Catechin–Folate nano-niosomes from Osbeckia parvifolia Arn. induce apoptotic cell death in Ovarian cancer Authors: Benedict Mathews Paul, Gowtham Kannan, Yamuna Annadurai, Parimelazhagan Thangaraj Published: 2026 ABSTRACT The current study establishes the mechanism on how dietary flavonoid, catechin fractions of Osbeckia parvifolia Arn. could be used as a therapeutic treatment for ovarian cancer, using a combination of network pharmacology, molecular validation and nano-formulations. From network modelling analysis, 420 protein targets linked to ovarian cancer, including five that act as major regulatory hubs – AKT1, STAT3, IL6, ESR1 and CASP3 were found. Catechin had the strongest binding affinity to all of these five hub proteins, especially for ESR1 (-11.089 kcal/mol) and AKT1 (-9.221 kcal/mol). In molecular dynamics simulation studies over 100 ns, catechin bonded strongly with its target proteins. HPLC guided fractionation (RT: 3.887 min) LC-MS/MS (RT:6.51 min), HPTLC (Rf: 0.973) were performed, which resulted in the confirmation of the presence of both catechin and epicatechin and verified the biosynthetic pathway of both through the measurement of the expression of their respective biosynthetic genes LAR (0.3637 folds) and ANR (0.2746 folds) in O. parvifolia. The evaluation of the six different nano-formulations of phytoniosomes led to F5 being the most stable with uniform spherical architecture, having an optimal particle size (229.2 nm), a high positive zeta potential (+ 42.7mV), and having the highest cytotoxicity (93.17%) in SKOV3 cells, the mechanism of action being through apoptosis induction and cell-cycle arrest. Hence, this study supports the hypothesis that catechin extracted from O. parvifolia can serve as an effective therapeutic candidate for the targeted treatment of ovarian cancer.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-026-57651-7. Introduction Ovarian cancer (OC) remains a highly lethal malignancy with poor long‑term survival due to aggressive growth, late detection, and diagnostic challenges that limit effective early intervention. Standard treatments—surgery and chemotherapy—improve outcomes for some patients but still yield suboptimal five‑year survival and substantial toxicity, highlighting an urgent need for more effective, less toxic therapeutic strategies. Because many cancers arise from interactions between genetic alterations and environmental exposures such as diet, integrating nutrition‑based approaches into cancer care is gaining interest as a complementary route to reduce disease burden and improve patient outcomes1–6. Natural compounds derived from food, particularly flavonoids, have attracted attention as potential anticancer agents because they modulate multiple cancer‑relevant processes. Flavonoids can inhibit angiogenesis, alter proliferative signalling, induce apoptosis, and block metastasis in OC models, and structural analogs show chemopreventive and chemosensitizing effects when combined with conventional agents in preclinical studies7,8. Catechins, a flavonoid subgroup, have demonstrated specific activity against OC cells, but their clinical translation is limited by poor solubility, rapid metabolism, off‑target effects, and generally low bioavailability9,10. Nanotechnology‑based delivery systems offer a promising solution to these pharmacological barriers. Phytoniosomes—nonionic surfactant vesicles formulated with stabilizing excipients such as lecithin, chitosan, Poloxamer 407, Tween 80, and PEG6000—can improve solubility, stability, and circulation time while providing sustained release and enhanced tumour accumulation via the enhanced permeation and retention (EPR) effect11–13,16–20. Surface modification with folic acid enables active targeting to folate receptor‑overexpressing OC cells, increasing tumour selectivity, reducing systemic exposure, and potentially overcoming resistance mechanisms associated with conventional chemotherapy14,15,21–24. Complementing experimental approaches, network pharmacology and integrated multi‑omics analyses provide a systems‑level framework to predict multi‑target interactions and key regulatory nodes affected by natural products like catechins25–29. In‑silico network models and transcriptomic/proteomic/metabolomic integration can prioritise candidate catechin fractions, suggest synergistic pathways, and guide targeted molecular validation—thereby strengthening the translational rationale before costly in vivo and clinical work28,29. Building on this rationale, the present study sources catechin fractions from the ethnopharmacologically used medicinal plant Osbeckia parvifolia Arn. and addresses the translational barriers of catechins by formulating folate‑conjugated phytoniosomes to enhance bioavailability, tumour targeting, and sustained delivery30,31. The study evaluates these formulations for improved pharmacokinetics, targeted cytotoxicity against OC cells, and effects on oncogenic signalling and metabolic pathways using network pharmacology and experimental validation, thereby filling a key research gap between promising pharmacological flavonoid activity and viable therapeutic strategies for OC. Materials and methods Chemicals and Reagents The chemicals used in this study were obtained from Himedia Laboratories, Mumbai; Sisco Research Laboratories (SRL), Mumbai; Merk, Bengaluru and Sigma Aldrich, USA. All the chemicals were of analytical grade and solvents (High performance Liquid Chromatography; HPLC grade) used in this study were purchased from Merck Millipore (Bangalore, India). For all studies carried out, Milli-Q water with resistivity of 18.2 MΏ cm−1 at 25 °C and a total oxidizable carbon value below 5 ppb was used (Merck Millipore; Billerica, MA, USA)32. Plant material The aerial parts of O. parvifolia as a whole was collected31 from the Kotagiri region of the Western Ghats (11°27’06.3“N 76°54’15.4”E) during the month of December, where it was considered as a weed by local inhabitants. The taxonomic identity was formally confirmed by Dr. M.U. Sharief, Scientist F and Head, Botanical Survey of India (BSI), Southern Circle, Coimbatore, and Tamil Nadu comparing the deposited voucher specimens with accession nos. 85986, 80760, 14327 and 173713 in BSI. The plant specimen was then certified for its identity through Certificate No. BSI/SRC/5/23/2021/Tech/177. Prior permissions were obtained for plant collection adhering to the institutional, national and international guidelines and legislations. The plant collection complied with the International Union for Conservation of Nature (IUCN) Policy Statement on Research Involving Species at Risk of Extinction and the Convention on the Trade in Endangered Species of Wild Fauna and Flora. Network modelling Targets identification The compounds identified by GC-MS analysis of O. parvifolia extract31 were subjected to chemical identification for PubChem ID, SMILES, structure data files (.sdf) from PubChem database (https://pubchem.ncbi.nlm.nih.gov/). The targets common between the O. parvifolia compounds and OC were identified using STITCH (http://stitch.embl.de/), and SwissTargetPrediction tools (http://swisstargetprediction.ch/). The targets were sorted based on probability > 0.1. The targets identified through each tool were combined together and duplicates were removed from the list and further checked for accuracy validation through UniProt Knowledgebase (https://www.uniprot.org/)33,34. Compounds target network Cytoscape 3.10.2 was used to build the compound-target network. To determine how each target interacted with the network, node-degree analysis was performed. Protein–protein interactions The interaction between the proteins common between O. parvifolia phytocompound- targets and OC targets were analysed primarily by proteins identifiers that match with Homo sapiens on STRING database (https://string-db.org/). The confidence score ≥ 0.100 was set to create the functional protein association network also known as Protein-Protein interaction (PPI) network. The top 10 proteins that ranked in the highest degrees were derived from the PPI network. Using the Cytohubba plugin in CytoScape 3.10.2 the PPI network was analysed for hub nodes based on degree and shortest path33. Gene ontology enrichment analysis and MAF oncoplot analysis Gene Ontology (GO) enrichment analysis, including Biological Process (BP), Molecular Function (MF), and Cellular Component (CC) categories, was performed using the SRplot online platform (http://www.bioinformatics.com.cn/) for the common protein targets between O. parvifolia phytocompounds and OC. Significantly enriched GO terms were ranked based on gene count, fold enrichment, and false discovery rate (FDR) adjusted using the Benjamini–Hochberg method33. From the PPI network constructed previously, the top 10 hub proteins were identified according to degree centrality, representing the most highly connected nodes within the network. Mutation profiling of these hub proteins was conducted using Mutation Annotation Format (MAF) data from The Cancer Genome Atlas (TCGA) ovarian serous cystadenocarcinoma cohort (n = 436), and the mutational landscape was visualized using an oncoplot. Molecular docking To ascertain the interaction of O. parvifolia phytocompounds with the top 5 ranked proteins, molecular docking analysis was performed. The 3D structure of O. parvifolia phytocompounds in.sdf format (22 Ligands) were retrieved from PubChem database. The proteins AKT1 (PDB ID: 3O96), IL6 (PDB ID: 4CNI), STAT3 (PDB ID: 3CWG), ESR1 (PDB ID: 1XP1), and CASP3 (PDB ID: 3EDQ) were selected for in silico molecular docking analysis based on the following criteria35. High resolution of the crystal structure to ensure accurate docking (≤ 3.0 Å).Completeness of the binding domain or functional region relevant to ligand binding with no pre-existing mutations in the protein.Relevance of the co-crystallized ligand or conformational state (active/inactive) to our study. High resolution of the crystal structure to ensure accurate docking (≤ 3.0 Å). High resolution of the crystal structure to ensure accurate docking (≤ 3.0 Å). Completeness of the binding domain or functional region relevant to ligand binding with no pre-existing mutations in the protein. Completeness of the binding domain or functional region relevant to ligand binding with no pre-existing mutations in the protein. Relevance of the co-crystallized ligand or conformational state (active/inactive) to our study. Relevance of the co-crystallized ligand or conformational state (active/inactive) to our study. The protein preparation wizard was employed to prepare the proteins by removing all the water molecules and adding hydrogen ions. pH was fixed to be 7.0. Small molecules that were bound to the crystallized structure of proteins were removed. Energy minimization was done for all the protein structures. Grid box dimensions were generated using the axis coordinates listed in Supplementary Table 1. GLIDE Module, a subset that is vital for molecular docking experiments, which allows the user to predict how small molecules may dock into a protein target was used for the analysis. Electrostatic potential was calculated, and Gasteiger charges were added according to the free energy. The receptor grid generation panel was used to set the grid around the target proteins. Active site prediction was performed using PrankWeb (https://prankweb.cz/), which identifies ligand-binding pockets based on a combination of residue conservation, surface topology, and physicochemical properties. The pockets were ranked by a deep learning-based score (DL score), and the highest-scoring pocket was selected for molecular docking studies. Ligand docking was performed for five proteins with all the derived small molecule stereoisomers, and the results were visualized with XP-visualizer, which is an essential tool for the high precision results analysis and visualization of molecular docking. Molecular dynamics-simulation Top-ranked complexes from the molecular docking analysis were subjected to molecular dynamics analysis to investigate the stability and dynamic behavior using the Desmond module of Schrodinger (Version 2024-4) following the method of Kannan et al.35 The protein structure was prepared by adding mission hydrogens, assigned bond order, creating a disulfide bond and optimization of ionization states using Protein preparation workflow. OPLS force field employed to optimize geometry of the system and for energy minimization to remove steric clashes. Each complex (ESR1-73160; AKT1-73160) were solvated in an orthorhombic simulation box using the TIP3P water model, with a buffer distance of 