Computational screen identifies naringenin as a CXCR4-binding candidate

A multi-stage computational study has prioritized naringenin, a plant-derived compound, as a candidate for binding the human chemokine receptor CXCR4, a protein implicated in the invasive, inflammatory and pro-survival mechanisms shared by endometriosis and ovarian cancer. The finding is an in silico prioritization rather than proof of therapeutic activity: the researchers performed no receptor-binding assays, functional experiments or animal studies, so the candidate’s effects in either disease remain unverified.

The researchers built a human CXCR4 model with SWISS-MODEL, reporting 98.2% template coverage and 100% sequence identity, and used the co-crystallized ligand IT1t to define the receptor’s docking cavity. Eighteen phytochemicals were then screened through ligand-guided docking. Naringenin produced the most favorable predicted docking score, at −6.361 kcal/mol, compared with −5.927 kcal/mol for apigenin and −5.155 kcal/mol for IT1t. Its modeled interactions involved Tyr116, Arg188 and Tyr255, three amino-acid residues within the predicted ligand-binding site.

The analysis combined docking with pharmacokinetic and toxicity predictions, pharmacophore mapping, density functional theory and two independent 100-nanosecond molecular-dynamics simulations of CXCR4 complexes containing naringenin, apigenin or IT1t. Across measures including root-mean-square deviation of atomic coordinates, solvent-accessible surface area, radius of gyration, hydrogen bonding, principal component analysis, free-energy landscape analysis and dynamic cross-correlation maps, the naringenin complexes showed comparatively consistent structural behavior. MM/GBSA binding-energy estimates were also more favorable for naringenin than apigenin in both simulation replicas, while its calculated HOMO–LUMO energy gap was wider than that of apigenin.

Predicted drug-likeness, solubility and gastrointestinal absorption were described as acceptable for naringenin, but the study also modeled a median lethal dose of 2,000 mg/kg and assigned it to toxicity class 4. That estimate underscores a central limitation: favorable computational scores for receptor interactions do not establish safety or biological efficacy. Together, the results identify naringenin as the most coherent candidate to test experimentally among the 18 compounds examined, but the research does not demonstrate that it binds CXCR4 in cells, alters CXCR4 signaling, reduces endometriosis or inhibits ovarian cancer progression. Those questions require biochemical binding, functional and in vivo validation before the compound’s medical relevance can be assessed.

Subject of Research: Computational identification of phytochemical candidates predicted to bind human CXCR4.

Article Title: Computational identification of natural CXCR4-binding candidates relevant to endometriosis and ovarian cancer progression

Article References: Karageçili, S., Inan, D., & Ali, N. (2026). Computational identification of natural CXCR4-binding candidates relevant to endometriosis and ovarian cancer progression. PLOS One, 21(10), e0359833. https://doi.org/10.1371/journal.pone.0359833

Image Credits: AI Generated

DOI: 10.1371/journal.pone.0359833

Keywords: CXCR4, naringenin, apigenin, endometriosis, ovarian cancer, molecular docking, molecular dynamics, phytochemicals, computational biology, Computational, identification, natural