0% READ
FULL TEXT
INTRODUCTION
Breast cancer remains one of the foremost health challenges worldwide, with hormone receptor-positive subtypes accounting for the majority of diagnosed cases (Schettini et al., 2020). The development of cyclin-dependent kinase 4 and 6 (CDK4/6) inhibitors has fundamentally altered the therapeutic landscape for these patients, offering remarkable improvements in progression-free survival when combined with endocrine therapy (Piezzo et al., 2020; Zhou et al., 2023). However, the clinical success of current CDK4/6 inhibitors has been tempered by persistent challenges, including acquired drug resistance (Shanabag et al., 2025), dose-limiting toxicities that necessitate treatment modifications (Loibl and Furlanetto, 2022; Stanciu et al., 2023), and the absence of true isoform selectivity between CDK4 and CDK6 (Watt and Goel, 2022). These limitations underscore the urgent need for next-generation inhibitors with improved selectivity and the ability to circumvent resistance mechanisms.
The molecular basis for CDK4/6 inhibitor activity centres on preventing Retinoblastoma (Rb) protein phosphorylation, thereby blocking cell cycle progression from G1 to S phase (Topacio et al., 2019; Kim et al., 2022). While palbociclib, ribociclib, and abemaciclib have demonstrated clinical efficacy in the metastatic setting, their combined inhibition of CDK4 and CDK6 contributes to dose-limiting hematological adverse events, primarily neutropenia (Sherr et al., 2016; Scheicher et al., 2015). The rationale for targeting CDK6 in breast cancer derives not from an expectation of an inherent safety advantage, since CDK6 activity in normal haematopoietic cells is itself a recognised contributor to this toxicity, but from the documented overexpression and amplification of CDK6 in specific aggressive subtypes, in which tumour cells display marked dependency on CDK6-driven cell cycle progression (Tadesse et al., 2015). Identifying structurally distinct scaffolds that engage CDK6 therefore remains of interest, with the recognition that any safety differentiation relative to approved CDK4/6 inhibitors would require direct experimental verification.
Spiropyran derivatives represent an underexplored scaffold in kinase inhibitor development. These photochromic compounds possess a unique structural architecture characterised by a spiro junction connecting indoline and chromene moieties, creating a three-dimensional molecular framework amenable to pharmacological optimisation (Fagan et al., 2021). Despite extensive investigation in materials science and sensor applications, their potential as kinase inhibitors has not been systematically explored (Ali et al., 2020). The structural diversity accessible through synthetic modification of the spiropyran core makes this scaffold particularly attractive for structure-activity relationship studies aimed at achieving kinase selectivity (Keyvan Rad et al., 2022).
Computational modelling studies suggest that appropriately substituted spiropyran derivatives can exploit subtle differences in the ATP-binding site architecture between CDK4 and CDK6 (Durrant and McCammon, 2011; Li et al., 2023), potentially exhibiting computationally predicted differential binding toward CDK6 that would require direct comparative kinase assays to be established as true selectivity. This study describes the design, synthesis, and biological evaluation of novel spiropyran derivatives (S1-S8) as putative CDK6-targeting anticancer agents for breast cancer treatment. We hypothesised that structural modifications of the spiropyran core would enhance CDK6 binding affinity while maintaining favourable drug-like properties. The novelty lies in employing photoswitchable spiropyran scaffolds for CDK engagement, validated through comprehensive molecular dynamics simulations and MM-GBSA calculations.
MATERIALS AND METHODS
Materials
2,3,3-Trimethyl-3H-indole (98%) was procured from Sciquaint Chemicals (Pune, India). Substituted 2-hydroxybenzaldehydes (99%) were obtained from HiMedia Laboratories (Mumbai, India). Anhydrous ethanol (99.5%, HPLC grade) and piperidine (99%) were sourced from Neeta Chemicals (Pune, India). MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] (98%) and DMSO (99.9%) were purchased from Research Lab Fine Chem Industries (Pune, India). Palbociclib reference standard (99%, MW 447.54 g/mol) was obtained from Sciquaint Innovations Pvt. Ltd., (Pune, India). MCF-7 breast cancer cell line was procured from the National Centre for Cell Science (NCCS, Pune, India). DMEM culture medium, Fetal Bovine Serum (FBS), and phosphate-buffered saline (PBS, pH 7.4) were obtained from SRL Diagnostics (Mumbai, India) and Central Drug House (New Delhi, India), respectively. All other reagents were of analytical grade.