10.0 Å in all directions (A, B, and C). The steepest descent algorithm was used to system minimization and restraint equilibration. After minimization, the box volumes of protein 1XP1 (ESR1) and 3O96 (AKT1) were 345,944 Å and 472,227 Å, respectively. To ensure electroneutrality, 6 Cl - ions added to 3O96 system and 6 Na + ions added to 1XP1 protein system. The equilibrium protocol followed for NVT at 0–300 K temperature and NPT is 1 bar of pressure based on Berendsen thermostat. Long-range electrostatic interactions were calculated using the Particle Mesh Ewald (PME) method. For the simulation of 1XP1-73160 complex 32,996 atoms and 3O96-73160 complex 43,277 atoms were involved in the simulation analysis. Using the Simulation Interaction Diagram tool in Desmond, trajectory analysis was carried out to measure the radius of gyration (Rg), protein–ligand interactions, root mean square deviation (RMSD), and root mean square fluctuation (RMSF) in order to evaluate the complexes stability and conformational behavior throughout the simulation period. Fractionation of catechins by HPLC The methanolic extract of O. parvifolia was subjected to identification of phenolic compounds using semi-preparative HPLC system (Shimadzu, Japan). The extract was eluted using a reverse phase C18 column and a mobile phase optimized with a combination of Pump A (0.1% Trifluoroacetic acid in HPLC water) and Pump B containing Acetonitrile : Methanol (7:3). The flow rate was maintained at 0.8mL/min for a total run time of 25 min and column temperature was set at 40℃. The Pump B concentration was set at isocratic gradient as 38% (1 min); 55% (10 min); 60% (15 min); 80% (20 min); 30% (23 min) and terminated at 25 min. Fractions corresponding to each peak were collected in the collecting tray and each fraction was noted for their retention time. Standard Catechin hydrate (Sigma Aldrich, USA; Prod. No. C1251) was run at the same mobile phase and the retention time was compared with the derived fractions to qualitatively identify the catechin fraction from the extract35. Confirmation by LC-MS/MS The derived fraction identical to the retention time of standard catechin was diluted with 0.5% acetic acid and filtered using 0.45-micron Millipore filter. Filtrate of the fraction was injected into the Agilent 6200 Series TOF and 6500 Series Q-TOF LC/MS System. LC-MS analysis was done by using UPLC-MS equipped with a binary pump. LC is connected to QTOF mass spectrometer coupled with ESI. The MS used was Agilent 6500 QTOF with negative ionization mode. ESI parameters used capillary temperature 110 °C, gas atomizer 40 L/h, and source voltage + 2.7 kV. Full scan mode from m/z 50–1100 was done with source temperature 40 °C. UPLC column used was ZORBAX RRHD SB-C18 1.8 μm, 2.1 × 150 mm. The mobile phase used was water with acetic acid 0.5% (A) and acetonitrile (B), which was linearly programmed from 95 to 46.2% (A) for 30 min at a flow rate programmed at 0.2 mL min − 1. The injection volume was 3 µL. To obtain sensitive and stable responses the negative ionization mode was selected on account of the higher and more stable signal strengths. Subsequently, the MS/MS ion transitions were screened in MRM scanning mode to enhance the selectivity of detection using range of collision energy 10–35 ev and collision gas N2. Agilent MassHunter Workstation Software - Qualitative Analysis was used in the selected and listed compounds36–38. HPTLC analysis The HPTLC analytical method followed standard practices in carrying out the assay for catechins. Two separate applications of 100 µg/mL catechin fractions and standard catechins were made to a 20 cm x 10 cm size of the HPTLC plate with a Camag Linomat-V (Switzerland). The device used an air compressor and had a delivery rate of 150 nL/s from a syringe by means of nitrogen flow. Plates were run in a Camag horizontal developing chamber at room temperature. The mobile phase was an optimized combination for flavonoids namely, ethyl acetate: Acetic Acid (glacial): formic acid: Water in the ratio 10:0.5:0.5:1.3. Developments of HPTLC plates for scanning were performed at 254 nm using a Camag TLC Scanner III with the use of Wincats 1.2.3 software31. qPCR gene expression studies To analyse the synthesis of catechin in O. parvifolia, DNA was extracted from the plant material using HiPurA® Plant DNA Isolation Kit (Prod. No. MB502) following the manufacturer protocol. The isolated DNA was subjected to purity assessment using Nanodrop Spectrophotometer (Thermo Scientific, USA). The derived DNA (10 µL) was mixed with forward and reverse primers (Table 1) of catechin biosynthesis genes (LAR, ANR, F3’H, 4CL and internal control GAPDH) and subjected to qPCR amplification using following steps for 40 cycles: Initial denaturation (94℃ − 5 min), denaturation (94℃ − 30 s), annealing (57℃ − 30 s), extension (72℃ − 45 s) and final extension (72℃ − 10 min). The resultant PCR products were run on 0.5% Agarose gel to quantify the synthesis genes’ expression. ImageJ software was used to develop the densitogram and the results were expressed as relative expression in folds39–42. Table 1Forward and Reverse primer sequences of catechin synthesis genes.4CLForward TTCCATATTCACACAGAGAC Reverse CTTGACTGAAAGCACTGG F3′HForward GGTAGTTATAAGCGATGTGTTTG Reverse CGGAGCAGTTCCTGAAGG ANRForward CCTTCTAGCACTAAAGGGTTCAGG Reverse TGACTACTCCTTGAATTGCTGGTT LARForward TTTTGTTGCAGGCTCTGATATAGG Reverse ATTATATTCACTGCTGCTGCTGCT GAPDHForward GCTTCACATACCTTTGGAGCTTC Reverse TCTCTTCCTCTTGTGCTCTTG Table 1Forward and Reverse primer sequences of catechin synthesis genes.4CLForward TTCCATATTCACACAGAGAC Reverse CTTGACTGAAAGCACTGG F3′HForward GGTAGTTATAAGCGATGTGTTTG Reverse CGGAGCAGTTCCTGAAGG ANRForward CCTTCTAGCACTAAAGGGTTCAGG Reverse TGACTACTCCTTGAATTGCTGGTT LARForward TTTTGTTGCAGGCTCTGATATAGG Reverse ATTATATTCACTGCTGCTGCTGCT GAPDHForward GCTTCACATACCTTTGGAGCTTC Reverse TCTCTTCCTCTTGTGCTCTTG Forward and Reverse primer sequences of catechin synthesis genes. Development of phytoniosomes The phytoniosome formulation strategy was adopted to enhance the bioavailability of catechin fractions. As they tend to degrade easily phytoniosomes shall act as a protective barrier for targeted drug delivery. The phytoniosomes were formulated following the method of Wilson et al., 2021 with slight modifications. Phytoniosomes were prepared using thin film hydration technique by different non-ionic surfactants and cholesterol. Lyophilized catechin fraction, folic acid, Poloxamer 407, lecithin, Tween 80 and PEG 6000 were solubilized in a round bottom flask containing 1:1 mixture of ethanol and chloroform to form a solution. The organic solvent was evaporated using a rotary evaporator (Equitron Evator, India) to form a thin film which was rehydrated using 10 mL of PBS (pH-7.4) for 1 h at 50℃. The rehydrated solution was sonicated for 20 min using an ultra-probe sonicator to form uniform vesicles. The unsonicated formulation without was also developed for comparison. The developed formulations (F1-F6) were converted to powder using a Lyophilizer (Delvac, India). The composition of the formulations is given in Table 243–46. Table 2Composition of phytoniosomal formulations.FormulationOPF1 load (mg)Folic acid (mg)Lecithin (mg)Tween 80 (mL)Poloxamer 407 (mg)PEG 6000 (mg)Chitosan (%)F13015251.25--0.01F2301525-250-0.01F33015251.25-2500.01F4301525--2500.01F53015251.252502500.01F6--251.252502500.01 Table 2Composition of phytoniosomal formulations.FormulationOPF1 load (mg)Folic acid (mg)Lecithin (mg)Tween 80 (mL)Poloxamer 407 (mg)PEG 6000 (mg)Chitosan (%)F13015251.25--0.01F2301525-250-0.01F33015251.25-2500.01F4301525--2500.01F53015251.252502500.01F6--251.252502500.01 Composition of phytoniosomal formulations. Cell culture The SKOV3 and SH-SY5Y cell lines were procured from the National Centre for Cell Sciences, Pune, India, and were cultured in McKoy’s 5 A medium and DMEM F12 medium respectively supplemented with fetal bovine serum (10%). The cells were placed in a CO2 incubator (Thermo Fisher, USA) at 37 °C and 5% CO2 to obtain a confluent monolayer. Trypsin (0.25% EDTA) was used to detach the cells from the 25 mL T- flask, after which the cells were sub-cultured and maintained at the exponential phase for further analysis. Cytotoxicity using MTT assay Cytotoxicity of the formulations (F1-F6) in SKOV3 and SH-5YSY cells after 48 h incubation was tested using a 3-(4, 5, and dimethylthiazol-2-yl)2,5 diphenyl tetrazolium bromide (MTT) kit (Cat.No CT-02; Sigma Aldrich, St. Louis, USA) following the manufacturer’s instructions. The resultant absorbance at 570 nm was read using an ELISA plate reader (EPOCH, Biotek, USA). The half-maximal inhibitory concentration (IC50) was determined from the absorbance versus concentration curve. The formulation with lowest IC50 value was further characterized for their anti-proliferative and physicochemical properties47. Apoptosis induction assessment Bioactive formulation was tested against SKOV3 human OC cell line in a 48-h trial at each dose level (25, 50, and 100 ug/mL). Using a six-well plate set-up, after 8 h of incubation with each dose level of bioactive formulation, cells were trypsinized and centrifuged for 10 min at 5000 rpm. Supernatant was removed and 100 μL of binding buffer (1x) was added to the pellet resulting from centrifugation followed by a vigorous shaking to create a homogenous suspension. After creating the suspension, 5μL of Annexin V-FITC (Annexin V that is conjugated to a FITC molecule) and 5μL of propidium iodide was added to the cells and allowed to incubate in a dark at room temperature for 15 min. When this incubation period was complete, 400μL of binding buffer was added to the treatment sample and mixed thoroughly before observation utilizing the BD FACSverse Flow Cytometer (BD Biosciences, USA) for the identification of apoptotic populations of the treated cell line. In addition to the above-mentioned assays, a dual staining experiment was also conducted using acridine orange (to stain nucleolar, cytoplasmic and nuclear regions in treated cell lines) and ethidium bromide (to specifically identify cells undergoing apoptosis) to determine the percentage of apoptosis in cells treated with F5 at concentrations of 25, 50 and 100μg/mL relative to control samples, utilizing an Inverted Fluorescence Microscope (AccuScope EXI310, USA)48,49. Cell cycle arrest assay The SKOV3 cells were treated with 25, 50, and 100 μg/mL of the formulation, as well as with an untreated control group for 48 h in six well plates. Following trypsinization and centrifugation, the cells were resuspended in 1mL of a reagent mixture consisting of sodium citrate (100 mg), propidium iodide (50 mg/mL; 2.5 mg), RNase (4 mg), and Tween 20 (30µL). The cells were stained with propidium iodide for 30 min at 4 °C, then subjected to flow cytometry analysis to determine the numbers of sub-G0, G0, G1, S, and G2/M cells. Data were analyzed using the FACSuite software program (BD Bioscience USA)50,51. Characterization of bioactive phytoniosomal formulation DLS and zeta potential analysis DLS characterization of the formulations was conducted using the Malvern Zetasizer Ver7.13, which measures both the mean particle diameter and the polydispersity index (PI)—a parameter indicating the breadth of particle size distribution. For accurate measurement, samples were analyzed at a fixed scattering angle of 90° using 10 mm diameter cells, maintaining a controlled temperature of 25 °C to ensure consistency. Prior to data acquisition, the samples were diluted with double-distilled water to achieve an optimal scattering intensity, minimizing errors due to excessive particle concentration. The zeta potential measurements on F5 were conducted using Litesizer 500BM10 (Anton Paar, India). To ensure consistency and accurate readings, samples were appropriately diluted using deionized water. The run mode was set as manual and the samples were subjected to 100 runs. Refractive index was 1.33.3 while viscosity was maintained at 0.89.3 mPa.s; Omega cuvette was used as a measurement cell at 25℃ with Smoluchowski’s approximation and 1.50 H Factor. The maximum voltage was set as 200 V52. FTIR analysis The frequency range between 4000 and 600 cm−1 was set to acquire the FTIR spectral image of the formulations (F1-F6) using an Attenuated Total Reflectance – Fourier Transform Infrared (ATR – FTIR) Spectrometer (Model: JASCO FT/IR 4700, Japan) equipped with a DLaTGS detector with an aperture size of 50 × 50 μm; 100 times accumulation under transmission measurement method. The acquired spectrum was further analysed to find the chemical groups associated with the phytochemical constituents and to