RB1-I680T mutation potentiates tumor growth and chemotherapy sensitivity in non-small cell lung cancer via derepressing E2F1 transcription

Abstract

Background

Retinoblastoma Transcriptional Corepressor 1 (RB1) is a critical tumor suppressor restricting the malignant progression of cancer cells. Emerging evidence indicates that RB1 mutations typically promote tumorigenesis through loss of its tumor-suppressive functions. Yet, the biological significance of mutated RB1, specifically certain rare variants, in non-small cell lung cancer (NSCLC) remains elusive. Here, we first reported a rare and previously uncharacterized missense mutation in RB1, the 680th residue isoleucine replaced by threonine (RB1-I680T), in NSCLC.

Methods

To investigate the functional and mechanistic consequences of the RB1-I680T mutation in NSCLC, we first used CRISPR-Cas9 to knock out endogenous RB1 in NSCLC cells. Then, we generated cell models harboring the RB1-I680T mutation by infecting these knockout cells with lentivirus carrying either wild-type RB1 (RB1-WT) or RB1-I680T expression constructs. The biological phenotypes mediated by RB1-I680T were investigated using in vitro and in vivo experiments. The exploration of the molecular mechanism was performed primarily through co-immunoprecipitation, immunofluorescence, dual-luciferase reporter assays, western blot analysis, and protein docking and dynamics simulation.

Results

Our study demonstrated that the I680T mutation caused faster tumor growth and potentiated chemotherapy-induced tumor regression compared to RB1-WT control. Mechanistic studies illustrated that the I680T mutation in RB1 disrupted its inhibition of E2F1 transcriptional activity by weakening the physical interaction between RB1 and E2F1 in a manner dependent on conformational flexibility of RB1 pocket B domain, which is essential for sustaining the enhanced proliferation and chemosensitivity in NSCLC cells.

Conclusion

Our findings elucidate that the I680T mutation-induced loss-of-function of RB1 simultaneously confers invasive proliferation and chemotherapeutic vulnerability to tumor cells, suggesting that RB1-I680T could serve as a predictive biomarker for chemotherapy response in NSCLC. Stratifying patients based on the RB1-I680T mutation status may enable personalized therapeutic strategies, particularly for tumors with E2F1 dysregulation.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12964-026-02695-5.

Introduction

As one of the most common cancers worldwide, lung cancer has been the leading cause of cancerous deaths, affecting both sexes equally [1, 2]. Non-small cell lung cancer (NSCLC), which constitutes about 85% of all lung cancer cases, poses significant clinical challenges due to its aggressive nature, complex biology, and resistance to treatment [3]. Over the past decade, rapid advances in sequencing technology have reinforced the growing consensus that lung cancer is a genetically heterogeneous disease [4–6]. Some mutations in the common tumor-driving genes, such as EGFR, KRAS, ALK, and ROS1 have been functionally characterized, forming the solid foundation for the effective application of targeted drugs in personalized NSCLC therapy [7]. However, the biological functions and potential clinical translational values of certain mutations, particularly rare variants, in key tumor suppressor genes remain to be fully elucidated. Identifying such characterizations is indispensable for advancing the development of more precise and individualized therapeutic strategies for NSCLC.

Retinoblastoma Transcriptional Corepressor 1 (RB1) derives its name from its pivotal role in the pathogenesis of retinoblastoma [8]. As a prototype tumor suppressor gene, the discovery of RB1 pioneered the molecular understanding of cancer suppression mechanisms. Subsequent studies illustrated that RB1-encoded protein functions as a transcriptional corepressor, orchestrating gene silencing through direct interactions with key transcription factors, especially members of the adenoviral early region 2 binding factor (E2F) family [9, 10]. By recruiting chromatin-modifying complexes and modulating epigenetic landscapes, RB1 represses the transcription of its target genes essential for cell cycle progression, proliferation, and differentiation, thereby serving as a critical node in the molecular circuitry governing cellular homeostasis [11].

The RB1 gene is frequently altered in human cancers through various mutational mechanisms, primarily including point mutations, frameshift mutations, and large insertions/deletions (indels). These genetic and epigenetic changes commonly lead to the loss of RB1 tumor suppressor function, contributing to malignant transformation [12]. Dysfunctional RB1 plays a pathogenic role in multiple cancer types, such as retinoblastoma [13], small cell lung cancer (SCLC) [14], prostate cancer [15, 16], breast cancer [17], and osteosarcoma [18]. For instance, recent genomic analyses of retinoblastoma have identified the predominant mutation spectrum, including nonsense mutations, splice-site variants, and frame-shift alterations, which highlights the prevalence of protein-truncating events in RB1 inactivation in retinoblastoma pathogenesis [19–23]. In SCLC, a genetic signature of the loss of two critical tumor suppressors, RB1 and TP53, was observed in most cases [24]. The universal RB1 inactivation in SCLC has spurred extensive research into targeted therapies, among which Aurora kinase inhibitors emerged as a particularly promising approach for RB1-deficient tumors [25, 26]. Nonetheless, the biological functions of mutated RB1, especially certain rare variants, in NSCLC remain largely unknown.

In this study, we found a previously uncharacterized missense mutation in the RB1 gene, isoleucine at the 680th site mutated to threonine (RB1-I680T), in NSCLC, although this mutation was sporadically detected in breast cancer [27], thyroid follicular carcinoma [28], rectal neuroendocrine tumors [29], and penile squamous cell carcinoma [30] based on the COSMIC database [31]. Our results revealed that the RB1-I680T mutation accelerated tumor progression in vitro and in vivo, while it unexpectedly enhanced chemosensitivity, resulting in more pronounced chemotherapeutic drug-induced tumor regression compared to NSCLC cells with wild-type RB1 (RB1-WT). More importantly, the I680T mutation in RB1 impairs its binding to E2F1 by impeding the conformational flexibility of pocket B domain in RB1 protein, compromising RB1-mediated transcriptional repression. This defective interaction leads to sustained E2F1 activation, driving hyperproliferation and heightening chemosensitivity in NSCLC cells. These findings highlight a potential therapeutic window targeting E2F1-driven pathways for NSCLC harboring the RB1-I680T mutation.

Materials and specimens

Mice

Male BALB/c athymic nude mice aged 4 weeks were purchased from Shanghai Model Organisms Center, Inc. (Shanghai, China). The animals were housed at the Model Animal Research Center of Shandong University under controlled conditions with temperature maintained at 22 ± 1 °C, humidity at 50 ± 10%, and a 12-hour light-dark cycle. All animal procedures were approved by the Institutional Animal Care and Use Committee in The Second Hospital of Shandong University (KYLL-2023-438).

