Virtual screening and experimental validation of a METTL3-targeting peptide with in vitro antiproliferative activity against non-small cell lung cancer cells Authors: Ying Dai, Hanying Wu, Yahui Shen, Fan Yang, Xiaohao Liu, Ting Zhuang Published: 2026 ABSTRACT AbstractThe METTL3–METTL14 protein–protein interaction (PPI) plays an important role in tumour progression, and disruption of this interaction has been explored as a potential strategy. In this study, four peptides were identified from a peptide database through virtual screening and molecular docking. Among them, peptide-1 showed the lowest Kd value among the tested peptides in MST analysis (Kd = 0.76 ± 0.02 μM). Binding mode analysis, molecular dynamics simulations, and MM/PBSA calculations suggested that peptide-1 might form a binding-related conformation with METTL3 under the simulated conditions. In lung cancer cells, peptide-1 showed growth-inhibitory activity, whereas weaker effects were observed in BEAS-2B cells. Peptide-1 also reduced the METTL3–METTL14-associated NanoBRET signal, decreased cellular m6A levels and JUNB mRNA expression, and its antiproliferative effect was attenuated by METTL3 knockdown. These findings suggest that peptide-1 may represent a METTL3-targeting peptide candidate for further evaluation.Keywords: Biological evaluation, METTL3–METTL14 PPI, non-small cell lung cancer, peptide, virtual screening Introduction Lung cancer is one of the most common malignancies and remains among the leading causes of cancer incidence and mortality worldwide1. Histologically, lung cancer is mainly classified into small cell lung cancer (SCLC) and non-small cell lung cancer (NSCLC)2. Among these, NSCLC is the predominant subtype, accounting for approximately 85–90% of all lung cancer cases3. At present, the clinical management of NSCLC is largely determined by disease stage. Surgical resection is generally used for patients with resectable early-stage disease. In contrast, advanced-stage disease is commonly managed with systemic therapy, radiotherapy, targeted therapy, immunotherapy, or their combinations4. Although conventional chemotherapy remains widely used, its efficacy is often limited by poor target selectivity, low bioavailability, and acquired drug resistance5. In recent years, molecularly targeted therapy has become a major focus in NSCLC research6. However, its long-term efficacy may be affected by resistance associated with epigenetic alterations and tumour heterogeneity7,8. Therefore, the development of safer and more effective inhibitors based on novel molecular targets is of considerable importance for NSCLC intervention. Methyltransferase-like 3 (METTL3) is a highly conserved core component of the N6-methyladenosine (m6A) writer complex9. When acting alone, METTL3 exhibits only weak methyltransferase activity in vitro and usually needs to form a complex with METTL14 through its methyltransferase domain (MTD) to achieve higher catalytic activity10,11. Studies have shown that METTL3 serves as the catalytic core of this complex, whereas METTL14 primarily functions as an RNA-binding platform12. Increasing evidence indicates that, by forming a complex with METTL14, METTL3 is involved in biological processes associated with the initiation and progression of lung cancer13,14. Both METTL3 and METTL14 show elevated expression in non-small cell lung cancer tissues and cell lines15,16. In addition, the METTL3–METTL14 complex has been suggested to contribute to the malignant progression of lung cancer cells by increasing JUNB expression.16,17 Given that METTL3 serves as the catalytic core of this complex and that its oncogenic function depends on the formation of a stable complex with METTL1414, targeting the interfacial region of METTL3 and disrupting its interaction with METTL14 may represent a potential strategy for intervention against this complex18. From a drug-discovery perspective, the METTL3–METTL14 complex can be regarded as a protein-protein interaction (PPI)-based system19. PPIs are widely involved in essential biological processes, including cell proliferation, apoptosis, post-transcriptional regulation, and signal transduction, and have therefore emerged as important intervention targets in cancer therapy20. However, PPI interfaces are often large and relatively flat and usually lack deep pockets that can be readily recognised by conventional small molecules, making them challenging targets for drug discovery21. Compared with conventional small molecules, peptides and peptidomimetics can mimic native protein-binding motifs, cover extended interaction surfaces, and recognise key PPI “hot spots”; thus, they have become an important strategy for developing PPI inhibitors22. In particular, cyclic peptides, stapled peptides, and other conformationally constrained peptides may improve conformational stability, target affinity, and resistance to proteolytic degradation, providing new opportunities for targeting traditionally challenging PPIs23. Recent advances in m6A‑targeted therapeutics have highlighted the development of small‑molecule inhibitors against METTL324. Most of these inhibitors are designed to target the S‑adenosylmethionine (SAM) binding pocket of METTL325. For example, STM2457 has been shown to competitively inhibit the catalytic activity of the METTL3‑METTL14 complex by binding to the SAM pocket and to suppress tumour growth in acute myeloid leukaemia models26,27. Optimisation efforts have yielded STC‑15, which exhibits stronger enzymatic inhibition and has entered Phase I clinical trials in cancer patients (NCT05584111)28. This progression underscores ongoing efforts to translate METTL3 inhibition into clinical applications. However, these catalytic site inhibitors generally need to compete with high intracellular concentrations of SAM, and discrepancies between enzymatic potency and cellular activity may limit their broader applicability29. Concurrently, strategies for targeting the m6A epitranscriptomic machinery have expanded beyond conventional catalytic inhibition to include allosteric modulation, protein–protein interaction (PPI) disruption, targeted degradation, and peptide‑based inhibitors30. For instance, stapled peptides such as RSM3 have been reported to target the interface of the METTL3 subunit, disrupt METTL3–METTL14 complex assembly, and exhibit cytotoxic effects in multiple cancer cell lines31. Together, these findings suggest that peptide‑based PPI inhibitors may provide a feasible approach for intervening in METTL3–METTL14 function32. Accordingly, in the present study, RSM3 was selected as a positive control to evaluate the activity of newly designed METTL3–METTL14 targeting peptides. Although peptides offer high specificity and effective interface recognition for targeting PPIs, their development as therapeutics faces challenges such as susceptibility to proteolytic degradation in vivo, short plasma half lives, poor cellular permeability, low oral bioavailability, and limited tissue distribution33,34. Therefore, peptide candidates often require optimisation strategies, including sequence modification, incorporation of D‑amino acids, N/C‑terminal capping, cyclisation, lipidation, PEGylation, fusion with cell‑penetrating peptides, stapling, or formulation with nanodelivery systems to enhance stability and in vivo delivery35,36. Based on these considerations, identifying novel peptides that can effectively recognise the METTL3–METTL14 interface and inhibit proliferation in NSCLC cells may help to expand intervention strategies targeting the m6A writer complex. Such peptides could also provide useful starting points for the further optimisation of METTL3–METTL14-targeting peptide inhibitors. Computer-aided drug design (CADD) has become a widely used approach for the efficient identification