identify the molecule’s backbone or the functional groups connected inside the sample, including the type of functional groups, including distinguishing linear/branched or saturation/unsaturation/aromatic chains and rings in the chemical structure present in the fingerprint region and double, triple and single bond areas. The derived peaks were compared to the reference database35. X-ray diffraction study The crystalline characteristics of the prepared phytoniosomes were evaluated using an X-ray diffractometer (D/teX Ultra detector) operated under standard Cu Kα radiation (λ = 1.5406 Å). The instrument was operated at 40 kV and 30 mA, following the instrument’s default measurement parameters. The phytoniosomes were placed on a quartz sample holder and scanned in the 2θ range of 5°–90° using a continuous scan mode, with a scan speed of 15°/min and a step width of 0.02°. A K-beta filter was used to eliminate secondary radiation interference. The incident slit was kept open, and the receiving slit dimensions were set at 2° and 0.15 mm. Diffractograms were recorded digitally and processed to evaluate the crystalline and amorphous phases53,54. Scanning Electron Microscopy (SEM) analysis Phytoniosome morphology was examined using a Carl Zeiss EVO 18 scanning electron microscope (Carl Zeiss, Germany). Aliquots of the formulation were deposited onto carbon stubs, air-dried, and sputter-coated with a thin Au layer (~ 4–6 nm). Imaging was performed with a tungsten filament source at 5 kV (SE detector) and a working distance of ~ 8.5 mm to evaluate surface topology; uncoated specimens were imaged in variable pressure mode (VP, ~ 10–200 Pa) at 1–3 kV to minimize charging. Representative images were captured between ×1,000 and ×60,000 magnification; images were saved as TIFF files and minimally adjusted for brightness/contrast. Instrument settings and analytical options follow the manufacturer’s EVO 18 specifications55,56. Drug release profile The drug release profile was investigated into the PBS buffer solution at 7.4 pH at 37 °C as reported in the literature. 5 mL of the sample was taken at different intervals (1 h, 2 h, 3 h, 4 h, 5 h, 6 h, 12 h, and 24 h) and analysed via a double beam UV–vis spectrophotometer (Thermoscientific Genesis 180, USA) and the spectrum was recorded (200–800 nm). PBS solution was chosen as a reference standard45,57. Statistical analysis The results were expressed as Mean ± Standard Deviation (n = 3). Using SPSS version 20.0, the data were statistically analysed using one-way ANOVA and Duncan’s test. A value of p < 0.05 was deemed statistically significant for the mean values35. Results Network modelling-based analysis Targets identification Initially the common targets between O. parvifolia phytocompounds identified by GC-MS analysis and OC were scrutinized as shown in the Venn diagram (Fig. 1a). The Compounds-Target network was constructed and has been represented in Fig. 1b. Fig. 1(a) Venn Diagram of Common Targets, (b) Compounds Target Network, (c) Protein-Protein Interaction Network, (d) Degree Sorted Top 10 hub proteins, (e) Human eFP expression analysis. (a) Venn Diagram of Common Targets, (b) Compounds Target Network, (c) Protein-Protein Interaction Network, (d) Degree Sorted Top 10 hub proteins, (e) Human eFP expression analysis. Protein–protein interaction network From about 528 targets that were common, 420 common proteins were identified after setting the confidence score as > 0.1 and removing the duplicates. The Protein-Protein Interaction (PPI) network (Fig. 1c) created from this identified dataset was extremely complex with 420 nodes and 5733 edges. The average node degree was 27.3 meaning that each protein interacts with at least 27 other proteins. This indicated that the proteins are highly dependent on one another and therefore have a high level of interaction among them. The PPI network showed evidence of high levels of modularity (average local clustering coefficient = 0.432) indicating that many of the proteins were predominantly found in tightly-connected clusters or functional complexes or pathway groups. It is important to note that the density of the PPI Network was much higher than would be expected. If we were to generate a random network with the same density as the existing PPI Network (420 Nodes and 5733 Edges), only 2695 edges would exist. The P-value for the observed PPI Network (P-Value < 1.0e-16) indicates that these proteins are linked through biological means and not merely by random chance. Topology analysis based on degree of expression revealed top ten protein targets (Fig. 1d) namely, AKT1, IL6, STAT3, ESR1, CASP3, HIF1A, BCL2, PPARG, HSP90AB1 and MMP9 in the respective order of expression. Human eFP module (Fig. 1e) revealed that, among top five proteins STAT3 and AKT1 was expressed in 1151.72 and 173.86 folds respectively in expression levels in ovary which deemed the targeting of the compounds fit in the network model. About 14 functional protein hubs (Table 3) were identified in the PPI network based on degree and shortest paths. Gene ontology analysis GO enrichment analysis (Fig. 2) has identified common pathways within all of the biological processes, molecular functions, and cellular components that suggest ways that these proteins are associated with OC progression by regulating these biological mechanisms and may be involved in inhibiting tumour growth. The analysis revealed steroid metabolic process as the most enriched biological process; membrane raft as the enriched cellular component and steroid binding to be the molecular function enriched by the target proteins. Fig. 2Gene Ontology plot. Gene Ontology plot. Mutation annotation oncoplot analysis Mutation annotation analysis of 436 samples of serous adenocarcinomas of the ovaries (Fig. 3) found that 21 cases (4.82%) present as exhibiting somatic changes in at least one of the genes studied here. STAT3 had the most mutations (~ 2%), primarily a missense mutation; other genes in the panel had very low mutation rates (≤ 1%); PPARG, ESR1, MMP9, HSP90AB1, IL6, HIF1A, AKT1, BCL2, and CASP3 showed infrequent scattered missense, in-frame, frame-shift and infrequent nonsense mutations. None of these genes showed widespread recurrent truncating mutations. Fig. 3MAF Oncoplot. MAF Oncoplot. Molecular docking analysis Molecular docking studies was conducted on 22 phytochemicals derived from O. parvifolia with respect to five of the top ranked proteins (AKT1, IL6, STAT3, ESR1 and CASP3) and represented in Fig. 4. The analysis demonstrated catechin (PubChem ID: 73160) as the most potent binding ligand for each of the five proteins. Unlike other phytochemicals, catechin possessed consistently high binding affinity to each of the five hub proteins, demonstrating catechin’s ability to interact with numerous critical cell-signalling proteins that are directly implicated in the progression of OC. Of all the proteins studied, ESR1 exhibited the highest affinity binding interaction with catechin, with a docking score of −11.089 kcal/mol. The binding interactions between catechin and ESR1 included π-π stacking and hydrogen bonds (−1.965 A˚) between PHE404, GLU353, LEU387, and HIS524 within the ESR1 ligand-binding domain. Catechin bound strongly to AKT1, producing the docking score, −9.221 kcal/mol, through both hydrogen bonds (−2.026 A˚) and π–π stacking interactions with the residues TRP8C, GLN79, THR82 and THR28C. The interaction between catechin and the protein, STAT3, showed a binding affinity of − 7.103 kcal/mol, which demonstrated that catechin makes hydrogen bond and π–cation interactions with important residues (ASN A538, ARG A353, GLU392, and GLU B530) of the protein. To the essence, catechin also interacted with IL6 which produced a dockscore of −7.038 kcal/mol through hydrogen bonding between hydroxyl groups of catechin and ASP H62, SER 95, LYS A99, TYR A2C residues. Hydrogen bonding was the highest to an extent of −2.269 A˚ in this interaction. In addition, CASP3, a major enzyme that plays an important function in the apoptosis process, had a binding score of −6.209 kcal/mol with catechin based on docking studies, with TRP333, GLU343, GLU374, and ASP33 being some of the residues that interact with the catechin. Although these interactions are not as strong as those seen with ESR1 and AKT1, it may still have a role in stabilizing or activating the caspase for catechins. Fig. 4Interaction of catechin with top five hub proteins; A- AKT1; B – IL6; C- STAT3; D- ESR1; E- CASP3;. Interaction of catechin with top five hub proteins; A- AKT1; B – IL6; C- STAT3; D- ESR1; E- CASP3;. Molecular dynamics simulation studies Molecular Dynamics simulation analysis was conducted to find out the binding efficiency and changes between catechin and top two target proteins that showed highest dockscores (Fig. 5). Simulation of catechin binding with ESR1 showed that the binding complex was very stable over time and throughout a range of environmental conditions (i.e. 300 K) within the NPT regime of the MD simulation system. The entire MD simulation in NPT of catechin-ESR1 included 32,067 total atoms (9331 of which were water), as well as cationic (Na+) and anionic (Cl-) species, providing charge neutrality in the MD system. Catechin - a small molecule with a molecular weight of 290.275 g/mol remained in a stable position within the ESR1 binding pocket throughout the entire 100ns simulation, as indicated by the RMSD values of proteins within the globular protein class, which averaged ~ 1.8–3.2 Å are the lowest values that indicate acceptable stability levels for globular proteins. Catechin retained a stable orientation (ligand RMSD values < 4.5 Å) at the active site, suggesting continued occupancy by catechin in the ESR1 active site. RMSF analysis indicated that loop regions were the most flexible (~ 6.4 Å), while α-helical and β-strand structures are relatively stable, exhibiting < 1.5 Å fluctuations, indicative of preserving structural architecture during ligand binding. According to secondary structure analyses, indications are that during the course of the trajectory ESR1 sustained 55.68% of its total structure as an ‘α’-helix while only exhibiting 2.87% of its total structure being composed of β-strands, suggesting very little disruption of the ESR1 by the catechin. Additionally, catechin formed stable and prolonged (greater than 70% of the total simulation) hydrogen bonds with LEU387 and HIS524. There were also additional stabilizing interactions (hydrophobic, stacking and bridge connections) that involved multiple additional residues (e.g., PHE404, ARG394, and GLU353) creating a complex network of interactions that helped provide a ‘solid base’ for catechin to maintain contact with the protein’s active site. Furthermore, ligand dynamics provided further evidence that catechin was a stable ligand as it maintained stable low-energy conformational torsional states, a consistent radius of gyration and a solvent accessible surface area. The polar surface area (~ 150 A2) reflected the hydrophilic properties of catechin and allowed for stable hydrogen bonding and continued contact with the ESR1A 100-nanosecond (ns) molecular dynamics (MD) simulation of catechin background and it’s binding to AKT1 under the same conditions (NPT at 300 K), produced a well-equilibrated system of 49,954 atoms with 14,558 water molecules. The catechin was observed to be highly dynamic but to be stable in the binding modulator through the entire trajectory with six rotatable bonds. Fig. 5Molecular Dynamics Simulation with top two docked complexes. Molecular Dynamics Simulation with top two docked complexes. The stability of AKT1 over the course of the simulation can be seen by observing that the protein (domain) had a RMSD that stabilized around 3.5 Å. The RMSD of the catechin ligand resurging from AKT1 showed that it remained below 3.9 Å. This would indicate that the catechin ligand maintains continuous and stable binding with the active site of AKT1 throughout the simulation phase. The analysis of RMSF indicated that the regions within the protein that were loops had the greatest fluctuations (up to 4.8 Å), while resilient secondary structures like α-helices and β-strands exhibited very stable conformations of < 1.5 Å. This indicates that the catechin ligand appropriately accommodated to AKT1 without compromising any of the core secondary structures of AKT1 protein. An evaluation of the secondary structure (SSE) confirmed that the structure of the AKT1 protein (43.35% total SSE) included 20.78% α-helices and 22.57% β-strands; further, this confirmed that structural conservation of AKT1 has occurred in the presence of catechin. The