In silico repurposing of clinically approved drugs as potential BRAF inhibitors: a pharmacophore-based virtual screening and experimental investigation Authors: Hooriehsadat Amini, Negin Ebrahimi, Sayedali Mousavirahimi, Mozhdeh Zamani, Kowsar Bagherzadeh, Hamed Montazeri, Homa Azizian Published: 2026 ABSTRACT The BRAFV600E mutation is a key oncogenic driver in multiple cancers. While FDA-approved BRAF inhibitors exist, their efficacy is often limited by drug resistance and adverse effects, necessitating the discovery of novel therapeutic agents. This study aimed to computationally repurpose clinically approved drugs as potential BRAF inhibitors. A ligand-based pharmacophore model was generated using five known BRAF inhibitors and validated to identify a robust three-feature hypothesis (one hydrogen bond acceptor and two aromatic rings). This model was used to screen the E-Drug3D database of FDA-approved drugs, yielding 666 initial hits. Subsequent structure-based virtual screening, molecular docking, and molecular dynamics simulations against both wild-type BRAF and BRAFV600E identified several promising candidates. Ticagrelor emerged as a top hit, demonstrating a stable binding mode and favorable molecular dynamics profile comparable to the reference inhibitor regorafenib. Experimental validation in colorectal cancer cell lines with varying BRAF status revealed that ticagrelor, along with other identified drugs like pioglitazone and nilotinib, exhibited significant anti-proliferative and pro-apoptotic effects. These findings suggest that ticagrelor is a promising repurposing candidate for BRAF inhibition and provide a foundation for overcoming current challenges in targeted cancer therapy.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s12935-026-04378-0. Introduction Malignancies arise from gene mutations that disrupt key cellular processes including proliferation, differentiation, and apoptosis. One of the most frequently dysregulated signaling cascades in human cancers is the RAS/RAF/MEK/ERK/MAP kinase pathway, which plays a central role in transducing extracellular growth signals to intracellular targets [1]. This pathway governs numerous physiological responses, making its proper regulation essential for maintaining cellular homeostasis [2]. Among the components of this cascade, RAF kinases—particularly BRAF—are pivotal in controlling cell survival, growth, differentiation, and apoptosis. BRAF functions as an oncogene due to its direct interaction with RAS and growth factor receptors, which drive oncogenic transformation when aberrantly activated [3]. Mammals express three RAF isoforms—ARAF, BRAF, and CRAF—each containing three conserved regions: CR1 (a regulatory RAS-GTP binding domain), CR2 (a serine-rich hinge region), and CR3 (a catalytic serine/threonine kinase domain with a bilobed structure) [4–6]. While all RAF kinases contribute to physiological signaling, mutations in BRAF are particularly common in various cancers, including non-small cell lung cancer (~ 5–8%), colorectal cancer (~ 15%), melanoma (~ 60%), and thyroid cancer (~ 60%) [6]. BRAF typically forms dimers and functions as a serine/threonine kinase. The full-length protein comprises 766 amino acids, with CR1 spanning residues 120–280. Under normal conditions, CR1 autoinhibits CR3 to prevent unwanted activation. Additionally, the RAS-binding domain (RBD) and zinc finger motif facilitate membrane docking by associating with RAS-GTP [7–9]. Over 50 unique BRAF mutations have been identified, primarily in the kinase domain, activation segment, and glycine-rich loop [10]. The most prevalent mutation, V600E, results in constitutive activation, driving unregulated cell proliferation and cancer progression [11, 12]. Structural analyses of wild-type BRAF and BRAFV600E, including its complex with the kinase inhibitor sorafenib, have provided valuable insights into how this mutation activates the kinase [11, 12]. Mutations at position V600 typically involve substitution with bulky, charged residues, such as glutamic acid (V600E), lysine (V600K), or aspartic acid (V600D), frequently detected in tumors [13]. facilitated by advances in understanding BRAF structure and function. Sorafenib was the first BRAF inhibitor to enter clinical trials for melanoma. Since then, several potent ATP-competitive inhibitors targeting the catalytic domain have been developed [14]. FDA-approved drugs such as vemurafenib, dabrafenib, and encorafenib, have demonstrated clinical efficacy in metastatic melanoma. Nevertheless, treatment limitations including drug resistance and adverse effects highlight the ongoing need for novel, more selective inhibitors [15–18]. Despite initial therapeutic successes, resistance remains a major obstacle in treating BRAFV600E -driven cancers, especially in colorectal and thyroid cancers, where response rates are often low [19–21]. BRAF inhibitors act by stabilizing inactive kinase conformations, which can be classified based on the orientation of key structural motifs: type I (αC-helix-in / DFG-in), type II (αC-helix-in/ DFG-out), and type I½ (αC-helix-out/ DFG-in). Vemurafenib and dabrafenib, for example, are type I½ inhibitors that occupy the ATP-binding pocket and lock the kinase in an inactive conformation [22–27]. To identify next-generation BRAF inhibitors, we employed an integrated computational pipeline combining pharmacophore modeling, molecular docking, and molecular dynamics simulations (MDs). Ligand-based pharmacophore models were constructed and validated using known BRAF inhibitors, enabling virtual screening of chemical libraries for novel scaffolds. Promising hits were further evaluated through docking and MDs to assess binding affinity and dynamic stability. This strategy facilitated the discovery of potential BRAF inhibitors with improved binding characteristics, offering a basis for the development of more effective anticancer therapeutics. Materials and methods All in silico studies were performed using Maestro (version 11.8), developed by Schrödinger, LLC. The workflow included the use of Protein Preparation Wizard [28], LigPrep [29], Pharmacophore Hypothesis Development (Phase) [30], Induced fit docking [31], Virtual Screening Workflow [32], MM-GBSA calculations (Prime), and molecular dynamics simulations (Desmond) [33]. All simulations and computations were carried out on a Linux Ubuntu × 86–64 platform. Pharmacophore generation and validation Pharmacophore modeling was conducted using the Phase module. Multiple FDA-approved BRAF inhibitors; dabrafenib, encorafenib, regorafenib, sorafenib, and vemurafenib were selected as reference ligands to generate pharmacophore hypotheses for the BRAF kinase (Fig. 1). Prior to modeling, all ligand structures were prepared using the LigPrep tool, with parameters configured to apply the OPLS_2005 force field, generate all possible stereoisomers, assign chirality from 3D structures, and retain default settings for other options.Fig. 1The 2D structures of the FDA-approved BRAF inhibitors The 2D structures of the FDA-approved BRAF inhibitors In Phase, pharmacophore