Folate-Targeted polymeric micelles enhance the anti-glioma efficacy of pterostilbene in an orthotopic rat model Authors: Xiaobo Yang, Yuze Sun, Huan Jin, Hailin Zhou, Yunqiao Yang, Yanan Su, Jinan Wang, Zhiwei Zhao, Wenyao Yu, Moli Wu, Qian Wang, Conghongting Liu, Wen Sun, Changyuan Wang, Chong Wang, Li Lv, Mingzhan Li, Jing Ning, Brant E. Isakson, Yinan Wang, Xiaohong Shu Published: 2026 ABSTRACT The clinical translation of Pterostilbene, a neuroactive stilbene with broad therapeutic potential, is hampered by its poor aqueous solubility and insufficient brain accumulation. To address these limitations, we developed a folate receptor-targeted polymeric micelle system (F–Pt/M) using poly (ethylene glycol)-block-poly (ε-caprolactone) copolymers. The optimized F–Pt/M exhibited a small hydrodynamic diameter (∼133 nm), high drug loading, and increased the apparent solubility of pterostilbene by 25-fold. In folate receptor-expressing U251 and RG-2 glioma cells, F–Pt/M significantly enhanced cellular uptake via folate-mediated endocytosis and showed superior cytotoxicity compared to non-targeted micelles. In an orthotopic RG-2 glioma model, intravenous administration of F–Pt/M improved brain accumulation and prolonged median survival by 76.5% relative to free drug. Mechanistically, the enhanced anti-glioma efficacy was associated with induction of apoptosis and inhibition of the STAT3 signaling pathway. These findings establish folate-targeted micelles as a robust and versatile nanoplatform for delivering poorly soluble bioactive compounds to the brain, with pterostilbene representing a promising candidate for further neuro-oncological development. Introduction Glioblastoma (GBM), the most aggressive brain tumor, poses formidable therapeutic challenges due to its highly invasive nature and limited treatment options. Accounting for approximately 60% of primary intracranial tumors, gliomas contribute significantly to global cancer mortality [1]. While current standard therapy combines surgical resection with radiotherapy and chemotherapy [2], the efficacy of chemotherapeutics is severely constrained by the blood-brain barrier (BBB) and blood-brain tumor barrier (BBTB), which exclude >98% of small-molecule drugs from reaching the tumor site [3], frequently leading to treatment failure and tumor recurrence. Pterostilbene (3,5-dimethoxy-4′-hydroxy-trans-stilbene), a naturally occurring dimethylated analog of resveratrol predominantly found in blueberries and grapes [4], exhibits superior metabolic stability, potent neuroprotective [5,6] and enhanced antitumor activity [7] compared to its parent compound, along with a favorable safety profile reported in preclinical studies [8,9]. However, the translation of its potential into practical applications is severely hampered by significant pharmacokinetic challenges, most notably its extremely poor aqueous solubility [10] and consequent low systemic bioavailability [11]. This limitation is a common critical hurdle for many lipophilic bioactive phytochemicals, drastically reducing their effective concentration at the brain target sites and necessitating the development of advanced delivery strategies. To address these challenges, various drug delivery strategies have been developed, including polymeric micelles [12], nanoparticles [13], ultrasound/microbubbles [14], carbon nanotubes [15], and liposomal systems [16]. Of these, polymeric micelles have emerged as particularly promising candidates, with several formulations advancing to clinical trials [17]. These nanocarriers, formed by self-assembling amphiphilic block copolymers, feature a hydrophobic core for drug encapsulation and a hydrophilic shell for biocompatibility. Among the copolymers, methoxy-poly (ethylene glycol)-block-poly (ε-caprolactone) (MPEG-PCL) stands out as an ideal nanomaterial due to its non-toxic, biocompatible, and biodegradable properties [18]. Studies have shown that MPEG-PCL micelles can enhance the solubility of drugs, control drug release, prolong the duration of drug action, and mitigate macrophage uptake [19], showing promise for clinical applications. To further achieve site-specific delivery and enhanced cellular internalization, active targeting ligands can be conjugated to the nanocarrier surface. A recent comprehensive review has systematically surveyed the landscape of nanoparticle-based strategies for GBM, with particular emphasis on targeting systems that exploit receptors overexpressed on both the BBB and GBM cells, highlighting that surface functionalization with appropriate ligands enables both BBB transcytosis and tumor-specific accumulation [20]. In a recent study, Riccobelli et al. [21] demonstrated that biotin-decorated inulin-based polymeric micelles significantly enhanced cellular uptake in GBM cells and exhibited increased cerebral localization in zebrafish models via biotin receptor-mediated targeting, validating the targeting strategy for improving drug delivery across the BBB and into GBM tissues. Folic acid (folate), a vitamin with high affinity for a folate receptor (FR), represents an ideal targeting moiety. Notably, FR is frequently overexpressed on various metabolically active and diseased cells (e.g., cells of the blood-cerebrospinal