Cell culture

The human lung adenocarcinoma cell line A549 and human embryonic kidney cell line 293T (HEK293T) were obtained from Fuheng Biology Co., Ltd. (Shanghai, China). All cell lines underwent authentication through Short Tandem Repeat (STR) profiling performed by Tsingke Biotechnology Co., Ltd. (Beijing, China). Complete mediums were prepared by supplementing 10% fetal bovine serum (ExCell Bio, FSP500) and 1% penicillin-streptomycin (NCM Biotech, C100C5) into basal mediums. A549 cells and their derivatives were maintained in Ham’s F-12 K (Kaighn-Improved) (BasalMedia, L450KJ) while 293T cells were cultured in DMEM medium (Corning, 10-013-CV). All cells were incubated at 37 °C in a humidified atmosphere containing 5% CO₂.

Establishment of A549 cell lines with RB1 knockout

A549 cell lines with RB1 knockout were generated by Haixing Biotechnology Co., Ltd. (Suzhou, China). Target exons in RB1 for knockout were selected based on NCBI genomic data, with sgRNAs designed via the CRISPOR web tool (http://crispor.tefor.net) [32]. The sgRNA and Cas9 protein were incubated and combined to form an RNP complex. Then, the RNP complex was transferred into the cells via the Neon™ Transfection System (Thermo Fisher, MPK5000) following the official website instructions. After cell expansion, RB1 knockout clones were validated through PCR product electrophoresis and Sanger sequencing. Two independent clones (RB1-KO#1 and RB1-KO#2) with confirmed biallelic knockout were pooled to generate a mixed cell line (designated as RB1-KM) for subsequent experiments, thereby minimizing clone-specific effects. The sequences of sgRNAs used here were listed in Supplementary Table 1.

Obtaining A549-KM cell line with stable expression of exogenous RB1-WT or RB1-I680T

Lentiviral particles comprising the sequence encoding RB1-WT or RB1-I680T were generated according to our published protocols [33]. Then, RB1-KM cells were infected with the lentiviral particles. Following puromycin (Beyotime, ST551) selection (1 µg/mL for 72 h), we established two stable polyclonal cell lines, RB1-KM-WT (reconstituted with wild-type RB1) and RB1-KM-I680T (expressing the I680T mutant), for the follow-up experiments.

Cell transfection with SiRNA

Upon reaching 70% confluence in six-well plates, cells were transiently transfected with siRNA using the siTran 2.0 transfection reagent (OriGene, TT320002) following the manufacturer’s guidelines. The siRNA sequences are listed in Supplementary Table 2.

The electrophoresis for Polymerase Chain Reaction (PCR) production

Genomic DNA was extracted using the TIANamp Genomic DNA Kit (TIANGEN, DP304), and quantified using the Nanodrop 2000 (Thermo Fisher, USA). Amplification of target DNA segment was performed with primers and 2X Es Taq MasterMix from (Cwbio, CW0690) via polymerase chain reaction (PCR) under the standard cycling conditions: 94 °C for 2 min, followed by 35 cycles of 94 °C for 30 s, 60 °C for 30 s, and 72 °C for 30 s, with a final extension at 72 °C for 2 min. Products were treated with electrophoresis in 1% agarose gels with TAE buffer (Solarbio, T1060) at 100 V for 30 min and imaged using the Tanon Imaging System. Primer sequences are provided in the Supplementary Table 3. The complete image of electrophoresis is provided in the Supplementary Material.

Reverse Transcription-quantitative PCR (RT-qPCR)

The RNA isolation was performed using TRNzol Universal Reagent for total RNA extraction (TIANGEN, DP424) following the manufacturer’s protocol. RNA concentration and purity were determined using a NanoDrop 2000 spectrophotometer (Thermo Fisher, USA). Subsequently, cDNA synthesis was carried out with the Goldenstar™ RT6 cDNA Synthesis Kit (Tsingke, TSK301). Quantitative PCR was performed using 2×TSINGKE^® Master qPCR Mix (Tsingke, TSE201) on a QuantStudio Q5 system (Thermo Fisher, USA) under standard cycling conditions. GAPDH served as the endogenous control, and relative gene expression was calculated using the 2^−ΔΔCT method. Primer sequences are listed in Supplementary Table 3.

Western Blot (WB) analysis

To obtain the cell lysate, experimental cells were lysed in the lysis buffer containing 10 mM Tris-HCl at pH 7.4, 1% SDS, and 1 mM Na₃VO₄, and then incubated in a 100 °C metal bath for 5 min. Subcellular fractionation was performed using the Minute™ Cytoplasmic and Nuclear Extraction Kit (Invent, SC-003) according to the manufacturer’s instructions to isolate nuclear and cytoplasmic protein fractions. Detailed protocols about WB analysis can be found in our previously published study [34]. Antibodies applied in the present study are summarized in Supplementary Table 4. The complete, uncropped WB images corresponding to all figures are provided in the Supplementary Materials.

Immunofluorescence (IF) assay

Total 10,000 cells were seeded on glass-bottom dishes (Biosharp, BS-15-GJM) and cultured for 24 h. Cells were fixed with 4% paraformaldehyde for 15 min, permeabilized with 0.5% Triton X-100 for 15 min, then blocked with 5% goat serum for 1 h at room temperature. Then, the dish was incubated with primary antibodies overnight at 4 °C. After washing the dish with PBS, corresponding fluorescence-conjugated secondary antibodies were applied for 1 h at room temperature. Nuclei were counterstained using the antifade mounting medium containing DAPI (Beyotime, P0131), and images were captured using a confocal laser scanning microscope (Zeiss, LSM 800). All primary and secondary antibodies used for IF analysis are comprehensively listed in Supplementary Table 4.

Immunohistochemistry (IHC) staining and terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) assays

Mice xenograft tissues underwent fixation in 4% paraformaldehyde and embedded with paraffin. After dewaxing and rehydration, antigen retrieval was performed on 4 μm tissue sections using the Improved Citrate Antigen Retrieval Solution (Beyotime, P0083). Sections were incubated with primary antibodies as recommended dilution at 4 °C then incubated with secondary antibodies conjugated with horseradish peroxidase for 1 h at room temperature. DAB solution (ZSGB-BIO, ZLI-9018) was used to detect the levels of indicated antigens and the sections were counterstained with hematoxylin. In addition, the Colorimetric TUNEL Apoptosis Assay Kit (Beyotime, C1098) was used to detect the apoptosis of tumor tissues followed by counterstain with hematoxylin. All images of sections were captured by NanoZoomer S60 Digital Pathology system (HAMAMATSU, Japan). IHC scores were calculated using the semiquantitative scoring, which incorporates multiplication of two parameters: the scores of staining intensity (0: negative, 1: weak, 2: moderate, 3: strong) and the percentage of immunoreactive cells (0: 0%, 1: 1–25%, 2: 26–50%, 3: 51–75%, 4: 76–100%). The data from 6 replicate samples are presented as mean ± standard deviation (SD).