features including hydrogen bond acceptors (A), hydrogen bond donors (D), hydrophobic groups (H), negatively charged groups (N), positively charged groups (P), and aromatic rings (R) were defined using default settings. The pharmacophore model represents the essential three-dimensional spatial arrangement of functional groups necessary for molecular recognition by the receptor [34]. For hypothesis generation, three to seven features were selected, with each feature appearing in at least 25% of the ligands used [35, 36]. To assess the predictive performance of the pharmacophore hypotheses, a validation process was carried out using an internal library consisting of 10,098 decoy compounds from the DUD-E dataset (https://dude.docking.org/) [37], along with 251 known BRAF inhibitors and 25 additional active ligands retrieved from the ChEMBL database (https://www.ebi.ac.uk/chembl/) [38]. All compounds were prepared using LigPrep. The ability of each hypothesis to discriminate active compounds from decoys was evaluated through receiver operating characteristic (ROC) curve analysis. The pharmacophore model with the best performance was selected for subsequent virtual screening. Pharmacophore screenin The E-Drug3D database (https://chemoinfo.ipmc.cnrs.fr/MOLDB/index.php) [39], comprising 2028 FDA-approved small molecules and their active metabolites (approved between 1939 to 2021) with molecular weights under 2000 Da, was used in this study. Prior to screening, the molecular structures were prepared using the LigPrep tool applying the OPLS_2005 force field, generating all possible stereoisomers, assigning chirality based on the 3D structures, and maintaining default settings for all other parameters. After database preparation, the most validated BRAF pharmacophore hypothesis was applied for ligand-based screening using the Phase module in Maestro (version 12.8). The resulting hits were selected for further analysis in the virtual screening workflow. Virtual screening workflow The crystal structures of human wild-type BRAF (PDB ID: 6N0Q), co-crystallized with N-(4-methyl-3-(1-methyl-2-oxo-2,3-dihydro-1H-benzo[d]imidazol-5-yl)phenyl)-3-(trifluoromethyl) benzamide (K7S801) and the BRAFV600E (PDB ID: 6P3D), bound to ponatinib (0LI1001) were retrieved from Protein Data Bank (http://www.rcsb.org) [40]. Structural preparation was carried out using the Protein Preparation Wizard, included the addition of missing side chains and loops, removal of crystallographic water molecules, assignment of protonation states, and energy minimization using the OPLS_2005 force field. To enhance the reliability of docking predictions, a hierarchical docking strategy was employed using High-Throughput Virtual Screening (HTVS), Standard Precision (SP), and Extra Precision (XP) protocols. Ligand flexibility was fully enabled throughout. HTVS generated up to 10 poses per compound, retaining the top 80% of hits for SP docking. SP docking also produced up to 10 poses per ligand, and the best scoring poses were forwarded to XP docking. The XP protocol, combined with MM-GBSA calculations, was then used to estimate binding free energies, generating up to three poses per compound and retaining the top 10% of candidates for further analysis. Cell line and culture conditions The Colo205, SW48 and HT-29 colon cancer cell lines were obtained from Iranian Biological Resource Center (IBRC, Tehran, Iran). HT-29 and Colo205 cells were cultured in DMEM supplemented with 10% heat-inactivated fetal bovine serum (FBS) and 1% penicillin–streptomycin mixture at 37°C in a humidified atmosphere containing 5% of CO2. SW48 cells were maintained under the same conditions except RPMI 1640 medium was used instead of DMEM. All cell culture reagents were sourced from Bioidea (Iran) and SPL Life Sciences Co. (South Korea). Cell viability assay The cytotoxic potential of the investigated drugs was assessed using the MTT assay (Sigma). Briefly, cells were seeded in a 96-well culture plates at a density of 1 × 104 cells per well. After 24 h, the cells were treated with various concentrations of the selected drug. Following 48 h post-treatment, 20 μL MTT solution (5 mg/mL in PBS) was added to each well and the plates were incubated in the dark for 4 h. Subsequently, 60 μL of DMSO was added to each well to fully dissolve the resulting formazan crystals. The absorbance was measured at 570 nm with a reference wavelength of 690 nm using a BioTek microplate reader (Synergy H1, USA). Cell cycle analysis To analyze the cell cycle distribution following treatment, the cells, 5 × 10 5 cells were seeded into each well of a 6-well plate. After 48 h of treatment, the cells were harvested and fixed in 70% ethanol for 1 h. The fixed cells were then washed with PBS and resuspended in 1 mL of PBS containing 1 mg/mL of RNase A (previously boiled; Sigma) and 50 μg/mL of propidium iodide (PI; Sigma). Samples were incubated in the dark at room temperature for 30 min prior to cell analysis. Cell cycle distribution was assessed using a FACS Calibur flow cytometer (BD Biosciences). Statistical analysis Data were analyzed using GraphPad prism® (v8). Comparisons between groups were performed using one-way analysis of variance (ANOVA) followed by Dunnett’s post hoc test, or Student’s t-test where appropriate (treated vs. untreated). A p-value < 0.05 was considered statistically significant. Results are presented as the mean ± standard deviation (SD) from at least three independent experiments. Molecular dynamic simulations Molecular dynamics simulations (MDs) were carried out using the Desmond v5.3 module integrated within the Maestro interface [33]. Suitable docking poses were selected based on the Induced Fit Docking (IFD) protocol, with the top 10 poses generated using Van der Waals radii of 0.7 for the receptor and 0.5 for the ligand in Glide [31]. The resulting protein–ligand complexes were solvated in an explicit SPC water model and embedded within an orthorhombic box under periodic boundary conditions. To mimic physiological conditions, the systems were neutralized with counter-ions and supplemented with 0.15 M NaCl. The MDs protocol involved energy minimization, equilibration and production phases. Initial minimization was performed using the steepest descent algorithm for 2500 steps. Subsequently, the system temperature was gradually increased from 0 to 300 K while applying positional restraints to the protein to avoid structural distortion. Production simulations were conducted in the NPT ensemble at 300 K and 1.01325 bar, with temperature and pressure maintained using the Nose–Hoover chain thermostat (1.0 ps relaxation time) and the Martyna–Tobias–Klein barostat (2.0 ps relaxation time), respectively, employing isotropic coupling. Long-range electrostatic interactions were calculated using the Particle Mesh Ewald (PME) method, with a 9.0 Å cutoff for Coulombic interactions. A 50 ns MDs was performed, with trajectory snapshots recorded every 1000 ps. Structural stability and dynamics of the complexes were assessed by calculating root mean square deviation (RMSD) and root