fluid barrier [22,23] and glioma cells [24]), providing a strategic avenue for targeted delivery not only in oncology but also in enhancing the cellular uptake of bioactive compounds [25]. In this context, we hypothesized that preparing folate-decorated polymeric micelles would create a robust nano-delivery platform to overcome the pharmaceutical limitations of Pterostilbene and enhance its efficacy against glioma. This study aimed to: (i) prepare and characterize folate-targeted, Pterostilbene-loaded micelles (F–Pt/M); (ii) evaluate their ability to enhance solubility, cellular uptake, and in vitro cytotoxicity in glioma cells; (iii) assess their in vivo pharmacokinetics, brain biodistribution, and anti-glioma efficacy; and (iv) investigate the underlying molecular mechanism, focusing on STAT3 signaling. By successfully demonstrating this targeted nanoplatform, our work provides a strategy not only for improving the delivery of Pterostilbene but also for advancing the broader application of high-value hydrophobic therapeutic agents for glioma treatment. Materials and methods Materials Pterostilbene (Pt, purity ≥99%) was obtained from Baoji Herbest Bio-Tech Co., Ltd. (Shanxi, China). The block copolymers mPEG2000-PCL2000 and FA-PEG3400-PCL2000 were purchased from Xi'an Ruixi Biological Technology (Shanxi, China). High-performance liquid chromatography (HPLC)-grade solvents including methanol and acetonitrile, dimethyl sulfoxide (DMSO, ≥99.9%), Resveratrol (internal standard, ≥99%), and Coumarin-6 were acquired from Sigma-Aldrich (St. Louis, MO, USA). Nile Red was obtained from Yeasen Biotechnology (Shanghai China). The Cell Counting Kit-8 (CCK-8) was procured from Nanjing KeyGen Biotech (Jiangsu, China). Phosphate-buffered saline (PBS) was purchased from Beijing Solarbio Science & Technology (Beijing, China). All other chemicals and reagents used were of analytical grade. Animals Male Sprague-Dawley (SD) rats (200–250 g) and BALB/c nude mice (20 ± 2 g) were obtained from the Experimental Animal Center of Dalian Medical University (Dalian, China). Animals were housed under controlled environmental conditions (temperature 22 ± 1 °C, humidity 55 ± 5%) with a 14:10 h light-dark cycle and free access to standard laboratory chow and filtered water. All experimental protocols were approved by the Institutional Animal Care and Use Committee of Dalian Medical University (SCXK (Liao) 2020–0001) and conducted in strict accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH Publication No. 8023, revised 1978). Every effort has been made to reduce the number of animals used and the pain they experience. Cell line and cell culture The human glioblastoma U251 cell line was obtained from Nanjing KeyGen Biotech Co., Ltd. (Nanjing, China), while the rat glioma RG-2 cell line was kindly provided by the Medical School of the University of Lausanne (Lausanne, Switzerland). All cells were maintained in Dulbecco's Modified Eagle Medium (DMEM; Gibco, Thermo Fisher Scientific, USA) supplemented with 10% fetal bovine serum (FBS; Gibco) and 1% penicillin-streptomycin (100 U/mL penicillin and 100 μg/mL streptomycin; Gibco). Cells were cultured at 37 °C in a humidified incubator with 5% CO2 and routinely subcultured upon reaching 80–90% confluence using 0.25% trypsin-EDTA (Gibco). Preparation and characterization of folate-modified pterostilbene loaded PEG2000/PEG3400-PCL2000 polymeric micelles (F–Pt/M) Preparation of F–Pt/M The folate-modified Pterostilbene-loaded polymeric micelles (F–Pt/M) were developed using a thin-film hydration technique [26]. Briefly, predetermined quantities of Pterostilbene (8.3 mg), mPEG2000-PCL2000 (30.0 mg), and Folate-PEG3400-PCL2000 (3.0 mg) were co-dissolved in 6 mL of acetone within a round-bottom flask. The organic solvent was subsequently evaporated under reduced pressure using a rotary evaporator (Büchi R-300, Switzerland) at 45 °C to form a homogeneous drug-polymer film. To ensure complete solvent removal, the film was further vacuum-desiccated overnight. Hydration was performed by adding 15 mL of pre-warmed (55 °C) phosphate-buffered saline (PBS, pH 7.4) followed by vigorous vortexing and probe sonication in an ice bath. The sonication was performed using a JY92 sonicator (JY92-IIDN, China) at 30% amplitude with a 1-s pulse on and 2-s pulse off cycle for a total processing time of 5 min. Unencapsulated Pt was separated by ultracentrifugation at 14,000×g for 30 min at 4 °C, and the micellar suspension was subsequently extruded through 200 nm polycarbonate membranes using a laboratory extrusion device (LiposoFast Basic LF-1, Avestin Inc., Canada). Parallel preparations were conducted for control formulations: (1) blank micelles (BM) without drug loading; (2) non-targeted Pterostilbene-loaded micelles (Pt/M); (3) Coumarin-6-loaded micelles (C6/M) and their folate-conjugated counterparts (F–C6/M) for cellular uptake studies; and (4) near-infrared dye DiR-loaded micelles (DiR/M) and folate-modified DiR micelles (F-DiR/M) for in vivo biodistribution evaluation. All micelle preparations were stored at 4 °C protected from light and used within 24 h of preparation. Characterization of F–Pt/M Hydrodynamic diameter and zeta potential The particle size distribution and