Experimental Design
The study followed a rational drug design approach: eight spiropyran derivatives (S1-S8) were designed based on computational docking against CDK6 (PDB ID: 5L2I), synthesised and characterised spectroscopically, then evaluated computationally (ADME, toxicity, molecular dynamics) and experimentally (MTT cytotoxicity assay). Compounds with docking scores >-9.0 kcal/mol and favourable predicted ADME profiles were prioritised for in vitro evaluation. Sample size of eight derivatives was determined based on systematic variation of the nucleophilic substituent at the N-1 position of the spiropyran core to establish structure-activity relationships.
Drug-Likeness, ADME, and Toxicity Prediction
Pharmacokinetic properties and drug-likeness were evaluated using ForceADME (Geinforce Technology Pvt. Ltd., Pune, India; https://geinforce.com/force-adme/). SMILES notations generated in ChemDraw were submitted for Lipinski’s Rule of Five assessment, oral bioavailability prediction, intestinal absorption, Blood-Brain Barrier (BBB) penetration, P-glycoprotein substrate recognition, cytochrome P450 inhibition profiles, and synthetic accessibility scores. Toxicity profiles were assessed using Geintox-II (Geinforce Technology Pvt. Ltd., Pune, India; https://geinforce.com), a computational toxicology platform employing machine learning algorithms trained on curated toxicological databases (Bhandari et al., 2026). Endpoints evaluated included acute oral toxicity (LD₅₀, toxicity class), hepatotoxicity, carcinogenicity, mutagenicity, immunotoxicity, cytotoxicity, respiratory toxicity, and blood-brain barrier penetration.
Molecular Docking
Molecular docking was performed using GeinDock Suite v2.3.1.0 (Geinforce Technology Pvt. Ltd., Pune, India), with results validated against AutoDock Vina (correlation coefficient r²=0.89). The CDK6 crystal structure (PDB ID: 5L2I) was prepared by removal of water molecules and heteroatoms, followed by addition of polar hydrogens and Gasteiger charges. Spiropyran derivatives were energy-minimised using the MMFF94 force field for 2500 steps. The docking grid was centred at x: -13.7515, y: 28.1005, z: 9.7195 Å, with box dimensions of 25 × 25 × 25 ų. Protocol validation was confirmed by redocking the co-crystallised palbociclib, achieving RMSD<2.0 Å. Three-dimensional interactions and 2D ligand-protein diagrams were generated using BIOVIA Discovery Studio Visualizer v4.5.
Molecular Dynamics Simulation and MM-GBSA Analysis
Molecular Dynamics (MD) simulations of palbociclib-CDK6 and S5-CDK6 complexes were conducted using Desmond v2020.1 (Schrödinger, LLC) with the OPLS-2005 force field and TIP3P water model in a cubic periodic boundary box (Bowers et al., 2006; Shivakumar et al., 2010). The system was neutralised with sodium ions and supplemented with 0.15 M NaCl. Energy minimisation was followed by 10 ns NVT and 12 ns NPT equilibration phases using the Nosé-Hoover thermostat at 300 K and 1 bar (Ivanova et al., 2018). Production simulations were run for 100 ns; stability was monitored via RMSD, RMSF, Radius of gyration (Rg), hydrogen bonding patterns, and Solvent Accessible Surface Area (SASA). The 100 ns duration was justified based on observed RMSD convergence of both complexes by 20 ns, with stable plateau values maintained throughout the production phase. Binding free energies were calculated using the MM-GBSA method via the Prime MM-GBSA module (Dasmahapatra et al., 2022), processing 500 final trajectory frames using thermal_mmgbsa.py (Forouzesh and Mishra, 2021).