mean square fluctuation (RMSF). The final energy-minimized structure from the equilibrated trajectory was analyzed to characterize protein–ligand interactions. Result & discussion Pharmacophore screening Molecular structures of five FDA-approved BRAF inhibitors; dabrafenib, encorafenib, regorafenib, sorafenib, and vemurafenib were utilized to generate pharmacophore hypotheses comprising three to seven features (Fig. 1) The Pharmacophore validation module was used to evaluate the performance of the generated models by calculating the Receiver Operating Characteristic (ROC) curve using a set of 276 BRAF active ligands and 10,098 decoy molecules. Among the generated hypotheses, the best-performing model was a three-feature hypothesis designated as ARR, consisting of one hydrogen bond acceptor (A) and two aromatic ring (R) features (Fig. 2a). This model yielded the highest ROC value of 0.8 (Fig. 2b), indicating a strong ability to discriminate between active and decoy compounds. The validated ARR pharmacophore model was then employed as a query for database screening using the Phase ligand screening module (Maestro v12.8), targeting FDA-approved drugs with comparable pharmacophore features. The screening was conducted on the pre-processed e-Drug3D database [39], comprising 2028 FDA-approved small molecules. The resulting hits were aligned to the pharmacophore hypothesis, and several critical parameters including fitness, site, vector, volume, and alignment scores were computed to assess the quality of the matches.Fig. 2(a)Three-dimensional representation of the highest-ranked pharmacophore hypothesis for BRAF, designated as ARR. The model consists of one hydrogen bond acceptor (pink) and two aromatic ring features (orange), with inter-feature distances shown in angstroms. (b) Receiver Operating Characteristic (ROC) curve corresponding to the ARR hypothesis, demonstrating its ability to distinguish active BRAF ligands from decoys (a)Three-dimensional representation of the highest-ranked pharmacophore hypothesis for BRAF, designated as ARR. The model consists of one hydrogen bond acceptor (pink) and two aromatic ring features (orange), with inter-feature distances shown in angstroms. (b) Receiver Operating Characteristic (ROC) curve corresponding to the ARR hypothesis, demonstrating its ability to distinguish active BRAF ligands from decoys The screening identified 666 compounds with fitness scores ranging from 2.636 to 0.519. Among the known BRAF inhibitors, regorafenib showed the highest fitness score (2.63), followed by sorafenib (2.60), vemurafenib (2.16), dabrafenib (2.11), and encorafenib (1.93). Higher fitness scores indicate a stronger alignment to the pharmacophore model, highlighting the ARR model’s capability to accurately capture key features shared by known BRAF inhibitors. Validation of molecular docking process To ensure the reliability of the docking protocol, re-docking experiments were performed using the co-crystallized ligands 0LI1001 and K7S801 against their respective protein structures, BRAFV600E (PDB ID: 6P3D) and wild-type BRAF (PDB ID: 6N0Q), respectively. The heavy-atom root-mean-square deviation (RMSD) between the crystallographic pose and the re-docked pose was calculated to assess docking accuracy. For 0LI1001 in complex with BRAFV600E, the RMSD was 0.13 Å, well below the 2.0 Å threshold typically used to confirm docking reliability. The Glide XP docking score, MM-GBSA binding free energy, and XP GScore were –13.945, –143.25, and –14.699 kcal/mol, respectively, confirming strong binding affinity and stability. The superimposed binding pose is illustrated in Fig. 3a and detailed protein–ligand interactions are shown in Fig. 3b.Fig. 3(a) Superimposed view of the co-crystallized ligand (green) and re-docked pose (purple) within the active site of BRAFV600E, showing close alignment and validating the docking protocol. (b) Detailed 2D interaction diagram of ligand 0LI1001 with key residues in the active site of BRAFV600E, highlighting hydrogen bonds and hydrophobic contacts (a) Superimposed view of the co-crystallized ligand (green) and re-docked pose (purple) within the active site of BRAFV600E, showing close alignment and validating the docking protocol. (b) Detailed 2D interaction diagram of ligand 0LI1001 with key residues in the active site of BRAFV600E, highlighting hydrogen bonds and hydrophobic contacts Similarly, re-docking of K7S801 to the wild-type BRAF structure yielded an RMSD of 2.1004 Å. This value falls within an acceptable range for docking validation, as RMSD thresholds can be case-dependent and influenced by protein structure and binding site characteristics, particularly for flexible binding sites. The XP docking score, MM-GBSA score, and XP GScore were –11.941, –103.00 kcal/mol, and –11.941, respectively. The superimposed structures and molecular interactions are presented in Fig. 4a and b, respectively.Fig. 4 (a) Superimposed view of the co-crystallized ligand (green) and re-docked pose (purple) within the active site of wild-type BRAF, confirming docking accuracy. (b) Interaction profile of ligand K7S801 with the wild-type BRAF active site, showing key binding interactions including hydrogen bonds, π-π stacking, and hydrophobic contacts (a) Superimposed view of the co-crystallized ligand (green) and re-docked pose (purple) within the active site of wild-type BRAF, confirming docking accuracy. (b) Interaction profile of ligand K7S801 with the wild-type BRAF active site, showing key binding interactions including hydrogen bonds, π-π stacking, and hydrophobic contacts Importantly, analysis of the re-docked poses confirmed the preservation of key interactions within the active site, with no significant displacement affecting the binding mode. Furthermore, the stability of the docked complexes was supported by subsequent molecular dynamics simulations, reinforcing the reliability of the docking protocol for downstream virtual screening. Virtual screening workflow A total of 666 compounds identified from the pharmacophore-based screening were subjected to structure-based virtual screening using the prepared crystal structures of both wild-type BRAF and BRAFV600E, using the Virtual Screening Workflow module in Maestro, which sequentially applies three Glide docking protocols: High-Throughput Virtual Screening (HTVS), Standard Precision (SP), and Extra Precision (XP), followed by post-processing with the Prime Molecular Mechanics/Generalized Born Surface Area (MM-GBSA) method. To further refine binding conformations and account for receptor flexibility, Induced Fit Docking (IFD) [31] was applied to the top-ranked XP-docked ligands. As summarized in Table 1, a total of 25 ligands for BRAFV600E and 23 ligands for wild-type BRAF were prioritized based on their MM-GBSA binding free energies. For BRAFV600E, MM-GBSA scores ranged from –116.48 to –49.66 kcal/mol, while the corresponding range for wild-type BRAF was –116.04 to –26.61 kcal/mol. Additionally, Glide XP docking scores ranged from –13.486 to –9.278 for BRAFV600E and –15.371 to –9.600 for wild-type