ligand made multiple, consistent hydrogen (H) bond interactions with multiple key residues (GLN79, TRP80, THR81, THR 82, VAL 83, LYS 179, AND THR 211) on AKT1, many of which continue beyond 60% during the simulation phase (ARG 273, ASP 274, GLU 298, AND ASN 54, VAL 271, ASP 292). HPLC guided fractionation of catechins Bioactive polyphenolic compounds were isolated from the methanol extract of O. parvifolia using gradient-flow mobile phase liquid chromatography. In addition to catechins, there are also polyphenolic compounds present in the extracts. An acetonitrile-methanol system was used along with TFA acidified water to perform liquid chromatography of the extracts. The results from chromatography of the methanol extracts showed six distinct and reproducible peaks on the chromatograms (Table 4) that were separated based on their respective retention times for each of the six peaks. The six eluted fractions (OPF1-OPF6) were collected individually and air dried to provide the isolated and purified polyphenolic compounds. To assess the presence of catechin, a standard solution of catechin hydrate was chromatographed under the same conditions. The retention time for the standard was found to be 3.987 min which is very similar to retention time for OPF1 (3.887 min). This close correspondence provides further support for the presence of catechin or a closely related compound in OPF1. The high peak purity index for OPF1 (0.941233) suggests that OPF1 contains a chemically homogeneous constituent (Fig. 6). The remaining fractions (OPF2-OPF6) exhibited retention times ranging from 5.624 to 22.780 min and showed high values for peak purity index (0.86–0.99). The number of fractionations that needed to be performed in order to obtain enough material for downstream bioactivity testing necessitated 70 separate fractionations of this sample, yielding a total of 83.47 mg of material from a 25 µL injection per run from each fraction. Table 3Functional cluster hubs identified in PPI network.Cluster No.ClusterScore(Density*#Nodes)NodesEdgesNode IDs1 37.95950930PGR, APP, TNF, AKT1, HMOX1, ERBB2, GSK3B, MAPK8, ABL1, PTGS2, JAK2, MDM2, ABCB1, ESR2, PPARG, ESR1, HDAC6, MAP2K1, HDAC1, MAPK14, MAPK3, BCL2, BCL2L1, PIK3CA, PDGFRB, AR, PRKDC, EP300, CREBBP, HIF1A, TERT, PARP1, EGFR, DNMT1, PTPN11, IL6, RELA, CDK4, IGF1R, TLR4, STAT3, SRC, NR3C1, MMP2, MCL1, MET, CDK1, BRAF, KDR, MAPK12 16.39147377ALK, HSP90AA1, PDGFRA, BRD4, HSP90B1, FLT3, JAK1, ABCG2, TGFBR2, PTPRC, ACE, TNFRSF1A, HDAC4, HSPA8, PPARA, CDK6, PRKCD, LCK, NOS3, RAF1, FLT1, TGFBR1, CDK2, HDAC2, PIK3R1, SYK, FYN, CASP3, PTK2, MAPT, PIK3CG, PLG, XPO1, CHEK1, SELE, MMP9, KIT, MMP1, EPAS1, FGFR1, IKBKB, PRKCA, NTRK1, ICAM1, AXL, PLAU, HSP90AB13 7.5651189AGTR1, TLR8, HMGCR, CCR2, CCR3, PPARD, CSF1R, ADAM17, HDAC7, PTPN6, MMP3, MMP13, PLK1, CTSB, SHH, GSR, KAT2B, IDH1, CCR5, SMARCA4, HDAC3, XDH, CDC25A, IDO1, ALDH1A1, WEE1, DNMT3A, PLAUR, NLRP3, CXCR2, HTT, MDM4, JAK3, NOS2, CXCR3, KDM1A, GLI1, CYP19A1, NR4A1, SIRT2, NOS1, NQO1, FASN, TLR9, AURKB, ARG1, TOP2A, CHEK2, PLAT, MPO, NTRK24 7.1541143KLKB1, G6PD, LIPC, TSPO, CRHR1, CHRNA4, CETP, CFTR, SLC6A2, AKR1C4, STS, HSD17B3, HSD17B1, CES1, AKR1C2, APOB, SLC18A2, AKR1C3, DRD3, F9, NR1H4, F11, SOAT1, HTR3A, NPC1L1, DRD1, F10, CYP51A1, SQLE, ADH1B, SLC6A3, SRD5A1, CYP3A4, NR5A1, DHFR, DRD4, CHRM2, CYP17A1, HTR2C, HTR1A, CYP1A25 6.09842125HTR2B, GABBR1, NPY1R, COMT, SIGMAR1, KCNA5, LPAR4, KCNMA1, GRM5, DHCR7, PRKCG, HTR7, HTR2A, PRKCQ, DRD2, PRKCE, EBP, SSTR1, SCN5A, KCNN3, CYP27B1, KCNA3, MAOA, KCNN1, GPR55, MTNR1B, CHRM1, LPAR2, OPRD1, MCHR1, KCNK2, CACNA1C, ADRA2B, PNOC, LPAR3, ADRA2C, CYP1B1, OPRM1, SCN9A, MTNR1A, CYP24A1, ADORA2A6 5.8461438CYP1A1, HSD11B1, NR3C2, HSD17B2, HSD17B7, CYP11B2, CYP2J2, CYP11B1, SRD5A2, AKR1C1, PLA2G1B, ALOX12, HSD11B2, PLA2G2A7 5.1883383LIPE, NR1I2, ACACA, F3, VCP, ANPEP, ACHE, DGAT1, CPT1A, PTGS1, ELANE, PRMT1, FABP4, PIK3CB, CYP2D6, CYP2A6, CYP2C9, CARM1, NR1I3, CTSK, PAK1, INSR, UGT2B7, CYP2C19, CD38, CCR1, CXCR1, MSR1, NR1H3, HPGDS, EPHX1, SCD, DPP48 446NEK2, CDC25C, CDC7, CDC25B9 3.33345PDE10A, PDE3A, DCK, PDE4D10 333AKR1B1, MGLL, FABP511 333MMP12, COL18A1, CTSL12 333PRKCB, EGLN1, TYR13 356CBR1, ADH4, ALDH3A1, MAOB, GLO114 333CA4, CA2, CA1 Table 3Functional cluster hubs identified in PPI network.Cluster No.ClusterScore(Density*#Nodes)NodesEdgesNode IDs1 37.95950930PGR, APP, TNF, AKT1, HMOX1, ERBB2, GSK3B, MAPK8, ABL1, PTGS2, JAK2, MDM2, ABCB1, ESR2, PPARG, ESR1, HDAC6, MAP2K1, HDAC1, MAPK14, MAPK3, BCL2, BCL2L1, PIK3CA, PDGFRB, AR, PRKDC, EP300, CREBBP, HIF1A, TERT, PARP1, EGFR, DNMT1, PTPN11, IL6, RELA, CDK4, IGF1R, TLR4, STAT3, SRC, NR3C1, MMP2, MCL1, MET, CDK1, BRAF, KDR, MAPK12 16.39147377ALK, HSP90AA1, PDGFRA, BRD4, HSP90B1, FLT3, JAK1, ABCG2, TGFBR2, PTPRC, ACE, TNFRSF1A, HDAC4, HSPA8, PPARA, CDK6, PRKCD, LCK, NOS3, RAF1, FLT1, TGFBR1, CDK2, HDAC2, PIK3R1, SYK, FYN, CASP3, PTK2, MAPT, PIK3CG, PLG, XPO1, CHEK1, SELE, MMP9, KIT, MMP1, EPAS1, FGFR1, IKBKB, PRKCA, NTRK1, ICAM1, AXL, PLAU, HSP90AB13 7.5651189AGTR1, TLR8, HMGCR, CCR2, CCR3, PPARD, CSF1R, ADAM17, HDAC7, PTPN6, MMP3, MMP13, PLK1, CTSB, SHH, GSR, KAT2B, IDH1, CCR5, SMARCA4, HDAC3, XDH, CDC25A, IDO1, ALDH1A1, WEE1, DNMT3A, PLAUR, NLRP3, CXCR2, HTT, MDM4, JAK3, NOS2, CXCR3, KDM1A, GLI1, CYP19A1, NR4A1, SIRT2, NOS1, NQO1, FASN, TLR9, AURKB, ARG1, TOP2A, CHEK2, PLAT, MPO, NTRK24 7.1541143KLKB1, G6PD, LIPC, TSPO, CRHR1, CHRNA4, CETP, CFTR, SLC6A2, AKR1C4, STS, HSD17B3, HSD17B1, CES1, AKR1C2, APOB, SLC18A2, AKR1C3, DRD3, F9, NR1H4, F11, SOAT1, HTR3A, NPC1L1, DRD1, F10, CYP51A1, SQLE, ADH1B, SLC6A3, SRD5A1, CYP3A4, NR5A1, DHFR, DRD4, CHRM2, CYP17A1, HTR2C, HTR1A, CYP1A25 6.09842125HTR2B, GABBR1, NPY1R, COMT, SIGMAR1, KCNA5, LPAR4, KCNMA1, GRM5, DHCR7, PRKCG, HTR7, HTR2A, PRKCQ, DRD2, PRKCE, EBP, SSTR1, SCN5A, KCNN3, CYP27B1, KCNA3, MAOA, KCNN1, GPR55, MTNR1B, CHRM1, LPAR2, OPRD1, MCHR1, KCNK2, CACNA1C, ADRA2B, PNOC, LPAR3, ADRA2C, CYP1B1, OPRM1, SCN9A, MTNR1A, CYP24A1, ADORA2A6 5.8461438CYP1A1, HSD11B1, NR3C2, HSD17B2, HSD17B7, CYP11B2, CYP2J2, CYP11B1, SRD5A2, AKR1C1, PLA2G1B, ALOX12, HSD11B2, PLA2G2A7 5.1883383LIPE, NR1I2, ACACA, F3, VCP, ANPEP, ACHE, DGAT1, CPT1A, PTGS1, ELANE, PRMT1, FABP4, PIK3CB, CYP2D6, CYP2A6, CYP2C9, CARM1, NR1I3, CTSK, PAK1, INSR, UGT2B7, CYP2C19, CD38, CCR1, CXCR1, MSR1, NR1H3, HPGDS, EPHX1, SCD, DPP48 446NEK2, CDC25C, CDC7, CDC25B9 3.33345PDE10A, PDE3A, DCK, PDE4D10 333AKR1B1, MGLL, FABP511 333MMP12, COL18A1, CTSL12 333PRKCB, EGLN1, TYR13 356CBR1, ADH4, ALDH3A1, MAOB, GLO114 333CA4, CA2, CA1 Functional cluster hubs identified in PPI network. Fig. 6HPLC Chromatograms of standard catechin and OPF1. HPLC Chromatograms of standard catechin and OPF1. LC–MS/MS analysis of the OPF1 Using LC-MS/MS-QTOF, a negative ionization mode of the isolated methanolic fraction from O. parvifolia (OPF1) was performed for higher sensitivity and a more stable response on ions. The separation of these compounds by chromatographic methods achieved distinct peaks (Fig. 7) for flavonoid compounds through the use of a ZORBAX RRHD SB-C18 column with water/acetonitrile (with 0.5% acetic acid) gradient. The MRM chromatogram indicated two major ions at 6.51 min and 8.21 min retention times corresponding to catechin (6.51) and epicatechin (8.21). The accurate mass and MS/MS fragmentation patterns supported the identification of these ions. Catechin had an m/z of 289.1581 with fragmentation products at m/z of 205.1883 and 246.1153 (− 0.6 ppm mass error). Epicatechin had an m/z of 289.1566 with characteristic fragments at m/z of 109.1684 and 245.1508 (1.9 ppm mass error). The low mass errors were indicative of high confidence in the assignment of these compounds. The fragmentation patterns of the two phenolic compounds are in accordance with published LC-MS analyses of flavan-3-ols and confirm that this fraction contained polyphenols. Fig. 7LC–MS/MS profiling of catechin and epicatechin in the OPF1. (A) Base Peak Chromatogram (BPC) showing two major peaks corresponding to catechin (RT 6.51 min) and epicatechin (RT 8.21 min) detected in negative ESI mode. (B) MS spectrum of catechin (RT 6.51 min) showing the deprotonated molecular ion [M–H]⁻ at m/z 289.1581 with fragment ion at m/z 246.1463. (C) MS/MS fragmentation spectrum of catechin confirming characteristic product ions at m/z 205.1883 and 246.1153. (D) MS spectrum of epicatechin (RT 8.21 min) showing the [M–H]⁻ ion at m/z 289.1566 along with fragment ion at m/z 245.1461. (E) MS/MS fragmentation spectrum of epicatechin exhibiting diagnostic product ions at m/z 109.1684 and 245.1508. LC–MS/MS profiling of catechin and epicatechin in the OPF1. (A) Base Peak Chromatogram (BPC) showing two major peaks corresponding to catechin (RT 6.51 min) and epicatechin (RT 8.21 min) detected in negative ESI mode. (B) MS spectrum of catechin (RT 6.51 min) showing the deprotonated molecular ion [M–H]⁻ at m/z 289.1581 with fragment ion at m/z 246.1463. (C) MS/MS fragmentation spectrum of catechin confirming characteristic product ions at m/z 205.1883 and 246.1153. (D) MS spectrum of epicatechin (RT 8.21 min) showing the [M–H]⁻ ion at m/z 289.1566 along with fragment ion at m/z 245.1461. (E) MS/MS fragmentation spectrum of epicatechin exhibiting diagnostic product ions at m/z 109.1684 and 245.1508. HPTLC analysis of OPF1 To ensure that the polyphenol compound catechin was also present in the derived OPFIB fraction, HPTLC chromatographic analysis was performed. Based on HPTLC chromatograms comparing OPFIB to standard catechin, results demonstrate a close correspondence for compounds contained in both fractions that possess similar properties, including catechin. The Rf value was 0.969 for standard catechin and 0.973 for the OPFIB fraction, indicating that both fractions contain constituents with similar chromatographic migration characteristics (Fig. 8). The absorbance units (AUs) of both chromatograms were also consistent with the presence of catechin. The AU of standard catechin was 0.471, while that of the OPFIB fraction was 0.188. Although the amount of compound present in the OPFIB fraction is less than that in the standard catechin, it is evident that catechin has been successfully extracted and retained. The difference in absorbance values probably reflects differences in phytochemical abundance, Fig. 8HPTLC chromatogram of standard catechin and OPF1. HPTLC chromatogram of standard catechin and OPF1. extraction efficiency, and co-elution with lesser amounts of other constituents. Overall, HPTLC data strongly supports the presence of catechin derivatives in OPF1 and demonstrates that the phytochemicals within this fraction contribute to the biological activity seen in the assays conducted, as well as to the cytotoxic and therapeutic activities previously described. Additionally, the similarity between the Rf values and similarities between peak intensities further identify OPF1 as a distinct catechin-rich phytocomponent. qPCR analysis for synthesis genes expression Differential expression of key genes involved in catechin biosynthesis was determined by qPCR analysis of O. parvifolia genomic DNA. DNA was precisely isolated from the plant sample using CTAB method and exhibited a purity of 1.488 as per the nanodrop spectrophotometric quantification. The four genes analyzed, LAR (leucoanthocyanidin reductase), ANR (anthocyanidin reductase), F3’H, and 4CL (4-coumarate-CoA ligase) showed unique amplification patterns that were corroborated by the presence of distinct bands (intensities) in an agarose gel. As GAPDH was used as an internal reference gene, all relative expression values were normalized to this reference (Fig. 9). LAR demonstrated the highest level of expression of 0.3637 folds in comparison with GAPDH. Development of phytoniosomes To maximize the encapsulation of catechin fractions isolated from O. parvifolia, we developed six nano-phytoniosomal formulations using different types and combinations of surfactants, stabilizers, and polymers (F1 to F6) as mentioned in Table 2. These formulations incorporated lecithin as the primary lipid to create the niosomal bilayer structure, folic acid as a targeting ligand, and chitosan as a surface-modifying polymer. The addition of three different types of stabilizing agents: Tween-80; poloxamer-407; and PEG-6000 allowed to optimize the physical properties of the formulations and improve their stability, encapsulation efficiency, and dispersion characteristics. Table 4Fractions identified and isolated through semi-preparative HPLC.S.No.FractionRetention timeTime RangePeak Purity Index1OPF13.8873.50–4.500.9412332OPF25.6245.20–6.500.8647503OPF313.32312.90–13.900.9417314OPF418.46418.00–19.000.9941835OPF520.82219.80–21.000.9507676OPF622.78022.00–23.500.963356 