surface charge characteristics of the micellar formulations were systematically evaluated. Hydrodynamic diameter and polydispersity index (PDI) were determined by dynamic light scattering (DLS) using a Zetasizer Nano ZS90 system (Malvern, UK) at 25 °C. Zeta potential measurements were performed in folded capillary cells using the same instrument. All measurements were conducted in triplicate. Morphology analysis The nanostructural morphology of the micelles was examined by scanning electron microscopy (SEM). Samples were deposited onto a silicon wafer, air-dried, sputter-coated with gold, and imaged using an SU8100 cold field-emission SEM (Hitachi, Japan) at 3.0 kV. Critical aggregation concentration (CAC) measurement The critical aggregation concentration (CAC) of the F–Pt/M polymer was determined using Nile Red as a fluorescent probe [27]. Briefly, 4.0 × 10−5 mg of Nile Red was added to each 4 mL aliquot of the micellar solutions at concentrations ranging from 5.12 × 10−7 to 1 mg/mL (5.12 × 10−7, 2.56 × 10−6, 1.28 × 10−5, 6.4 × 10−5, 3.2 × 10−4, 1.6 × 10−3, 8.0 × 10−3, 4.0 × 10−2, 0.2, and 1 mg/mL), yielding a final Nile Red concentration of 1.0 × 10−5 mg/mL in each sample. The mixtures were sonicated for 30 min and then shaken at room temperature for 60 min to allow partitioning of Nile Red into the hydrophobic domains of the polymer. Fluorescence intensities were measured using a Tecan SPARK microplate reader (Tecan Group Ltd., Austria) in 96-well plates, with excitation at 550 nm and emission at 650 nm. The CAC value was determined by plotting fluorescence intensity against the logarithm of polymer concentration, and the CAC was defined as the inflection point of the curve. Thermal analysis The thermal properties of formulation components were investigated using a differential scanning calorimeter (DSC 60, Shimadzu, Japan). For this purpose, the micelle formulations (Pt/M and F–Pt/M) were first lyophilized to obtain solid samples. Briefly, 5 mL of freshly prepared micellar dispersion (without a cryoprotectant) was frozen at −80 °C for 4 h and then transferred to a freeze-dryer (Scientz-10 N/A, Ningbo, China) at −62 °C under a vacuum pressure of <20 Pa for 48 h. The resulting lyophilized powder was stored in a desiccator until analysis. For DSC measurements, samples (3–5 mg) including pure Pterostilbene, copolymer materials (FA-PEG3400-PCL2000 and mPEG2000-PCL2000), and the lyophilized micelle formulations (Pt/M and F–Pt/M) were put in aluminum pans and analyzed under nitrogen purge (20 mL/min). Thermograms were recorded from 25 °C to 300 °C at a heating rate of 10 °C/min [28]. Entrapment efficiency (EE) and drug loading (DL) To determine the EE and DL of F–Pt/M, the drug entrapped micelles were dissolved in ethanol, sonicated for 10 min to disrupt their structure. The fluorescence intensity of the resulting solutions was then measured using a Varioskan LUX multimode microplate reader (Thermo Fisher Scientific, Finland) with an excitation wavelength of 330 nm and an emission wavelength of 374 nm. The EE and DL of micelles were subsequently calculated using the following equations.(1)EE(%)=(weightofPterostilbeneinmicelles)/(weightoffeedingPterostilbene)×100(2)DL(%)=(weightofPterostilbeneinmicelles)/(dryweightofthefinalmicellarformulation)×100 The dry weight of the final micellar formulation was determined by freeze-drying a known volume of the micellar dispersion and weighing the resulting solid residue. The mass yield (%) of the micellar systems was calculated as:(3)Massyield(%)=(actualdryweightofmicellesrecovered)/(totalweightoffeedingdrugsandpolymers)×100 LC-MS/MS analysis for quantification of pterostilbene in plasma Instrumentation and chromatographic conditions The quantitative analysis of Pterostilbene in plasma was performed using a validated LC-MS/MS method. The HPLC system (Agilent 1200 series, Agilent Technologies, USA) comprised a quaternary pump, degasser, thermostated autosampler, and column compartment. Chromatographic separation was achieved on a Kromasil C18 column (150 mm × 2.1 mm, 5 μm) maintained at 22 ± 1 °C. The mobile phase, consisting of acetonitrile with 0.1% formic acid (95:5, v/v), was delivered isocratically at a flow rate of 0.4 mL/min. The injection volume was set at 10 μL with a total run time of 5.5 min per sample. Mass spectrometric detection was performed using an API 3200 triple quadrupole instrument (Applied Biosystems MDS Sciex, USA) equipped with an electrospray ionization (ESI) source operating in negative ionization mode. The optimized ion source parameters were as follows: Ion spray voltage: 4500 V, Source temperature: 500 °C, Nebulizer gas (Gas 1): 25 L/min of nitrogen, Heater gas (Gas 2): 35 L/min of nitrogen, Curtain gas: 25 MPa (Nitrogen), Collision gas: 9 MPa (Nitrogen), Declustering potential (DP): 55 eV, and Entrance potential (EP): 10 eV. Quantification was performed using multiple reaction monitoring (MRM) of the following transitions: m/z 254.9 → 240.0 for Pterostilbene (collision energy: 32 eV) and m/z 227.0 → 142.8 for Internal Standard (IS, Resveratrol) (collision energy: 35 eV). The product ions of these compounds were shown in Fig. 1A and B.Fig. 1Product ion mass spectra of (A) Pterostilbene and (B) Resveratrol. Representative (C) particle size distribution, (D) zeta potential analysis, (E) a typical SEM image (red arrows point to the micelles), and (F) DSC curve of F–Pt/M confirmed the successful encapsulation of Pterostilbene within spherical micelles, which displayed uniform morphology and a narrow size distribution. (Pt/M, Pterostilbene loaded micelles; F–Pt/M, folate modified Pterostilbene loaded micelles.) Product ion mass spectra of (A) Pterostilbene and (B) Resveratrol. Representative (C) particle size distribution, (D) zeta potential analysis, (E) a typical SEM image (red arrows point to the micelles), and (F) DSC curve of F–Pt/M confirmed the successful encapsulation of Pterostilbene within spherical micelles, which displayed uniform morphology and a narrow size distribution. (Pt/M, Pterostilbene loaded micelles; F–Pt/M, folate modified Pterostilbene loaded micelles.) Preparation of standard solutions and quality controls A comprehensive solution preparation protocol was established to ensure analytical reliability. Primary stock solutions of Pterostillben (1 mg/mL) were prepared in HPLC-grade methanol using amber glass vials to minimize light exposure. Working standard solutions were subsequently prepared by serial dilution with the mobile phase (acetonitrile: 0.1% formic acid, 95:5 v/v) to obtain concentrations spanning the analytical range. All solutions were aliquoted and stored at −70 °C in light-protected containers to ensure long-term stability. For calibration standards, appropriate volumes of working solutions were spiked into 200 μL aliquots of drug-free rat plasma to yield final concentrations of 20, 40, 100, 200, 500, 1000, and 2000 ng/mL. Quality control (QC) samples were independently prepared at three concentration levels of 30, 400, and 1600 ng/mL. All sample preparation procedures were performed under dimly lit conditions to prevent photo-isomerization of the analytes [29]. Freshly prepared calibration standards and QC samples were processed alongside test samples in each analytical run to ensure method validity. Sample preparation Pterostilbene was extracted from plasma samples using an optimized liquid-liquid extraction protocol where 20 μL of internal standard working solution (Resveratrol, 250 ng/mL) was added to 200 μL plasma in microcentrifuge tubes, followed by vortex mixing for 30 s. Subsequently, 1 mL of methyl tert-butyl ether (MTBE) was added as extraction solvent, and the mixture was vigorously vortexed for 1 min before centrifugation at 1000 rpm for 5 min at 4 °C. The upper organic layer was carefully transferred to tubes and evaporated to dryness under a gentle nitrogen stream at 40 °C, with the residue reconstituted in 400 μL mobile phase, vortex-mixed for 30 s, and centrifuged at 12,000 rpm for 10 min at 4 °C. The resulting supernatant was transferred to amber auto-sampler vials, with 10 μL injected into the LC-MS/MS system for analysis. Validation of methodology The analytical method was validated in accordance with the Guidelines for Bioanalytical Method Validation issued by China SFDA, using quality control (QC) samples at three concentration levels (30, 400, and 1600 ng/mL) analyzed in 5 replicates for each validation parameter. Selectivity Method selectivity was verified by analyzing blank plasma samples from six individual rats, demonstrating no endogenous interference at the retention times of Pterostilbene and IS. Linearity The linearity of the method was evaluated by constructing calibration curves based on the peak area ratio of Pterostilbene to IS. Calibration standards were prepared in blank rat plasma and analyzed in triplicate over three consecutive validation days. Linear regression analysis with 1/x2 weighting was performed to establish the relationship between analyte concentration and detector response. Precision and accuracy The precision and accuracy of the method were rigorously evaluated through comprehensive intra-day and inter-day validation studies. For intra-day assessment, six replicates of QC samples at three concentration levels (30, 400, and 1600 ng/mL) were analyzed. Inter-day precision was determined by analyzing the same QC levels over three separate analytical runs conducted on different days. Method performance was considered acceptable when meeting the following criteria: accuracy expressed as relative error (RE) within ±15% of nominal concentrations, and precision expressed as relative standard deviation (RSD) ≤15% for all QC levels. Matrix effect evaluation The potential matrix effects were evaluated using a post-extraction spiking approach. Six different batches of blank rat plasma were processed through the extraction procedure, and the resulting extracts were spiked with QC solutions (30, 400, and 1600 ng/mL) and IS (250 ng/mL). These samples were then compared with reference solutions prepared in mobile phase at equivalent concentrations. The matrix factor (MF) was calculated as the ratio of peak areas in plasma extracts to those in mobile phase, normalized by the internal standard response. Extraction recovery assessment The extraction efficiency of the liquid-liquid extraction method was quantitatively evaluated at three QC concentration levels (30, 400, and 1600 ng/mL). Recovery was determined