of lead compounds37,38. This approach integrates virtual screening and molecular docking to enable efficient identification of potential bioactive molecules from large-scale compound libraries, with the advantages of high computational efficiency and good predictive capability39,40. In this study, potential active peptides were screened based on the structural features of the METTL3–METTL14 complex. Four targeted peptides (peptides 1–4) were hit from the database and further evaluated for in vitro activity. Among these peptides, peptide‑1 exhibited a relatively favourable overall profile and showed measurable activity in non-small cell lung cancer-related models. Materials and methods Reagents Peptides 1–4, RSM3, the scrambled peptide, the unrelated peptide YELALE, and FITC-labelled peptide-1 were purchased from GL Biochem (Shanghai, China). The MonolithTM RED-NHS Protein Labelling Kit (Cat. No. MO-L011) and standard MST capillaries (Cat. No. MO-K022) were purchased from NanoTemper Technologies GmbH (Munich, Germany). Sterile phosphate-buffered saline (PBS, 1×, calcium- and magnesium-free; Cat. No. 10010023), high-glucose DMEM containing l-glutamine (Cat. No. 11965092), penicillin-streptomycin solution (100×; Cat. No. 15140122), 0.25% trypsin-EDTA solution (Cat. No. 25200056), Opti-MEMTM I Reduced Serum Medium (Cat. No. 31985070), heat-inactivated foetal bovine serum (FBS, Premium; Cat. No. A5670502), High-Capacity cDNA Reverse Transcription Kit (Cat. No. 4368814), and SYBRTM Green qPCR Master Mix (Cat. No. A66732) were purchased from Thermo Fisher Scientific (Waltham, MA, USA; Grand Island, NY, USA). Dimethyl sulfoxide (DMSO, HPLC grade; Cat. No. 34869), human serum, Tris–HCl, NaCl, TCA, rhodamine B isothiocyanate-dextran (Cat. No. R9379), bisBenzimide H 33342 trihydrochloride (Hoechst 33342; Cat. No. B2261), fibronectin from human plasma (Cat. No. F0895), TRI Reagent® (Cat. No. T9424), and lentiviral particles carrying METTL3 shRNA were obtained from Sigma-Aldrich (St. Louis, MO, USA). The MTT Cell Proliferation and Cytotoxicity Assay Kit (Cat. No. C0009S) and Crystal Violet Staining Solution were purchased from Beyotime Biotechnology (Shanghai, China). NanoBRETTM Nano-Glo® Detection System (Cat. No. N1661) was purchased from Promega (Madison, WI, USA). PI/RNase Staining Buffer (Cat. No. 550825) was obtained from BD PharmingenTM (San Jose, CA, USA). Human Cleaved Caspase-3 (Asp175) ELISA Kit (Cat. No. ab220655) and Human Bcl-2 ELISA Kit (Cat. No. ab272102) were purchased from Abcam (Cambridge, UK). The EpiQuikTM m6A RNA Methylation Quantification Kit (Colorimetric; Cat. No. P-9005) was purchased from EpiGentek Group Inc. (Farmingdale, NY, USA). Cell culture Non-small cell lung cancer cells A549 (Catalog No.: CRM-CCL-185), NCI-H1975 (Catalog No.: CRL-5908), and H23 (Catalog No.: CRL-5800), as well as bronchial epithelial cells BEAS-2B (Catalog No.: CRL-3588), were purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA). Cells were cultured in high-glucose DMEM containing l-glutamine supplemented with 10% heat-inactivated FBS and 1% penicillin–streptomycin at 37 °C in a humidified atmosphere containing 5% CO2. Cells were passaged every 2–3 days, and experiments were performed using cells in the logarithmic growth phase. Virtual peptide library construction The method for the construction of the peptide library was adopted from published literature41. In this study, the QuaSAR-CombiGen module in the Molecular Operating Environment (MOE; Chemical Computing Group Inc., Montreal, Quebec, Canada) was used to perform full combinatorial enumeration of three peptide fragment sets, generating a virtual library containing 59,319 peptides. Each peptide in the library was composed of six amino acids. Briefly, the connection sites between peptide fragments were defined as terminal atoms labelled “An”, where n is a positive integer. For the first dipeptide fragment set, the C-terminal oxygen atom was defined as the “A0” site, yielding 39 fragments. For the second dipeptide fragment set, the N-terminal hydrogen atom and the C-terminal oxygen atom were assigned as the “A1” and “A2” sites, respectively, yielding 39 fragments. For the third dipeptide fragment set, the N-terminal hydrogen atom was defined as the “A0” site, yielding 39 fragments. The “A0” sites of the first and third dipeptide fragment sets were linked to the “A1” and “A2” sites of the second dipeptide fragment set, respectively. Full combinatorial enumeration of the three dipeptide fragment sets generated a virtual library of 59,319 hexapeptides. The resulting two-dimensional (2D) structures were converted into three-dimensional (3D) conformations and energy-minimized in MOE prior to molecular docking. Preprocessing of the METTL3–METTL14 complex crystal structure Crystal structure preprocessing was performed as previously described42. The crystal structure of the METTL3–METTL14 complex (PDB ID: 5IL1; resolution, 1.71 Å) was obtained from the Protein Data Bank (PDB) and imported into MOE. To target the METTL3–METTL14 interaction interface and identify potential bioactive peptides capable of disrupting their association, METTL14 was removed from the complex structure, and only METTL3 was retained as the docking receptor. The resulting structure was then prepared using the QuickPrep workflow under the Amber10:EHT force field. Hydrogen atoms were added using Protonate 3D, and water molecules located more than 4.5 Å from the methyltransferase domain were removed. Finally, energy minimisation was performed on the system until the root-mean-square (RMS) gradient decreased 0.1 kcal·mol−1·Å−1. Virtual screening Virtual screening was conducted according to the strategy reported in the literature41. The preprocessed structure of METTL3 was employed as the receptor for molecular docking. Native residues within the binding pocket were retained, and no mutations were introduced. Protonation states under physiological conditions were assigned and verified using the Protonate 3D module in MOE. The docking region was defined based on the METTL3–METTL14 interaction interface, and a docking sphere with a 5 Å radius was generated to cover the selected interface region. Ligand conformations were sampled within the binding pocket at a density of 1.0. During docking, ligand poses were generated using the Triangle Matcher placement method, with up to 100 poses generated for each ligand. The receptor was treated as rigid. Initial ranking was performed using the London dG scoring function, followed by rescoring with the GBVI/WSA dG method. All calculations were repeated three times under identical conditions, and the mean predicted binding free energy (docking score) was used for ligand prioritisation. The four top-ranked peptides were selected for subsequent experimental evaluation. In addition, the highest-scoring peptide was randomly rearranged to generate a scrambled peptide. A scrambled peptide with the same amino acid composition and sequence length as the original peptide, but with a different residue arrangement, was used as a negative control in subsequent experiments to evaluate non-specific effects. An unrelated peptide, YELALE, was also included as an additional negative control. This peptide has a different amino acid composition from peptide-1, while retaining a similar short linear hexapeptide format and comparable physicochemical properties, including molecular weight, net charge, and hydrophobicity. It was tested under the same experimental conditions to help assess non-specific peptide-related effects. Microscale thermophoresis (MST) assay The MST assay was performed according to a previously reported protocol, with minor modifications43. With RSM3 and the scrambled peptide serving as the positive and negative controls, respectively, METTL3 was