Synthesis of Spiropyran Derivatives (S1-S8)
The spiropyran derivatives were synthesised via a multi-step route. 2,3,3-Trimethyl-3H-indole (Sa, 1.0 mmol) was treated with 2-chloroacetyl chloride (1.2 mmol) in anhydrous DMF (5 mL) at 0ºC for 2 hr to afford compound Sb (Mandal et al., 2022). Compound Sb was treated with triethylamine (1.5 mmol) in anhydrous DCM (10 mL) at room temperature for 1 hr to generate compound Sc. The key spiropyran intermediate (Sd) was synthesised by reacting Sc (1.0 mmol) with 2-hydroxy-5-nitrobenzaldehyde (1.0 mmol) in anhydrous ethanol (15 mL) under reflux at 78ºC for 4-6 hr. Final derivatives S1-S8 were obtained by refluxing Sd (1.0 mmol) with appropriate nucleophiles in anhydrous DCM (10 mL) containing pyridine (0.3 mL) at 40-45ºC for 8-12 hr (Hughes-Whiffing and Perry, 2023). The synthesis scheme and proposed mechanism are illustrated in Figures 1 and 2, respectively. All compounds were characterised by IR, ¹H NMR (400 MHz), ¹³C NMR (100 MHz, where applicable), ESI-MS, and elemental analysis (C, H, N).
In vitro Cytotoxicity Assay
Antiproliferative activity of compounds S2, S5, and S7, selected based on docking scores >-9.0 kcal/mol and favourable predicted ADME profiles, along with palbociclib, was evaluated against MCF-7 breast cancer cells using the MTT assay (Ramazi et al., 2023). MCF-7 cells (1×10⁴ cells/well) were seeded in 96-well plates in DMEM supplemented with 10% FBS and incubated at 37ºC in 5% CO₂ for 24 hr (Razak et al., 2019). Cells were treated with concentrations ranging from 1.56 to 100 μM (two-fold serial dilutions) for 72 hr. MTT solution (5 mg/mL, 20 μL/well) was added and incubated for 4 hr; formazan crystals were dissolved in 100 μL DMSO and absorbance measured at 570 nm (reference: 630 nm) using a microplate reader (BioTek, USA) (Ghasemi et al., 2021). Each experiment comprised three independent biological replicates on separate days, each containing technical triplicates (n=3, expressed as Mean ± SD). IC₅₀ values were determined by non-linear regression analysis using GraphPad Prism v9.0.
Statistical Analysis
Cytotoxicity data are expressed as Mean ± SD (n=3 independent experiments). IC₅₀ values with 95% Confidence Intervals (CI) were derived from non-linear regression (GraphPad Prism v9.0). Statistical comparisons between compound IC₅₀ values and palbociclib were performed using one-way Analysis of Variance (ANOVA) followed by Dunnett’s post-hoc test. A p-value<0.05 was considered statistically significant. MM-GBSA binding free energy values are reported as Mean ± SD across 500 trajectory frames.
RESULTS
Drug-Likeness, ADME, and Toxicity Assessment
Drug-likeness and pharmacokinetic profiles of spiropyran derivatives S1-S8 are presented in Table 1. Compound S1 exhibited one violation of Lipinski’s Rule of Five (MlogP=5.36), indicating increased lipophilicity relative to palbociclib (LogP=2.74) (Lipinski, 2004). Hydrogen bond acceptor counts (5-7) and donor counts (0-1) for all derivatives fell within acceptable ranges. Total Polar Surface Area (TPSA) values for compounds S3-S7 (98.98-127.85 Ų) exceeded the optimal range (60-90 Ų), which may limit membrane permeability (Ertl et al., 2000), while S1, S2, S6, and S8 showed more favourable TPSA values (81.91-107.93 Ų), comparable to palbociclib (105.04 Ų). All compounds showed low predicted gastrointestinal absorption and were predicted to penetrate the blood-brain barrier (Table 1), a property warranting in vivo confirmation and consideration of potential central nervous system exposure. Compounds S5 and S7 were identified as P-glycoprotein substrates, similar to palbociclib, potentially affecting cellular accumulation. Toxicity assessment via Geintox-II (Table 2) classified all compounds as Toxicity Class 4 (prediction accuracy 85%). Compounds S3, S4 and S6 showed LD₅₀ values (1151.7-1267.4 mg/kg) marginally lower than palbociclib (1382.5 mg/kg), with all derivatives classified within Toxicity Class 4. All spiropyran derivatives showed active predictions for hepatotoxicity, carcinogenicity, and respiratory toxicity, warranting careful genotoxicity evaluation in preclinical studies (Banerjee et al., 2024). The low gastrointestinal absorption observed for all compounds except S8 represents a limitation that requires formulation strategies, such as nanoparticle-based or prodrug approaches, for oral delivery optimisation (Adon et al., 2021).
