epidermal growth factor – tyrosine kinase inhibitor; TLE1: transducin-like enhancer of split 1.
These results provide a functional proof-of-concept for the Bit1/AES/TLE1 axis as a viable therapeutic target. By delivering the Bit1 apoptosis domain to bypass the requirement for endogenous mitochondrial release, we demonstrated the utility of the Bit1 cell death pathway to potentiate EGFR-TKI sensitivity, even in settings of significant TLE1-mediated resistance mechanism or upregulation.
Discussion
Acquired resistance to EGFR-TKIs remains a clinical challenge in treating EGFR-mutated lung adenocarcinoma, limiting the long term effectiveness of these potent targeted agents (17-19). The present study identifies the Bit1/TLE1 axis as a critical, MOMP-independent regulator of EGFR-TKI-induced apoptosis, thereby identifying a novel therapeutic vulnerability. We found that effective EGFR-TKI treatment triggers the release of mitochondrial Bit1 into the cytosol in drug-sensitive LUAD cells, initiating a pro-apoptotic cascade. This effect is mediated by the cytosolic assembly of the Bit1/AES complex, which promotes the nuclear-to-cytoplasmic translocation and sequestration of the oncogenic co-repressor TLE1, effectively neutralizing its pro-survival transcriptional program (Figure 10).
Working model of the Bit1/TLE1/HES1 axis in EGFR-TKI sensitivity and resistance. In sensitive cells, EGFR-TKIs trigger the mitochondrial release of Bit1, which forms a complex with AES to sequester TLE1 in the cytoplasm, leading to apoptosis. In resistant cells and DTPs, TLE1 is significantly upregulated and localized to the nucleus. There, it interacts with transcription factors such as ZEB1 and HES1 to activate pro-survival programs such as EMT and Notch signaling. Pharmacological delivery of the Bit1 cell death domain restores the inhibitory Bit1/AES/TLE1 complex and re-sensitize cells to EGFR-TKIs. Figure was generated using Biorender. Bit1: Bcl-2 inhibitor of transcription 1; TLE1: transducin-like enhancer of split 1; EGFR-TKI: epidermal growth factor – tyrosine kinase inhibitor; AES: amino-terminal enhancer of split; DTP: drug tolerant persister; EMT: epithelial-mesenchymal transition.
A primary finding of this work is the MOMP-independent nature of the Bit1 pathway, which provides a unique rationale for circumventing traditional resistance mechanisms. As an OMM-tethered protein facing the cytosol (11), Bit1 is positioned to be released through mechanisms that do not require full, irreversible mitochondrial poration unlike cytochrome c which requires irreversible MOMP. Consequently, the Bit1/AES axis acts as an “alternate exit ramp” for cell death. By bypassing the need for MOMP and remaining caspase-independent, this pathway can theoretically nullify the protective effects of Bcl-2 family proteins or IAP overexpression, which typically block canonical apoptosis (20-23).
While this release occurs without complete OMM disruption, it may be regulated by post-translational modifications or proteolytic cleavage which are mechanisms currently under investigation in our laboratory. For example, Protein Kinase D (PKD), a critical upstream activator of Bit1 mitochondrial release (24), is frequently co-mutated with TP53 in EGFR-mutant NSCLC. Clinical cohorts indicate that this co-mutation is associated with significantly shorter disease-free survival (25). It remains an exciting possibility that the TP53/PKD mutational landscape represents a distinct clinical subgroup where the Bit1-mediated apoptotic axis is functionally disrupted, contributing to the aggressive phenotype of these tumors.
This axis appears particularly pivotal in the emergence of adaptive drug resistance. We found that TLE1 is significantly upregulated in DTP cells and is required for their survival, while Bit1 activation significantly attenuates DTP formation. Interestingly, clinical cohorts indicate that co-mutations in TP53 and protein kinase D, a known activator of Bit1 release (24), are associated with significantly shorter disease-free survival in NSCLC (25). It is an exciting possibility that these mutational landscapes disrupt the Bit1 apoptotic axis, contributing to the aggressive phenotype of persister populations.
The functional significance of TLE1 as a diagnostic and prognostic marker is well established in several mesenchymal and epithelial malignancies, most notably in synovial sarcoma, where its overexpression is a hallmark of the disease (26). Our findings extend the oncogenic relevance of TLE1 to the context of EGFR-mutant lung adenocarcinoma, where it specifically functions as a master regulator of adaptive resistance to TKI therapy. Our results expand the known regulatory landscape of TLE1 beyond the previously established ZEB1 interaction (6). Current evidence suggests that TLE1 also interacts with HES1, thus potentially expanding TLE1’s regulatory landscape. This suggests that TLE1 does not operate in a single pathway but plays roles in multiple transcription factor networks. However, the specific anti-apoptotic genes regulated by the TLE1/HES1 complex remains to be elucidated and validated to fully map the pro-survival program.