BRAF, with more negative values indicating stronger predicted binding affinities.Table 1Results of the structure-based virtual screening workflow on BRAFV600E and wild-type BRAFBRAFV600EWild-type BRAFNoDrug nameMM-GBSA score (kcal/mol)Dock scoreNoDrug nameMM-GBSA score(kcal/mol)Dock score1Lapatinib-116.48-10.5221Ripretinib-116.04-15.3712Ticagrelor-108.17-10.0212m-Ripretinib-101.07-10.7453Nilotinib-102.45-12.5173Nilotinib-98.95-12.2384Regorafenib-94.72-9.7634Revefenacin-98.29-10.3225m-Ticagrelor-94.18-9.2785Lapatinib-94.51-11.7616Sorafenib-92.28-9.7006Pioglitazone-90.85-10.7227Entrectinib-90.42-10.0097Salmeterol-88.12-11.0408Salmeterol-90.29-10.2898Canagliflozin-88.12-9.8369Mirabegron-88.21-12.6809Mitoxantrone-87.71-10.34310Carvedilol-82.22-10.06010Carvedilol-86.35-10.16611Axitinib-81.22-9.89411Axitinib-81.52-11.15112Canagliflozin-80.10-9.47412Empagliflozin-74.66-9.67313Labetalol-79.52-9.96713Umeclidinium-72.30-9.62514Vemurafenib-78.67-13.48614Bicalutamide-71.33-10.15215Arbutamine-73.85-11.28115Mebendazole-70.88-10.75816Protokylol-72.17-10.69016Labetalol-69.10-9.80917Dobutamine-71.98-9.42217Riboflavin-63.53-11.46318Vorapaxar-71.29-10.41218Masoprocol-63.50-11.04919Pexidartinib-70.12-10.90619Baricitinib-62.27-9.60020m-Amodiaquine-69.60-11.37220Fenoldopam-57.21-9.80021Darolutamide-68.15-10.33221Antazoline-46.21-10.19722Ritodrine-63.30-9.76022Balsalazide-46.08-9.74023m-Netarsudil-57.00-9.66723Levoleucovorin-26.61-10.74524Masoprocol-49.73-12.06125Balsalazide-49.66-9.906 Table 1Results of the structure-based virtual screening workflow on BRAFV600E and wild-type BRAFBRAFV600EWild-type BRAFNoDrug nameMM-GBSA score (kcal/mol)Dock scoreNoDrug nameMM-GBSA score(kcal/mol)Dock score1Lapatinib-116.48-10.5221Ripretinib-116.04-15.3712Ticagrelor-108.17-10.0212m-Ripretinib-101.07-10.7453Nilotinib-102.45-12.5173Nilotinib-98.95-12.2384Regorafenib-94.72-9.7634Revefenacin-98.29-10.3225m-Ticagrelor-94.18-9.2785Lapatinib-94.51-11.7616Sorafenib-92.28-9.7006Pioglitazone-90.85-10.7227Entrectinib-90.42-10.0097Salmeterol-88.12-11.0408Salmeterol-90.29-10.2898Canagliflozin-88.12-9.8369Mirabegron-88.21-12.6809Mitoxantrone-87.71-10.34310Carvedilol-82.22-10.06010Carvedilol-86.35-10.16611Axitinib-81.22-9.89411Axitinib-81.52-11.15112Canagliflozin-80.10-9.47412Empagliflozin-74.66-9.67313Labetalol-79.52-9.96713Umeclidinium-72.30-9.62514Vemurafenib-78.67-13.48614Bicalutamide-71.33-10.15215Arbutamine-73.85-11.28115Mebendazole-70.88-10.75816Protokylol-72.17-10.69016Labetalol-69.10-9.80917Dobutamine-71.98-9.42217Riboflavin-63.53-11.46318Vorapaxar-71.29-10.41218Masoprocol-63.50-11.04919Pexidartinib-70.12-10.90619Baricitinib-62.27-9.60020m-Amodiaquine-69.60-11.37220Fenoldopam-57.21-9.80021Darolutamide-68.15-10.33221Antazoline-46.21-10.19722Ritodrine-63.30-9.76022Balsalazide-46.08-9.74023m-Netarsudil-57.00-9.66723Levoleucovorin-26.61-10.74524Masoprocol-49.73-12.06125Balsalazide-49.66-9.906 Results of the structure-based virtual screening workflow on BRAFV600E and wild-type BRAF In vitro evaluation of the extracted agents In this study, we employed three colorectal cancer (CRC) cell lines; HT-29 and Colo205, both harboring heterozygous BRAFV600E mutation, and SW-48 which expresses wild-type BRAF. While Lapatinib and Ripretinib showed the most favorable docking scores, they were not advanced to in vitro validation. Docking scores alone were not used as the sole selection criterion. In addition, both compounds are well-characterized kinase inhibitors with established anticancer activity, including reported interactions with related signaling pathways, and were therefore considered to offer limited novelty within a drug repositioning framework. Instead, compounds were prioritized based on a combination of factors, including novelty, chemical diversity, and practical considerations such as availability. Consequently, Ticagrelor, Nilotinib, and Pioglitazone were selected for experimental evaluation. Although the observed anti- proliferative and pro-apoptotic effects of tested compounds are consistent with interference in BRAF/MAPK cascade(s), direct enzymatic inhibition of BRAF kinase activity was not experimentally assessed in this study. Moreover, these repurposed drugs have well known primary targets; such as P2Y12 receptor for ticagrelor, β-adrenergic receptors for salmeterol and carvedilol and PPARγ for pioglitazone. Therefore, off-target effects and promiscuity binding cannot be excluded, particularly at the tested concentrations. Future investigations using direct kinase profiling, cellular thermal shift assays (CETSA), or target engagement assays are warranted to confirm on-target BRAF inhibition. In vitro evaluation of ticagrelor against CRC cell lines with different BRAF mutational status Ticagrelor, a widely used antiplatelet agent, has been previously reported to potentiate the effects of chemotherapeutics such as gemcitabine in pancreatic ductal adenocarcinoma, modulate transcription factors involved in epithelial–mesenchymal transition (EMT), and demonstrate a degree of selectivity toward cancer cells over normal diploid cells [41]. In our MTT assay, ticagrelor exhibited a dose-dependent growth-inhibitory effect across all three CRC cell lines, albeit with varying sensitivity (Fig. 5a). The wild-type BRAF cell line SW-48 showed growth inhibition at a relatively low concentration of 20 μM. In contrast, the BRAFV600E cell lines Colo205 and HT-29 demonstrated sensitivity starting at 30 μM and 75 μM, respectively. The greater resistance observed in HT-29 cells may be attributed to two plausible factors: (i) the presence of an oncogenic PIK3CA mutation, which is known to confer resistance to targeted agents [42], and (ii) the possibility that ticagrelor engages unknown molecular targets at higher concentrations (≥ 75 μM) in HT-29 cells, leading to increased vulnerability compared to Colo205 at 75 μM and to both Colo205 and SW-48 at 100 μM.Fig. 5Cytotoxic effects of selected drugs on three different colorectal cancer (CRC) cell lines. (A) Ticagrelor, (B) Pioglitazone, (C) Nilotinib, (D) Empagliflozin, (E) Salmeterol, and (F) Carvedilol. Cells were treated with increasing concentrations of each compound for 48 h, and cell viability was assessed using the MTT assay. Results are expressed as the percentage of viable cells relative to the untreated control group (vehicle-treated). Data represent the mean ± SD from three independent experiments performed in octuplicate. Calculated IC50 values for each compound and cell line are provided in Supplementary Table S1. Statistical significance: *p < 0.05, ** p < 0.01, and *** p < 0.001 compared with untreated cells. # p < 0.05, ## p < 0.01, and ### p < 0.001 compared with another cell line at the same concentration Cytotoxic effects of selected drugs on three different colorectal cancer (CRC) cell lines. (A) Ticagrelor, (B) Pioglitazone, (C) Nilotinib, (D) Empagliflozin, (E) Salmeterol, and (F) Carvedilol. Cells were treated with increasing concentrations of each compound for 48 h, and cell viability was assessed using the MTT assay. Results are expressed as the percentage of viable cells relative to