fluorescently labelled to a final concentration of 10 μM using the Monolith™ RED-NHS 2nd Generation Protein Labelling Kit in accordance with the manufacturer’s instructions. The labelled METTL3 protein samples were wrapped in aluminium foil and kept on ice. The test peptides, RSM3, and the scrambled peptide were dissolved in PBST (phosphate-buffered saline containing Tween) and serially diluted two-fold to generate 16 concentration gradients. For peptides with high lipophilicity, an appropriate amount of DMSO was added to facilitate solubilisation, with the final DMSO concentration controlled at ≤5%. The mixtures were subsequently kept in the dark for 5 min. Following capillary loading, fluorescence signals were recorded on a Monolith NT.115 instrument (Nano Temper Technologies GmbH). Structure–activity relationship (SAR) analysis Structure–activity relationship (SAR) analysis was performed according to the previously reported method44. To further examine the interaction features between each peptide and METTL3, the top-ranked docking conformation of each peptide was imported into Molecular Operating Environment (MOE) for interaction analysis and visualisation. For each METTL3–peptide complex, METTL3 residues surrounding the peptide-binding interface were inspected, and protein-peptide interaction maps were generated using the interaction analysis module in MOE. Based on the spatial relationship between each peptide and the surrounding METTL3 residues, hydrogen-bond interactions were primarily analysed and displayed in the interaction maps. Two-dimensional interaction diagrams and three-dimensional binding models were used to compare the binding orientations and interaction patterns of peptides 1–4 at the METTL3 binding interface. Molecular dynamics (MD) simulations MD simulations were performed as previously described, with minor modifications45. A 100 ns molecular dynamics (MD) simulation was carried out using GROMACS (GROningen MAchine for Chemical Simulations, version 2021.5) to preliminarily evaluate the predicted conformational behaviour of the METTL3–peptide-1 complex. The initial structure of the METTL3–peptide-1 complex was taken from the top-ranked docking pose. After preprocessing and parameterisation, the complex was placed in a cubic box with a 1.0 nm margin, solvated using the SPC/E water model, and neutralised with Na+ and Cl– ions. The system was energy-minimized for 5000 steps using the steepest descent method and subsequently equilibrated under the NVT (300 K) and NPT (1.0 bar) ensembles. During the simulation, the leap-frog algorithm was used with a time step of 2 fs, and energy and trajectory data were saved every 10 ps. Root mean square deviation (RMSD), root mean square fluctuation (RMSF), secondary structure, radius of gyration (Rg), and hydrogen-bond profiles were subsequently analysed using GraphPad Prism. Free energy landscape Based on the RMSD and Rg data of the METTL3–peptide-1 complex, Gibbs free energy landscape data were generated via GROMACS, with 2D and 3D visualisation analyses conducted using Python (version 3.13.3). MM/PBSA calculations According to the strategies reported in previously published literature46, the binding free energy of the METTL3–peptide-1 complex was calculated using the single-trajectory approach. The protein-ligand binding affinity was quantitatively estimated from the difference between the total free energy of the complex and those of the unbound components (METTL3 and peptide-1). Inhibition of cell growth The experimental procedure was performed as previously described41. A549, NCI-H1975, H23, and BEAS-2B cells were seeded into 96-well plates at 5 × 103 cells per well in 100 μL of high-glucose DMEM containing l-glutamine and supplemented with 10% FBS. After 12 h, the cells were exposed to serial twofold dilutions of the indicated test compounds and peptides, including peptides 1–4, RSM3, STM2457, the scrambled peptide, and the unrelated peptide, at concentrations ranging from 0.23 μM to 60.00 μM (0.23, 0.47, 0.94, 1.88, 3.75, 7.50, 15.00, 30.00, and 60.00 μM) for 72 h at 37 °C in 5% CO2. MTT solution (10 μL, 5 mg/mL in PBS) was added to each well and incubated for 4 h. The supernatant was then removed, and 100 μL of DMSO was added to dissolve the formazan crystals. Absorbance was measured at 570 nm using a BioTek Synergy HT microplate reader, and IC50 values were calculated accordingly. All experiments were performed in triplicate and repeated independently at least three times. Serum stability assay The serum stability assay was performed according to previously reported methods with minor modifications31. Briefly, peptide-1 was incubated at 200 μM in a human serum/Tris–HCl buffer mixture at a ratio of 1:3 (v/v). The buffer contained 20 mM Tris–HCl and 100 mM NaCl, pH 7.4. At 0, 0.5, 1, 2, 4, and 6 h, aliquots were collected and mixed with an equal volume of 15% TCA to precipitate serum proteins. After incubation on ice for 15 min, samples were centrifuged at 13,000 rpm for 10 min at 4 °C using an Eppendorf 5424 R refrigerated centrifuge (Eppendorf, Hamburg, Germany). The supernatants were analysed by HPLC. The remaining peptide-1 was quantified based on the corresponding peak area and normalised to the 0 h value, which was set as 100%. Data are presented as mean ± SD from three independent experiments. Cellular uptake assay by confocal microscopy The cellular uptake assay was performed according to previously reported methods with minor modifications47. Considering that METTL3/METTL14 have been reported to be mainly localised in the nucleus10,48,49, confocal imaging was used to observe the intracellular distribution of FITC-labelled peptide-1 in A549 cells. Briefly, A549 cells were seeded in 35-mm glass-bottom dishes and cultured overnight. After washing with PBS, cells were incubated with FITC-labelled peptide-1 at 8 μM and rhodamine-labelled dextran at 1 mg/mL for 2 h at 37 °C. DextranRho was used as an endocytic tracer. Cells were then washed with PBS, stained with Hoechst 33342 at 2 μg/mL for 15 min, and imaged using an LSM 700 laser scanning confocal microscope (Carl Zeiss, Jena, Germany). FITC-labelled peptide-1, DextranRho, and Hoechst 33342 were detected in the green, red, and blue channels, respectively. Scale bars represent 50 μm. Colony formation assay The colony formation assay was performed according to previously reported methods with minor modifications50. Briefly, A549 cells were seeded into 12-well plates at 2,000 cells per well. After attachment, cells were treated with vehicle control or peptide-1 (2, 4, and 8 μM), with the treatment-containing medium replaced every 2–3 days. After 10–14 days, visible colonies were fixed with methanol for 15 min and stained with 0.1% crystal violet for 15 min. Images were captured using an Olympus IX73 inverted microscope (Olympus, Tokyo, Japan). Colonies containing at least 50 cells were counted using ImageJ software (National Institutes of Health, Bethesda, MD, USA), and colony formation inhibition was calculated as follows: inhibition (%) = [1 – colony number in the treated group/colony number in the vehicle control group] × 100%. Each group included three technical replicates, and the experiment was independently repeated three times. Wound healing assay The wound healing assay was performed according to previously reported methods with minor modifications31. Briefly, A549 cells were seeded into 12-well plates and cultured until a nearly confluent monolayer was formed. A straight scratch was generated using a sterile 200 μL pipette tip, and detached cells were removed by PBS washing. Cells were then cultured in serum-free high-glucose DMEM containing l-glutamine and treated with vehicle control or peptide-1 at 2, 4, or 8 μM. Images were captured at 0 and 12 h using an Olympus