The investigation began with a sweeping chemical census. Drawing on the IMPPAT 2.0 database, a manually curated repository built from more than 100 traditional Indian medicinal texts and over 7,000 peer-reviewed publications, the researcher retrieved 123 phytochemicals associated with Camellia sinensis. Canonical SMILES structures were cross-referenced through PubChem, and each compound was then pushed through a battery of in silico filters: admetSAR 3.0, SwissADME, the artificial intelligence-driven Deep-PK platform, and the graph-based predictor pkCSM. The gauntlet evaluated molecular weight, lipophilicity, hydrogen bonding capacity, topological polar surface area, gastrointestinal absorption, blood-brain barrier permeation, cytochrome P450 inhibition, clearance, mutagenicity, hepatotoxicity, and acute oral toxicity.

Only 14 compounds survived the full screening cascade, and their identities are telling. The list includes familiar catechins such as epicatechin and cianidanol, phenolic acids like caffeic acid and gallic acid, vitamins and cofactors including ascorbic acid and pantothenic acid, and a striking contingent of brassinosteroid-related sterols: typhasterol, teasterone, brassinolide, and castasterone, alongside the triterpenoid saponin theasapogenol B and the sapogenin A1-barrigenol. Notably, several high-profile tea polyphenols, including theasinensins and heavily galloylated derivatives, failed Lipinski’s rule of five because their sheer molecular size and polar surface area would sabotage oral bioavailability. The survivors, by contrast, showed high predicted gastrointestinal absorption, minimal interference with major CYP450 drug-metabolizing enzymes, and largely non-mutagenic, non-hepatotoxic profiles.

With the shortlist established, the study turned to target prediction. SwissTargetPrediction, a reverse-screening engine built on chemical similarity principles, assigned up to 100 putative human protein targets to each of the 14 phytochemicals, generating 1,400 raw predictions that collapsed to 262 unique proteins after deduplication. These were fed into the STRING database to construct a protein-protein interaction network of 260 nodes and 2,504 edges, with an average node degree of 19.3 and a PPI enrichment p-value below 1.0 × 10⁻¹⁶, confirming that the connectivity reflects genuine biology rather than statistical noise. Applying a stringent combined-score threshold above 0.9 retained 488 high-confidence interactions for downstream analysis.

Clustering algorithms then carved the network into eight functional modules, each a dense island of cooperating proteins. The top-scoring module, with an MCODE score of 10.824, was dominated by the PI3K/AKT and receptor tyrosine kinase machinery, including PIK3CA, AKT1 through AKT3, EGFR, ERBB2, JAK1 through JAK3, and IGF1R. Other modules captured cell cycle regulators such as CDK1, AURKA, and PLK1; GABA receptor subunits tied to neurotransmission; MAPK stress-signaling proteins; a neurodegeneration-and-apoptosis cluster featuring PSEN1, PSEN2, GSK3B, and HDAC1; cell cycle checkpoint proteins; matrix metalloproteinases involved in tissue remodeling; and cholesterol biosynthesis enzymes including HMGCR and SQLE. The breadth of these modules hints at why tea has been linked to such a bewildering variety of health benefits.

To separate the true regulatory heavyweights from peripheral players, the study applied four independent centrality algorithms in the cytoHubba plugin: Degree, Betweenness, Closeness, and Maximal Clique Centrality. Only three proteins ranked among the top ten under every single method: PIK3CA, the catalytic subunit of phosphatidylinositol-3-kinase; AKT1, the master survival kinase; and ESR1, the estrogen receptor alpha. The convergence is biologically compelling. The PI3K/AKT axis governs proliferation, apoptosis, glucose metabolism, and inflammatory signaling, and its dysregulation is a hallmark of cancer, insulin resistance, and neurodegeneration, while ESR1 sits at the intersection of hormonal signaling, neuroprotection, and breast cancer biology.