These findings, however, require in vivo validation using either animal models or patient derived xenografts (PDX) to account for tumor microenvironmental influences and systemic factors. Therefore, future research must prioritize confirming that pharmacological activation of Bit1 enhances TKI efficacy and suppresses DTP mediated recurrence in a relevant physiological setting. Furthermore, for rational drug design purposes, the upstream signal responsible for triggering the MOMP independent release of Bit1 from the mitochondria might need to be identified.
Despite these limitations, our functional data using the iRGD-CDD peptide provide a translational proof-of-concept for this axis. By delivering the Bit1 functional domain to bypass the requirement for endogenous mitochondrial release, we successfully diminished TLE1-mediated resistance in PC9 models. Notably, the observation that iRGD-CDD more effectively sensitized control cells compared to TLE1-overexpressing cells underscores the stoichiometric nature of the Bit1-AES-TLE1 interaction. This suggests that while exogenous Bit1 can neutralize oncogenic TLE1, high levels of TLE1 expression may act as a biochemical buffer, requiring optimized dosing strategies to fully overcome the resistant phenotype. This demonstrates that the Bit1–TLE1 axis is a dual point of intervention: activating Bit1 can restore apoptosis independently of mitochondrial permeabilization, while sequestering TLE1 reinstates the expression programs required for drug sensitivity. These strategies bypass the limitations of targeting individual transcription factors or relying solely on reactivating canonical apoptosis.
Conclusion
In summary, our findings identified the Bit1/TLE1 axis as an additional mediator affecting cellular response following EGFR-TKI treatment. By providing evidence that EGFR-TKI-induced cytosolic Bit1 release physically sequesters and inhibits the pro-survival effect of the co-repressor TLE1, we have uncovered a previously unrecognized mechanism governing both initial drug sensitivity and adaptive resistance state. Furthermore, the identification of the TLE1/HES1 interaction and its regulation of Notch signaling, alongside the transcriptomic confirmation of a definitive epithelial-mesenchymal transition program, extends our understanding of the molecular landscape of TKI resistance. This discovery highlights a vulnerable node in LUAD survival signaling, suggesting that pharmacological strategies to recapitulate Bit1 activity could bypass TLE1-mediated survival programs and improve outcomes for patients with EGFR-mutated lung adenocarcinoma.
Footnotes
Authors’ Contributions
Ma Carmela Dela Cruz: Conceptualization, Methodology, Investigation, Visualization, Formal Analysis, Writing – Original Draft; Xin Yao: Conceptualization, Methodology, Investigation, Visualization. Alajah Nealy: Investigation. James Bailey: Investigation. Micah Nalls: Investigation. Paul Mark Medina: Conceptualization, Supervision, Writing-Reviewing and Editing. Renwei Chen – Reviewing and Editing, Hector Biliran: Conceptualization, Methodology, Validation, Formal Analysis, Resources, Writing – Reviewing and Editing, Supervision, Funding acquisition.
Conflicts of Interest
The Authors have no conflicts of interest to declare in relation to this study.
Funding
This research was supported by the National Institutes of Health (NIH) Grant 1R16GM145484-01 (to H.B.), which provided funding for laboratory resources, equipment, reagents, and research assistance. Additional support was provided in part by the National Science Foundation (NSF) HBCU EiR Grant 2502119 (to H.B.) for mechanistic analysis of the TLE1 transcriptional program and the purchase of reagents. The primary author was supported by a research fellowship from the Philippine Council for Health Research and Development (PCHRD), Department of Science and Technology.
Artificial Intelligence (AI) Disclosure
During the preparation of this manuscript, a large language model (Grammarly) was used solely for language editing and stylistic improvements in select paragraphs. No sections involving the generation, analysis, or interpretation of research data were produced by generative AI. All scientific content was created and verified by the authors. Furthermore, no figures or visual data were generated or modified using generative AI or machine learning–based image enhancement tools.
- Received January 8, 2026.
- Revision received January 25, 2026.
- Accepted February 16, 2026.
- Copyright © 2026 The Author(s). Published by the International Institute of Anticancer Research.
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.