IX73 inverted microscope. Wound widths were measured using ImageJ software, and the wound closure rate was calculated as follows: wound closure rate (%) = [(W0–W12)/W0] × 100%, where W0 and W12 represent wound widths at 0 and 12 h, respectively. The migration index was normalised to the vehicle control group. At least three fields were analysed for each group, and the experiment was independently repeated three times. Transwell invasion assay The Transwell invasion assay was performed according to previously reported methods with minor modifications51. Briefly, A549 cells were suspended in serum-free high-glucose DMEM containing l-glutamine with vehicle control or peptide-1 at 2, 4, or 8 μM and seeded into Matrigel-coated Transwell inserts with 8.0 μm pores. High-glucose DMEM containing l-glutamine and 10% FBS was added to the lower chambers. After 24 h, non-invasive cells were removed, and invaded cells were fixed with methanol and stained with 0.1% crystal violet. Images were captured using an Olympus IX73 inverted microscope. Invaded cells were counted using ImageJ software, and the invasion index was normalised to the vehicle control group. Each group included three technical replicates, and the experiment was independently repeated three times. Cell cycle analysis The cell cycle assay was performed according to previously reported methods with minor modifications50. Briefly, A549 cells were seeded into 6-well plates and cultured until they reached approximately 60–70% confluence. The cells were synchronised in serum-free high-glucose DMEM containing l-glutamine for 12 h and then treated with vehicle control or peptide-1 at 2, 4, or 8 μM for 24 h. After treatment, both floating and adherent cells were collected to minimise the loss of detached cells. The culture medium containing floating cells was first collected, and the adherent cells were gently washed with cold PBS and detached using trypsin. The detached cells were combined with the floating cells and centrifuged at 1000 rpm for 5 min. After discarding the supernatant, the cells were washed twice with cold PBS and fixed in ice-cold 75% methanol at −20 °C for at least 2 h. After fixation, the cells were centrifuged to remove the fixative and washed with PBS to remove residual methanol. The cell pellets were then resuspended in PI/RNase Staining Buffer and incubated for 15–30 min at room temperature in the dark to allow RNase digestion and PI staining of cellular DNA. PI fluorescence was detected using a BD FACSCanto II flow cytometer (BD Biosciences, San Jose, CA, USA). Debris and cell aggregates were excluded by appropriate forward- and side-scatter gating, and doublets were excluded using single-cell gating. At least 1 × 104 single-cell events were collected for each sample. The percentages of cells in the G0/G1, S, and G2/M phases were analysed using FlowJo software (FlowJo LLC, Ashland, OR, USA). NanoBRET assay for METTL3–METTL14 interaction Based on previously reported strategies52, the effect of peptide-1 on the METTL3–METTL14 interaction was evaluated using a NanoBRET-based protein-protein interaction assay in living cells. A549 cells were co-transfected with plasmids encoding METTL3 fused to NanoLuc luciferase (METTL3–NanoLuc) and METTL14 fused to HaloTag (METTL14-HaloTag). After 20–24 h, cells were harvested and resuspended in phenol red-free medium containing 4% FBS. Cells were prepared in parallel with HaloTag 618 ligand-labelled wells and corresponding no-ligand background control wells according to the manufacturer’s protocol. peptide-1 was added at indicated concentrations, with vehicle and scrambled peptide controls included. NanoBRET substrate was added according to the manufacturer’s protocol. Donor (450 nm) and acceptor (610 nm) emissions were measured using a Varioskan LUX multimode microplate reader (Thermo Fisher Scientific, Waltham, MA, USA). Corrected BRET ratios were calculated by subtracting background (wells without ligand) and expressed in milliBRET units. Fusion protein expression and cell viability were monitored in parallel to exclude potential artefacts. IC50 values were derived from normalised BRET signals using GraphPad Prism. Overall m6A quantification in A549 cells Based on previously reported methods53, overall m6A RNA methylation levels were quantified in A549 cells using the EpiQuikTM m6A RNA Methylation Quantification Kit (Colorimetric; EpiGentek Group Inc., Farmingdale, NY, USA; Cat. No. P-9005). Briefly, A549 cells were seeded in 6-well plates and treated for 24 h with vehicle control, scrambled peptide (8 μM), RSM3 (8 μM), or peptide-1 (2, 4, and 8 μM). Total RNA was extracted using TRIzol reagent, and RNA concentration and purity were assessed using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). Samples with A260/A280 ratios of 1.8–2.1 were used for m6A quantification. Equal amounts of RNA were immobilised in assay wells and sequentially incubated with m6A capture and detection antibodies according to the kit instructions. Absorbance was measured at 450 nm using a Varioskan LUX microplate reader (Thermo Fisher Scientific, Waltham, MA, USA). Relative m6A levels were normalised to RNA input, with vehicle-treated samples set to 100%. Reverse transcription quantitative real-time polymerase chain reaction (RT-qPCR) The experimental protocol was designed with reference to previously published studies16. Total RNA was extracted from control A549 cells and cells treated with 2, 4, or 8 μM peptide-1 using TRIzol reagent. RNA concentration and purity were assessed using a NanoDrop 2000 spectrophotometer, and samples with an A260/A280 ratio of 1.8–2.0 were used for cDNA synthesis. Reverse transcription was performed with 1 μg of total RNA using the RT Reverse Transcription Kit according to the manufacturer’s instructions. Quantitative real-time PCR was carried out on an Applied Biosystems QuantStudio 3 system using SYBR Green qPCR Master Mix. Each 20 μL reaction contained 10 μL SYBR Green qPCR Master Mix, 0.4 μL each of forward and reverse primers (10 μM), 2 μL cDNA template, and 7.2 μL nuclease-free water. Amplification was performed at 95 °C for 30 s, followed by 40 cycles of 95 °C for 5 s and 60 °C for 34 s. Melting curve analysis was used to confirm amplification specificity. The primer sequences were as follows: JUNB forward, 5′-CAAGGTGAAGACGCTCAAGG-3′, reverse, 5′-TCATGACCTTCTGTTTGAGCTG-3′; GAPDH forward, 5′-GAAGGTGAAGGTCGGAGTC-3′, reverse, 5′-GAAGATGGTGATGGGATTTC-3′. Relative JUNB mRNA expression was calculated by the 2−ΔΔCt method using GAPDH as the internal reference and normalised to the control group. ELISA analysis of apoptosis-related proteins The ELISA assay was performed according to previously reported methods with minor modifications54. Briefly, A549 cells were seeded into 6-well plates and treated with vehicle control or peptide-1 at 2, 4, and 8 μM for 24 h when cells reached approximately 70% confluence. After treatment, cells were washed with cold PBS and lysed using the cell extraction buffer provided in the ELISA kits. Cell lysates were centrifuged at 12,000×g for 10 min at 4 °C, and the supernatants were collected. Total protein concentrations were determined using a BCA protein assay. The levels of cleaved caspase-3 and Bcl-2 were measured using the corresponding ELISA kits according to the manufacturers’ instructions. Absorbance was read at 450 nm using a BioTek Synergy HT microplate reader (BioTek Instruments, Winooski, VT, USA). The results were normalised to total protein concentration and expressed relative to the vehicle control group. Transfection of METTL3 shRNA The experimental procedures were carried out according to previously published methods55. For shRNA transfection, lentiviral particles carrying METTL3 shRNA were diluted in Opti-MEM medium containing 6 μg/mL