by comparing the mean peak areas of pre-extraction spiked samples (n = 6) that underwent the complete extraction procedure with those of post-extraction spiked samples (n = 6). Stability evaluation A comprehensive stability assessment of Pterostilbene in rat plasma was conducted under various storage and processing conditions using triplicate determinations at three QC levels (30, 400, and 1600 ng/mL). The stability studies included: (1) short-term stability evaluated after 8-h exposure at ambient temperature (25 ± 2 °C); (2) freeze-thaw stability through three complete cycles (frozen at −70 °C for ≥12 h per cycle, then thawed at room temperature); and (3) long-term stability assessed after 15-day storage at −70 °C. Samples were deemed stable if results met acceptance criteria of ±15% RE for accuracy and ±15% RSD for precision. Pharmacokinetic study The validated LC-MS/MS method was employed to characterize the pharmacokinetic profile of Pterostilbene following intravenous administration in Sprague-Dawley rats (n = 3 per group). After an overnight fasting period with ad libitum access to water, animals were briefly anesthetized with diethyl ether for the duration of drug administration. Two formulations were evaluated: free Pterostilbene solution and F–Pt/M, both administered at a dose equivalent to 3 mg/kg Pterostilbene via tail vein injection. Serial blood samples (300 μL) were collected from the jugular vein into heparinized tubes at precisely timed intervals: 5, 15, 30, 45, 60, 90, 120, 180, and 270 min for the free Pterostilbene solution, and 5, 10, 20, 30, 45, 90, 120, and 180 min for the F–Pt/M formulation. Plasma was immediately separated by refrigerated centrifugation (4 °C, 3500 rpm, 10 min) and stored at −70 °C in cryovials until analysis. In vitro cytotoxicity and cellular uptake studies Cell viability assessment using CCK-8 assay The cytotoxic effects of various Pterostilbene formulations were evaluated on human U251 and rat RG-2 glioma cell lines using a standardized CCK-8 assay protocol. Cells were seeded in 96-well plates at a density of 5 × 103 cells/well in complete growth medium (DMEM supplemented with 10% FBS and 1% penicillin-streptomycin) and allowed to adhere for 24 h. Following attachment, cells were exposed to free Pterostilbene solution, non-targeted micelles (Pt/M), Folate-conjugated micelles (F–Pt/M) and Blank micelles (BM, polymer control) across a concentration gradient (5–100 μM) for 48 h. After treatment, 10 μL of CCK-8 reagent was added to each well and plates were incubated for 2 h under standard culture conditions. Absorbance was measured at 450 nm using a Varioskan LUX multimode microplate reader (Thermo Fisher Scientific, Finland). Cell viability was calculated as:(4)Cellviability(%)=(ODtreatment-ODblank)/(ODcontrol-ODblank)×100where ODblank represented medium-only wells and ODcontrol represented untreated cells. IC50 values were determined using SPSS Statistics 22.0 (IBM, USA). In vitro cellular uptake assay The cellular uptake behavior was evaluated using coumarin-6 (C6) as a fluorescent surrogate for Pterostilbene due to its superior optical properties while possessing comparable hydrophobicity and micelle incorporation characteristics. C6-loaded micelles (C6/M) and folate-conjugated C6 micelles (F–C6/M) were prepared following identical procedures to the Pt/M or F–Pt/M formulations. For qualitative assessment, U251 and RG-2 cells were seeded on cover slips in 24-well plates (1 × 104 cells/well) and cultured for 24 h. Cells were then exposed to free C6, C6/M, or F–C6/M (100 ng/mL C6 equivalent) for 4 h, followed by three cold PBS washes, fixation with 4% paraformaldehyde (30 min), and DAPI nuclear staining (5 min at 37 °C). Cellular uptake was visualized using an Olympus IX83 fluorescence microscope with standardized imaging parameters. Quantitative cellular uptake analysis was performed by seeding cells in 6-well plates (1 × 106 cells/well) and treating with equivalent formulations for 1 h or 4 h. After thorough PBS washing, cells were trypsinized, pelleted, and lysed with RIPA buffer for fluorescence measurement (λex = 456 nm, λem = 500 nm) using a Varioskan LUX plate reader. To confirm folate receptor-mediated uptake, competitive inhibition studies were conducted by pre-treating cells with excess free folate (10 μg/mL, 30 min) prior to F–C6/M exposure, with all other procedures maintained identically. In vivo imaging The biodistribution of different formulations was evaluated in BALB/c nude mice using near-infrared fluorescence imaging. Mice received intravenous injections of Free DiR, DiR/M, or F-DiR/M at a dose equivalent to 500 μg/kg of DiR. Whole-body imaging was performed at 1, 4, and 24 h post-injection using the Carestream In-Vivo FX Professional Imaging System (Bruker MI SE, USA). The imaging parameters were set as follows: excitation wavelength 748 nm, emission wavelength 780 nm, exposure time 60 s, with maximum camera gain. At 24 h post-injection, mice were euthanized and major organs (heart, liver, spleen, lungs, kidneys, and brain) were harvested for ex vivo imaging. Excised organs were rinsed with ice-cold PBS, blotted dry, and arranged on a black imaging plate for fluorescence scanning using the same parameters