Dual-luciferase reporter assay

Experimental cells in 24-well plates were co-transfected with pE2F-TA-Luc reporter plasmid (Beyotime, D4054) and pRL-TK Renilla control plasmid (Beyotime, D2760) using LipoMax™ (SUDGEN, 32012). After 48 h, all samples were measured using the dual-luciferase assay kit (Beyotime, RG027) and signals were detected using the Cytation 5 plate reader (BioTek, USA). Transfection efficiency was monitored and calculated by using the pRL-TK Renilla luciferase as a control. Signals were normalized with the control group, and fold changes from three replicate wells are expressed as mean ± SD.

Co-Immunoprecipitation (Co-IP) assay

Cells at 90% confluency were lysed in the cell lysis buffer (Beyotime, P0013). Co-IP assays were performed following our established protocol [35], which details the complete experimental procedures and bead specifications. All information about antibodies or antibody-conjugated magnetic beads for all Co-IP assays is provided in Supplementary Table 4.

Protein docking and dynamics simulation

The molecular dynamics simulations were performed using the GROMACS software [36]. The CHARMM36 force field was employed for the protein, and the TIP3P water model was used to solvate the system. The initial structure was obtained from the X-ray study (PDB ID: 1O9K) [37]. Missing residues were modeled using the AlphaFold 3 online server [38]. The protein was placed in a cubic box with a minimum distance of 1.5 nm between the protein and the box edges. The system was neutralized by adding Na^་ and Cl^- ions to achieve a physiological salt concentration of 0.1538 M. Energy minimization was carried out using the steepest descent algorithm for 10, 000 steps to remove any steric clashes. Subsequently, the system was equilibrated using an annealing procedure: The temperature was maintained at 298.15 K using the velocity-rescale thermostat with a coupling constant of 0.2 ps. The pressure was maintained at 1.0 bar using the C-rescale barostat with a coupling constant of 0.5 ps. The LINCS algorithm was used to constrain all bonds involving hydrogen atoms. The particle mesh Ewald (PME) method was employed for long-range electrostatic interactions, and a cutoff of 1.2 nm was applied for both Van Der Waals and Coulomb interactions. After equilibration, productive molecular dynamics was performed for 400 ns without any restraints. The integration time step was set to 2 fs. Coordinates were saved every 10 ps for further analysis. The root mean square deviation (RMSD), radius of gyration (Rg), and number of hydrogen bonds were calculated using GROMACS tools. Free energy surfaces (FES) were generated from Rg and RMSD data via the sham tool of GROMACS. Molecular dynamics results were visualized by using the DulvyTools (https://zenodo.org/doi/10.5281/zenodo.6339993), and protein structures were rendered using ChimeraX [39].

Cell viability assay

Total 2,000 cells per well were seeded in 96-well plates. At 0, 1, 2, and 3 days, 100 µL medium containing 10 µL Cell Counting Kit-8 (CCK-8) reagent (NCM Biotech, C6005) was added to each well. After 2 h of incubation at 37 °C, absorbance at 450 nm was measured using the Infinite M200 PRO NanoQuant microplate reader (TECAN, Switzerland). The relative viability was normalized with values from day 0.

Calculating the half-maximal Inhibitory Concentration (IC₅₀)

The cells were plated in 96-well plates (2,000 cells per well). The indicated drugs with concentration gradient were applied to treat cells for 48 h. Cell viability was detected to obtain the dose-response curves. The values were normalized with untreated groups as a control. The fitting curves were generated to calculate the IC₅₀ values of drugs using the GraphPad Prism version 8.0.0 for Windows (GraphPad Software, San Diego, California, USA).

Plate colony formation assay

Cells were seeded at a density of 1,000 cells per well in six-well plates and cultured for 14 days, with the medium replaced every 3 days. Colonies were fixed with 4% paraformaldehyde, followed by being stained with 0.5% crystal violet. The colony formation rate was calculated as the percentage of colonies relative to 1,000 seeded cells. The relative rates of colony formation were obtained by normalizing all experimental groups with the control group.

Soft agar colony formation assay

Complete medium containing 0.5% low-melting-point agarose (Sigma-Aldrich, A4018) was pre-coated as a base layer in six-well plates. Subsequently, 10,000 cells suspended in complete medium with 0.35% agarose were overlaid per well. After 21 days of culture, colonies were enumerated. Relative colony formation rates were normalized to the untreated control group.

Sphere formation assay

The experimental cells were cultured using serum-free DMEM/F12 medium (Gibco, C11330500BT) supplemented with epidermal growth factor (EGF) (20 ng/mL) (MCE, HY-P7109) and basic fibroblast growth factor (bFGF) (10 ng/mL) (MCE, HY-P7004) in ultra-low attachment plates (Corning, 3471). Spheres were counted after 21 days. Detailed experimental procedures and sphere counting methodology were performed as previously described [40].

Wound healing assay

Scratches were created in confluent monolayer cells in six-well plates using 200 µL pipette tips (KIRGEN, KG1212). Images were captured at 0 and 24 h using the MShot Image Analysis System (MShot, Guangzhou, China) and quantified using the ImageJ software [41]. The following formula was used to calculate the rate of wound closure: 100% - (wound area after 24 h/wound area in the 0 h) × 100%.

Transwell assay

The detailed information about the protocol was described in our previous study [33]. In brief, the chambers for transwell (Corning, 3422) were coated with or without Matrigel^® Matrix (Corning, 356234) to conduct migration or invasion assays, respectively. Total 20,000 cells in serum-free medium were added to the upper chambers, with complete medium used in the lower chambers. Following 36 h of incubation, the migrated cells on the lower membrane surface were fixed with 4% paraformaldehyde for 15 min, then stained with 0.1% crystal violet for 30 min at room temperature. The images were obtained using a microscope with the MShot Image Analysis System (MShot, Guangzhou, China).

