Magnolia Bark Compounds Evaluated for Antioxidant Activity and Apoptosis Protein Interactions

Researchers investigating the phytochemical composition and antioxidant properties of Magnolia officinalis bark extract found that methanolic preparations yielded the highest concentrations of phenolic and flavonoid compounds among tested solvents, while computational modelling explored predicted interactions between candidate bark constituents and apoptosis-associated proteins. Across laboratory measurements, the methanolic extract demonstrated a total phenolic content of 64.55 ± 0.65 milligrams of gallic acid equivalents per gram of dry extract and a total flavonoid content of 21.90 ± 0.55 milligrams of quercetin equivalents per gram of dry extract. In benchtop chemical assays, the extract exhibited concentration-dependent scavenging of DPPH and ABTS radical species, displaying half-maximal inhibitory concentrations of 143.02 micrograms per milliliter and 326.01 micrograms per milliliter, respectively. However, subsequent molecular docking and molecular dynamics simulations examining magnolol and honokiol against caspase-3 and BCL-2 represent exploratory structural hypotheses that do not demonstrate functional target regulation, cytoprotection, or therapeutic efficacy against cellular injury.

The study established its context around cisplatin, a widely utilized platinum-containing anticancer chemotherapy agent prescribed for diverse solid malignancies, including testicular, ovarian, bladder, cervical, lung, and head and neck cancers. Despite its established therapeutic efficacy, clinical application of cisplatin is significantly limited by severe dose-dependent toxicities and acquired resistance. The pathophysiology of cisplatin-mediated tissue injury involves multifactorial cellular cascades driven by excessive generation of reactive oxygen species, redox disruption, mitochondrial dysfunction, and activation of apoptotic signalling pathways. Caspase-3 operates as a principal executioner protease within apoptotic machinery, whereas B-cell lymphoma 2 (BCL-2) functions as an anti-apoptotic protein that helps maintain mitochondrial membrane integrity. Because programmed cell death serves dual roles across biological contexts—enabling necessary malignant cell destruction during chemotherapy while driving unwanted organ toxicity in normal tissues—predicted binding cannot be assumed to be cytoprotective or chemosensitizing without direct experimental testing in living models.

To systematically evaluate phytoconstituent extraction, dried bark powder obtained from an authenticated commercial supplier was subjected to 72 hours of ambient maceration with intermittent agitation across four solvents: distilled water, methanol, ethanol, and acetone. Following filtration through Whatman paper and concentration under reduced pressure via rotary evaporation, quantitative analyses were conducted in triplicate. One-way analysis of variance demonstrated significant solvent-dependent differences in total phenolic content (F(3,8) = 2761, p < 0.0001). The methanolic preparation exhibited the highest phenolic content at 64.55 ± 0.65 mg GAE/g dry extract, significantly exceeding aqueous extract at 38.74 ± 0.51 mg GAE/g, ethanolic extract at 30.30 ± 0.54 mg GAE/g, and acetone extract at 26.48 ± 0.54 mg GAE/g.

The antioxidant capacity of the methanolic bark extract was confirmed through two cell-free radical-scavenging assays across concentrations from 100 to 500 micrograms per milliliter. In the DPPH assay, percentage radical inhibition increased progressively from 47.01% ± 0.52% at 100 μg/mL to 53.96% ± 0.62% at 200 μg/mL, 59.93% ± 0.25% at 300 μg/mL, 68.56% ± 0.43% at 400 μg/mL, and 73.93% ± 0.86% at 500 μg/mL. The positive control butylated hydroxytoluene (BHT) demonstrated greater scavenging efficacy, rising from 55.23% ± 0.04% to 85.59% ± 0.15% across corresponding concentrations and recording an IC50 of 31.78 μg/mL compared to 143.02 μg/mL for the extract. In the ABTS radical cation assay, extract inhibition increased from 24.39% ± 1.12% at 100 μg/mL to 65.14% ± 0.14% at 500 μg/mL, generating an estimated IC50 of 326.01 μg/mL based on the linear regression equation y = 0.105x + 15.771 (R2 = 0.9654). The ascorbic acid reference standard yielded an IC50 of 275.87 μg/mL via y = 0.1159x + 18.035 (R2 = 0.9755).