Functional annotation through the DAVID platform painted the pathways these hubs inhabit. Gene Ontology analysis linked them to apoptosis, glucose metabolic processes, insulin receptor signaling, kinase activity, and PI3K signal transduction, with cellular localization concentrated in the cytosol, plasma membrane, and lamellipodia. KEGG pathway enrichment pulled in an impressive roster of disease-relevant cascades: pathways in cancer, TNF signaling, HIF-1 signaling, AMPK signaling, FoxO signaling, VEGF signaling, estrogen signaling, Toll-like receptor signaling, prolactin signaling, and thyroid hormone signaling. A phytochemical-target-pathway network then visualized how the 14 compounds converge on AKT1, ESR1, and PIK3CA, which in turn fan out into these interconnected pathways, a textbook illustration of the multitarget, multi-pathway logic that distinguishes network pharmacology from the classical one-drug-one-target paradigm.

The structural validation stage delivered the study’s most eye-catching numbers. Using AutoDock Vina through PyRx, with docking protocols verified by re-docking co-crystallized ligands to RMSD values between 1.0 and 1.2 angstroms, several tea phytochemicals outperformed their reference inhibitors. Epicatechin and cianidanol bound AKT1 at −9.8 kcal/mol, comfortably beating the reference ligand IQO at −6.9. For the estrogen receptor ESR1, typhasterol and theasapogenol B reached −8.9 kcal/mol against OHT’s −6.5. And castasterone posted −9.7 kcal/mol against PIK3CA, far surpassing the 2Q7 reference at −6.5. Interaction maps showed the compounds engaging the same catalytic residues as the native ligands: epicatechin and cianidanol contacting Thr211, Lys268, and Val270 in AKT1; epicatechin hydrogen-bonding with Asp351 and Glu353 in ESR1; and multiple compounds anchoring to Lys802, Arg992, and Leu1028 in PIK3CA.

The authors are careful to frame these findings as hypothesis-generating rather than definitive. Docking scores estimate relative interaction strength but do not substitute for measured binding affinities, the enrichment analyses relied on unadjusted p-values vulnerable to false positives, and no ligand pose superposition or molecular dynamics simulations were performed. Experimental validation in vitro and in vivo remains the essential next step. Even so, the study provides a rigorous, systems-level rationale for centuries of empirical enthusiasm about tea, pinpointing epicatechin, cianidanol, castasterone, typhasterol, and theasapogenol B as the most promising candidates and PIK3CA, AKT1, and ESR1 as the molecular crossroads where a humble cup of tea may exert its most consequential effects.

Subject of Research: Multitarget therapeutic potential of Camellia sinensis phytochemicals analyzed by network pharmacology and molecular docking

Article Title: Elucidating the multitarget therapeutic potential of Camellia sinensis (Tea) phytochemicals using network pharmacology, functional annotation, and molecular docking

Article References: Hossain, M. M. (2026). Elucidating the multitarget therapeutic potential of Camellia sinensis (Tea) phytochemicals using network pharmacology, functional annotation, and molecular docking. Discover Chemistry, 3(1), Article 541. https://doi.org/10.1007/s44371-026-01000-0

Image Credits: AI Generated

DOI: 10.1007/s44371-026-01000-0

Keywords: Camellia sinensis, tea, network pharmacology, molecular docking, phytochemicals, PIK3CA, AKT1, ESR1, ADMET, drug discovery, cancer, molecular targets

Liver fibrosis arises when chronic injury from viral hepatitis, alcohol, fatty liver disease, or other insults pushes hepatic stellate cells into an activated, myofibroblast-like state. In their quiescent form, these cells store vitamin A and quietly regulate blood flow through the liver’s sinusoids. When activated, they proliferate, migrate, and deposit extracellular matrix faster than it can be degraded, gradually choking the organ’s architecture into the stiff, nodular tissue of cirrhosis. Despite decades of research, no approved therapy reverses established fibrosis; treatment has largely meant removing the underlying cause and hoping the liver’s own regenerative capacity keeps pace. Praziquantel had already shown hints of anti-fibrotic activity in experimental settings, but how a drug best known for paralyzing schistosome worms could calm scar-forming liver cells remained a mystery.

To crack that mystery, the research team, led by Zhongkui Lu and Guoying Zhang of Nanjing Integrated Traditional Chinese and Western Medicine Hospital affiliated with Nanjing University of Chinese Medicine, together with colleagues at Xuzhou Medical University and Jinling Hospital, assembled potential praziquantel targets from public pharmacological databases and cross-referenced them against genes implicated in liver fibrosis. The overlap yielded 137 candidate genes. Enrichment analyses of this set pointed toward pathways involving xenobiotic metabolism and neuroactive ligand-receptor interactions, a signature consistent with the drug’s known pharmacology but also hinting at receptor-mediated effects beyond simple parasite membrane disruption.