polybrene and added to A549 cells. After 3 days, stably transduced cells were selected with 5 μg/mL puromycin. The selected cells were then seeded into 96-well plates, and peptide IC50 values were determined by MTT assay. Statistical analysis Statistical analysis was performed using GraphPad Prism 10.0 (GraphPad Software Inc., San Diego, CA, USA). Quantitative data are presented as mean ± SD unless otherwise indicated. Selected experiments were independently repeated at least three times, as specified in the corresponding methods or figure legends. Comparisons between two groups were analysed using an unpaired Student’s t test, while comparisons among multiple groups were performed using one-way ANOVA followed by Tukey’s post hoc test. A p values <0.05 was considered statistically significant (*p < 0.05, **p < 0.01, ***p < 0.001). Results Virtual screening The virtual screening workflow used to identify peptides with the potential to disrupt the METTL3–METTL14 interaction is shown in Figure 1. In this study, virtual screening and molecular docking were combined to identify peptides from a 59,319-member database that were predicted to interact favourably with the METTL3 binding interface. Each peptide in the database was individually docked into the corresponding binding interface of the preprocessed METTL3 structure. As shown in Figure 2, all peptides were ranked according to their predicted binding free energies (docking scores), from lowest to highest. In general, a lower predicted binding free energy suggests a more favourable interaction pattern between the peptide and the target protein. Ultimately, the top four peptides (peptides 1–4) were selected for subsequent interaction analysis, and their sequences are listed in Table 1. peptides 1–4 all exhibited low predicted binding free energies. Among them, peptide-1 showed the lowest predicted binding free energy, suggesting that it may adopt the most favourable interaction mode with METTL3 in the docking model. According to a SciFinder search, peptides 1–4 have not been previously reported. In addition, to assess the sequence specificity of peptide-1 binding, a scrambled control peptide was generated by randomly rearranging its amino acid sequence. The workflow of multi-step virtual screening of METTL3-targeting peptides. Calculated binding free energies (kcal/mol) of peptides 1–4. Docking calculations were repeated three times under identical conditions, and data are presented as mean ± SD (n = 3). Table 1.Peptide sequences and binding affinities towards METTL3 determined by MST analysis.NamePeptide sequenceKd ± SD (μM)Peptide-1NVDIDF0.76 ± 0.02Peptide-2NASIDL2.37 ± 0.15Peptide-3NCTVSI5.04 ± 0.38Peptide-4SASIDI9.52 ± 0.63RSM3R8RCMELGRECLNLW3.06 ± 0.17Scrambled peptideDVINFD>30Unrelated peptideYELALE>30Notes: Data are presented as mean ± SD from three independent experiments (n = 3). Kd values were determined by MST assays. Values reported as greater than 30 μM indicate that the exact Kd value was not determined within the tested concentration range. Peptide sequences and binding affinities towards METTL3 determined by MST analysis. Notes: Data are presented as mean ± SD from three independent experiments (n = 3). Kd values were determined by MST assays. Values reported as greater than 30 μM indicate that the exact Kd value was not determined within the tested concentration range. MST analysis Because peptides 1–4 all showed relatively low predicted binding free energies in the docking analysis, microscale thermophoresis (MST) assays were further performed to evaluate their binding affinities for METTL3, and the results are summarised in Table 1. The dissociation constants (Kd) were determined using a Monolith NT.115 instrument, with lower Kd values generally indicating higher binding affinity. Compared with the positive control peptide RSM3 (Kd = 3.06 ± 0.17 μM) and other peptides, peptide-1 showed the lowest Kd value (0.76 ± 0.02 μM), suggesting that peptide-1 exhibited the strongest METTL3-binding affinity among the peptides examined in this assay. In addition, both the scrambled peptide and the unrelated peptide showed Kd values greater than 30 μM under the tested conditions, indicating that definitive binding affinities could not be determined within the tested concentration range. These results support the possibility that the interaction between peptide-1 and METTL3 is related to its specific sequence features, although further validation is still needed to fully define the binding determinants. SAR analysis To further compare the interaction patterns of peptides 1–4 with key METTL3 residues, their docking conformations were systematically analysed. As shown in Figure 3 and Table 2, all four peptides could be accommodated within the METTL3 binding groove and formed hydrogen-bonding interactions with key residues, including Arg468, Arg471, Trp475, and Asn477. Specifically, peptide-4 was able to insert into the binding pocket and form hydrogen bonds with Arg468, Trp475, Asn477, Gly479, and Lys480, while also exhibiting nonbonded interactions with Val458, Gly470, Arg471, and Gly473, which may help stabilise its positioning at the binding interface. By comparison, peptide-3 retained these core interactions and extended further into the pocket, and its Asn residue formed additional interactions with Gly473 and Asn477, suggesting improved spatial complementarity relative to peptide-4. The binding mode of peptide-2 appeared to be further optimised. In addition to forming multiple hydrogen bonds with Arg468, Arg471, Gly473, Trp475, Asn477, and Gly479, its Asp residue at position 5 formed two additional hydrogen bonds with Lys480, resulting in broader coverage of the binding interface than that observed for peptide-3. Among the four peptides, peptide-1 showed the most favourable binding features. This peptide adopted a relatively extended conformation within the binding groove and formed a more continuous hydrogen-bonding network with Arg471, Gly473, Trp475, Asn477, Gly479, and Lys480, together with good spatial complementarity. In addition, the Asp residue at position 3 formed multiple hydrogen-bonding interactions with Arg468, which may further strengthen its anchoring at the binding interface. Taken together, peptide-1 showed advantages over the other candidate peptides in terms of hydrogen-bonding interactions, interface coverage, and pocket complementarity, suggesting a relatively more favourable binding mode in the docking model. This interpretation is also consistent with its lower predicted binding free energy. Docking poses of peptides 1–4 at the METTL3 binding interface. Peptide-1 is coloured yellow in panels A and B; peptide-2 is coloured cyan in panels C and D; peptide-3 is coloured purple in panels E and F; and peptide-4 is coloured blue in panels G and H. METTL3 is shown as a gray surface model, and residues surrounding the peptide-binding interface are shown as pink sticks. Nitrogen and oxygen atoms are coloured blue and red, respectively. Dashed lines indicate hydrogen-bond interactions. Table 2.Key binding interactions of peptides 1–4 within the METTL3 active site.NameHydrogen bonding interactionsHydrophobic interactionsPeptide-1Arg468: 4 H-bonds,Arg471: 1 H-bond,Gly473: 1 H-bond,Trp475: 1 H-bond,Asn477: 3 H-bonds,Gly479: 1 H-bond,Lys480: 2 H-bondsVal458,Gly470Peptide-2Arg468: 2 H-bonds,Arg471: 1 H-bond,Gly473: 1 H-bond,Trp475: 1 H-bond,Asn477: 3 H-bonds,Gly479: 1 H-bond,Lys480: 2 H-bondsVal458,Gly470Peptide-3Arg468: 2 H-bonds,Arg471: 1 H-bond,Gly473: 1 H-bond,Trp475: 1 H-bond,Asn477: 3 H-bonds,Gly479: 1 H-bondVal458,Gly470,Lys480Peptide-4Arg468: 2 H-bonds,Trp475: 1 H-bond,Asn477: 2 H-bonds,Gly479: 1 H-bond,Lys480: 2 H-bondsVal458,Gly470,Arg471,Gly473 Key binding interactions of peptides 1–4 within the METTL3 active site. MD