Analytical characterisation by gas chromatography–mass spectrometry (GC–MS) yielded tentative library-based assignments for eleven phytochemical constituents within the methanolic bark extract. The detected constituents included isomagnolol (relative peak area 2.36%), 4-O-methylhonokiol (2.70%), β-eudesmol (3.05%), honokiol (4.30%), magnolol (4.30%), obovatol (5.38%), 2-isopropenyl-4a,8-dimethyl-1,2,3,4,4a,5,6,7-octahydronaphthalene (5.45%), D-limonene (5.58%), caryophyllene oxide (6.37%), pentadecanoic acid (6.87%), and heptadecane (7.21%). Peak areas denote relative chromatographic abundance rather than absolute concentrations, and assignments lacked validation through authentic standards or retention indices. Complementary liquid chromatography–mass spectrometry (LC–MS) provided supportive molecular-ion evidence for select compounds. A deprotonated molecular ion detected at m/z 265 between 12.352 and 12.664 minutes corresponded to C18H18O2, compatible with the structural isomers magnolol, honokiol, and isomagnolol without distinguishing them. Signals at m/z 279 and m/z 222 supported assignments for 4-O-methylhonokiol and β-eudesmol, respectively, while m/z 137 supported D-limonene; remaining constituents lacked diagnostic LC–MS confirmation. Magnolol and honokiol were selected for computational screening based on their established occurrence in Magnolia officinalis rather than definitive analytical proof.

In silico molecular docking conducted via AutoDock Vina evaluated the binding compatibility of selected phytoconstituents against the crystallographic structures of human caspase-3 (PDB ID: 1QX3) and BCL-2 (PDB ID: 6QGG). When docked against caspase-3, magnolol yielded the most favourable computational docking score at −7.1 kcal/mol, followed by honokiol at −6.1 kcal/mol and β-eudesmol at −5.6 kcal/mol. Against BCL-2, magnolol produced the most negative score at −7.4 kcal/mol, followed by honokiol at −6.9 kcal/mol, obovatol at −6.1 kcal/mol, and 2-isopropenyl-4a,8-dimethyl-1,2,3,4,4a,5,6,7-octahydronaphthalene at −5.9 kcal/mol. Interaction analyses identified non-covalent contacts, including hydrogen bonding, hydrophobic interactions, and van der Waals forces. However, the docking poses did not demonstrate specific occupancy of the caspase-3 catalytic dyad comprising His121 and Cys163, nor did they occupy the canonical BH3-binding groove of BCL-2. Furthermore, because co-crystallized ligand redocking and benchmark validation were omitted, these docking scores provide exploratory estimations of structural compatibility rather than experimental proof of binding affinity or biological regulation.

Computational pharmacokinetic profiling and toxicity predictions provided initial drug-likeness assessments for five representative constituents. Analysis on the SwissADME web platform indicated high predicted gastrointestinal absorption and blood–brain barrier permeability for magnolol, β-eudesmol, and obovatol, all adhering to Lipinski’s rule of five without violations. Honokiol demonstrated identical high gastrointestinal absorption and blood–brain barrier permeation but recorded one violation due to elevated lipophilicity (XLOGP3 > 3.5). In contrast, 2-isopropenyl-4a,8-dimethyl-1,2,3,4,4a,5,6,7-octahydronaphthalene displayed low predicted gastrointestinal absorption, lacked predicted blood–brain barrier permeability, and showed one Lipinski violation (MLOGP > 4.15). In silico toxicological modeling via ProTox-II classified magnolol (predicted oral LD50 of 2200 mg/kg) and the octahydronaphthalene derivative (5300 mg/kg) into toxicity class 5, while honokiol (1649 mg/kg), β-eudesmol (2000 mg/kg), and obovatol (1930 mg/kg) were assigned to class 4. All five compounds were predicted to be inactive for hepatotoxicity, carcinogenicity, mutagenicity, immunotoxicity, and cytotoxicity endpoints, providing preliminary prioritization data rather than clinical safety evidence.