the untreated control group (vehicle-treated). Data represent the mean ± SD from three independent experiments performed in octuplicate. Calculated IC50 values for each compound and cell line are provided in Supplementary Table S1. Statistical significance: *p < 0.05, ** p < 0.01, and *** p < 0.001 compared with untreated cells. # p < 0.05, ## p < 0.01, and ### p < 0.001 compared with another cell line at the same concentration Interestingly, results obtained from cell cycle analysis revealed a distinct response pattern. Ticagrelor treatment significantly increased the Sub-G1 population, indicative of apoptotic cell death, across all three cell lines (Fig. 6b). Notably, the increase in Sub-G1 fraction was more pronounced in BRAFV600E cells compared to the wild-type BRAF SW-48 cells, supporting our initial hypothesis of mutant-specific vulnerability. Furthermore, in Colo205 cells, ticagrelor treatment led to a marked increase in the G1-phase population relative to untreated controls, suggesting a G1-phase arrest that may reflect BRAF pathway inhibition [43].Fig. 6Cell cycle analysis of SW-48, HT-29, and Colo205 cell lines following treatment with potential BRAF inhibitors for 48 h at their respective IC50 concentrations (see Supplementary Table S1). (A) Control; (B) Ticagrelor; (C) Pioglitazone; (D) Nilotinib; (E) Empagliflozin; (F) Salmeterol; (G) Carvedilol. Cells were treated with each compound for 48 h, stained with propidium iodide, and analyzed by flow cytometry (FACS) to determine the distribution of cells across different cell cycle phases. Histograms represent the mean ± SD of three independent experiments. Statistical significance: # p < 0.05, ## p < 0.01, and ### p < 0.001 compared with another cell line at the ic50 concentration Cell cycle analysis of SW-48, HT-29, and Colo205 cell lines following treatment with potential BRAF inhibitors for 48 h at their respective IC50 concentrations (see Supplementary Table S1). (A) Control; (B) Ticagrelor; (C) Pioglitazone; (D) Nilotinib; (E) Empagliflozin; (F) Salmeterol; (G) Carvedilol. Cells were treated with each compound for 48 h, stained with propidium iodide, and analyzed by flow cytometry (FACS) to determine the distribution of cells across different cell cycle phases. Histograms represent the mean ± SD of three independent experiments. Statistical significance: # p < 0.05, ## p < 0.01, and ### p < 0.001 compared with another cell line at the ic50 concentration In vitro evaluation of pioglitazone against CRC cell lines with different BRAF mutational status Pioglitazone, a well-known antidiabetic agent, functions as an agonist of peroxisome proliferator-activated receptor gamma (PPARγ). Beyond its metabolic effects, pioglitazone has been reported to exert anti-neoplastic activity across various malignancies, including melanoma [44], glioblastoma [45], and hematologic cancers [46]. Based on molecular simulation results suggesting a preferential binding affinity for wild-type BRAF, we evaluated the in vitro effects of pioglitazone on colorectal cancer cell lines with different BRAF statuses. The MTT assay results revealed that pioglitazone inhibits the growth of wild-type BRAF-expressing SW-48 cells more effectively than that of BRAFV600E mutant HT-29 and Colo205 cells, particularly at concentrations ≥ 100 μM (Fig. 5b). At concentrations below 100 μM, however, a comparable degree of growth inhibition was observed in both SW-48 and HT-29 cells, suggesting that BRAF-specific inhibitory effects of pioglitazone emerge at higher concentrations. Cell cycle analysis further clarified these effects (Fig. 6c). Following pioglitazone treatment, HT-29 and SW-48 cells exhibited clear cell cycle perturbations. Notably, in HT-29 cells, pioglitazone induced a significant accumulation in the G1 phase, with minimal increase in the Sub-G1 population, indicating that the primary mechanism of action in this cell line is growth arrest rather than apoptosis. This observation aligns with previous reports that BRAF inhibition can trigger G1-phase arrest [44]. In contrast, no substantial changes in cell cycle distribution were detected in Colo205 cells. These findings collectively suggest that pioglitazone exerts BRAF-dependent growth-inhibitory activity, with preferential effects on wild-type BRAF-expressing cells, in accordance with our computational predictions. In vitro evaluation of nilotinib against CRC cell lines with different BRAF mutational status Nilotinib is a second-generation tyrosine kinase inhibitor that acts through ATP-competitive inhibition and exhibits significantly higher in vitro binding potency for Bcr-Abl compared to its predecessor, imatinib [47]. Beyond Bcr-Abl, nilotinib inhibits a wide range of kinases, including KIT, MAPK, ZAK, and PDGF receptors with lower potency. These features have expanded its potential applications to other malignancies such as gastrointestinal stromal tumors, breast cancer, and melanomas [48]. Based on our molecular simulation data, nilotinib was predicted to exert a greater inhibitory effect on BRAFV600E than on the wild-type form. However, experimental results contradicted this prediction. SW-48, the wild-type BRAF cell line, exhibited the highest sensitivity to nilotinib, as shown by the MTT assay (Fig. 5c). Although Colo205 and HT-29 (both harboring BRAFV600E) demonstrated moderate sensitivity, with growth inhibition starting at 30 μM and 40 μM, respectively, the extent of cytotoxicity observed in SW-48 was significantly greater. These results suggest that nilotinib may inhibit both wild-type and mutant BRAF-expressing cells, but its anti-proliferative effects are more pronounced in wild-type BRAF cells. Cell cycle analysis following nilotinib treatment (Fig. 6d) supported these observations. A substantial increase in the Sub-G1 population, indicative of apoptosis, was observed in SW-48 and Colo205 cells, while HT-29 exhibited a more modest increase. Thus, nilotinib appears capable of inducing apoptosis in all three cell lines, with the strongest apoptotic response seen in SW-48, followed by Colo205. The relative resistance of HT-29 cells to nilotinib-induced apoptosis may be explained by the presence of an oncogenic mutation in the PIK3CA gene [44]. This mutation is known to enhance PI3K activity, promoting AKT phosphorylation, which in turn sustains pro-survival and growth signaling pathways, potentially counteracting the pro-apoptotic effects of nilotinib. In vitro evaluation of empagliflozin against CRC cell lines with different BRAF mutational status Empagliflozin is an anti-hyperglycemic agent belonging to the sodium-glucose cotransporter2 (SGLT2) inhibitors class, approved for the treatment of type 2 diabetes mellitus via an insulin-independent mechanism. Beyond its metabolic effects, empagliflozin has demonstrated in vitro anticancer activity across several cancer models [49, 50]. In the performed MTT analysis we observed SW-48, the wild-type BRAF harboring cell line, sensitive to empagliflozin. However, this effect was only evident at the highest tested concentration of 200 μM (Fig. 5d). In