| Compound | MLogP | MW (g/mol) | HBA | HBD | Violations | TPSA (Ų) | GI Abs. | BBB Pen. |
|---|---|---|---|---|---|---|---|---|
| S1 | 5.36 | 442.47 | 5 | 0 | 1 | 81.91 | Low | Yes |
| S2 | 3.89 | 441.49 | 5 | 1 | 0 | 84.71 | Low | Yes |
| S3 | 1.58 | 422.44 | 6 | 0 | 0 | 98.98 | Low | Yes |
| S4 | 1.48 | 408.41 | 6 | 0 | 0 | 98.98 | Low | Yes |
| S5 | 1.08 | 470.48 | 7 | 1 | 0 | 114.14 | Low | Yes |
| S6 | 3.22 | 457.49 | 6 | 1 | 0 | 107.93 | Low | Yes |
| S7 | 0.41 | 486.48 | 7 | 1 | 0 | 127.85 | Low | Yes |
| S8 | 3.5 | 428.44 | 5 | 0 | 0 | 84.59 | High | Yes |
| Palbociclib | 2.74 | 447.54 | 9 | 2 | 0 | 105.04 | Low | Yes |
| Compound | LD₅₀ (mg/kg) | Tox. Class | Hepatotox. | Carcinogen. | Mutagenicity | BBB Penetration | Cytotox. |
|---|---|---|---|---|---|---|---|
| Palbociclib | 1382.5 | 4 | I (40%) | I (54%) | A (73%) | A (92%) | I (57%) |
| S1 | 626.4 | 4 | A (99%) | A (99%) | A (99%) | A (99%) | A (71.5%) |
| S2 | 490.4 | 4 | A (99%) | A (99%) | A (99%) | A (99%) | A (71.9%) |
| S3 | 1238.6 | 4 | A (87.5%) | A (99%) | A (99%) | A (99%) | A (74%) |
| S4 | 1267.4 | 4 | A (81.1%) | A (97.4%) | A (97.4%) | A (97.4%) | I (68%) |
| S5 | 571.2 | 4 | A (98.6%) | A (98.6%) | A (98.6%) | A (84.6%) | A (73.2%) |
| S6 | 1151.7 | 4 | A (97.4%) | A (97.4%) | A (97.4%) | A (97.4%) | A (75.7%) |
| S7 | 556.4 | 4 | A (99%) | A (99%) | A (99%) | A (85.1%) | I (68.5%) |
| S8 | 500 | 4 | A (54%) | A (59%) | A (83%) | A (69%) | A (56%) |
Molecular Docking Results
Molecular docking against CDK6 (PDB ID: 5L2I) yielded binding energies ranging from -7.6 to -9.8 kcal/mol (Table 3). Compound S5 demonstrated the highest binding affinity (-9.8 kcal/mol), marginally higher than palbociclib (-9.4 kcal/mol), attributed to dual hydrogen bonding interactions with GLY20 (3.72 Å) and TYR24 (3.65 Å) and a long-range electrostatic association with LYS43 (4.91 Å) (Martin et al., 2017). Compounds S2 and S7 also exhibited strong binding energies (-9.1 and -9.0 kcal/mol, respectively), forming critical hydrogen bonds with TYR24 and electrostatic interactions with ASP163 (Rana et al., 2019; Susanti et al., 2021). Compounds S3 and S4 relied primarily on hydrophobic and van der Waals forces (-8.2 kcal/mol each) without direct hydrogen bonding. In contrast, palbociclib primarily engages VAL101 (2.13 Å) and GLU99 (3.77 Å) through hydrogen bonds, indicating binding motifs for the spiropyran derivatives that are distinct from those of palbociclib (Wells et al., 2020; Wander et al., 2022). Molecular docking interaction images for S5, S2, and palbociclib are presented in Figure 3.