The next step was to find the critical nodes within this network. Using the STRING database to construct a protein-protein interaction map and Cytoscape to visualize and prune it, the researchers identified six hub genes at the center of the praziquantel-fibrosis intersection: EGFR, ALB, TP53, PTGS2, ESR1, and CYP3A4. These genes span a striking range of functions, from growth factor signaling and tumor suppression to drug metabolism and hormone reception. But which of them actually matters causally for fibrosis, rather than merely being correlated with it? To answer that question, the team turned to Mendelian randomization, a statistical technique that uses naturally occurring genetic variants as instruments to test whether an exposure, here the expression or function of a candidate gene, has a causal effect on an outcome.

The Mendelian randomization analysis delivered a clear verdict for one gene. ESR1, the estrogen receptor alpha gene, showed genetically supported evidence of a protective causal role against liver fibrosis. A colocalization analysis, which tests whether the same genetic variant drives both the gene signal and the disease association in a genomic region, nominated a specific variant, rs3020404, as a plausible functional basis for the link. In other words, the population genetics did not merely suggest that ESR1 expression tracks with fibrosis severity; it suggested that inherited differences in ESR1 activity genuinely shift fibrosis risk, making the receptor a credible therapeutic target rather than a bystander.

Genetic plausibility still needed a physical mechanism, and for that the researchers turned to molecular modeling. Molecular docking placed praziquantel within ESR1’s ligand-binding pocket, and molecular dynamics simulations confirmed that the drug-receptor complex remains stable over simulated time, with the small molecule maintaining consistent contacts with the receptor. The modeling cannot prove binding in a living cell on its own, but it established that praziquantel and ESR1 are chemically compatible partners, setting the stage for functional tests.

The most revealing layer of the study came from single-cell RNA sequencing of liver tissue. Analyzing the data with the Seurat framework, the researchers mapped ESR1 expression across the liver’s cellular ecosystem and found it broadly present, but with a telling pattern: quiescent hepatic stellate cells and a cytokine-producing stellate cell subset, dubbed cyHSCs, expressed significantly higher levels of ESR1 than activated myofibroblastic stellate cells, or myHSCs. The receptor that praziquantel appears to target is most abundant precisely in the cell states that fibrosis threatens to destroy or corrupt, suggesting the drug may act by reinforcing the quiescent, non-fibrogenic identity of these cells.

To probe what ESR1 actually does inside stellate cells, the team ran virtual knockout experiments using scTenifoldKnk, a computational method that predicts how silencing a gene rewires a single-cell gene regulatory network. Removing ESR1 in silico disrupted a network whose most prominent casualties included RXFP1, EGFLAM, and several mitochondrial genome components such as MT-CO1, MT-CO2, and MT-ND4L. Pathway analysis of the perturbed genes showed strong enrichment in oxidative phosphorylation and immune signaling, including T cell receptor signaling. The picture that emerges is of ESR1 as an orchestrator of mitochondrial metabolic homeostasis and immunoregulatory signaling in stellate cells; when it is lost, the cells’ energy metabolism falters and inflammatory programs gain ground, conditions that favor fibrogenic activation.

Computational predictions, however convincing, demand wet-lab confirmation, and the researchers provided it. Working with LX-2 cells, a widely used human hepatic stellate cell line, they silenced ESR1 and tested whether praziquantel could still exert its anti-fibrotic effects. It could not, at least not fully. The loss-of-function experiments confirmed that ESR1 is functionally required for the drug’s benefit, closing the loop between network prediction, genetic causality, structural modeling, and cellular mechanism. The authors propose that praziquantel activates ESR1, which in turn maintains a protective gene network preserving mitochondrial function and immune balance in stellate cells, thereby blocking their transition into collagen-secreting myofibroblasts.