Table 4Fractions identified and isolated through semi-preparative HPLC.S.No.FractionRetention timeTime RangePeak Purity Index1OPF13.8873.50–4.500.9412332OPF25.6245.20–6.500.8647503OPF313.32312.90–13.900.9417314OPF418.46418.00–19.000.9941835OPF520.82219.80–21.000.9507676OPF622.78022.00–23.500.963356 Fractions identified and isolated through semi-preparative HPLC. Fig. 9qPCR densitogram of catechin synthesis gene expression. Gel image cropped for clarity; Whole gel image is available in Supplementary information figure. qPCR densitogram of catechin synthesis gene expression. Gel image cropped for clarity; Whole gel image is available in Supplementary information figure. Cytotoxicity analysis of phytoniosomal formulations The cytotoxicity of the phytoniosomal formulation on SKOV3 OC cells using phytoniosomal formulations F1 through F5 showed clear variations in how effective they were due to the excipient ingredient used in each formulation. The percent of cytotoxicity for F1, F2, F3, F4, and F5 were 75.07 ± 5.75c, 83.47 ± 2.23b, 79.19 ± 0.51c, 83.31 ± 3.61b, 93.17 ± 9.8a, respectively. Of these values, F5 was the most cytotoxic, while F1 was the least cytotoxic. Apoptosis induction by bioactive formulation Flow cytometry analysis showed that F5 treatment caused a significant increase in the percentage of SKOV3 OC cells that underwent apoptosis in a concentration-dependent manner (Fig. 10a). As expected, untreated control cells maintained almost complete viability (99.98%) indicating no spontaneous apoptosis occurred, thus confirming that the flow cytometry assay is a reliable method of evaluating apoptosis. The detection of 80.73% early apoptosis in SKOV3 OC cells treated with F5 at 25 µg/mL indicates the ability of F5 to activate apoptotic signaling pathways even at low concentrations. However, the detection of considerably less total apoptosis (62.44%) when the concentration of F5 was increased to 50 µg/mL suggests that this intermediate range may activate cellular stress pathways that temporarily delay the progression of cells to apoptosis. Fig. 10(a) Flow cytometry plots of apoptosis induction by F5 in SKOV3 cell lines. (b) Dual staining fluorescence images of F5 treated SKOV3 cell lines showing apoptosis. (a) Flow cytometry plots of apoptosis induction by F5 in SKOV3 cell lines. (b) Dual staining fluorescence images of F5 treated SKOV3 cell lines showing apoptosis. The concentration of 100 µg/mL resulted in the highest amount of apoptosis being detected; 97.99% of the SKOV3 OC cells underwent apoptosis. This concentration was also associated with a great deal of late-stage apoptosis (41.97%), which implies that higher concentrations have the capacity to accelerate the transition of cells from early apoptotic stages into later stages of apoptosis. The nearly complete elimination of viable cells (2.01%) at this dosage exemplifies the cytotoxic potential of F5 at their maximum dose (Supplementary Table 2). The lack of necrosis observed in all the treated groups indicates that F5 primarily induce programmed (non-necrotic) cell death, which is advantageous for cancer therapy as it results in less inflammation. Dual staining with Acridine orange and Ethidium Bromide revealed the cellular architecture during apoptosis in treated cell samples (Fig. 10b). Cell cycle arrest determination of bioactive formulation Following treatment with F5, SKOV3 cellular populations were evaluated through flow cytometric analysis utilizing PI staining and assessed for the G0/G1, S and G2/M cell cycle phases (Fig. 11). Analysis revealed a concentration-dependent effect of F5 on the distribution of cells among the three phases of the cell cycle in all treatment groups. At a concentration of 25 µg/mL, there was a higher percentage of cells in G0/G1 (55.50%) with a decrease in the number of G2/M cells when compared to cells from the control. At 50 µg/mL, there was a decrease in the percentage of cells in S phase when compared to control despite an increase in the number of cells in G0/G1 (51.90%); however, the percentage of cells in G2/M was about the same as control. The highest concentration of 100 µg/mL resulted in the greatest shift in distribution with 92.81% of the population in G0/G1 and no G2/M cells (0%); Thus, there were no mitotic cells detected (Supplementary Table 3). Fig. 11Flow cytometry-based Cell Cycle Arrest induced by F5 in SKOV3 cell lines. Flow cytometry-based Cell Cycle Arrest induced by F5 in SKOV3 cell lines. Characterization of bioactive phytoniosomal formulation (F5) Dynamic light scattering analysis Dynamic Light Scattering (DLS) of the phytoniosomal formulation (F5) showed a hydrodynamic particle size average of 229.32 nm, indicating that the formulation was capable of forming nanosized vesicles which are desirable for targeted drug delivery. The polydispersity index (PDI), which measures the uniformity of the size distribution, was found to be 0.318. Zeta potential analysis The assessment of Zeta Potential (ZP) for the developed phytoniosomal formulation (F5) resulted in a very strongly positive ZP of + 42.7 mV, which demonstrated that excellent colloidal stability, and adequate opportunity for development of charge by the combined excipients were achieved. FTIR analysis Analysis of the phytoniosome formulation via FTIR demonstrated that the spectral characteristics (Fig. 12) of each of the major components of the formulation, as well as the catechin fraction being encapsulated, are similar to those of other types of niosomes. Fig. 12Comparative FTIR spectra of constituents and optimized formulation (F5). Comparative FTIR spectra of constituents and optimized formulation (F5). The FTIR spectra of the individual constituents and the optimized phytoniosomal formulation (F5) demonstrated characteristic functional group vibrations confirming the successful incorporation and intermolecular interaction of catechin fraction, folic acid, and the nano-vesicular excipients within the developed phytoniosomal system. Comparative spectral analysis revealed several significant peak shifts, peak broadening patterns, and intensity changes, suggesting effective conjugation and encapsulation of the phytocompounds within the niosomal matrix. The catechin fraction exhibited a broad absorption band in the region around 3200–3500 cm⁻¹ corresponding to the stretching vibration of hydroxyl (–OH) groups associated with polyphenolic compounds. Prominent peaks observed near 1600 cm⁻¹ represented aromatic C = C stretching vibrations, while peaks between 1200 and 1000 cm⁻¹ were attributed to C–O and C–O–C stretching of flavonoid glycosidic and phenolic structures. Folic acid displayed characteristic absorption bands corresponding to N–H stretching and carboxylic acid-associated vibrations, particularly within the regions around 3300 cm⁻¹ and 1650–1700 cm⁻¹, confirming the presence of amide and carboxyl functional groups. Lecithin demonstrated distinct aliphatic C–H stretching bands near 2850–2920 cm⁻¹ together with phosphate-associated vibrations around 1230 cm⁻¹ and 1080 cm⁻¹, characteristic of phospholipid structures. Similarly, chitosan exhibited broad hydroxyl and amino group absorption bands near 3400 cm⁻¹ along with amide-associated peaks around 1650 cm⁻¹, indicating its polysaccharide backbone structure. PEG 6000, Tween 80, and Poloxamer 407 displayed characteristic ether-linked C–O–C stretching vibrations predominantly within the fingerprint region between 1000 and 1150 cm⁻¹, confirming the presence of non-ionic surfactant and polymeric stabilizing components within the formulation. In the optimized F5 phytoniosomal formulation, noticeable spectral modifications were observed compared with the individual constituents. The broad hydroxyl stretching band became comparatively broadened and shifted, indicating extensive hydrogen bonding interactions between catechin polyphenols, folic acid, chitosan, and phospholipid components within the vesicular matrix. Furthermore, attenuation and slight shifting of the characteristic catechin aromatic and phenolic peaks suggested successful encapsulation of the catechin-rich fraction within the nano-vesicular system. The phosphate-associated peaks of lecithin and the ether-linked vibrations of PEG 6000, Tween 80, and Poloxamer 407 were retained in the F5 formulation, confirming structural integrity of the niosomal bilayer and stabilization components following formulation. Additionally, minor peak broadening and reduced intensity in the carbonyl and amide-associated regions indicated possible intermolecular interactions between folic acid and the surfactant-polymer matrix, supporting successful folate conjugation onto the phytoniosomal surface. Importantly, the absence of any major additional peaks or complete disappearance of characteristic functional groups suggested that the encapsulation process did not result in chemical degradation of the active phytoconstituents or excipients. Collectively, the comparative FTIR spectral analysis confirmed the successful incorporation of catechin-rich fractions, folic acid conjugation, and stabilization of the bioactive compounds within the optimized F5 phytoniosomal formulation through hydrogen bonding, electrostatic interactions, and vesicular entrapment mechanisms. X-ray diffraction analysis Figure 13 depicts the solid-state characteristics (XRD analysis) of the bioactive phytoniosomal formulation (F5). The crystalline structure of the catechin fraction as evidenced by the presence of numerous, narrow and distinct diffraction peaks (10–30°) was confirmed by the native catechin XRD pattern. Likewise, no peaks characteristic of catechin are observed on the phytoniosomal catechin diffractogram. The diffractogram obtained from F5 identifies diffraction peaks of approximately 17° to 19° and 31°. These peaks are likely due to the presence of components associated with the formulation, the peaks correlate with the known diffraction signatures for PEG-6000 and for folic acid. In particular, there was a small additional contribution due to the semi-crystallinity of chitosan. No peaks were present that were attributed to crystalline catechin within the composition of the phytoniosomal formulation. Fig. 13XRD diffractogram of F5. XRD diffractogram of F5. Scanning electron microscopy analysis SEM analysis showed that the phytoniosome formulation contained spherical to slightly oval-shaped nanocarriers and that these were all uniform in size (Fig. 14). All measured particle sizes ranged from 150 to 230 nm and were typically 151.2 nm, 179.4 nm, 213.7 nm, 229.2 nm, and thus form consistent and uniform nanocarriers for drug-delivery products. Fig. 14SEM image of F5. SEM image of F5. Drug release profile A bi-phasic pattern was established in the in-vitro analysis of drug release studies of phytoniosome formulation (F5). This study was conducted by allowing a dialysis membrane 150 (HiMedia, India) to conduct the drug into a PBS buffer of pH 7.4 and monitored for a full 24-hour period (Fig. 15). The full 24 hours exhibited an initial ‘burst’ of release for up to 4 hours due to the rapid hydration of the vesicle, and the subsequent diffusion of the catechin molecules that are associated with the vesicle’s surface. Once this initial ‘burst’ was completed, there was then a gradual, sustained ‘release’ of drug over the rest of the 24 h, where the rate of diffusion decreased steadily until it reached an equilibrium plateau (which reflects a slow rate of diffusion). Fig. 15Drug Release profiling of F5 through dialysis model. Drug Release profiling of F5 through dialysis model. Discussion Network pharmacology and molecular target identification The Compounds Target network highlights the relationships between key constituents of O. parvifolia, as well as the proteins of the cells that participate in the development of ovarian cancer, providing information about how these compounds may exert their anti-cancer effects through the same pathway58,59. These types of analyses are essential to determine which molecular targets will be the most effective and also how the phytochemicals in natural products may interact with each other to produce a therapeutic effect in treating a complex disease such as ovarian cancer60. PPI network indicates that the dataset is reflective of a biologically relevant interactive environment. The combination of a very high density of Interaction, almost exclusively as clustered blocks of proteins and a P-value that is statistically significant for the presence of enriched protein categories indicates that these proteins interact together to perform complementary functions within a coordinated biological