Cell cycle analysis

Cells were serum-starved for 24 h, followed by the recovery in complete medium for 24 h. Subsequently, cells were fixed in pure ethanol overnight at − 20 °C. The following day, staining was performed using the Cell Cycle Assay Kit (Elabscience, E-CK-A351). Flow cytometry analysis was conducted on a CytoFLEX flow cytometer (Beckman Coulter, USA). Cell populations in G0/G1, S, and G2/M phases were quantified based on propidium iodide fluorescence intensity.

Apoptosis assay

After 48 h of drug treatment, cells were stained using the Annexin V-FITC/PI Apoptosis Kit (Elabscience, E-CK-A211) according to the manufacturer’s protocol. The cell apoptosis analysis was conducted on a CytoFLEX flow cytometer (Beckman Coulter, USA), which enabled the quantitative assessment of three distinct cell populations: viable (Annexin⁻/PI⁻), early apoptotic (Annexin⁺/PI⁻), and late apoptotic (Annexin⁺/PI⁺) populations.

Mice xenograft models

The suspension containing 5 × 10⁶ experimental cells in 100 µL PBS was injected subcutaneously into the axillary fossa of 4-week-old BALB/c nude mice. Tumor volume and mouse body weight were monitored every 4 days, with tumor volume calculated using the formula: (length × width²)/2. Once the volume of tumors reached about 100 mm³, intraperitoneal injection of chemotherapeutic agents was initiated. Drug dosages were determined by converting clinical doses of human [42] to mouse based on body surface area normalization. Each drug was diluted in PBS to the appropriate concentration, then mice were administered through intraperitoneal injections. This injection regimen was repeated every 4 days. The experimental endpoint was defined as when any individual tumor volume reached the threshold of 1,000 mm³, at which point all mice were humanely euthanized.

Statistical analysis

All data from three independent experiments are presented as mean ± SD. We performed all statistical analyses by applying the Prism software version 8.0.0 (GraphPad; San Diego, CA, USA). Statistical significance for the measurements of cell viability and xenograft growth was assessed by a two-way analysis of variance (ANOVA) test. Other parametric comparisons were analyzed using either unpaired t-tests for two groups or one-way ANOVA for multiple groups. Non-parametric statistical analysis was performed by a Mann-Whitney test. Statistical methods for each experiment are detailed in the corresponding figure legends. A P-value < 0.05 was considered statistically significant for all analyses.

Results

NSCLC cellular models harboring RB1-I680T mutation were successfully established

During our investigation of mutations within hotspot regions in key tumor suppressors and oncogenes, we identified a rare, previously unreported RB1 missense mutation, RB1-I680T, in NSCLC. Interestingly, this specific mutation is infrequently documented in cancer databases, with few reports limited to breast cancer [27], thyroid follicular carcinoma [28], rectal neuroendocrine tumors [29], and penile squamous cell carcinoma [30] in the COSMIC database [31] (Fig. 1A). Notably, the residue at position 680 in human RB1 protein is highly conserved among species and another two members of human pocket proteins (RBL1 and RBL2), which indicates its key biological functions (Fig. 1B). To elucidate the potential phenotypes driven by the RB1-I680T mutation in NSCLC, A549 cells were selected to establish the cellular model harboring RB1-I680T. First, to eliminate the interference from endogenous RB1 protein (wild-type), we generated two A549 cell lines with the RB1 gene knocked out (RB1-KO#1 and RB1-KO#2) by using CRISPR-Cas9-mediated gene editing (Fig. 1C), with successful clone selection confirmed by the Sanger sequencing and PCR product electrophoresis (Fig. 1D, E). Next, in order to minimize potentially cell-specific confounding factors, we pooled the two independent clones to obtain the mixed cell line (designated as RB1-KM) for subsequent experiments (Fig. 1C). WB analysis further determined the depletion of endogenous RB1 proteins in RB1-KO#1, RB1-KO#2, and RB1-KM cell lines (Fig. 1F). Finally, we constructed the lentivirus containing the plasmid sequences encoding RB1-WT-Flag or RB1-I680T-Flag to infect RB1-KM cells. After screening with puromycin, we generated stable RB1-KM cell lines expressing exogenous RB1-WT-Flag or RB1-I680T-Flag, named RB1-KM-WT and RB1-KM-I680T cell variants (Fig. 1C). Stable clones were further screened to ensure exogenous RB1 expression matched endogenous RB1 protein levels in their parental A549 cells, as verified by WB analysis (Fig. 1G). Additionally, the I680T mutation did not alter RB1 accumulation in the nucleus, as evidenced by comparable nuclear localization patterns between RB1-WT and RB1-I680T in WB analysis and IF assays (Fig. 1H, I). Together, our results validate the successful generation of RB1-I680T mutant NSCLC cell lines.

Fig. 1.

Successful establishment of NSCLC cellular models harboring the RB1-I680T mutation. A Missense mutation landscape of RB1 across pan-cancer (upper) and lung cancer (lower) cohorts in the COSMIC database. B Protein sequences alignment showing conservation of isoleucine (I) at position 680 of human RB1 in mice, rats, and another two human pocket protein family members (RBL1, RBL2). C Schematic workflow for generating A549 with RB1 knockout (RB1-KO#1, RB1-KO#2, RB1-KM) and RB1-KM cell lines stably expressing RB1-WT-Flag or RB1-I680T-Flag (RB1-KM-WT and RB1-KM-I680T). D Sanger sequencing was applied to confirm the knockout of RB1 gene in RB1-KO#1 and RB1-KO#2 clones. Red lines indicate the deletion of DNA segment. E The electrophoresis of PCR production was performed to validate the deletion of RB1 DNA segment in RB1-KO#1 (middle), RB1-KO#2 (right), while their parental A549 cells (left) were used as a control. F, G Western blot (WB) analysis was applied to detect endogenous RB1 protein levels in RB1-KO#1, RB1-KO#2, and RB1-KM cells (F) and exogenous RB1 expression in RB1-KM-WT and RB1-KM-I680T cells (G), with A549 cells as the positive control. β-Actin was served as a loading control. H Cytoplasmic and nuclear protein fractions were isolated from RB1-KM-WT and RB1-KM-I680T cells, followed by WB analysis to assess RB1 protein levels. GAPDH and Lamin B1 were selected as cytoplasmic and nuclear markers, respectively. I Immunofluorescence (IF) assays were applied to reveal RB1 protein localization across stable RB1-KM-WT and RB1-KM-I680T cell lines with the antibody against RB1. The scale bar is as indicated