| Compound | Binding Energy (kcal/mol) | Hydrogen Bonds (Residue, Dist. Å) | Electrostatic Interactions (Residue, Dist. Å) | Key Hydrophobic Residues |
|---|---|---|---|---|
| S1 | -7.6 | LYS29 (3.35) | ASP104 (3.84) | ILE19, LEU152, HIS100, ALA41 |
| S2 | -9.1 | TYR24 (3.24) | ASP163 (4.46) | ILE19, LEU152, PHE98, VAL27, ALA41, VAL77, VAL101, LYS43 |
| S3 | -8.2 | None | ASP163 (4.33) | ILE19, LEU152, VAL27, ALA41, ALA162, VAL77, PHE98, VAL101, LYS43 |
| S4 | -8.2 | None | None | ILE19, LEU152, VAL27, ALA41, ALA162, VAL77, PHE98, VAL101, LYS43 |
| S5 | -9.8 | GLY20 (3.72), TYR24 (3.65) | LYS43 (4.91) | LEU152, VAL27, ALA41, ALA162, VAL77, PHE98, ILE19 |
| S6 | -8.1 | TYR24 (3.30) | ASP163 (4.34) | ILE19, LEU152, VAL27, ALA41, ALA162, VAL77, PHE98, VAL101 |
| S7 | -9 | TYR24 (3.23) | ASP163 (4.25) | ILE19, LEU152, PHE98, HIS100, VAL27, ALA41, VAL77, ALA162, VAL101 |
| S8 | -8.8 | TYR24 (3.17) | ASP163 (4.40) | ILE19, LEU152, HIS100, VAL27, ALA41, ALA162, VAL77, PHE98, VAL101 |
| Palbociclib | -9.4 | VAL101 (2.13), GLU99 (3.77) | None | ILE19, VAL27, LEU152, ALA41, PHE98 |
Molecular Dynamics Simulation and MM-GBSA Analysis
RMSD analysis (Figure 4A) revealed that the palbociclib-CDK6 complex maintained values averaging 1.78±0.32 Å after equilibration (20-100 ns), while S5-CDK6 showed slightly higher flexibility (2.05±0.48 Å), with fluctuations remaining within acceptable limits for stable protein-ligand complexes (<3 Å) (Farhadi et al., 2025). Both complexes exhibited similar RMSF fluctuation patterns (Figure 4B); the palbociclib complex showed marginally higher fluctuations (average 1.11 Å) around residues 37-38 and 215 (flexible loop regions), while S5-CDK6 exhibited slightly lower RMSF values (average 1.0 Å). Radius of gyration analysis (Figure 4C) indicated maintained structural integrity for both complexes, with the S5-CDK6 complex showing a slightly more compact structure (Rg: 4.24-5.02 Å vs 5.0-5.66 Å for palbociclib), suggestive of stronger stabilising intramolecular interactions (Sarac et al., 2025). Hydrogen bond formation (Figure 4D) was consistently higher for palbociclib-CDK6 (169-232 contacts) compared to S5-CDK6 (161-225 contacts). MM-GBSA binding free energy analysis (Table 4) indicated a marginally more favourable predicted binding free energy for the S5-CDK6 complex (∆Gᵇᵉᵀ=-131.12±7.76 kcal/mol) relative to palbociclib-CDK6 (-127.66±5.26 kcal/mol); given the overlapping standard deviations, the two complexes are best regarded as comparable, with the difference attributable primarily to enhanced lipophilic interactions (∆Gᵇᵉᵀ_Lipo: -28.55 vs -26.58 kcal/mol) (Dasmahapatra et al., 2022).
| Energy Component (kcal/mol) | Palbociclib-CDK6 Mean ± SD | S5-CDK6 Mean ± SD |
|---|---|---|
| ∆Gᵇᵉᵀ (Total) | -127.66±5.26 | -131.12±7.76 |
| ∆Gᵇᵉᵀ Coulomb | 20.59±12.28 | 20.43±18.56 |
| ∆Gᵇᵉᵀ Covalent | 5.88±2.56 | 3.66±1.68 |
| ∆Gᵇᵉᵀ H-bond | -5.71±0.99 | -5.78±1.15 |
| ∆Gᵇᵉᵀ Lipo | -26.58±1.41 | -28.55±2.53 |
| ∆Gᵇᵉᵀ Packing | -1.74±0.37 | -1.68±0.92 |
| ∆Gᵇᵉᵀ SolvGB | -20.05±12.14 | -22.90±15.94 |
| ∆Gᵇᵉᵀ vdW | -99.99±3.18 | -96.34±3.08 |
In vitro Cytotoxicity
In vitro cytotoxicity evaluation against MCF-7 breast cancer cells demonstrated promising antiproliferative activity for the spiropyran derivatives. As shown in Figure 4E, compound S5 was the most active among the tested compounds, with an IC₅₀ of 15.58±1.24 μM, which was comparable to palbociclib (IC₅₀=17.33±1.38 μM) (p>0.05, one-way ANOVA with Dunnett’s test). Compound S2 showed moderate activity (IC₅₀=21.76±1.79 μM), while compound S7 (IC₅₀=24.18±2.03 μM) demonstrated lower potency, consistent with its docking score (-9.0 kcal/mol). High correlation coefficients (R²>0.99) for all compounds confirmed robust dose-response relationships (Ji et al., 2020). These results are consistent with the molecular docking rankings, with S5’s highest predicted binding affinity (-9.8 kcal/mol) corresponding to the lowest IC₅₀ value. The IC₅₀ values observed in the micromolar range are consistent with early-stage lead identification and are commonly reported for structurally novel scaffolds before SAR-guided potency optimisation (Bora et al., 2021; Manzoor et al., 2025). Morphological analysis of MCF-7 cells revealed characteristic changes, including cell shrinkage, membrane blebbing, detachment, and debris formation, following treatment with S5 and palbociclib at their respective IC₅₀ concentrations, consistent with CDK-inhibitor-induced cell-cycle arrest (Watt and Goel, 2022; Shanabag et al., 2025).