process61,62. Proteins commonly act together to accomplish multi-step processes, which is typical of networks that support physiological processes, or pathways that respond to stressors63. Therefore, these proteins likely work together to mediate biological processes that are important for the ovarian cancer regulation. The selection of the top 10 proteins from the protein–protein interaction (PPI) network was primarily based on degree centrality analysis, which identifies highly connected hub proteins within the interaction network. Proteins with higher degree values are considered biologically significant because they participate in multiple signaling interactions and frequently function as central regulatory nodes governing cancer-associated cellular processes64. In ovarian cancer, hub proteins are often implicated in the coordination of proliferation, apoptosis resistance, inflammatory signaling, angiogenesis, and metastatic progression. Therefore, prioritizing proteins with the highest degree centrality enabled the identification of biologically influential targets most likely to contribute to ovarian cancer pathogenesis and therapeutic response65. Furthermore, the selection strategy was intended to maintain computational feasibility and analytical focus during downstream molecular docking and molecular dynamics analyses while ensuring representation of key oncogenic pathways66. The final target selection was additionally supported by biological relevance reported in ovarian cancer literature, including involvement in PI3K/AKT signaling, inflammatory cytokine regulation, apoptosis, estrogen receptor signaling, and tumor microenvironment modulation. Although other proteins within the network may also contribute to ovarian cancer progression, the present study prioritized the top-ranking hub proteins as representative molecular targets for evaluating the multi-target therapeutic potential of catechin fractions67. Biological significance of mutation oncoplot and GO enrichment analysis The Gene Ontology analysis especially as in biological processes, it is evident that the pathways enriched are the steroid, fatty acid, and alcohol metabolic processes that are the main contributors to the biology of ovarian cancer. The rate at which tumor cells grow is exponentially dependent upon how often they engage in lipid re-modelling for membrane production, rapid cell growth, and oncogenic signalling68. For example, when there is downregulation and/or modulation of key enzymes involved in the biosynthesis of steroids, the level of estrogen-dependent proliferation also decreases which slows the rate of tumor growth. Additionally, changes that occur to the metabolism of fatty acid derivatives can lead to decreased production of pro-oncogenic lipid mediators and can interrupt the inflammatory processes that occur within the tumor microenvironment69. Collectively, the enriched metabolic pathways may be used by the proteins contained in these pathways to help restore proper lipid homeostasis, therefore slowing progression of ovarian cancer70. The effects of drug response and the phosphorylation of serine residues indicate a role of these proteins in a drug-sensitive response and cellular intracellular signalling71. Chemoresistant forms of ovarian cancer typically have alterations in the phosphorylation-dependent pathways (e.g. PI3K/Akt, MAPK, NF-κB) as a result of alterations in the regulation of oncogenic kinases. Therefore, by affecting those pathways, the proteins in this network may be able to restore drug sensitivity to chemotherapy72. Furthermore, the presence of proteins in the pathways regulating the modifications of serine residues provides further evidence that they may have an inhibitory effect on oncogenic kinase signalling pathways73. The cellular component data indicated significant enrichment for the presence of membrane raft-related proteins and their ability to facilitate cell–cell communication, clustering of receptors, and propagation of signals (very exuberant in ovarian cancer). Thus, any activity associated with receptors located in rafts may inhibit oncogenic pathways such as EGFR, HER2, and VEGFR74. The large amount of postsynaptic/presynaptic membrane-related proteins associated with neurotransmitter-like signaling axes indicates that some of these signalling axes have a regulatory role in ovarian microenvironments and metastases75. Enrichment for molecular function depicted the proteins that are involved in the steroid-binding, drug-binding, nuclear receptor activity and transcription factor activity, resulted in an overall enrichment for functions that have potential to mediate and/or inhibit hormone signal transduction, as well as the induction of transcriptional programs responsible for promoting the proliferation, angiogenesis and survival of ovarian cancer cells76. Enrichment for oxidoreductase activity, as well as redox-associated proteins, suggests functions that regulate an antioxidant and redox environment77. The accumulation of oxidative mutations and occurrence of chemoresistance due to oxidative stress highlights the need for regulation of redox homeostasis to create less aggressive tumours and improve the response of patients to therapy78. The results of MAF oncoplot suggests that, for serous adenocarcinomas of the ovaries, the examined genes in the panel do not present high frequency rates of mutation, despite their established biological significance as mediators of inflammation, adaptation to hypoxia, metabolic regulation and apoptotic regulation. The low frequency of the found mutations in the case of STAT3 is consistent with findings that activation of STAT3 in serous adenocarcinoma is often driven more by mediators in the cytokine pathway (i.e. IL6/JAK signal transduction) rather than by direct alterations at the level of the genome. Hence, although a few mutations may have an effect in specific cases, they would not represent a predominant molecular mechanism for the entire patient group studied here79. Likewise, the low mutation rates of PPARG and ESR1 show that ovarian serous adenocarcinoma does not occur in the context of hormonal dysregulation at the genetic level. Thus, modification of estrogen or metabolic signaling pathways will primarily result from differentially expressed genes, epigenetic alterations, or interactions with the tumor microenvironment than from somatic mutations80. The low level of change that has been measured in HIF1A, MMP9, and IL6 further supports our assertion that the major determinants of hypoxia adaptation, extracellular matrix remodeling, and inflammatory response representing the hallmark features of advanced ovarian cancer will be transcriptional activation and resulting activation of signaling cascades rather than mutations in the corresponding genes, which have infrequently been found on repeated analysis81. Similarly, the infrequent forms of AKT1, BCL2, and CASP3 indicate that the dysregulated apoptotic process associated with ovarian serous adenocarcinoma results primarily from an alteration at the pathway level rather than at the structural or genomic level within these genes82. The analysis has also revealed distinct mutational patterns among the selected target genes associated with ovarian carcinoma. In cancer biology, genomic alterations are generally categorized into “driver mutations” which directly contribute to tumor initiation and progression, and “passenger mutations” which accumulate during tumor evolution without exerting significant oncogenic effects83. In ovarian cancer, classical driver mutations such as TP53 are highly prevalent and are strongly associated with genomic instability, aggressive tumor behavior, and poor prognosis84. In contrast, several proteins identified in the present study, including STAT3, AKT1, IL6, ESR1, and VEGFA, are more frequently dysregulated through aberrant signaling activation, inflammatory responses, epigenetic modulation, or tumor microenvironment interactions rather than through recurrent structural mutations. Therefore, some of the comparatively lower-frequency mutations observed in the present analysis may represent passenger mutations with limited direct oncogenic influence85. Nevertheless, despite lower mutational frequencies, these proteins remain biologically significant because their abnormal expression and pathway hyperactivation play critical roles in ovarian cancer progression, angiogenesis, immune evasion, and resistance to apoptosis. This observation highlights the importance of integrating mutation profiling with functional pathway analyses to identify therapeutically actionable targets beyond highly mutated driver genes alone86. Molecular docking and dynamics interpretation The molecular docking analysis with AKT1 which plays a central role in the phosphatidylinositol 3-kinase (PI3K) pathway, and becomes frequently activated in ovarian cancer shows a strong binding of catechin to the active region of AKT1 suggests that catechin may inhibit AKT1 kinase activity and block the downstream prosurvival signals that keep cancer cells alive and growing. Therefore, the combination of both catechin’s ability to bind to AKT1 and to inhibit its activity may make ovarian cancer cells more susceptible to therapeutic agents used in the treatment of cancer87. STAT3 is a key regulator of the inflammatory microenvironment in ovarian cancer and plays a role in promoting metastasis and the maintenance of cancer stem-like cells. The inhibition of STAT3 would and does disrupt oncogenic signaling via IL6, reducing tumor progression. The affinity of catechin for STAT3 indicates that catechin may disrupt the dimerization or DNA-binding activity of STAT3 and block the transcriptional programs that contribute to tumor growth in the ovaries88. As IL6 is a major activator of the IL6/JAK/STAT3 pathway involved in inflammation and cancer development, it appears that catechin may help block this pathway if it could form stable hydrogen bonds with proteins that bind to its active site. If catechin can block the IL6 receptor site, then it has the potential to decrease STAT3 activation down the line from the receptor to the nucleus. Blocking IL6 receptors may also decrease pro-coagulant factors, which are linked with many types of tumors, including ovarian cancer. In addition to blocking IL6 at the receptor level, catechin may also disrupt other important pathways involved in tumor formation89,90. In ovarian cancer, one of the hallmarks of the disease is the resistance to the normal process of cell death, or apoptosis, which leads to an increase in disease progression and recurrence. By interacting with CASP3, catechins could serve to restore some apoptotic sensitivity, thereby allowing for the controlled death of the cancer-cell population91. The results of simulation dynamics indicate that catechin interacts strongly and stably with ESR1, preserving the structural integrity of ESR1 while generating stable intermolecular bonds. Due to the role of ESR1 in ovarian cancer proliferation, these results also support the potential for catechin to be used as a selective modulator to disrupt estrogen induced oncogenic pathways92. The stability and the binding of catechin can be attributed to its physicochemical properties, including low-energy torsional states, compact radius of gyration, adequate exposure to hydrophilicity, and favorable polar surface properties. The overall results support catechin’s strong, stable, and sustained mode of binding with AKT1 within its active site. Given that AKT1 is a major regulator of survival and proliferation and chemoresistance in ovarian cancer, catechin’s ability to bind stably is indicative of its potential as a natural AKT1 modulator to disrupt oncogenic PI3K/AKT signalling93. Collectively, these findings suggest that catechin may exert multitarget inhibitory effects against ovarian carcinoma by simultaneously modulating survival signaling, inflammatory pathways, angiogenesis, hormone-mediated proliferation, and apoptosis-associated proteins. The multitarget nature of catechin interactions identified through docking analysis supports the subsequent in vitro observations of cytotoxicity, apoptosis induction, and cell cycle arrest in SKOV3 ovarian cancer cells. However, these observations represent predictive computational insights, and further biochemical and protein expression studies in vivo are required to experimentally confirm the precise inhibitory or activating effects of catechin on these molecular targets. Phytochemical isolation and catechin validation HPLC guided fractionation indicates the successful separation of catechins with additional polyphenolic