contrast, HT-29 and Colo205 cell lines, both of which harbor the BRAFV600E mutation, did not exhibit significant growth inhibition at any of the tested concentrations. These results are consistent with our molecular docking and simulation data, which predicted a preferential binding of empagliflozin to wild-type BRAF. This selective inhibition gains further relevance when considering that empagliflozin has been reported to inhibit mTOR activity [51]. Both mTOR and wild-type BRAF can be activated downstream of receptor tyrosine kinases such as EGFR [52], suggesting a possible convergent inhibitory effect in SW-48 cells. Cell cycle analysis following empagliflozin treatment (Fig. 6e) revealed a significant increase in the Sub-G1 and G1 populations in SW-48 and HT-29 cells, indicating that empagliflozin can induce both apoptosis and G1 phase arrest in these cell lines. Interestingly, no notable alterations were observed in the Colo205 cell line. These findings suggest that while growth inhibition in SW-48 may be driven by dual targeting of BRAF and mTOR, HT-29 sensitivity could be mediated through mTOR inhibition alone, independent of BRAF status. In vitro evaluation of salmeterol against CRC cell lines with different BRAF mutational status Anti-proliferative effects of long-acting β2-adrenergic receptor agonists, including salmeterol, across various cancer cell lines, have been reported in several studies, [53, 54]. In our current study, MTT assay results revealed that SW-48 (wild-type BRAF) and Colo205, (BRAFV600E) cell lines showed comparable sensitivity salmeterol at the lowest tested concentration of 5μM (Fig. 5e). In contrast, HT-29 cell line, -also harboring BRAFV600E exhibited reduced viability from 50μM, consistent with the presence of an oncogenic PIK3CA mutation known to confer drug resistance. Notably, HT-29 demonstrated the highest sensitivity at 100 μM, which may reflect the inhibition of additional, unidentified molecular targets or signaling pathways specific to this cell line. Cell cycle analysis following salmeterol treatment (Fig. 6f) indicated a marked increase in the Sub-G1 population in both SW-48 and Colo205, indicative of enhanced apoptosis induction. Conversely, no significant alterations in the cell cycle distribution were observed for HT-29 cells, further supporting the hypothesis that this line’s resistance to salmeterol-induced apoptosis may be related to its PIK3CA mutation or alternative survival mechanisms. In vitro evaluation of carvedilol against CRC cell lines with different BRAF mutational status Carvedilol, a non-selective β-adrenoreceptor with known pleiotropic activities has previously been reported to exhibit anti-proliferative effects across various malignancies [55, 56]. In the current study, MTT assay results demonstrated that exerts cytotoxic effects on all three colorectal cancer cell lines examined. As anticipated from our molecular modeling studies, the SW-48 cell line (wild-type BRAF) displayed the highest sensitivity (Fig. 5f). However, this sensitivity was not dose-dependent, suggesting a more complex mechanism of action. In the case of BRAFV600E cell lines, Colo205 exhibited reduced viability beginning at 10 μM, while HT-29 required a higher concentration (75 μM) to elicit similar effects, indicating a greater resistance. The increased resistance in HT-29 cells may be due to the presence of an oncogenic PIK3CA mutation, which has been implicated in therapeutic resistance. Additionally, it is plausible that carvedilol targets yet unidentified molecular components in HT-29 cells at 75 μM and above, contributing to the observed cytotoxicity at higher concentrations when compared to Colo205 and SW-48. Cell cycle analysis (Fig. 5f) further corroborated the cytotoxic effects of carvedilol. A notable increase in the Sub-G1 population, indicative of apoptosis, was observed in both SW-48 and HT-29 cell lines. In contrast, Colo205 did not exhibit significant changes in cell cycle distribution, suggesting resistance to carvedilol-induced apoptosis at the concentrations tested. Molecular dynamic simulations studies Molecular Dynamic simulations were performed to gain insight into the effect of ticagrelor on the BRAFV600E enzyme active site [33]. For this purpose, structural perturbations induced by ticagrelor were evaluated by analyzing root means square deviation (RMSD), root mean square fluctuation (RMSF), and the compound’s effect on the active site. Comparisons were made with the BRAFV600E–regorafenib complex (as a reference inhibitor) and the apo form of BRAFV600E. The enzyme backbone RMSD was monitored over 50 ns of MDs to assess the stability of the protein–ligand complex trajectories (Fig. 7). In all three studied systems, the RMSD value increased during the initial 4 ns and then remained relatively stable up to 25 ns. Between 25 and 40 ns, the RMSD of the BRAFV600E–regorafenib complex fluctuated between 3.0 and 3.5 Å, indicating a higher degree of structural variation compared to the other systems. During the same period, the BRAFV600E–ticagrelor complex maintained an RMSD between 2.5 and 3.0 Å, closely resembling that of the apo-BrafV600E. The systems reached equilibrium state in the final 10 ns of the simulation with RMSD values stabilizing around 2.8 Å for the regorafenib complex, 2.6 Å for the ticagrelor complex, and 2.3 Å for the apo-BRAFV600E. These results suggest that apo-BRAFV600E exhibits the highest structural stability, while ticagrelor induces less deviation than regorafenib, indicating a potentially stable binding mode within the active site.Fig. 7Root mean square deviation (RMSD) of BRAFV600E backbone during a 50 ns MDs. The RMSD trajectories are shown for the BRAFV600E–regorafenib complex (green), the BRAFV600E–ticagrelor complex (red), and the apo-BRAFV600E (yellow) Root mean square deviation (RMSD) of BRAFV600E backbone during a 50 ns MDs. The RMSD trajectories are shown for the BRAFV600E–regorafenib complex (green), the BRAFV600E–ticagrelor complex (red), and the apo-BRAFV600E (yellow) The flexibility of protein structure was evaluated by analyzing the root mean square fluctuation (RMSF) of evaluate the fluctuations of the Cα atoms which estimates residue-specific deviations from the average position during the entire simulation period. As it is shown in Fig. 8, the most notable differences in fluctuations occurred within loop regions. Sixteen loops are present in the BRAFV600E structure, and several of them displayed considerable variations in flexibility between the apo form and the complexes with regorafenib and ticagrelor. These regions of interest are highlighted as boxed zones in Fig. 7. Compared to the apo-BRAFV600E, complexation with ticagrelor (green color) generally resulted in reduced RMSF value BRAFV600E with, indicating increased structural rigidity and stability, except in zones 4 and 5, where slight increases were observed. Notably, zone 6 encompasses the DFG motif (residues 594, 595, and 596), a critical regulatory element in kinase function. In this region, both ticagrelor and regorafenib significantly reduced backbone