as in vivo imaging. In vivo anti-glioma activity evaluation of F–Pt/M in orthotopic RG-2 glioma rat models Orthotopic glioma model establishment and treatment Male SD rats were randomly allocated into four experimental groups (n = 5 per group): (1) Control group (RG-2 cell implantation + 0.9% normal saline treatment), (2) Sham group (sterile PBS implantation + 0.9% normal saline treatment), (3) Free Pterostilbene group (RG-2 cell implantation + free drug solution treatment), and (4) F–Pt/M group (RG-2 cell implantation + F–Pt/M micellar solution treatment). Orthotopic glioma models were established through stereotactic implantation following standardized procedures [2]. Briefly, rats were anesthetized via intraperitoneal injection of 3% pentobarbital sodium (0.1 mL/100 g) and secured in a stereotaxic frame. After shaving and disinfecting the surgical site, a midline scalp incision was made using sterile surgical scissors. A burr hole (1 mm below the dura) was drilled 1 mm posterior and 3 mm lateral to bregma. RG-2 glioma cells (3 × 106 cells in 10 μL sterile PBS) were slowly injected (1 μL/min) to a depth of 6 mm below the dura using a microsyringe. The needle remained in place for 10 min post-injection before being gradually withdrawn over 1 min to prevent cell reflux. The surgical site was closed with absorbable sutures. Treatment regimens were initiated 72 h post-surgery and maintained until endpoints were reached. The administration protocols were as follows: (1) Control and Sham groups received 2 mL/kg of 0.9% normal saline (i.v., daily); (2) Pterostilbene group was administered 1 mg/kg of free drug solution (i.v., daily); and (3) F–Pt/M group received 1 mg/kg of Pterostilbene-equivalent micellar formulation (i.v., daily). All mortality events were meticulously recorded, and Kaplan-Meier survival curves were generated using GraphPad Prism software. Following decapitation, brains were rapidly extracted, rinsed with ice-cold 0.9% saline, and immediately flash-frozen in liquid nitrogen until further analysis. Magnetic resonance imaging (MRI) visualization of tumor growth Tumor progression was assessed through high-field MRI at the Radiology Core Facility (Second Affiliated Hospital, Dalian Medical University). Experimental subjects underwent scanning 7 days post-RG2 implantation using a 7T/30 cm bore clinical MRI system (SIEMENS Verio 3.0 T, Germany). Prior to scanning, rats received intraperitoneal administration of 10% chloral hydrate anesthetic (0.3 mL/100 g body weight) and were positioned prone in the magnet isocenter. MR signal acquisition was achieved through a dedicated cranial surface coil, while radiofrequency pulses were delivered via a quadrature volume coil (internal diameter: 72 mm; Bruker BioSpin). For contrast enhancement, gadopentetate dimeglumine (Gd-DTPA) was intravenously administered at a dose of 0.3 mmol/kg through the caudal vein. Histological analysis Brain tissue samples were fixed in 4% paraformaldehyde solution at 4 °C for 24 h, followed by transfer to 10% neutral buffered formalin for long-term storage. After complete fixation, tissues were processed through a standard dehydration series, cleared in xylene, and embedded in paraffin wax. Serial sections (5 μm thickness) were cut using a Leica RM2235 rotary microtome (Leica Microsystems, Germany). For histological evaluation, sections were stained with hematoxylin and eosin (Sigma-Aldrich, USA) using standard protocols, and coverslipped for microscopic examination. Histopathological assessment was performed to evaluate tumor invasion patterns and tissue architecture disruption. Immunohistochemical (IHC) analysis Standard IHC staining was performed on 5-μm paraffin-embedded brain tissue sections following antigen retrieval [30]. Antibodies against STAT3 (1:400), p-STAT3 (1:400), Bax (1:500), Bcl-2 (1:400), Survivin (1:500), CyclinD1 (1:400) and C-Myc (1:500) were obtained from Proteintech Group, Inc. (USA) and utilized in accordance with the manufacturer's instructions. Apoptosis detection by TUNEL immunofluorescence assay Apoptotic cells were identified using the TUNEL (Terminal deoxynucleotidyl transferase dUTP Nick-End Labeling) assay according to the manufacturer's protocol (Promega Corporation, USA). For comprehensive nuclear visualization, the tissue sections were counterstained with DAPI (4′,6-diamidino-2-phenylindole, 1 μg/mL in PBS) in a light-protected environment. Following three gentle washes with PBS, sections were mounted with anti-fade mounting medium to prevent fluorescence quenching. Fluorescence imaging was performed using an Olympus BX63 microscope equipped with FITC filter set (Ex: 490 nm, Em: 520 nm) for TUNEL-positive cells and DAPI filter set (Ex: 358 nm, Em: 461 nm) for nuclear staining. Western blotting analysis Peri-tumoral brain tissues were carefully dissected from the Control, Pterostilbene, and F–Pt/M groups. Tissues from corresponding anatomical regions were collected from the Sham group to ensure comparability. Total proteins were extracted following a detailed protocol [2]. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was conducted using 15 μg of protein samples loaded into the gels. The resultant membranes were then subjected to electrochemiluminescence and captured using a chemiluminescence system from Bio-Rad Laboratories, USA. The grayscale intensity of the bands was analyzed using Image Lab 3.0 software. STAT3 (1:1000), p-STAT3 (1:1000), Bax (1:5000), Bcl-2 (1:2000), and GAPDH (1:10,000) antibodies were obtained from Proteintech Group, Inc., USA. Molecular docking and molecular dynamics (MD) simulation In this study, molecular docking simulations were performed using Maestro 13.5 software to elucidate the binding mode between Pterostilbene and STAT3. The process utilized preprocessed receptor proteins, ligand molecules, and a predefined docking grid, with molecular interactions simulated and analyzed through the software's built-in algorithms. For ligand preparation, the 2D structure of Pterostilbene (PubChem ID: 5,281,727) was obtained from the PubChem database (URL: https://pubchem.ncbi.nlm.nih.gov) and processed using Maestro's LigPrep module. This included protonation state adjustment, desalting, hydrogen atom addition, tautomer and stereoconformation generation, and energy minimization, yielding an optimized ligand structure for docking. For receptor preparation, the STAT3 crystal structure (PDB, ID: 6NJS, URL: http://www.rcsb.org) [31] was optimized using Maestro 13.5, involving bond order correction, removal of water molecules, hydrogen addition, amino acid residue optimization, missing side chain completion, and energy minimization with the OPLS4 force field to achieve precise hydrogen atom coordination and hydrogen bond network optimization. Meanwhile, a docking grid was then generated using the STAT3-selective small-molecule degrader KQV as a reference ligand [31], establishing a spatial positioning benchmark for docking calculations. Molecular Dynamics (MD) simulations were conducted using GROMACS 2023.1 (single-precision version) on an Ubuntu 20.04.01 platform with a DELL T3680 workstation equipped with an Nvidia RTX-4070TI SUPER GPU. Heavy atoms of proteins and small molecules were modified using SPDBV 4.1.0 software, with protein topology generated using the AMBER99SB-ILDN force field and ligand topology prepared through the AMBER force field combined with the Acpype online tool (https://bio2byte.be/acpype) [32]. The complexes were dissolved in a TIP3P water model within a cubic box (each side of the box extending at least 1 nm beyond the dissolved portion), neutralized with Na+/Cl− ions, and supplemented with 0.15 M NaCl to mimic physiological conditions. System preparation included energy minimization via steepest-descent algorithm (5000 steps, force <1000 kJ/mol/nm) followed by sequential 100-ps NVT and NPT equilibrations (310 K, 1 bar). Production MD runs were performed for 100 ns (5 × 107 steps) using the Verlet cut-off scheme and Leap-frog integrator, with trajectory data saved every 10 ps under periodic boundary conditions. The binding stability between Pterostilbene and STAT3 was analyzed by calculating the root-mean-square deviation (RMSD), radius of gyration (Rg), and solvent-accessible surface area (SASA, probe radius = 1.4 Å). In addition, the effect of Pterostilbene on the SH2 domain of STAT3 was analyzed by calculating the root-mean-square fluctuation (RMSF). Statistical analysis All quantitative data were analyzed using SPSS Statistics software (IBM Corporation, USA). Statistical significance between the groups was achieved via one-way analysis of variance (ANOVA) with Dunnett's post hoc test. A P value of less than 0.05 was considered to be significant. Results and discussion Preparation and characterization of F–Pt/M F–Pt/M was successfully prepared using an optimized thin-film hydration method, providing a nano-enabled delivery platform for the poorly water-soluble natural polyphenol Pterostilbene. Dynamic light scattering analysis revealed a mean hydrodynamic diameter of 133.2 ± 1.2 nm with excellent monodispersity (PDI = 0.19 ± 0.02), as evidenced by the unimodal size distribution profile (Fig. 1C). The hydrodynamic diameter is a critical determinant of nanocarrier biodistribution and passive targeting efficacy through the enhanced permeability and retention (EPR) effect. Micelles with sizes smaller than 200 nm demonstrate optimal biological performance, as they can enhance blood-brain barrier penetration via both passive diffusion and carrier-mediated transport, and facilitate efficient cellular uptake in tumor tissues through endocytic pathways [33]. Zeta potential was measured to assess the surface charge of F–Pt/M micelles. The formulation exhibited a zeta potential of −8.93 ± 0.22 mV (Fig. 1D). This negative surface charge is consistent with the presence of folate carboxyl groups at the micelle surface. In comparison, non-targeted Pt/M micelles showed a value of −1.11 ± 0.1 mV under the same conditions. The colloidal stability of F–Pt/M is primarily governed by steric hindrance from the hydrated PEG shell, which effectively prevents particle aggregation [34]. In addition, storage stability studies showed that the hydrodynamic diameter and PDI of F–Pt/M remained stable over 30 days at 4 °C (Supplementary Table S2), further validating the physical stability of the formulation. Physicochemical characterization of all dye-loaded micelles (Supplementary Table S3) confirmed that the