DISCUSSION
The present study reports the design, synthesis, and evaluation of novel spiropyran derivatives as putative CDK6-targeting anticancer agents for breast cancer treatment. Compound S5 emerged as the lead candidate, demonstrating the highest predicted CDK6 docking affinity (-9.8 kcal/mol) in the series, marginally exceeding the approved CDK4/6 inhibitor palbociclib (-9.4 kcal/mol). These findings are consistent with published computational studies identifying dual hydrogen bond formation with the CDK6 hinge region as a key determinant of binding affinity; specifically, the GLY20 and TYR24 interactions observed for S5 are analogous to the interaction patterns reported for potent CDK6-binding compounds in recent structure-based drug design investigations (Martin et al., 2017; Luo et al., 2021; Debnath et al., 2025). A long-range electrostatic association with LYS43 and extensive hydrophobic engagement of LEU152, PHE98, and ALA41 further distinguish S5 from the reference compound, which relies predominantly on VAL101 and GLU99 hydrogen bonds. These distinct binding motifs may be of pharmacological interest, since resistance mutations have been reported at conventional hinge-region interaction sites in patients treated with CDK4/6 inhibitors (Shanabag et al., 2025); as resistance mechanisms were not investigated in the present study, this remains a hypothesis requiring dedicated experimental validation.
Molecular dynamics simulations provided dynamic validation extending beyond static docking poses. Although S5 exhibited slightly higher RMSD flexibility (2.05±0.48 Å vs 1.78±0.32 Å for palbociclib), these values remained well within the threshold for stable protein-ligand complexes (<3 Å), consistent with productive binding (Farhadi et al., 2025; Grewal et al., 2025). The more compact radius of gyration observed for S5-CDK6 relative to palbociclib-CDK6 suggests stronger stabilising intramolecular interactions following ligand binding. MM-GBSA analysis indicated a more favourable predicted binding free energy for S5, driven primarily by enhanced lipophilic and van der Waals interactions (∆G_Lipo:-28.55 vs -26.58 kcal/mol), consistent with the nitrosoaniline moiety of S5 making deeper contacts within the hydrophobic subpocket of the CDK6 ATP-binding site (Dasmahapatra et al., 2022). Notably, the net Coulombic component of the decomposition was positive for both complexes (S5:+20.43±18.56 kcal/mol; palbociclib:+20.59±12.28 kcal/mol), reflecting an unfavourable electrostatic desolvation contribution; binding was therefore governed by lipophilic and van der Waals terms rather than by net electrostatics, and the LYS43 contact observed in docking represents a long-range association that does not translate into a net favourable Coulombic contribution. The convergence of RMSD values by 20 ns and their stability for the remaining 80 ns justified the 100 ns simulation duration as sufficient to capture the equilibrium binding behaviour of both complexes.