constituents, such as other flavonoids phenolic acids, and their derivatives, which have been previously documented to exist in O. parvifolia31,94. The gradients that provided this separation have indicated that the mobile phase at each point has allowed for the separation of compounds over a wide range of polarities. The derived yield not only indicates a good representation of the various types of polyphenols found in the methanol extract, but also demonstrates the high degree of efficiency achieved in the chromatographic separation. It was also found that the fraction OPF1 produced high amounts of catechin compounds, which are associated with antioxidant, anti-inflammatory, and potential anticancer effects attributed to O. parvifolia. HPLC has been used for many years to isolate, separate and purify polyphenolic compounds from plant extracts95. Polyphenol structures are diverse, ranging from simple phenolic acids to complex flavonoids and tannins. Due to the diversity of polyphenols, it is necessary for a separation technique to be able to resolve, separate, and produce identical results regardless of quantity or batch used. The results presented in this study support that HPLC is effective at isolating individual compounds from complex botanical matrices96. Gradient elution techniques utilizing ‘acidified’ aqueous, organic solvents (i.e., acetonitrile or methanol) for separation allow for a higher degree of separation efficiency97. Gradient elution takes advantage of polarity and hydrophobicity differences of the compounds being separated to achieve the highest separation efficiency98. Catechin/epicatechin (flavan-3-ols) are examples of structurally similar chemical entities separated effectively with HPLC. The narrow range in peak purity indices for the flavan-3-ols indicates that HPLC provides excellent resolution resulting in the exaction of defined fractions that can be utilized in future structural and biochemical investigations99. The ability to detect catechin and epicatechin using LC-MS/MS-QTOF illustrates the ability of this method to effectively identify heavily structurally similar flavonoids that commonly co-elute when using more traditional analysis techniques100. Negative ESI mode was shown to be beneficial for the detection of phenolic compounds because of the acidic functional groups associated with the compounds, which improved sensitivity for the identification of such compounds101. The ability of this type of analysis to provide accurate mass determination and structural information in the same analysis allows for the determination of the bioactive components of complex plant matter102. As a whole, the results obtained from this analytical process confirmed that the components isolated from O. parvifolia are primarily flavanols, specifically catechin and epicatechin, which possess several bioactivities. The results obtained through this analytical approach demonstrate the power and capability of LC-MS/MS-QTOF to perform phytochemical characterization, providing a high degree of accuracy with regard to mass determination, and providing reliable identification of phenolic metabolites found in O. parvifolia103. Gene expression analysis of catechin biosynthesis In qPCR quantification of catechin synthesis genes, LAR demonstrated the highest relative expression value of 0.3637, which is approximately fivefold greater than the expression levels of the other three genes, indicating that there is increased conversion of leucoanthocyanidins into catechins; furthermore, the detection of catechins in isolated fractions by LC-MS aligns with this finding. With a relative expression value of 0.2746, ANR had the second highest relative expression value and thus supports its role in synthesising epicatechin104, another significant flavanol identified by LC-MS detection of isolated fractions. LAR and ANR function as centre points in the last few steps of catechin biosynthesis and their coordinated expression indicates a strong flow of metabolites to form catechin and epicatechin105. Moderate expression (0.2466-fold) of F3′H (Flavonoid 3′-Hydroxylase) indicates enough precursor modification to produce dihydroquercetin, a major feedstock for both the LAR and ANR pathways106. Conversely, expression of 4CL (4-Coumarate-CoA Ligase) was lowest (0.1331-fold), indicating that phenylpropanoid pathway early steps could be rate-limiting for catechin production107 from O. parvifolia. However, while lower, 4CL activity is required to generate 4-coumaroyl-CoA, which is the starter compound for the flavonoid pathway108. The qPCR data support the findings from LC-MS by confirming the presence of catechin and epicatechin, both of which demonstrate active expression of the necessary genes for flavanol biosynthesis. The significantly higher expression of LAR and ANR demonstrates highly efficient terminal step of the pathway and directs metabolic flux towards the accumulation of catechin109. These findings collectively indicate that O. parvifolia is a potential natural source of catechin-related bioactive polyphenols and provide a molecular understanding of flavonoid biosynthesis. Development of phytoniosomes With respect to the phytoniosomes, the structural integrity of the niosomal bilayer was provided by lecithin, which also ensured that catechin would remain chemically stable in a protective environment. Due to its amphiphilic structure, lecithin efficiently entrapped both hydrophilic and limitations lipophilic phenolics19,110. The addition of Tween-80 in some of the formulations (F1 F3, and F5) dramatically improved the emulsification process by providing greater reduction of interfacial tension as the vesicles form, leading to the production of smaller particles with improved homogeneity. Tween-80 also aided in the dissolution of catechins in the system, leading to an increase in the entrapment efficiency111,112. The Poloxamer-407 used in the formulations (F2, F5, and F6) provided steric stabilization of the vesicle colloidal form, reducing the tendency to agglomerate and providing improved long-term colloidal stability113. Poloxamer-407 produced a hydrophilic “coat” around the nanoparticle vesicles, which reduces the degree to which any pair of nanoparticles will come into close contact. This leads to reduced polydispersity index values, which is essential for retaining the aggregation characteristics necessary for vesicles in the nano range to enter cells46,114. Inclusion of PEG-6000 (F3, F4, F5, F6) gave extra steric stability and controlled release to the Phytoniosomes, as the PEG chains cause more hydration to occur around the vesicles, preventing the fusion and subsequent leakage of the encapsulated catechins. The addition of PEGylation has been shown to limit the amount of opsonization, allowing the phytoniosomes to remain in circulation longer and have improved bioavailability115–117. Chitosan was included in all formulations at a concentration of 0.1% by weight. Chitosan has a positive charge on its surface due to protonated amine groups, which enhance mucoadhesion, stability, and interaction with biological membranes that are negatively charged. The presence of chitosan will also slow the diffusion rate of catechins from the vesicles, contributing to the sustained release effect45,116,118–120. Folic acid is an active targeting ligand and thus, increases the likelihood of the phytoniosomes being taken up preferentially by ovarian cancer cells that overexpress folate receptors. This provides an additional layer of functional specificity to the phytoniosomes and enhances their therapeutic potential, while reducing potential off-target effects121–123. The combination of Tween-80, Poloxamer-407, and PEG-6000 in the formulations (particularly in F5) possessed the best combination of benefits by achieving optimal physicochemical properties and maximizing the reduction of particle size, increased encapsulation efficiency and stability, and targeted delivery of catechin-rich fractions. The reason the F1 formulation exhibited a lower level of activity is related to its simple excipient composition; F1’s excipient contained only Tween-80 as a surfactant. While Tween-80 enhances the solubility of a phytofraction, the lack of stabilizing excipients likely resulted in less stability of the vesicle and, therefore, less efficient cellular uptake124. Conversely, the F2 and F4 formulations exhibited a greater percentage of cytotoxicity than the F1 formulation because of the presence of excipients that stabilise the vesicles and allow for more effective interaction with the cancer cell membranes125; Poloxamer 407 was added to the F2 formulation, and PEG 6000 was added to the F4 formulation. The values for the F2 and F4 formulations are very similar; therefore, it is indicated that both stabilisers had a similar effect independently in improving the delivery of phytoniosomes. The F3 formulation contained Tween-80 as the surfactant and PEG 6000 as the stabiliser, which resulted in a moderate level of cytotoxicity toward SKOV3 cells compared to the F1 formulation. The combination of excipients in the F3 formulation did provide a greater level of solubility and steric stability than Tween-80 alone; however, the lack of Poloxamer 407 was likely a limiting variable in terms of achieving optimal vesicle organisation and internalisation of the phytofraction, thereby resulting in a lower level of activity for F3 relative to F2 and F4126. The enhanced bioactive delivery of F5 is due to the synergistic effects of Tween-80, Poloxamer 407 and PEG 6000, creating a more uniformly stable and easily absorbed triple-component vesicle compared to Tween-80 and Poloxamer 407. The combination of these three components provided a much more stable and uniform vesicle, allowing enhanced bioactive delivery to SKOV3 cells. Anticancer activity and mechanistic implications Furthermore, as F5 exhibited most cytotoxicity in SKOV3 cells, it acted as a neuroprotective formulation (IC50 – 266.38 µg/mL) that did not cause any cytotoxicity in SH-SY5Y cell lines. This shows that the formulation would not affect the CNS metabolism and hence deemed neuroprotective in nature. The results of apoptosis induction assay demonstrate a strong pro-apoptotic effect of F5 on SKOV3 cells. The highest concentration (100 µg/mL) produced the greatest efficacy. The increased involvement of the intrinsic apoptotic pathways as evidenced by the increase in late-stage (compared to early-stage) apoptosis leads to the conclusion that phytoniosomal formulations have significant potential for use as an anticancer agent in the treatment of ovarian malignancies. Reasons for the observed results include the ability of catechin and folic acid to target cancer cells and induce apoptosis through various mechanisms, including the activation of caspases and the modulation of cell cycle regulatory proteins127. Moreover, the use of a phytoniosomal formulation, particularly F5, enhances the delivery and bioavailability of these bioactive compounds, thus maximizing their therapeutic efficacy against SKOV3 ovarian cancer cells. This enhanced delivery can lead to increased oxidative stress and mitochondrial dysfunction in ovarian cancer cells, ultimately contributing to apoptosis128. The treatment with F5 caused a dose-dependent blockade of SKOV3 cells’ ability to progress through the cell cycle, which primarily happened within the G0/G1 phase. In addition, this substantial increase in the G0/G1 population at a concentration of 25 µg/mL indicates that early G1 phase cell-cycle checkpoints may have been triggered, thereby preventing the transition into S-phase. The decrease in G2/M phase cells, along with the substantial increase in the G0/G1 cell population also indicates that once a G1 arrest has occurred, further progression down the cell cycle to mitosis has been reduced significantly. At 50 µg/mL, while G0/G1 continued to increase as did the percentage of G1 with respect to total number of cells, the change in S-phase (decreased) certainly indicates that the level of DNA replication activity has decreased significantly. However, the G2/M phases’ ratio remained unchanged indicating that at this concentration G1 had partially arrested, and therefore there remains a subpopulation of SKOV3 cells that can still complete progression through the cell cycle. A different pattern of cell distribution occurred after 100 µg/mL, with nearly complete distribution to G0/G1 and absent distribution to G2/M. A total block of mitosis or severe will occur as a result of either cellular stress response or DNA damage response. This suggests either very high p53 or p21 activity levels or targeted