H-Trp-Glu-OH and H-Glu-Trp-OH dipeptides as potential anticancer agents against glioblastoma multiforme: Multi-target molecular docking, DFT analysis, molecular dynamics simulation, ADMET profiling, and cell viability validation
Authors: Demirag, Aliye Demet
Source journal: Computational biology and chemistry (2026-11-11)
DOI: 10.1016/j.compbiolchem.2026.109574
Abstract
BACKGROUND: Glioblastoma multiforme (GBM) is the most common and lethal primary brain malignancy in adults. Despite the Stupp protocol, median overall survival remains approximately 15 months, and the inadequacy of current therapies combined with the restrictive permeability of the blood-brain barrier (BBB) underscore the need for new therapeutic candidates. This study aimed to evaluate the sequence isomeric dipeptides H-tryptophan-glutamic acid-OH (H-Trp-Glu-OH) and H-glutamic acid-tryptophan-OH (H-Glu-Trp-OH), hereafter abbreviated Trp-Glu and Glu-Trp, as potential anticancer agents against GBM through an integrated computational and experimental approach. METHODS: Geometry optimization and electronic structure analyses were performed in Gaussian at the B3LYP/6-311 ++G(d,p) level. A conformational search over multiple starting geometries preceded optimization, and vibrational frequency calculations confirmed that all reported structures are true minima. Two reference states were used, each applied identically to both sequence isomers so that every comparison is strictly like-for-like. The electronic structure calculations and the ADMET predictions were carried out on the neutral form, which is the interpretable reference state for Koopmans-based reactivity descriptors and the input convention of the ADMET predictor; molecular docking and molecular dynamics were carried out on the anionic species that predominates at pH 7.4 (NH3 +, two COO-; net charge -1), which is the species that actually engages the binding sites. Molecular docking against six GBM-associated target proteins (CHI3L1-YKL-40/8R41, PI3K-alpha/4JPS, VEGFR-2/4ASD, IDH1 R132H/4UMX, STAT3/6NJS, and System xc-/7P9U) was performed using AutoDock Vina with grid-box centres, box dimensions, exhaustiveness, pose count and random seed reported in full; the protocol was validated by redocking each co-crystallized ligand into its own site and computing the heavy-atom RMSD against the crystallographic pose. Three targets (CHI3L1/YKL-40, VEGFR-2 and IDH1 R132H) were carried forward to 100 ns molecular dynamics (MD) simulations in Schrodinger Desmond using the OPLS4 force field, one trajectory per complex, with both sequence isomers built in the same ionization state. Every docking calculation was repeated with five independent random seeds and is reported as mean +/- SD; the redocking validation was performed as a single run per target. No MM-GBSA binding free-energy estimation with per-residue decomposition. Comprehensive ADMET profiling was carried out using ADMETlab 3.0. In vitro anticancer activity was evaluated in U-87 GBM cells using the MTT assay at 10-100 mug/mL over 24, 48, and 72 h, with temozolomide as the reference agent; all concentrations are reported in both mass and molar units, and 72 h IC50 values with 95% confidence intervals were obtained by non-linear regression. RESULTS: In the corrected screen, every docking calculation was repeated with five independent random seeds and scores are reported as mean +/- SD. Glu-Trp gave the strongest score of the panel against CHI3L1/YKL-40 (8R41, -8.60 +/- 0.05 kcal/mol). Trp-Glu scored highest against CHI3L1/YKL-40 and VEGFR-2 (-8.19 +/- 0.24 and -8.17 +/- 0.02 kcal/mol), two values that are indistinguishable from one another. The two isomers ranked the six targets almost identically, but the difference between them lay within the seed-to-seed variability of the scoring function at only one of the six targets (System xc-, 0.08 +/- 0.21 kcal/mol); at the remaining five the separation exceeded that variability. VEGFR-2 was the exception of mechanistic interest: there the isomers occupied two distinct sub-pockets, Trp-Glu at the ATP-site hinge and Glu-Trp in the type-II back pocket, sharing no contact residue. Redocking of the co-crystallized ligand reproduced the crystallographic pose within the conventional 2.0 A criterion for four of the six targets (VEGFR-2 0.54 A, PI3K-alpha 0.62 A, CHI3L1 1.02 A, IDH1 R132H 1.20 A), with STAT3 at 2.65 A and System xc- at 3.91 A. The HOMO-LUMO energy gap was 4.818 eV for Trp-Glu and 4.952 eV