The implications extend well beyond one drug and one receptor. Repurposing praziquantel, whose safety profile is established through mass administration programs across the tropics, could dramatically shorten the path to clinical testing for an anti-fibrotic indication compared with developing a novel molecule from scratch. More broadly, the study showcases an integrative strategy, combining network pharmacology, Mendelian randomization, colocalization, molecular dynamics, single-cell transcriptomics, virtual knockout, and in vitro validation, that can elevate a computational hypothesis to a mechanistically grounded candidate therapy. ESR1 modulation itself may prove a fruitful therapeutic direction independent of praziquantel, and the identification of rs3020404 as a candidate functional variant offers a genetic handle for stratifying patients most likely to benefit. Much work remains: the findings rest heavily on human cell lines and public datasets, and animal models and clinical trials will be needed to confirm that the mechanism operates in scarred livers in living patients. But the study reframes a familiar antiparasitic as a plausible antifibrotic and hands hepatology a genetically validated, druggable target at the heart of the stellate cell’s decision to scar or stay quiet.

Subject of Research: Network pharmacology and experimental validation identifying ESR1 as the target through which praziquantel alleviates liver fibrosis

Article Title: Praziquantel targeting ESR1 to alleviate liver fibrosis: a comprehensive network analysis insight

Article References: Lu, Z., Kong, D., He, F., Lv, H., Guo, Y., Xia, X., & Zhang, G. (2026). Praziquantel targeting ESR1 to alleviate liver fibrosis: a comprehensive network analysis insight. Journal of Translational Medicine. https://doi.org/10.1186/s12967-026-08941-1

Image Credits: AI Generated

DOI: 10.1186/s12967-026-08941-1

Keywords: praziquantel, liver fibrosis, ESR1, hepatic stellate cells, Mendelian randomization, molecular docking, single-cell RNA sequencing, drug repurposing, mitochondrial function, hepatology, gene regulatory network, LX-2 cells

The scale of the problem they targeted is enormous. More than nine million osteoporotic fractures occur worldwide each year, and people living with diabetes mellitus face a 20 to 50 percent higher risk of fragility fractures than the general population. Diabetic osteoporosis is increasingly described as a dual pandemic, driven by mechanisms that differ fundamentally from ordinary age-related bone loss. Advanced glycation end products, insulin resistance, oxidative stress, and elevated inflammatory cytokines all conspire to disrupt bone metabolism, lowering osteoprotegerin while raising RANKL, a signal that fuels osteoclast-driven bone resorption. At the same time, chronic inflammation and hyperglycemia push bone-forming osteoblasts toward apoptosis and steer mesenchymal stem cells away from bone formation and toward fat production, a molecular switch governed by activated PPAR-gamma and suppressed Runx-2 expression.

Because existing antidiabetic drugs manage blood glucose imperfectly and carry side effects without curing the underlying disease, the research team turned to bioinformatics to hunt for multi-target natural molecules. Their strategy began with predicting genistein’s pharmacokinetic profile using the SwissADME web tool and its oral toxicity using ProTox-II and ProTox-3.0, which returned encouraging drug-likeness scores and low predicted toxicity. They then mined the GeneCards and Comparative Toxicogenomic databases for disease targets, retrieving a staggering 32,133 genes associated with diabetes mellitus and osteoporosis, and used Swiss Target Prediction to identify 105 human protein targets for genistein itself.

Overlaying the two gene sets produced a protein-protein interaction network of 104 nodes and 606 edges, visualized in Cytoscape and analyzed through topological parameters including degree, betweenness centrality, and closeness centrality. Five hub genes rose decisively above the rest: EGFR, the epidermal growth factor receptor; ESR1, the estrogen receptor alpha; MMP9, matrix metalloproteinase-9; PTGS2, the inflammatory cyclooxygenase-2 enzyme; and PPARG, the nuclear receptor governing fat and glucose metabolism. Gene ontology and KEGG pathway enrichment analyses revealed that these targets converge on the PPAR signaling pathway and EGFR tyrosine kinase inhibitor resistance pathways, alongside biological processes spanning apoptosis regulation, stress response, growth factor signaling, and catecholamine metabolism.

The computational deep-dive then moved to the atomic scale. Molecular docking using Schrödinger’s Glide module showed genistein binding strongly to all four top targets, with the strongest standard-precision score of minus 10.752 kcal/mol against ESR1, followed by minus 8.381 against EGFR, minus 7.096 against MMP9, and minus 6.495 against PPARG. Each complex was anchored by specific hydrogen bonds and hydrophobic contacts within the binding pockets, indicating that the soy isoflavone nestles into the same active regions as purpose-built synthetic drugs.