I680T mutation induces the loss-of-function of RB1 as a tumor suppressor

To figure out the potential effects of I680T mutation on RB1’s tumor-suppressing function, we carried out a series of cellular function experiments. The CCK-8 and plate colony formation assays demonstrated that RB1 deficiency remarkably enhanced cell proliferation, while ectopic RB1-WT almost reversed the proliferative capacity of RB1-KM cells to the level of their parental cells (A549) (Fig. 2A, B). Interestingly, RB1-KM-I680T cells displayed partially weakened cell proliferation when compared to RB1-KM cells but still grew faster than RB1-KM-WT cells (Fig. 2A, B). Similar results were found in soft agar and sphere formation assays evaluating cellular capacity of anchorage-independent growth and stemness, respectively (Fig. 2C, D). Flow cytometric analysis demonstrated that RB1-KM-I680T cells exhibited a significantly higher proportion of cells in the S and G2/M phases but a reduced fraction in the G0/G1 phase compared with RB1-KM-WT controls, supporting that the I680T mutation compromises RB1’s role in restricting G1-S phase transition (Figure S1A, B). However, based on the results from transwell and wound healing assays, RB1-KM, RB1-KM-WT, and RB1-KM-I680T cells exhibited similar migration and invasion capacities compared to A549 cells, indicating that RB1 depletion or I680T mutation induces no significant alteration of the cell motility (Fig. 2E, F). Consistent to RB1-WT- or RB1-I680T-mediated phenotypes, WB analysis revealed that knocking out RB1 remarkably upregulated the protein levels of PCNA (an indicator for proliferation) and two markers for stemness (c-MYC and SOX2), while reintroduction of RB1-WT, but not RB1-I680T, significantly reduced these protein levels in RB1-KM cells. Notably, the expression of these proteins in RB1-KM-I680T cells remained significantly elevated compared to RB1-KM-WT cells (Fig. 2G). Meanwhile, neither RB1 depletion nor the I680T mutation has significant effects on epithelial-mesenchymal transition (EMT) markers, as evidenced by unchanged E-Cadherin (an epithelial marker) and N-Cadherin (a mesenchymal marker) expression levels (Fig. 2G). In addition, the AlphaFold predicted that the RB1-I680T mutation likely possesses pathogenicity, providing computational support for the observed phenotypic alterations (Figure S2). Collectively, our comprehensive functional analyses confirm that the I680T mutation induces a significant loss of RB1’s tumor suppressor function.

Fig. 2.

The I680T mutation impairs RB1’s tumor suppressor function. A-D CCK-8 (A) and plate colony formation assays (B) were applied to evaluate the proliferative capacity of A549, RB1-KM, RB1-KM-WT, and RB1-KM-I680T cells, while soft agar colony formation (C) and sphere formation (D) assays were performed to assess the anchorage-independent growth and stemness of these cells, respectively. E, F Transwell (E) and wound healing assays (F) were used to detect the migration and invasion capabilities of the indicated cell lines. G The indicated protein levels in A549, RB1-KM, RB1-KM-WT, and RB1-KM-I680T cells were analyzed by WB analysis with β-Actin as a loading control. Scale bars were included in all relevant figure panels to indicate magnification. P-values were calculated by a two-way ANOVA (A) or one-way ANOVA (B-F)

RB1-I680T mutation improved the chemotherapeutic efficacy in NSCLC cells

Previous studies have established a strong association between RB1 loss-of-function and altered sensitivity of cancer cells to anti-tumor drugs [43]. Strikingly, analysis of the Genomics of Drug Sensitivity in Cancer (GDSC) database [44] revealed that lung cancer cells with mutant RB1 show a significantly positive correlation with increased sensitivity of several clinically used chemotherapeutic drugs including the Cisplatin (CDDP), Docetaxel (DTX), and Vinorelbine (VNR) compared to RB1-WT controls (Fig. 3A). It is worth noting that the missense mutation I680T, whose functional phenotypes were first clearly characterized in this study, was absent from the GDSC database’s mutant RB1 cohort. Therefore, we explored the potential impacts of the RB1-I680T mutation on the chemosensitivity of NSCLC. To determine the IC₅₀ values of these chemotherapeutic agents (CDDP, DTX, and VNR), we treated the RB1-KM, RB1-KM-WT, and RB1-KM-I680T cell lines with progressively increasing concentrations of these chemotherapeutic agents. Our results revealed significantly intercellular variability in the sensitivity to the three agents: RB1-KM-I680T cells exhibited relatively low IC₅₀ values (1.39 µg/mL for CDDP, 3.16 nM for DTX, and 7.59 nM for VNR, respectively), whereas RB1-KM-WT cells showed markedly higher resistance to these drugs (IC₅₀ values: 2.21 µg/mL for CDDP, 11.76 nM for DTX, and 17.52 nM for VNR, respectively) (Fig. 3B-D). Given that chemotherapy for NSCLC typically employs platinum-based doublet regimens, we further evaluated the therapeutic efficacy of CDDP in combination with either DTX or VNR. As illustrated in Fig. 3E, RB1-KM-I680T cells displayed significantly enhanced chemosensitivity when treated with CDDP combined with DTX therapy. Similar results were also found in RB1-KM-WT and RB1-KM-I680T cells under the treatment with CDDP-VNR combination (Fig. 3F). Interestingly, compared with RB1-KM-WT and RB1-KM-I680T cells, the RB1-KM cells displayed markedly proliferative suppression in both monotherapy (IC₅₀ values: 0.72 µg/mL for CDDP, 1.18 nM for DTX, and 2.99 nM for VNR, respectively) or platinum-based doublet regimens (CDDP combined with DTX or VNR), which might be related to its higher cell proliferation ability (Fig. 3B-F). Consistently, flow cytometry-based analysis of apoptotic cell populations demonstrated that RB1-KM-I680T cells showed a markedly higher apoptosis rate than RB1-KM-WT cells in the condition of monotherapy (CDDP, DTX, and VNR) or platinum-based doublet regimens (CDDP combined with DTX or VNR), while RB1-KM cells exhibited the highest rate of apoptotic cell populations among these three cell lines (Figs. 3G, I, and S3A, B), which was further supported by levels of cleaved PARP1 (Fig. 3H, J). Together, these results identify RB1-I680T as a sensitizing mutation that potentiates chemotherapy efficacy in NSCLC.

Fig. 3.