analysis of the METTL3–peptide-1 complex As shown in Figure 4(A–E), a 100 ns MD simulation was performed to preliminarily predict the conformational behaviour of the METTL3–peptide-1 complex under the simulated conditions. The RMSD curve showed fluctuations during the early stage and then gradually reached a relatively stable range, mainly remaining around 0.30–0.40 nm. This result suggests that the overall conformation of the complex may have reached a relatively equilibrated state during the simulation. RMSF analysis showed relatively low fluctuations in most residues, whereas several local regions displayed higher flexibility. The Rg curve remained generally stable at approximately 1.70–1.80 nm, suggesting that the overall compactness of the complex did not show marked changes during the simulation. The number of hydrogen bonds mainly fluctuated between 10 and 15, indicating that hydrogen-bonding interactions may contribute to the predicted association between peptide-1 and METTL3. Secondary structure analysis further showed that the major structural elements were largely preserved, with only minor local variations. Overall, these computational results provide preliminary predictive support that the METTL3–peptide-1 complex may maintain a relatively stable binding-related conformation under the simulated conditions, although direct experimental validation is still required to confirm its binding mode and stability. Molecular dynamics (MD) simulation analysis of the METTL3–peptide-1 complex. (A) RMSD of the METTL3–peptide-1 complex and peptide-1. (B) RMSF of METTL3 Cα residues. (C) Rg of METTL3. (D) Evolution of the secondary structure content of METTL3. (E) Number of intermolecular hydrogen bonds formed between METTL3 and peptide-1 (distance < 0.35 nm, standard angle criteria). (F) Residue contributions of potential hot residues of Peptide-1 binding to METTL3. Data are presented as mean ± SEM. (G) Energy decomposition of potential hot residues of Peptide-1 binding to METTL3. Data are presented as mean ± SEM. Residue contribution and energy decomposition analysis of METTL3–peptide-1 binding As shown in Figure 4(F–G), energy decomposition analysis was further performed to evaluate the potential contribution of individual residues to the METTL3–peptide-1 interaction. In Figure 4(F), several residues showed negative contributions to the binding free energy, suggesting that these sites may participate in stabilising the interaction between peptide-1 and METTL3. Among them, Arg468, Leu476, His478, and Lys480 appeared to contribute relatively more strongly and might represent important local residues involved in complex stabilisation. In Figure 4(G), the energy components of key residues were further decomposed. The results suggest that favourable contributions from some residues may mainly arise from electrostatic energy and van der Waals energy. Notably, Arg468 and Lys480 showed relatively pronounced electrostatic contributions, indicating that charged interactions might be involved in maintaining peptide-1 binding to METTL3. Overall, these results suggest that the interaction between peptide-1 and METTL3 may be mediated by multiple residues and jointly influenced by van der Waals interactions, electrostatic effects, and solvation-related contributions. FEL analysis of the METTL3–peptide-1 complex As shown in Figure 5(A,B), the free energy landscape of the METTL3–peptide-1 complex was constructed based on RMSD and Rg values to further analyse its conformational distribution during the simulation. Figure 5(A) presents the two-dimensional free energy landscape, in which the x axis represents RMSD, the y axis represents Rg, and the colour scale indicates the relative free energy level. Figure 5(B) shows the corresponding three-dimensional free energy landscape, providing a more intuitive view of the energy distribution among different conformational states. The lower-energy regions were mainly concentrated around RMSD values of approximately 0.25–0.35 nm and Rg values of approximately 1.73–1.76 nm, suggesting that the complex might preferentially populate these conformational regions during the simulation. Together with the RMSD and Rg analyses, these results suggest that the METTL3–peptide-1 complex might maintain a relatively stable binding conformation under the simulated conditions, while still exhibiting some local conformational fluctuations. Gibbs free energy landscape (FEL) of the METTL3–peptide-1 complex. (A) Two-dimensional (2D) contour map and (B) three-dimensional (3D) surface plot. The blue global minimum energy basin represents the dominant binding conformation of the complex. Antiproliferative activity of peptides 1–4 in A549 cells Considering that peptide-1 showed a relatively favourable profile in virtual screening and MST binding assays, we further evaluated the antiproliferative activities of peptides 1–4 in A549 cells using MTT assays. As shown in Table 3, peptide-1 showed the lowest IC50 value among the tested peptides (1.13 ± 0.08 μM). The dose–response curve further showed that peptide-1 reduced A549 cell viability in an apparent concentration-associated manner, whereas the scrambled peptide showed no obvious inhibitory effect within the tested concentration range (Figure S1). peptide-2, peptide-3, and peptide-4 exhibited IC50 values of 4.96 ± 0.42 μM, 9.80 ± 0.61 μM, and 16.05 ± 1.19 μM, respectively. In contrast, the scrambled peptide and the unrelated peptide did not show obvious growth-inhibitory effects within the tested concentration range; therefore, their IC50 values were reported as greater than 30 μM in Table 3. The reference peptide RSM3 and the METTL3 inhibitor STM2457 showed IC50 values of 7.24 ± 0.53 μM and 13.76 ± 1.45 μM, respectively. These results suggest that peptide-1 may have relatively favourable antiproliferative activity in A549 cells under the present experimental conditions and was therefore selected for subsequent cellular experiments. Table 3.Growth-inhibitory activity of peptides 1–4, RSM3, STM2457, and control peptides in A549 cells.NameA549IC50 ± SD (µM)Peptide-11.13 ± 0.08Peptide-24.96 ± 0.42Peptide-39.80 ± 0.61Peptide-416.05 ± 1.19RSM37.24 ± 0.53STM245713.76 ± 1.45Scrambled peptide>30Unrelated peptide>30Notes: Data are presented as mean ± SD from three independent experiments (n = 3). IC50 values were determined by MTT assays in A549 cells. Values reported as greater than 30 μM indicate that the exact IC50 value was not determined within the tested concentration range. Growth-inhibitory activity of peptides 1–4, RSM3, STM2457, and control peptides in A549 cells. Notes: Data are presented as mean ± SD from three independent experiments (n = 3). IC50 values were determined by MTT assays in A549 cells. Values reported as greater than 30 μM indicate that the exact IC50 value was not determined within the tested concentration range. Serum stability of peptide-1 The serum stability of peptide-1 was evaluated in a human serum-containing system at 37 °C. As shown in Figure S2, the percentage of peptide-1 remaining showed a gradual decrease over the tested period. The peptide level was relatively close to the initial value at early time points and decreased to approximately 75% at 60 min. After 120 min of incubation, about 68% of peptide-1 was detected, and approximately 59% remained detectable after 360 min. These results indicate that peptide-1 could still be detected in the serum-containing system during the observation period, although a time-dependent decrease was observed. Cellular uptake and inhibitory effects of peptide-1 in A549 cells To further evaluate the cellular effects of peptide-1, its intracellular distribution and effects on colony formation, migration, invasion, and cell-cycle progression were examined in A549 cells. Confocal microscopy