Docking, however, captures only a frozen snapshot. To test whether these interactions survive the thermal chaos of a living cell, the team ran 100-nanosecond molecular dynamics simulations in triplicate for each protein-ligand complex using the Desmond engine at 300 Kelvin under constant pressure and temperature. All four systems equilibrated within 20 nanoseconds and remained stable throughout. Backbone root-mean-square deviations stayed between 1.8 and 2.5 angstroms, ligand RMSD values remained below 2 angstroms, and active-site residues fluctuated less than 1.5 angstroms. The ESR1 and PPARG complexes proved especially robust, maintaining three to four hydrogen bonds for roughly 80 to 88 percent of the trajectory and retaining high alpha-helical content of about 55 to 57 percent, signatures of thermodynamically stable, persistent binding.

MM/GBSA free energy calculations sealed the computational case. The ESR1 complex posted the most favorable binding free energy at minus 51.86 kcal/mol, followed by EGFR at minus 49.23 and PPARG at minus 42.82, with van der Waals, electrostatic, and nonpolar solvation terms driving the favorable energetics. These numbers confirmed that genistein’s grip on its targets is not an artifact of rigid-receptor scoring but a genuinely stable molecular partnership sustained by the same forces that govern real drug binding.

Crucially, the researchers did not stop at the computer. In a dexamethasone-induced insulin resistance rat model, a well-established experimental mimic of type 2 diabetes metabolic dysfunction, genistein was formulated as a solid dispersion with PVP-K30 to improve solubility and administered orally at 1, 2, and 4 mg/kg daily for 25 days. Post-treatment, genistein-treated rats showed statistically significant reductions in fasting blood glucose and serum insulin compared with untreated positive controls, with the highest dose performing best, indicating restored insulin sensitivity.

The skeletal results were equally compelling. Scanning electron microscopy of rat femurs revealed that diabetic control animals had porous, microcracked, eroded trabecular surfaces and visible resorption pits, while genistein-treated bones appeared dense, compact, and structurally organized. Nanoindentation showed that treated animals maintained tissue-level hardness and reduced modulus close to normal values, and three-point bending tests demonstrated dramatic mechanical recovery: maximum load capacity in the highest-dose group reached 67.4 newtons, exceeding even the normal control value of 60.81 newtons, while the untreated diabetic group collapsed to just 10.33 newtons. Ultimate stress, stiffness, and toughness all followed the same restorative pattern.

Mechanistically, the findings weave a coherent story. EGFR dysregulation impairs the PI3K/AKT insulin signaling axis and undermines osteoblast survival, while genistein’s selective affinity for estrogen receptor beta and modulation of NF-kB and MAPK pathways counteracts inflammation-driven bone resorption. MMP9, overexpressed under hyperglycemic oxidative stress, chews through bone matrix and is partially responsible for skeletal degradation in diabetic animals, and PPAR-gamma overactivation diverts bone marrow stem cells into fat rather than bone. By binding all of these targets simultaneously, genistein appears to act as a dual-action agent, lowering blood glucose while defending bone microarchitecture and mechanical strength. The authors caution that further clinical and translational work is needed, but their integrated evidence positions this inexpensive soy-derived phytoestrogen as a promising template for evidence-based functional foods and tailored therapeutics against a complication that diabetes medicine has long undermanaged.

Subject of Research: Genistein as a multi-target phytoestrogen therapy for diabetic osteoporosis, investigated through network pharmacology, molecular docking, molecular dynamics simulation, and rat model experiments

Article Title: Genistein potential and mechanisms against diabetes osteoporosis: An integrated study of network pharmacology, molecular docking, and molecular dynamics simulation

Article References: Sharma, S., Chaudhary, R., Hooda, T., Sharma, C., Dabral, S., Kumar, A., Bansal, S., & Gupta, S. (2026). Genistein potential and mechanisms against diabetes osteoporosis: An integrated study of network pharmacology, molecular docking, and molecular dynamics simulation. Results in Chemistry, 30, Article 103833. https://doi.org/10.1016/j.rechem.2026.103833

Image Credits: AI Generated

DOI: 10.1016/j.rechem.2026.103833

Keywords: genistein, diabetic osteoporosis, network pharmacology, molecular docking, molecular dynamics simulation, MM/GBSA, EGFR, ESR1, MMP9, PPARG, insulin resistance, bone metabolism