To characterize time-dependent structural dynamics, 100-nanosecond molecular dynamics simulations were conducted using GROMACS 2024.3 with the CHARMM36 force field and TIP3P solvation. Rather than maintaining uniform stability, the simulated systems displayed complex-specific conformational adjustments. In the caspase-3–magnolol trajectory, protein backbone root mean square deviation (RMSD) exhibited limited fluctuations, whereas magnolol RMSD displayed a marked excursion during approximately 45 to 65 nanoseconds, rising to about 1.2 nanometres. This pronounced increase reflects substantial displacement and conformational rearrangement of magnolol relative to its starting docking pose, demonstrating that initial docking accommodation was not continuously preserved. In comparison, the BCL-2–honokiol system displayed smaller RMSD variations across its 100-nanosecond run. Root mean square fluctuation (RMSF) analysis revealed regional alterations in residue flexibility between ligand-bound and apo proteins, while radius of gyration (Rg) profiles indicated retention of overall protein compactness. Time-dependent hydrogen-bond tracking documented between zero and four hydrogen bonds for caspase-3–magnolol (occasionally five) and zero to three for BCL-2–honokiol (occasionally four).

Binding energetics were estimated using Molecular Mechanics/Generalized Born Surface Area (MM/GBSA) calculations executed across equilibrated trajectory frames. The caspase-3–magnolol complex yielded an estimated total binding free energy of −18.40 kcal/mol, comprising −38.11 kcal/mol van der Waals energy, +1.57 kcal/mol electrostatic energy, +23.15 kcal/mol Generalized Born polar solvation energy, and −5.00 kcal/mol surface energy. The caspase-3–honokiol complex produced a total binding free energy of −16.20 kcal/mol, driven by −24.01 kcal/mol van der Waals, −4.51 kcal/mol electrostatic, +15.33 kcal/mol Generalized Born, and −3.00 kcal/mol surface energies. For BCL-2, the magnolol complex generated −14.70 kcal/mol total binding energy, featuring −14.91 kcal/mol van der Waals, −27.66 kcal/mol electrostatic, +30.67 kcal/mol Generalized Born, and −2.79 kcal/mol surface energies; this large electrostatic term warrants particular caution given magnolol’s neutral structure. The BCL-2–honokiol complex yielded −14.55 kcal/mol, with −18.03 kcal/mol van der Waals, −1.35 kcal/mol electrostatic, +6.89 kcal/mol Generalized Born, and −2.07 kcal/mol surface energies. Because configurational entropy was omitted and statistical convergence was not evaluated, these calculations represent comparative computational estimates rather than experimental free energies.

The investigators explicitly highlighted several key scientific boundaries and methodological limitations that frame the scope of their conclusions. The study did not experimentally administer cisplatin to cell cultures or animal tissues, nor did it directly measure caspase-3 or BCL-2 catalytic activity, cellular apoptosis, chemosensitisation, or organ cytoprotection. Consequently, no conclusions can be drawn regarding the therapeutic efficacy of Magnolia officinalis bark extract or its constituents in preventing chemotherapy-induced toxicity. Furthermore, phytochemical identifications from GC–MS and LC–MS were tentative library matches lacking verification with authentic reference standards or orthogonal structural techniques. Molecular dynamics simulations were conducted as single 100-nanosecond trajectories without independent replicates, leaving the reproducibility of observed ligand rearrangements unverified. Known translational challenges associated with magnolol and honokiol, including poor aqueous solubility and limited in vivo bioavailability, continue to constrain their clinical application. Future investigations require authentic standard validation, target-specific biochemical enzyme assays, replicated computational simulations, and cisplatin-challenged biological models to determine whether these candidate constituents produce genuine cytoprotective effects.

Subject of Research: Phytochemical profiling and computational modelling of Magnolia officinalis bark extract against apoptosis proteins

Article Title: Integrated phytochemical profiling and molecular modelling of Magnolia officinalis bark-derived bioactive compounds interacting with apoptosis-associated proteins relevant to cisplatin-induced cellular injury

Article References: Vijiyakumar, N., & Evan Prince, S. (2026). Integrated phytochemical profiling and molecular modelling of Magnolia officinalis bark-derived bioactive compounds interacting with apoptosis-associated proteins relevant to cisplatin-induced cellular injury. Frontiers in Bioinformatics, 6, Article 1952073. https://doi.org/10.3389/fbinf.2026.1952073

Image Credits: AI Generated

DOI: 10.3389/fbinf.2026.1952073

Keywords: Magnolia officinalis, phytochemical profiling, antioxidant activity, molecular docking, molecular dynamics, caspase-3, BCL-2, magnolol, honokiol, computational biology, Integrated, phytochemical