showed that FITC-labelled peptide-1 fluorescence was detectable in A549 cells after treatment, with fluorescence signals observed in the nuclear region and partially overlapping with Hoechst 33342 staining in the merged images. Considering that METTL3 has been reported to be mainly localised in the nucleus, these results suggest that peptide-1 may enter A549 cells and show partial nuclear-region fluorescence distribution, which may provide a spatial opportunity for interaction with nuclear METTL3 (Figure S3). peptide-1 treatment increased the inhibition of colony formation in a concentration-associated manner, indicating that it may suppress the long-term proliferative capacity of A549 cells under the present experimental conditions (Figure S4A). In addition, peptide-1 reduced the migration and invasion indices of A549 cells compared with the control group, and these effects appeared to become more evident with increasing concentrations (Figure S4B,C). Cell-cycle analysis further showed an increased proportion of cells in the G2/M phase after peptide-1 treatment, accompanied by a reduction in the G0/G1 and S-phase populations, suggesting that peptide-1 may contribute to G2/M-phase cell-cycle accumulation in A549 cells (Figure S4D). Together, these results suggest that peptide-1 may inhibit A549 cell proliferation and motility and may affect cell-cycle progression under the present experimental conditions. Growth-inhibitory profile of peptide-1 in lung cancer cells and BEAS-2B cells To further examine whether the growth-inhibitory effect of peptide-1 could be observed in other lung cancer cell lines, MTT assays were performed in NCI-H1975 and H23 cells, with BEAS-2B cells included as a normal lung epithelial cell model. As shown in Table 4, peptide-1 showed IC50 values of 1.97 ± 0.14 μM and 2.03 ± 0.16 μM in NCI-H1975 and H23 cells, respectively. In contrast, the IC50 value of peptide-1 in BEAS-2B cells was reported as greater than 30 μM within the tested concentration range. Both the scrambled peptide and the unrelated peptide showed IC50 values greater than 30 μM in all three cell lines. The METTL3 inhibitor STM2457 showed IC50 values of 18.05 ± 2.26 μM, 20.17 ± 3.04 μM, and 24.26 ± 3.81 μM in NCI-H1975, H23, and BEAS-2B cells, respectively. These results suggest that peptide-1 may retain growth-inhibitory activity in multiple lung cancer cell lines, while showing relatively weaker effects on BEAS-2B cells under the present experimental conditions. The limited activity of the scrambled peptide and unrelated peptide further supports the possibility that the observed effects of peptide-1 may be related to its specific sequence features, although this conclusion requires further validation. Table 4.IC50 values of peptide-1, STM2457, and control peptides in additional lung cancer cell lines and normal lung epithelial cells.NameIC50 (μM) ± SDNCI-H1975H23BEAS-2BPeptide-11.97 ± 0.142.03 ± 0.16>30Scrambled peptide>30>30>30Unrelated peptide>30>30>30STM245718.05 ± 2.2620.17 ± 3.0424.26 ± 3.81Notes: Data are presented as mean ± SD from three independent experiments (n = 3). IC50 values were determined by MTT assays. Values reported as greater than 30 μM indicate that the exact IC50 value was not determined within the tested concentration range. IC50 values of peptide-1, STM2457, and control peptides in additional lung cancer cell lines and normal lung epithelial cells. Notes: Data are presented as mean ± SD from three independent experiments (n = 3). IC50 values were determined by MTT assays. Values reported as greater than 30 μM indicate that the exact IC50 value was not determined within the tested concentration range. Effects of peptide-1 on the METTL3–METTL14 interaction and m6A-related cellular responses To further investigate whether peptide-1 affects the METTL3–METTL14 interaction and m6A-related cellular responses, NanoBRET, m6A quantification, RT-qPCR, and ELISA assays were performed in A549 cells. Given that NanoBRET has been reported as a proximity-based live-cell approach for monitoring protein-protein interactions56–58, we employed this assay to evaluate the effect of peptide-1 on the METTL3–METTL14 interaction in A549 cells. In the NanoBRET assay, peptide-1 reduced the normalised BRET signal in a concentration-dependent manner, with an IC50 value of 0.98 ± 0.05 μM, suggesting that peptide-1 may interfere with the METTL3–METTL14 interaction and/or spatial proximity in living cells (Figure 6A). Consistently, peptide-1 decreased the relative m6A level in A549 cells, whereas the scrambled peptide showed little effect at 8 μM. The positive control RSM3 also reduced m6A levels, supporting the responsiveness of this assay system (Figure 6B). peptide-1 treatment was also associated with a concentration-related reduction in JUNB mRNA expression, suggesting a possible effect on METTL3/m6A-related downstream gene expression (Figure 6C). Cleaved caspase-3 and Bcl-2 were selected as apoptosis-related protein markers because their levels have been commonly assessed by ELISA in previous apoptosis studies50,59,60. ELISA analysis further showed that peptide-1 increased the relative level of cleaved caspase-3 and decreased the relative level of Bcl-2, suggesting that peptide-1 treatment may be associated with apoptosis-related changes in A549 cells (Figure 6D). Taken together, these results suggest that peptide-1 may interfere with the METTL3–METTL14 interaction and/or spatial proximity, reduce cellular m6A levels, and affect apoptosis-related signalling under the present experimental conditions. Peptide-1 affects METTL3–METTL14-associated m6A signalling and related cellular readouts in A549 cells. (A) Dose–response curve of peptide-1 measured by BRET assay, showing modulation of METTL3–METTL14 interaction (IC50 = 0.98 ± 0.05 μM). (B) Overall m6A RNA methylation levels in A549 cells following treatment with Peptide-1, scrambled peptide, or the positive control peptide RSM3, as determined using an ELISA-based assay. (C) Relative JUNB mRNA expression in A549 cells following peptide-1 treatment, as determined by RT-qPCR. (D) Protein levels of cleaved Caspase‑3 and Bcl‑2 after peptide‑1 treatment in A549 cells, measured by ELISA. For quantitative analyses, values are presented as mean ± SD from at least three independent experiments. Statistical significance was assessed by one-way ANOVA with Tukey’s multiple-comparison correction. *p < 0.05, **p < 0.01, ***p < 0.001; p > 0.05 was considered not significant. Effect of METTL3 knockdown on peptide-1-mediated growth inhibition To assess whether the antiproliferative activity of peptide-1 is related to METTL3, METTL3 expression was knocked down in A549 cells using short hairpin RNA (shRNA), and the peptide’s activity was evaluated in both shControl and shMETTL3 groups. As shown in Table 5, peptide-1 exhibited inhibitory effects in shControl A549 cells, with an IC50 value of 1.13 ± 0.08 μM. In contrast, when METTL3 was knocked down, the growth-inhibitory effect of peptide-1 was markedly reduced, and its IC50 value was reported as greater than 30 μM within the tested concentration range. The scrambled peptide and the unrelated peptide showed IC50 values greater than 30 μM in both shControl and shMETTL3 cells. These results suggest that the growth-inhibitory effect of peptide-1 may be, at least in part, related to METTL3 expression, indicating that its activity may involve a METTL3-related mechanism rather than being solely due to non-specific cytotoxicity. However, further target-binding assays and mechanistic studies will still be required to clarify the extent to which peptide-1 acts through a METTL3-related mechanism. Table 5.The IC50 values of peptide-1, its scrambled control peptide and STM2457 in A549 cells with METTL3 knockdown (shMETTL3).NameIC50 (μM) ± SDShControlshMETTL3Peptide-11.13 ± 0.08>30Scrambled peptide>30>30Unrelated peptide>30>30STM245713.76 ± 1.45>30Notes: Data are presented as mean ± SD from three independent experiments (n = 3). IC50 values were determined by MTT assays. Values reported as greater than 30 μM indicate that the exact IC50 value was not determined within the tested concentration range. The IC50 values of peptide-1, its scrambled control peptide and STM2457 in A549 cells with METTL3 knockdown (shMETTL3). Notes: Data are presented as mean ± SD from three independent experiments (n = 3). IC50 values were determined by MTT assays. Values reported as greater than 30 μM indicate that the exact IC50 value was not determined within the tested concentration range. Conclusion In conclusion, peptide-1 was identified as a potential METTL3-targeting peptide through virtual screening followed by experimental validation. peptide-1 showed detectable binding affinity towards METTL3, whereas the scrambled peptide exhibited no obvious binding signal, suggesting that this interaction may be related to its specific sequence. Computational analyses further suggested that peptide-1 might form a relatively stable binding-related conformation with METTL3. In lung cancer cells, peptide-1 displayed growth-inhibitory activity and showed partial nuclear-region fluorescence distribution, and its treatment was associated with reduced colony formation, migration and invasion, G2/M-phase cell-cycle accumulation, and apoptosis-related changes. Mechanism-related assays indicated that peptide-1 may interfere with the METTL3–METTL14 interaction, reduce cellular m6A levels, and downregulate JUNB mRNA expression, while METTL3 knockdown attenuated its antiproliferative effect. Overall, these results suggest that peptide-1 may represent a potential METTL3-targeting peptide and provide a reference for the further optimisation and evaluation of METTL3/m6A-related peptide inhibitors. Discussion Taken together, the present results suggest that peptide-1 may exert its effects in lung cancer cells through a METTL3-related mechanism. Computational analyses first suggested that peptide-1 might bind to the METTL3 interaction interface and form a relatively stable binding-related conformation, which was consistent with the METTL3 binding affinity observed in the MST assay. On this basis, the cellular NanoBRET results further showed that peptide-1 reduced the signal associated with the METTL3–METTL14 interaction, suggesting that it may affect the METTL3–METTL14 interaction and/or spatial proximity. Consistent with this observation, peptide-1 treatment decreased cellular m6A levels and reduced JUNB mRNA expression, suggesting that METTL3–METTL14-associated changes may be linked to downstream molecular responses. These molecular changes were accompanied by reduced proliferation, colony formation, migration and invasion, G2/M-phase cell-cycle accumulation, and apoptosis-related alterations in A549 cells. Furthermore, the attenuation of peptide-1-mediated growth inhibition after METTL3 knockdown also suggests that its cellular activity may be at least partly associated with METTL3 expression. However, the current results are still insufficient to fully clarify the detailed process by which peptide-1 affects cellular m6A levels, JUNB mRNA expression, and cellular phenotypic changes. Therefore, these findings are more appropriately interpreted as suggesting that peptide-1 may target METTL3 and interfere with the METTL3–METTL14 interaction. This effect may further influence m6A-related cellular responses and contribute to the observed changes in lung cancer cell phenotypes. The proposed working model is summarised in Figure S5. It should also be noted that Kd and IC50 values are derived from different experimental systems and therefore should not be treated as directly interchangeable parameters61. The Kd values measured by MST mainly reflect the apparent in vitro binding affinity between the peptides and METTL3 under defined assay conditions. In contrast, the IC50 values obtained from MTT assays represent cell-based phenotypic responses, which may be influenced by peptide uptake, intracellular distribution, peptide stability, target accessibility, and downstream cellular effects62,63. Therefore, the relationship between METTL3-binding affinity and cellular growth-inhibitory activity is better viewed as indirect rather than strictly proportional. The previously reported peptide inhibitor RSM3 provides an important reference for the present study31, particularly because it was rationally designed from the METTL3–METTL14 interaction interface and further optimised through stapling to improve its helical conformation and pharmacological properties. peptide-1 differs from RSM3 in both origin and structure, as it was identified from a virtual hexapeptide library and remains a short linear peptide without conformational stapling. Under the present experimental conditions, peptide-1 showed favourable binding to METTL3 and a lower IC50 value in A549 cells than the reference peptide RSM3, suggesting that this structurally simpler peptide may have relatively favourable in vitro activity in the current assay system. However, these findings mainly reflect the activity profile of peptide-1 within the current experimental system and should not be taken to indicate that it is overall superior to RSM3, since the two peptides are not directly equivalent in sequence length, structural modification, conformational stability, or stage of validation. At present, the mechanistic evidence for RSM3 is more extensive, as previous studies have linked its activity to disruption of the METTL3–METTL14 complex, METTL3 degradation, reduced global m6A modification, and tumour growth inhibition in vivo. In comparison, the present results suggest that peptide-1 may bind METTL3, reduce the METTL3–METTL14 NanoBRET signal, decrease cellular m6A levels, affect JUNB mRNA expression, and induce apoptosis-related changes; meanwhile, METTL3 knockdown attenuated its antiproliferative effect. These observations suggest that peptide-1 may act through a METTL3-related mechanism, although its detailed mechanism of action and in vivo activity remain to be further clarified. Therefore, peptide-1 may represent a shorter and structurally simpler METTL3-targeting peptide for further optimisation, whereas RSM3 remains an important reference with stronger evidence for peptide stabilisation, degradation-related mechanism, and in vivo activity. Based on the above differences, further direct mechanistic studies are still needed to more clearly evaluate whether peptide-1 affects METTL3–METTL14 complex formation and related molecular processes. In addition, the in vivo activity of peptide-1 may be examined in lung cancer xenograft models and compared with vehicle, scrambled peptide, and RSM3 controls. In this context, analyses of tumour growth, systemic toxicity, pharmacokinetic properties, METTL3-related and m6A-related signalling in tumour tissues, as well as proliferation and apoptosis markers, would allow a more comprehensive assessment of the in vivo activity of peptide-1 and its METTL3-related mode of action. These studies may also provide a basis for the subsequent structural optimisation and pharmacological evaluation of peptide-1. Supplementary Material Supplementary material for review.docx Funding Statement This work was supported by Taizhou Science and Technology Support Program (grant TS202412). Disclosure statement The authors report no conflicts of interest. Data availability statement The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author. References Associated Data Supplementary Materials Supplementary material for review.docx Data Availability Statement The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.