The I680T mutation in RB1 increases chemosensitivity in NSCLC cells. A The half-maximal inhibitory concentration (IC₅₀) values of Cisplatin (CDDP), Docetaxel (DTX), and Vinorelbine (VNR) in RB1-WT and RB1-mutant lung cancer cells were analyzed based on the Genomics of Drug Sensitivity in Cancer (GDSC) database. B-F RB1-KM, RB1-KM-WT, and RB1-KM-I680T cells were exposed to the monotherapy: CDDP (B), DTX (C), or VNR (D) or platinum-based doublet regimens: CDDP combined with DTX (E) or VNR (F) with varying concentrations as indicated for 48 h, followed by CCK-8 viability assays. Dose-response curves were generated, and IC₅₀ values were calculated using nonlinear regression analysis. G-J RB1-KM, RB1-KM-WT, and RB1-KM-I680T cells were treated with single agent: CDDP (2 µg/mL), DTX (8 nM), or VNR (16 nM) (G, H) or platinum-based doublet regimens: CDDP (1 µg/mL) + DTX (4 nM) or CDDP (1 µg/mL) + VNR (8 nM) (I, J). PBS (dissolving CDDP) or DMSO (dissolving DTX or VNR) was used as a vehicle control. Apoptosis rates were quantified 48 h post-treatment by the flow cytometry using Annexin V-FITC/PI staining (G, I), while WB analysis was performed to evaluate cleaved PARP1 levels (H, J). Protein loading was normalized using β-Actin as an internal control. Statistical significance was determined by the unpaired t-test (A) or one-way ANOVA (G, I) with P-values as shown

RB1-I680T mutation causes faster tumor growth but potentiates the chemotherapeutic efficacy in vivo

To identify the in vivo phenotypic consequences driven by RB1-I680T mutation, we established mice subcutaneous xenograft models by using RB1-KM-WT and RB1-KM-I680T cell lines, followed by treatment with clinically relevant platinum-based doublet therapies (CDDP combined with either DTX or VNR) (Fig. 4A). As expected, RB1-KM-I680T cells formed subcutaneous tumors with significantly faster growth kinetics, larger tumor volumes, and greater tumor weights compared to RB1-KM-WT controls (Fig. 4B-D). Meanwhile, the administration with CDDP-DTX or -VNR combination led to higher tumor inhibition rates (IR) in xenograft tumors derived from RB1-KM-I680T cells (Fig. 4D). Furthermore, the body weight of mice in the treatment and the control groups exhibits no significant difference, suggesting that the applied dosages of chemotherapeutic agents did not cause evident toxicity or adverse effects (Fig. 4E). Consistently, IHC analysis determined a marked upregulation in Ki-67 and PCNA protein levels in tumor tissue slices derived from RB1-KM-I680T xenografts compared to controls (Fig. 4F, G). Furthermore, TUNEL assays revealed that treatment with CDDP-DTX or -VNR combinations induced significantly higher apoptosis rates in xenograft tumors derived from RB1-KM-I680T cells than RB1-KM-WT controls (Fig. 4H). Collectively, our results indicate that the RB1-I680T mutation accelerated tumor growth in vivo, whereas it enhanced the efficiency of chemotherapy-mediated tumor suppression.

Fig. 4.

RB1-I680T mutation promotes tumor growth while enhancing chemosensitivity in vivo. A Schematic illustration indicates detailed information about mice subcutaneous xenograft establishment and drug administration. B-E Subcutaneous RB1-KM-WT and RB1-KM-I680T xenografts were treated with vehicle control (PBS), CDDP (0.43 mg per mouse) + DTX (0.43 mg per mouse), or CDDP (0.43 mg per mouse) + VNR (0.14 mg per mouse) (n = 6 per group) every 4 days starting from the 22nd day. The tumor growth kinetics is evidenced by the measurements of tumor volumes (B), while the representative excised tumor images at the endpoint (C), final tumor weights with inhibition rates (IR) (D), and monitoring of mice body weight (E) are displayed. F, G IHC images were applied to determine Ki-67 (F) and PCNA (G) protein expression in the xenograft tumors derived from RB1-KM-WT and RB1-KM-I680T cells. H TUNEL assays were performed to assess apoptosis rates in RB1-KM-I680T or RB1-KM-WT xenograft tumors from mice treated with platinum-based doublet regimens (CDDP + DTX or CDDP + VNR). Scale bars are included in all relevant panels. Data was analyzed by a two-way ANOVA for (B, E), one-way ANOVA for (D), Mann-Whitney test for (F, G), and unpaired t-test for (H) with P-values as shown

RB1-I680T mutation sustains the phenotypes by derepressing E2F1 transcriptional activity

It has been clearly demonstrated that RB1 functions as a tumor suppressor via restricting the transcriptional activity of E2F family members, among which RB1 exhibits preferential binding to E2F1 in a cell cycle phase-specific manner [45–47]. Hence, we first evaluated whether I680T mutation impacts E2F1-mediated transcription. As shown in Fig. 5A, the dual-luciferase reporter assays with the pE2F-TA-Luc reporter (detecting E2F’s transcriptional activity) clarified that knocking out RB1 remarkably enhanced the luciferase activity, while ectopic RB1-WT almost restored the fluorescence intensity in RB1-KM cells to the level comparable to its parental A549 cells. Intriguingly, luciferase activity in RB1-KM-I680T cells was lower than that in RB1-KM cells but remained higher than that in RB1-KM-WT cells (Fig. 5A). Then, to determine whether RB1-I680T mutation exerts its phenotypic effects through regulating E2F1 transcriptional activity, we performed rescue experiments by knocking down E2F1 in RB1-KM-I680T cells, followed by functional assays. Our results demonstrated that siRNA targeting E2F1 dramatically reduced endogenous expression of E2F1 (Figure S4A, B), which inhibited the luciferase activity in RB1-KM-I680T cells (Fig. 5B). Moreover, siRNA-induced E2F1 depletion markedly attenuated the capacities of proliferation and stemness of RB1-KM-I680T cells (Fig. 5C-F), which was further supported by the decreased protein levels of PCNA, c-Myc, and SOX2 (three downstream targets of E2F1 signaling pathway) caused by E2F1 knockdown (Fig. 5G). Meanwhile, flow cytometry analysis demonstrated that apoptotic ratio of RB1-KM-I680T cells is higher than that in RB1-KM-WT cells but lower than that in RB1-KM cells, while E2F1 knockdown significantly reduced apoptosis rates in RB1-KM-I680T cells treated with platinum-based doublet regimens (CDDP combined with DTX or VNR) (Figs. 5H and S5). These results were further corroborated by the trend of cleaved PARP1 levels (Fig. 5I), implying that E2F1 deficiency impairs chemosensitivity induced by RB1-I680T mutation. In summary, our findings strongly support that the phenotypes induced by the I680T mutation were maintained through abolishing RB1-mediated inhibition of E2F1-driven transcription.