activation of various checkpoint regulatory proteins to produce G1 arrest and totally prevent entry into S and mitosis. In summary, the profound inhibition of the entry of SKOV3 cells into S-phase and their entry into mitosis represent a potent and effective cytostatic activity on SKOV3 cells leading to a disruption in their proliferative potential129–131. Physicochemical characterization of bioactive phytoniosomes The DLS experiment indicates that there is moderate uniformity within the size distribution. While generally a PDI below 0.3 is considered ideal for monodispersity and is an indicator that most of the phytoniosomes have approximately the same size with only limited aggregation. The DLS data provides substantial evidence that the F5 phytoniosomes provide ideal size, homogeneity and colloidal stability, thus, ensuring efficient drug delivery and targeted transport of catechins to ovarian tumours. The F5 formulation presents significant evidence of reliable measurement quality with a high-count rate of 932.8 kcps and a correlation intercept of 0.8872. In addition, there is a narrow peak with no secondary populations suggesting that the F5 formulation does not contain large particles or debris. Nanocarriers that measure 150 to 250 nm have shown to have greater stability in circulation, low clearance via kidney, and allow greater absorption through the enhanced permeability and retention (EPR) effect into tumours. A consistent size distribution also allows for predictable cellular absorption, consistent biodistribution and reproducible therapeutic response132–134. The combination of non-ionic surfactants and PEG-stabilizers provide steric hindrance, which prevents vesicle aggregation, while possibly facilitating electrostatic stabilization of the vesicles. Chitosan appears to provide additional electrostatic stabilization of the vesicles and thus, acts to maintain vesicle integrity135,136. Based on the ± 30mV electrostatic stability threshold, the obtained ZP of + 42.7 mV indicates that the vesicles exhibited a sufficient quantity of repulsive forces to prevent aggregation and sustain their total structural integrity throughout the duration of their storage and biological circulatory process. The formulation consisted of lecithin-based bilayers stabilized by Tween 80, PEG6000, and Poloxamer 407 and included the addition of chitosan early in the preparation process, resulting in vesicles that exhibited well-defined cationic surface environments137,138. Each component contributed to the totality of the electrokinetic profile exhibited by this formulation. Lecithin provided the core bilayer matrix for encapsulation of catechin fraction, while Tween 80 increased fluidity of the membrane and decreased the tendency for fusion of vesicles. PEG6000 and Poloxamer 407 were employed as steric stabilizers to enhance dispersion and minimise protein adsorption in vivo. Chitosan, which is a cationic biopolymer, imparted strong positive charge through protonated amino groups, thus ensuring consistent electrostatic repulsive forces among vesicles, as well as improved dispersion stability45,120. Beyond providing an alternative charge for use in targeted therapies against ovarian cancer, the cationic phytoniosomal carriers take advantage of the negative charge on the membranes of ovarian cancer cells (high-grade serous) due to increased sialylation and irregular glycosylation. The interaction between the positively charged phytoniosomal nanocarriers and the negatively charged cancer cell membranes enhances adsorption and penetration of the phytoniosomal nanocarriers and increases their endocytic uptake, thus maximizing their usefulness as drug delivery vehicles. The positive charge also enhances drug delivery through folate receptor-α (FRα) targeted therapy, which is expressed at much higher levels in ovarian carcinoma than in normal healthy cells139,140. The enhanced level of positive charge also provides a stabilizing effect on the catechin fractions that would otherwise be rapidly degraded due to oxidation. This added layer of electrostatic stability, sterically shielding, and receptor-specific targeting contributes to reducing the chances for premature release of drugs and to the maintenance of the sustained release of drugs to the tumor. Additionally, these physicochemical attributes are conducive to enhancing the concentration and localization of the phytoniosomal carriers at the tumor site. As a result, all of the integrated physicochemical properties of the phytoniosomal system suggest that it is a unique and highly effective and selective method for delivering drugs specifically to ovarian cancer cells21,141. The disappearance of catechin specific crystal reflections for the F5 formulation in the diffractogram implies that the catechin-containing fraction has been transformed from a well-ordered crystalline (3D structure) form to an amorphous or randomly-distributed (i.e., molecularly dispersed) state as it becomes encapsulated within a phytoniosome system (i.e., combination of lecithin and chitosan). This loss of long-range order indicates that catechin molecules are uniformly distributed throughout the lecithin bilayer and polymer coating. The high degree of dispersion prevents catechin from forming a new crystalline solid during and after dispersion. The narrow peaks observed in the diffractogram corresponding to PEG-6000, folic acid, and some of the diffracted intensity from the chitosan used to prepare the physical solution for F5 represent the specific structural properties of each excipient. Each excipient retains some degree of crystallinity even when hydrated and self-assembled in the presence of the other excipients. While the presence of the excipient-related diffraction peaks in the diffractogram does not hinder the amorphous state of the enclosed drug, they do reflect specific structural characteristics that aid in the vesicle stability, targeting and migration via folate moieties (folic acid), and adhesion to mucosal membranes (chitosan or any other polysaccharides)142,143. The semi-amorphous nature of the F5 formulation described earlier is consistent with reports regarding interaction between polymer–surfactant in niosomal systems (e.g. poloxamer 407 and Tween 80) whereby each of these components intercalate into the lipid bilayers disrupting the crystalline packing of the bilayer lipids and leading to broadening of the diffracted intensity144–146. The presence of structures with this degree of disorder may provide the potential for enhancing the solubility, dissolution, and absorption behaviours of the formulations produced from polysaccharides or with polysaccharide-based methods. The FTIR results confirm that the catechins were contained within the composite phytoniosomal system and was stabilized through the utilization of lecithin and nonionic surfactants, and then coated/functioned by chitosan and folic acid. The shifts and broadened peaks within the FTIR spectrums indicate that any hydrogen bondings, as well as possible electrostatic contacts, were responsible for promoting the interaction between the two before being chemically modified through attachment. Therefore, it may be concluded that the active phytochemicals remain chemically unchanged, while being contained/stabilized within the carrier. Furthermore, there were no new bands due to oxidation (such as removing the ester’s carbonyl group (C = O) or forming carbonyls outside of their expected ranges). This indicates that when they formulated the product, they maintained both elements of the formulation chemically intact - both the lipids and the bioactive compound147. The surface textures of most vesicles were smooth, indicating that lecithin, Tween-80, Poloxamer-407, PEG6000, and a coating with chitosan–folic acid created a single-layer lipid/polymeric bilayer. The absence of surface granulation, cracks, and fragmentation demonstrates the continued structural integrity of the vesicles, as well as their continued stable formation. Several smaller particles located around the larger vesicles correspond to either secondary vesicles or residual polymer aggregates that do not affect the morphological characteristics of a hybrid nanoformulation144,148,149. These results demonstrate the effectiveness and control associated with the preparation method by the relative uniformity of particle shapes and size distributions, thus providing evidence for the successful control of vesicle self-assembly. The nanoscale size range also provides significant advantages for cancer-targeted delivery because nanoparticles between 100 and 250 nm exhibit enhanced permeability and retention (EPR) effects and allow accumulation in tumour tissues. In addition, the appearing resemblance of continuous and smooth surfaces resulted in aggregating less, thus providing better colloidal stability and more reliability in release behavior consistent with the observed sustained-release kinetic profiles of these formulations. Also, a review of the SEM data confirms that these phytoniosome formulations have desirable nanoscale morphology, as well as high-level structural stability, supporting its use as an innovative nancarrier system for catechin delivery in anticancer purposes22. The characteristic ‘long-tail’ shape of the released cumulative drug versus time curves from study performed using dialysis model on multiple replicates indicates that the F5 formulation exhibits a slow, controlled diffusion of the encapsulated catechins from within the internal lipid-polymer membranes, with the rate of drug migration determined by matrix (Higuchi type) diffusion through the lipid polymer membranes as demonstrated by the smooth transition of the release rate from an initial ‘burst’ to a gradually decreasing rate of diffusion to an equilibrium level. The uniformity of the cumulative drug versus time curves generated from the multiple replicates substantiate the reproducibility and stability of the F5 formulations. The extended time of the slow-release phase represents the controlled diffusion of the encapsulated catechins through the respective (non-porous) lipid-polymer membranes, and polymeric matrices formed from lecithin, Tween 80, Poloxamer, PEG6000, and the Chitosan-folic acid coating all of which greatly reduce the permeability of the vesicles and enhance the stability of the encapsulated catechins. As a result, the findings on release kinetics indicate that F5 phytoniosomes are an acceptable method of sustained-release drug delivery, which is capable of providing long-lasting drug concentrations throughout the course of the release study, thereby supporting their clinical applicability for long-term drug administration in oncology46,150–152. Conclusion In conclusion, the folate conjugated catechins from O. parvifolia extract encapsulated in a phytoniosomal formulation can be utilized as a targeted approach to deliver anticancer therapeutics. An advantage of this mode of delivery is that it is based upon the enhanced permeability and retention effect, which allows for the selective accumulation of therapeutic substances within the tumor microenvironment through nanoscale carriers. As a result, the efficacy of these products will be maximized while their toxicity will be eliminated from the systemic circulation. The phytoniosomal formulation of catechin represents a significant future opportunity for developing drugs specifically for treating cancer, as it utilizes natural catechins to develop targeted anticancer therapies. Future efforts should aim to examine the in vivo pharmacokinetics and efficacy of the catechin phytoniosomal formulation in models of cancer as a means of further validating the preferential accumulation of the formulation within the tumor based upon the EPR (enhanced permeability and retention) phenomenon, and the development of production methodologies that meet the requirements for GMP-compliance related to the production of drug products. The development of GMP-compliance manufacturing methodologies will permit the initiation of human clinical trials and evaluation of the bioavailability, tolerability, and selective targeting of the phytoniosomal formulation of catechin to cancers known to be responsive to catechin therapy. To enhance the efficacy of the phytoniosome formulations of catechin, additional preclinical studies should evaluate the potential of utilizing the formulation in conjunction with other anticancer agents such as Doxorubicin and immuno-therapeutics, to enhance the anti-tumor effects of these agents while decreasing the off-target toxicities of these agents. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1