for Glu-Trp, a difference of 0.134 eV that is too small to be treated as evidence of differing biological activity. Molecular dynamics was extended from one target to three and each complex was simulated for 100 ns with both isomers in the same ionization state. The two sequences behaved differently, and in opposite directions at different targets. At VEGFR-2 Glu-Trp held its docked pose, with ligand RMSD confined to roughly 2-3.5 A and hydrogen bonds to Glu885 and Arg1027 persisting for most of the trajectory, whereas Trp-Glu left its starting pose within the first 20 ns and settled into an alternative surface site anchored by Arg929 and Arg932. At IDH1 R132H the ordering reversed: Trp-Glu showed the smaller positional drift and a persistent hydrogen-bond network centred on Ile128, Ser287 and Ala282. At CHI3L1/YKL-40 Trp-Glu also relaxed out of its initial pose after approximately 65-70 ns; the Glu-Trp ligand of that system had been built without explicit bond-order information, so the Glu-Trp results at this target are reported as provisional and the comparison between the isomers there is not treated as quantitative. Each complex was simulated once rather than in replicate, and MM-GBSA free energies were not computed. ADMET profiling returned a mixed rather than a uniformly favourable picture: low predicted hERG and AMES liabilities, but one Lipinski violation (six hydrogen-bond donors), poor predicted Caco-2 and MDCK permeability, poor human liver microsome stability, very low predicted BBB permeability, and elevated nephrotoxicity, ototoxicity and genotoxicity probabilities. In the MTT assay, both dipeptides exerted negligible cytotoxicity at 24-48 h but produced marked, significant reductions in viability at 72 h. Trp-Glu displayed a monotonic concentration-response with its strongest effect at the highest concentration tested (100 mug/mL, 300 muM; IC50 82.9 mug/mL), whereas Glu-Trp displayed a non-monotonic concentration-response with its strongest effect at the lowest concentration tested (10 mug/mL, 30 muM). Because temozolomide was included as a single-concentration reference (100 muM, equivalent to 19.4 mug/mL) rather than as part of a dose-response series, the comparison is presented on a molar basis and no claim of superiority over temozolomide is made. CONCLUSIONS: Because Trp-Glu and Glu-Trp are sequence isomers of identical molecular formula and identical ionizable groups, every difference observed between them is attributable to residue order and to nothing else. On that controlled comparison the study makes one substantive finding, supported independently by two methods: sequence directionality changes where and how these dipeptides engage a protein. In docking it relocated the ligand between two distinct sub-pockets of VEGFR-2 with no contact residue in common, and in molecular dynamics it determined which isomer retained its initial pose, with the advantage reversing between VEGFR-2 and IDH1 R132H. The same directionality is visible in the cells, where the two isomers produced opposite concentration-response patterns at 72 h. The present work does not establish blood-brain barrier penetration, target engagement, mechanism of action or in vivo efficacy; the very low predicted BBB permeability and several elevated in silico toxicity flags are substantial liabilities. These dipeptides are therefore best regarded as starting structures whose sequence-dependent behaviour merits further mechanistic investigation, rather than as candidates ready for preclinical development.
Related topics
Related open access articles
- Development and Validation Molecular Docking Analysis of Human serum albumin (HSA)
- Facing COVID-19 via anti-inflammatory mechanism of action: Molecular docking and Pharmacokinetic studies of six anti-inflammatory compounds derived from Passiflora edulis
- Computational analysis by molecular docking of thirty alkaloid compounds from medicinal plants as potent inhibitors of SARS-CoV-2 main protease
- Curcumin and pipernonaline to curb diabetes the natural way: A molecular modeling, docking and dynamics simulation study
- DENSITY FUNCTIONAL THEORY ANALYSIS AND MOLECULAR DOCKING EVALUATION OF 8-CHLOROQUINOLINE 2-CARBALDEHYDE
- Thermochemical, DFT, Molecular Docking and Pharmacokinetic Studies of Methyl β-D-galactopyranoside Esters
Abstract and bibliographic data from PubMed (U.S. National Library of Medicine). Copyright of the abstract belongs to the original publisher or authors.