Fig. 5.

RB1-I680T mutation maintains phenotypes through E2F1 transcriptional derepression. A, B The pE2F-TA-Luc reporter and pRL-TK Renilla control plasmids were co-transfected into A549, RB1-KM, RB1-KM-WT, and RB1-KM-I680T cells (A) or RB1-KM, RB1-KM-WT, RB1-KM-I680T, and RB1-KM-I680T cells with siRNA (si-E2F1#1)-induced E2F1 knockdown (B), followed by the dual-luciferase reporter assays measuring the luciferase activities. C-G RB1-KM, RB1-KM-WT, RB1-KM-I680T cells, and RB1-KM-I680T cells with E2F1 knockdown (RB1-KM-I680T + si-E2F1#1) were selected for the CCK-8 (C) and plate colony formation (D) assays assessing cell proliferation, soft agar colony formation assay detecting the anchorage-independent growth (E), sphere formation evaluating stem cell properties (F), and WB analyzing PCNA, c-Myc, and SOX2 protein expression levels (G). H, I RB1-KM, RB1-KM-WT, RB1-KM-I680T cells, and RB1-KM-I680T cells with E2F1 knockdown (RB1-KM-I680T + si-E2F1#1) were treated with platinum-based doublets: CDDP (1 µg/mL) + DTX (4 nM) or CDDP (1 µg/mL) + VNR (8 nM). PBS (dissolving CDDP) or DMSO (dissolving DTX or VNR) was applied as the vehicle control. Then, cell apoptosis was assessed after 48 h of treatment via flow cytometry with Annexin V-FITC/PI dual staining (H), while WB analysis was conducted to detect the protein levels of cleaved PARP1 with β-Actin as a loading control I. All scale bars are as indicated. Data is presented as mean ± SD. The one-way ANOVA was performed to identify the statistical significance (A, B, D, E, F, H), while a two-way ANOVA was applied in (C)

The I680T mutation disrupts the physical interaction between RB1 and E2F1 via weakening the conformational flexibility of RB1 pocket B domain

To explore the potential molecular mechanisms underlying RB1-I680T-mediated derepression of E2F1 transcriptional activity, we first detected the effects of RB1-I680T mutation on E2F1 expression. The results from the RT-qPCR and WB analysis exhibited no significant difference in E2F1 mRNA and protein levels between RB1-KM-WT and RB1-KM-I680T cells (Fig. 6A, B). Then, we assessed whether RB1-I680T regulates the nuclear and cytoplasmic distribution of E2F1 protein. Our results identified that RB1-I680T mutation did not affect E2F1 nuclear accumulation, as demonstrated by the WB and IF analyses (Fig. 6C-E). Therefore, we reasoned that I680T mutation may attenuate the physical interaction between RB1 and E2F1 proteins. In accordance with this hypothesis, Co-IP assays determined the significantly decreased amount of E2F1 protein in the complex pulled down by the antibody against RB1-I680T-Flag compared to RB1-WT-Flag control (Fig. 6F). Similar consequences were demonstrated in HEK293T cells (Fig. 6G). Notably, RB1-KM-I680T cells exhibited significantly fewer RB1-E2F1 co-localization foci compared to RB1-KM-WT controls (Fig. 6E). To further illustrate how the I680T mutation disrupts RB1-E2F1 binding, we performed all-atom molecular dynamics simulations using GROMACS [36]. The RB1-I680T/E2F1 complex demonstrated significantly higher RMSD compared to the complex containing RB1-WT, with a pronounced peak from the 270 ns (Fig. 6H). FES analysis revealed that the I680T-mutant RB1/E2F1 complex sampled multiple potential wells, indicating greater conformational instability than the wild-type complex (Fig. 6I). Trajectory snapshots captured progressive dissociation of the transcription activation domain (TAD) in E2F1 from RB1-I680T (Fig. 6J), corresponding to increased E2F1-specific RMSD in the RB1-I680T/E2F1 complex (Fig. 6K). Structural visualization confirmed movement of RB1’s pocket B domain away from E2F1 in the I680T-mutant complex (Fig. 6L). Structure analysis indicated the backbone of leucine at the 676th site (L676) acted as an acceptor for hydrogen bonds, forming more interactions with mutant T680 rather than wild-type I680 (Fig. 6M). Trajectory analysis demonstrated these persistent hydrogen bonds throughout 400 ns simulation (Fig. 6N), which likely facilitates RB1 pocket B domain movement. Collectively, the I680T mutation impairs RB1-E2F1 binding by reducing the structural flexibility of the pocket B domain in RB1, thereby destabilizing their interaction.

Fig. 6.

The I680T mutation impairs RB1-E2F1 binding by reducing conformational flexibility in the pocket B domain of RB1. A-B The expression of E2F1 mRNA and protein levels in RB1-KM-WT and RB1-KM-I680T cells was determined by the RT-qPCR (A) and WB analysis (B). GAPDH and β-Actin are selected as internal controls, respectively. C, D Subcellular fractionation was performed to isolate cytoplasmic and nuclear proteins from RB1-KM-WT and RB1-KM-I680T cells (C) or HEK293T cells expressing exogenous RB1-WT-Flag and RB1-I680T-Flag (D). WB analysis of these fractions confirmed E2F1 protein levels with GAPDH as the cytoplasmic marker and Lamin B1 as a nuclear marker. E The subcellular localization of E2F1 protein and its binding to RB1 in RB1-KM-WT and RB1-KM-I680T cells were illustrated by the IF analysis with the antibody against E2F1 or RB1. DAPI was used for nuclear stainings. Co-localization of E2F1 and RB1 was indicated by white arrows. Scale bars: 5 μm. F, G Co-IP assays with anti-Flag beads were performed to obtain the protein complexes in RB1-KM-WT and RB1-KM-I680T cells (F) or HEK293T cells transfected with the plasmids expressing RB1-WT-Flag or RB1-I680T-Flag (G). Then, the immunoblotting was performed to detect the levels of indicated proteins. H The root mean square deviation (RMSD) analysis throughout the 400 ns simulations was applied to reveal the conformational fluctuations in complexes of RB1-WT/E2F1 or RB1-I680T/E2F1. I Free energy surfaces (FES) analysis was used to assess the conformational instability in RB1-WT/E2F1 or RB1-I680T/E2F1complex. J Representative structural snapshots (from 200 to 400 ns) were captured to illustrate the progressive E2F1 dissociation from RB1-I680T (initiated from 270 ns). Transcriptional activation domain (TAD) of E2F1 is depicted in blue or orange, respectively, while RB1 is shown in gray. K The RMSD analysis was performed to assess the fluctuations of E2F1 when it bound to RB1-WT or RB1-I680T during molecular dynamics simulations. L Structural alignment was used to highlight the RB1 pocket B domain moving away from E2F1 in RB1-I680T complex compared to RB1-WT/E2F1 complex. RB1 is displayed in gray, while TAD of E2F1 is depicted as a blue mesh. Orange arrows indicate the direction of pocket B domain movement after being structurally aligned to RB1-WT. M, N Hydrogen bond occupancy (M) and the persistence of bond formation (N) between I680 and L676 in RB1-WT/E2F1 or between T680 and L676 in RB1-I680T/E2F1 complex were displayed throughout simulations. Data in (A) was displayed as mean ± SD and was statistically analyzed using an unpaired t-test with P-values as shown