fluctuations relative to the apo structure. Zone 7 includes functionally important residues such as T599 (from the activation segments), V600 (the key V600E mutation site), and S602 (end of the activation segment). RMSF values at these positions were also markedly decreased in the presence of either ligand, supporting their stabilizing effects on the activation segment. Interestingly, residue G606 at the end of zone 7 showed a spike in flexibility for the regorafenib-bound complex compared to the others. Zone 8 containing residues P622, E623 and V624 at the end of the activation segment area followed the same trend, where both ligand-bound complexes exhibited reduced fluctuations relative to the apo form. Overall, these results suggest that ticagrelor mimics the stabilizing effects of the known BRAF inhibitor regorafenib, particularly in regions critical for kinase activation and signaling.Fig. 8Root Mean Square Fluctuation (RMSF) analysis of the BRAFV600E backbone in its apo form (yellow), and in complex with regorafenib (green) and ticagrelor (red) over a 50-ns MDs. RMSF values reflect the flexibility of each residue throughout the simulation. Boxed regions (zones) highlight loop segments with notable differences in flexibility, including functionally critical motifs such as the DFG motif (zone 6) and the activation segment (zones 7 and 8). Both ligands, particularly ticagrelor, reduced fluctuations in these key regions, suggesting enhanced structural stabilization compared to the apo form Root Mean Square Fluctuation (RMSF) analysis of the BRAFV600E backbone in its apo form (yellow), and in complex with regorafenib (green) and ticagrelor (red) over a 50-ns MDs. RMSF values reflect the flexibility of each residue throughout the simulation. Boxed regions (zones) highlight loop segments with notable differences in flexibility, including functionally critical motifs such as the DFG motif (zone 6) and the activation segment (zones 7 and 8). Both ligands, particularly ticagrelor, reduced fluctuations in these key regions, suggesting enhanced structural stabilization compared to the apo form According to Fig. 9b and c, in comparison to Fig. 9a (apo-BRAFV600E) the regions involved in ligand binding exhibited a notable reduction in RMSF values and structural fluctuations. This decrease suggests enhanced local rigidity and stabilization of the protein upon ligand binding, particularly within loop regions that were otherwise more flexible in the apo structure. As summarized in Table 2, twenty-five residues of BRAFV600E interacted with ticagrelor and nineteen with regorafenib during the 50 ns MDs. RMSF analysis of these interacting residues across all three structures—apo, ticagrelor-bound, and regorafenib-bound—revealed that both ligands significantly decreased flexibility at key binding sites. Notably, several of these residues were located in functionally critical regions, including the αC helix (red pocket) and the DFG motif (blue pocket). These findings support the hypothesis that ticagrelor, like regorafenib, effectively stabilizes BRAFV600E by reducing the dynamic motion of essential structural elements. Figure 10 illustrates the detailed binding orientation and key interactions that persisted for more than 30% of the simulation time during the equilibrated phase for BRAFV600E in complex with ticagrelor and regorafenib. These stable interactions highlight the critical structural moieties responsible for anchoring both ligands within the active site. To further assess the strength of these interactions, Prime/MM-GBSA calculations were performed on representative structures obtained from clustering the MDs trajectories. The estimated binding free energies (ΔGbind) for the BRAFV600E/ticagrelor and BRAFV600E/regorafenib complexes were − 108.17 kcal/mol and − 94.72 kcal/mol, respectively. These results indicate that ticagrelor forms a more stable and energetically favorable complex with BRAFV600E compared to the reference inhibitor regorafenib. The spatial arrangement of the two critical structural elements—the DFG motif and the αC Helix—plays a pivotal role in determining the efficacy and classification of BRAF inhibitors. To investigate how ligand binding influences the conformation of these regions, we measured the average distance between them over three MDs intervals: 0 ns (initial structure), 25–40 ns (mid-equilibrated phase), and 40–50 ns (final equilibrated phase). As shown in Fig. 11 and Fig. 12, a reduction in the distance between the DFG motif and the αC Helix is associated with effective BRAFV600E inhibition, reflecting a more compact and less accessible active site. While regorafenib induced a greater decrease in this distance compared to ticagrelor, ticagrelor still caused a notable reduction relative to the apo-BRAFV600E structure. This suggests that ticagrelor contributes to structural stabilization of the kinase domain and supports its inhibitory potential against BRAFV600E.Fig. 9Ligand binding location of BRAFV600E in bound and unbound states over a 50 ns MDs time. Ligand binding location of BRAFV600E in bound and unbound states over a 50 ns MDs. (a) Apo-BRAFV600E (unbound), (b) BRAFV600E complexed with regorafenib, and (c) BRAFV600E complexed with ticagrelor. Highly flexible regions corresponding to the αC Helix and the DFG motif are highlighted in yellow and orange, respectively (c)Table 2The RMSF values of BRAFV600E residues interacting with ticagrelor and regorafenib during 50 ns MDsRMSF (Å)NoResidueapo-BRAFV600EBRAFV600E /RegorafenibBRAFV600E /Ticagrelor1ILE 4631.261.021.172PHE 4682.471.221.403VAL 4711.100.710.834ALA 4810.810.560.575LYS 4831.010.540.706PHE 4981.69-1.187ASN 5001.86-1.348GLU 5011.700.631.339VAL 5021.21-1.1710VAL 5041.090.721.0711LEU 5050.840.610.8812ILE 5130.760.460.5513LEU 5140.800.480.5814ILE 5270.91-0.6615THR 5290.62-0.4516TRP 5310.850.7107517CYS 5321.031.10-18LEU 5670.820.500.6219ILE 5720.97-0.6520ILE 5730.88-0.5921HIS 5740.860.490.5422ARG 5750.86-0.6023PHE 5830.640.590.5424ILE 5920.67-0.4825ASP 5941.680.560.6026PHE 5951.590.780.7527GLY 5962.331.30-28LEU 5972.081.37-The RMSF values of the residues were analyzed and compared across three structures: apo-BRAFV600E, BRAFV600E–ticagrelor, and BRAFV600E–regorafenib complexes. Residues highlighted in red correspond to the αC Helix, while those in blue represent the DFG motifFig. 10Detailed orientation and ligand atom interactions that occurred for more than 30% of the simulation time during the equilibrated phase of the 50 ns MDs for BRAFV600E in complex with ticagrelor and regorafenib. These interactions highlight the key structural features contributing to the stabilization of each ligand within the active siteFig. 11Average distances between residues of the DFG motif and the αC Helix in BRAFV600E during three-time intervals (0 ns, 25–40 ns, and 40–50 ns) of the 50 ns MDs. The figure compares the apo- BRAFV600E structure with BRAFV600E complexed with regorafenib and ticagrelor, indicating how inhibitor binding influences the spatial relationship between these two functional regionsFig. 