The in vitro cytotoxicity results showed qualitative agreement with the computational rankings. S5 exhibited an IC₅₀ of 15.58±1.24 μM against MCF-7 cells, statistically comparable to palbociclib (17.33±1.38 μM; p>0.05). The ordering of docking scores broadly tracked the observed cytotoxicity; however, no formal quantitative structure-activity relationship was derived, as a statistical correlation between binding free energies and biological activity across the series was not established, and any such relationship would require evaluation of a larger compound set. The micromolar IC₅₀ values observed are characteristic of early-stage lead identification for structurally novel scaffolds and are in line with IC₅₀ values reported for structurally related spiropyran and spiro-heterocyclic compounds evaluated at comparable stages (Bora et al., 2021). Morphological observations of cell shrinkage, membrane blebbing, and detachment in S5-treated MCF-7 cells are consistent with CDK inhibition-mediated G1/S arrest and apoptosis, corroborating the proposed mechanism of action (Shanabag et al., 2025; Hong et al., 2024). Optimisation of the spiropyran scaffold through targeted structural modifications guided by SAR findings from the current study is expected to yield derivatives with improved potency in subsequent rounds of development.
From a drug-likeness perspective, S5 complied with all Lipinski Rule of Five criteria with no violations, demonstrating a favourable pharmacokinetic profile with acceptable TPSA (114.14 Ų) and rotatable bonds (7). This favourable drug-likeness pertains to physicochemical parameters and does not extend to the predicted toxicity profile, as S5, in common with the series, returned active predictions across several toxicity endpoints and a predicted LD₅₀ (571.2 mg/kg) in the lower range of the series, underscoring the need for experimental genotoxicity and safety evaluation. The low predicted gastrointestinal absorption identified for all derivatives represents the primary formulation challenge for this series. Prodrug strategies or nanoparticle-based delivery platforms represent viable approaches to overcome this limitation, an avenue that warrants systematic investigation in future work (Adon et al., 2021). The photoswitchable nature of the spiropyran scaffold presents an additional opportunity for spatiotemporally controlled drug activation in tumour tissue, a concept with growing precedent in oncology drug delivery (Fagan et al., 2021).
Several limitations of the present study warrant acknowledgement. First, cytotoxicity was evaluated against a single breast cancer cell line (MCF-7), and generalisation of findings to other breast cancer subtypes requires additional evaluation across a panel of cell lines. Second, CDK4 versus CDK6 isoform selectivity was not confirmed by experimental kinase profiling assays; the selectivity evidence remains computational, and in vitro kinase inhibition assays against both CDK4/Cyclin D1 and CDK6/Cyclin D3 are essential to substantiate selectivity claims. Third, in vivo pharmacokinetic and efficacy data are absent, and the compound’s stability and tolerability require evaluation in appropriate animal models. Finally, the toxicity predictions were based on computational tools, and definitive safety profiling through standard preclinical genotoxicity and regulatory toxicology studies is required. The strengths of this study include the multi-modal integrated approach combining rational design, in silico profiling, synthesis, and in vitro validation, the use of a validated docking platform cross-checked against AutoDock Vina, and the inclusion of 100 ns MD simulations providing robust dynamic assessment of CDK6 engagement. These findings establish spiropyran derivative S5 as a promising lead warranting advancement to comprehensive in vivo efficacy studies, kinase selectivity profiling, and structural optimisation campaigns.
CONCLUSION
This study presents the design, synthesis, and biological evaluation of novel spiropyran derivatives (S1-S8) as putative CDK6-targeting anticancer agents for breast cancer treatment. Compound S5 emerged as the lead candidate, exhibiting the highest predicted CDK6 docking affinity (-9.8 kcal/mol) in the series and cytotoxic potency (IC₅₀=15.58±1.24 μM) comparable to palbociclib against MCF-7 cells, while complying with all Lipinski Rule of Five criteria. Molecular docking and 100 ns MD simulations indicated stable CDK6 engagement through distinct binding interactions involving GLY20, TYR24, and LYS43, with MM-GBSA analysis indicating a more favourable predicted binding free energy for S5 relative to palbociclib. Given that only three compounds were evaluated experimentally and that no kinase inhibition assays or quantitative correlation analyses were performed, these findings support preliminary agreement between computational and biological data rather than validation of the spiropyran scaffold for CDK6 inhibitor development. They nonetheless identify the spiropyran scaffold as a candidate framework worth further investigation, warranting direct CDK4 and CDK6 kinase inhibition and selectivity assays, evaluation across additional breast cancer and normal cell lines, in vivo efficacy studies, and structural optimisation before the series can be advanced toward preclinical development.