Discussion

Mutations in RB1 gene commonly contribute to malignant biological behaviors of many tumors via disrupting its original anti-cancer effects [12]. Here, we focused on a previously unreported RB1 missense mutation (RB1-I680T) in NSCLC that demonstrates its dual functions, oncogenic and chemosensitivity properties. Mechanistic exploration reveals that the RB1-I680T mutation induces a partial loss-of-function that disrupts RB1’s canonical repression of E2F1-mediated transcription, leading to two distinct but interconnected consequences, invasive proliferation and chemotherapeutic vulnerability (Fig. 7). This interesting duality challenges the conventional understanding of RB1 loss-of-function in cancer biology.

Fig. 7.

Graphical summary illustrates the molecular mechanism of RB1-I680T mutation-mediated phenotypes

Our study first identified a novel missense mutation in the RB1 gene (RB1-I680T) in NSCLC. While this genetic alteration has been occasionally reported in other malignancies, including breast cancer [27], thyroid follicular carcinoma [28], rectal neuroendocrine tumors [29], and penile squamous cell carcinoma [30], its functional significances in NSCLC remains unexplored. Therefore, we successfully generated NSCLC functional models carrying the RB1-I680T mutation using CRISPR-Cas9-regulated genome editing, followed by the transfection with plasmids expressing either RB1-WT or I680T variant. The RB1-I680T mutation exhibits dual oncogenic/therapeutic effects: while it accelerates tumor progression in vitro and in vivo, it concurrently enhances chemosensitivity to conventional anticancer agents. Interestingly, considering that Aurora kinase inhibitors exhibit the potential value in clinical therapy of SCLC with RB1 deficiency [25, 26], the anti-tumor effects of these inhibitors and their combination strategies in NSCLC harboring RB1-I680T mutation merit systematic exploration.

In quiescent cells, the RB1 protein suppresses E2F-mediated transcription (including E2F1’s transcriptional activity) through direct complex formation, by which RB1 coordinately regulates both proliferative and apoptotic pathways [9, 10, 48]. In tumorigenesis, RB1 function is frequently compromised either through mutational inactivation or hyperphosphorylation [12]. In the present study, I680T mutation partially abrogates RB1-mediated suppression of E2F1 transcriptional activity, which leads to significant dysregulation of cell cycle and apoptosis. I680T is located in RB1 pocket B domain, which is responsible for its physical interaction with the E2F family [9]. We demonstrated that this genetic alteration (RB1-I680T mutation) attenuates the binding between RB1 and E2F1 to derepress the transcription of E2F-responsive genes. However, whether I680T mutation affects the interaction between RB1 and other E2F family members, such as E2F2 and E2F3, was not explored in this study. Given dual consequences induced by RB1-I680T, we reasoned that upon this mutation, the majority of E2F1 transcription factors are released, enabling activation of cell cycle-promoting genes, while the E2F1 may remain partially bound to mutant RB1 protein, forming a repressive complex that selectively suppresses the transcription of anti-apoptotic target genes, which deserved further investigation.

It is particularly important to note that RB1-I680T might regulate or intervene other key signaling pathways in NSCLC progression, although we identified the essential role of E2F1’s transcriptional activity in maintaining RB1-I680T-driven phenotypes. For instance, beyond its canonical interaction with E2F transcription factors, RB1 engages with FOXM1 to restrict its transactivation [49]. Another study revealed the E2F-independent mechanisms that RB1 modulates chromatin remodeling by dysregulating the activity of KDM5A, a demethylase specific for H3K4, and further enhancing the mitochondrial function and differentiation mediated by PGC-1α transcriptional coactivator complex [50]. In mouse subretinal angiomatous proliferation and hemangioblastoma, RB1 plays a critical role in regulating angiogenesis through its ability to repress the transcription of hypoxia-inducible factor (HIF) target genes [51]. In addition, the hyperphosphorylation of RB1 typically attenuates its suppression of E2F’s transcriptional activity via accelerating the degradation of RB1 protein [9, 48]. Given that the I680T mutation introduces a potential phosphorylation site in RB1 protein, we raise a critical question of whether this alteration modulates RB1 phosphorylation dynamics, despite our observation that RB1-I680T retains wild-type-like expression levels and subcellular localization.

In summary, our study provides the first evidence that the RB1-I680T mutation exerts a dual oncogenic-therapeutic role that it simultaneously drives tumor progression through E2F1-dependent transcriptional dysregulation and confers vulnerability to conventional chemotherapy in NSCLC. These findings position RB1-I680T as both a prognostic marker and a potential predictor of treatment response in NSCLC.

Authors’ contributions

PCL, YPZ, and XGZ conceived and designed this project. YLZ, FYG, YL, and JFW performed most of the experiments and collected the data. FRL, BW, BXW, YW, YFZ, and NM assisted in the experiments. YLZ, FYG, YL, and ZXT performed data analysis and visualization. PCL and YLZ drafted this original manuscript. YPZ, XLZ, and XGZ revised the manuscript. PCL, YPZ, YLZ, FYG, and YL checked the data. All authors read and approved of the final manuscript.

Funding

This work was financially supported by the Nurturing and Development Fund from The Second Hospital of Shandong University (No. 2022YP62, Peichao Li), the Shandong Provincial Natural Science Foundation for Young Scholars (No. ZR2022QH285, Peichao Li) and National Natural Science Foundation for Young Scholars (No. 82403079, Peichao Li), China.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with international guidelines for biomedical research and was approved by the Ethics Committee and the Institutional Animal Care and Use Committee in The Second Hospital of Shandong University (KYLL-2023-438).

Competing interests

The authors declare no competing interests.

Conflict of interest

The authors have no conflict of interest.