12Close-up representation of the BRAFV600E active site highlighting the spatial relationship between the αC Helix, DFG, HRD motifs and Glycine-rich loop Ligand binding location of BRAFV600E in bound and unbound states over a 50 ns MDs time. Ligand binding location of BRAFV600E in bound and unbound states over a 50 ns MDs. (a) Apo-BRAFV600E (unbound), (b) BRAFV600E complexed with regorafenib, and (c) BRAFV600E complexed with ticagrelor. Highly flexible regions corresponding to the αC Helix and the DFG motif are highlighted in yellow and orange, respectively (c) Table 2The RMSF values of BRAFV600E residues interacting with ticagrelor and regorafenib during 50 ns MDsRMSF (Å)NoResidueapo-BRAFV600EBRAFV600E /RegorafenibBRAFV600E /Ticagrelor1ILE 4631.261.021.172PHE 4682.471.221.403VAL 4711.100.710.834ALA 4810.810.560.575LYS 4831.010.540.706PHE 4981.69-1.187ASN 5001.86-1.348GLU 5011.700.631.339VAL 5021.21-1.1710VAL 5041.090.721.0711LEU 5050.840.610.8812ILE 5130.760.460.5513LEU 5140.800.480.5814ILE 5270.91-0.6615THR 5290.62-0.4516TRP 5310.850.7107517CYS 5321.031.10-18LEU 5670.820.500.6219ILE 5720.97-0.6520ILE 5730.88-0.5921HIS 5740.860.490.5422ARG 5750.86-0.6023PHE 5830.640.590.5424ILE 5920.67-0.4825ASP 5941.680.560.6026PHE 5951.590.780.7527GLY 5962.331.30-28LEU 5972.081.37-The RMSF values of the residues were analyzed and compared across three structures: apo-BRAFV600E, BRAFV600E–ticagrelor, and BRAFV600E–regorafenib complexes. Residues highlighted in red correspond to the αC Helix, while those in blue represent the DFG motif The RMSF values of BRAFV600E residues interacting with ticagrelor and regorafenib during 50 ns MDs The RMSF values of the residues were analyzed and compared across three structures: apo-BRAFV600E, BRAFV600E–ticagrelor, and BRAFV600E–regorafenib complexes. Residues highlighted in red correspond to the αC Helix, while those in blue represent the DFG motif Detailed orientation and ligand atom interactions that occurred for more than 30% of the simulation time during the equilibrated phase of the 50 ns MDs for BRAFV600E in complex with ticagrelor and regorafenib. These interactions highlight the key structural features contributing to the stabilization of each ligand within the active site Average distances between residues of the DFG motif and the αC Helix in BRAFV600E during three-time intervals (0 ns, 25–40 ns, and 40–50 ns) of the 50 ns MDs. The figure compares the apo- BRAFV600E structure with BRAFV600E complexed with regorafenib and ticagrelor, indicating how inhibitor binding influences the spatial relationship between these two functional regions Close-up representation of the BRAFV600E active site highlighting the spatial relationship between the αC Helix, DFG, HRD motifs and Glycine-rich loop Notably, the concentrations required to achieve significant growth inhibition in our in vitro experiments exceed the therapeutic plasma levels attained with standard clinical dosing of these drugs for their original indications. This is a common observation in drug-repurposing studies and should be explained as providing preliminary biological conceptual validation for potential BRAF related anti-cancer activity rather than immediate clinical translatability. The relatively high concentrations may reflect moderate binding affinity toward BRAF, contribution of off-target mechanisms, or poor cellular permeability. These findings support further medicinal chemistry optimization, structure activity relationship (SAR) studies, or advanced formulation strategies to enhance potency and achieve therapeutically relevant plasma level. From a mechanistic perspective, the identified compounds, particularly ticagrelor, may exhibit binding characteristics distinct from classical BRAF inhibitors. Clinically approved BRAF inhibitors are typically classified as type I or type I½ inhibitors, which target specific conformational states of the kinase domain. These agents are associated with resistance mechanisms arising from pathway reactivation and conformational flexibility within the MAPK signaling cascade [12, 57]. In contrast, the binding behavior observed in the present study suggests that repurposed compounds may engage the kinase domain differently, potentially offering alternative interaction profiles that could be less susceptible to conventional resistance pathways. Furthermore, these findings open the possibility for combinational therapeutic strategies. In colorectal cancer, BRAF inhibition alone has shown limited efficacy due to compensatory activation of parallel signaling pathways, particularly through EGFR [57]. Therefore, combining such repurposed agents with anti-EGFR therapies or MEK inhibitors may enhance therapeutic outcomes and overcome resistance mechanisms. Finally, while the current study provides initial evidence of anti-proliferative and pro-apoptotic effects, further validation is required to confirm direct BRAF targeting. Future studies should include biochemical kinase inhibition assays, target engagement analyses, and in vivo validation using appropriate animal models to establish the therapeutic potential and mechanistic basis of these compounds. Conclusion To identify effective BRAF inhibitors, a pharmacophore-based drug design study was conducted employing in silico approaches to screen FDA-approved drugs for their potential to inhibit BRAFV600E. Following pharmacophore hypothesis generation, virtual screening, molecular docking, molecular dynamic simulations (MDs), ticagrelor emerged as a promising candidate for further evaluations. Experimental validation in colorectal cancer cell lines with varying BRAF status revealed that ticagrelor, along with other identified drugs like pioglitazone and nilotinib, exhibited significant anti-proliferative and pro-apoptotic effects. These findings suggest that ticagrelor is a promising repurposing candidate for BRAF inhibition and provide a foundation for overcoming current challenges in targeted cancer therapy. The efficacy of ticagrelor was further assessed through RMSD, RMSF, and interaction analysis over BRAFV600E active site, in comparison to regorafenib, a standard BRAF inhibitor, and the apo-BRAFV600E. During the final 10 ns of the MDs, the RMSD values of BRAFV600E complexed with regorafenib, ticagrelor, and apo-BRAFV600E stabilized at approximately 2.8 Å, 2.6 Å, and 2.3 Å, respectively. RMSF analysis revealed that both regorafenib and ticagrelor reduced backbone fluctuations across critical regions, particularly within the DFG motif and αC Helix, compared to apo-BRAFV600E. These reductions in flexibility indicate that ticagrelor stabilizes the active site similarly to regorafenib, suggesting a comparable inhibitory mechanism. Furthermore, the inter-residue distance between the DFG motif and the αC Helix—an important determinant of BRAF inhibition was reduced upon ticagrelor binding, albeit not to the same extent as with regorafenib. Nevertheless, this distance contraction was significantly greater than in the apo form, supporting the potential of ticagrelor as a BRAFV600E inhibitor. Supplementary Information Additional file 1. Additional file 1.