incorporation of C6 and DiR did not significantly alter the colloidal properties relative to the Pterostilbene-loaded micelles. SEM analysis of the F–Pt/M formulation (Fig. 1E) confirmed the formation of monodisperse, spherical micelles with uniform morphology. Notably, the SEM-derived particle diameters were consistently smaller than those obtained by DLS (133.2 ± 1.2 nm, Fig. 1C). This discrepancy arose from fundamental differences in measurement principles: DLS determines the hydrodynamic diameter, which includes the hydrated PEG corona, while SEM visualization captures the dried core-shell structure after solvent evaporation during sample preparation [35]. The EE and DL values represent critical quality attributes for evaluating nanocarrier systems. The developed F–Pt/M formulation demonstrated exceptional drug encapsulation performance, achieving an EE of 98.6 ± 0.8% and a DL capacity of 20.8 ± 0.4 %. This high loading efficiency resulted in a remarkable enhancement of Pterostilbene's apparent solubility. Based on the formulation composition (8.3 mg Pt, 15.0 mL hydration PBS, 30.0 mg PEG2000-PCL2000, and 3.0 mg Folate-PEG3400-PCL2000), the resulting micellar dispersion exhibited a final solids concentration of 2.63 ± 0.06 mg/mL. Within this system, the encapsulated Pt concentration reached approximately 545.6 μg/mL, corresponding to a 25-fold increase over its intrinsic aqueous solubility (21 μg/mL) [36]. Such substantial solubility enhancement confirmed the formulation's ability to effectively solubilize this hydrophobic therapeutic agent through micellar encapsulation, addressing a major pharmaceutical challenge for Pterostilbene delivery. We acknowledge that the solubility enhancement reported herein was derived from a single formulation, and a systematic phase solubility study (e.g., Higuchi-Connor method) was not conducted. Future investigations will be performed to fully characterize the solubilization capacity of this micellar system across a broader range of polymer concentrations. In addition, the mass yield for F–Pt/M was determined to be 95.5 ± 2.1%, and for Pt/M, it was 94.0 ± 3.3%. The CAC value of the F–Pt/M polymer was determined to be 2.34 μg/mL (see Supplementary Fig. S1) using Nile Red as a fluorescent probe. This low CAC value indicates that the developed micelles possess high stability against dilution, a favorable attribute for maintaining integrity upon intravenous administration, where substantial dilution occurs in the systemic circulation. DSC analysis was conducted to elucidate the physicochemical state of Pterostilbene within the micellar formulations (Fig. 1F). The thermogram of pure Pterostilbene displayed a characteristic sharp endothermic peak at 95.9 °C, corresponding to its crystalline melting transition. Notably, this crystalline melting peak was absent in both micellar formulations (Pt/M and F–Pt/M), confirming the successful disruption of Pterostilbene's crystalline lattice upon encapsulation. Instead, the lyophilized Pt/M and F–Pt/M formulations exhibited broad endothermic transitions at 76.1 °C and 80.8 °C, respectively, suggesting that Pterostilbene existed predominantly in an amorphous state or as a molecular dispersion within the micellar core [37], with the slightly higher transition temperature observed for F–Pt/M potentially indicating stronger drug-polymer interactions in the folate-conjugated system. The absence of the native drug's melting endotherm provided conclusive evidence of successful molecular encapsulation and crystalline-to-amorphous phase transformation. Together, these results confirmed the successful preparation of folate receptor-targeted polymeric micelles (F–Pt/M) with favorable physicochemical properties, including a small hydrodynamic diameter, high drug loading, and an enhancement in aqueous solubility, effectively addressing the delivery challenges of the natural dimethylated resveratrol analog Pterostilbene. LC–MS/MS method validation Assay selectivity The selectivity of the chromatographic method was evaluated by analyzing blank rat plasma samples, blank plasma samples spiked with Pterostilbene at the limit of quantification (LOQ, 20 ng/mL) and internal standard (IS, 500 ng/mL), as well as plasma samples collected after intravenous administration of Pterostilbene to rats. Representative chromatograms were shown in Supplementary Fig. S2. The method demonstrated excellent specificity, with Pterostilbene and IS eluting at 2.8 min and 1.5 min, respectively. No interfering peaks from endogenous plasma components were observed at the retention times of either analyte, confirming the method's ability to selectively quantify Pterostilbene. The signal-to-noise ratio at the LOQ concentration exceeded 10:1, meeting international bioanalytical method validation guidelines. These results verified that the chromatographic conditions effectively separate Pterostilbene from potential matrix interferences while maintaining robust detection sensitivity for pharmacokinetic studies. Linearity and LOQ A linear regression model was established to evaluate the correlation between the peak area ratio (Pterostilbene to IS) and the corresponding plasma concentrations over the range of 20–2000 ng/mL. The calibration curve exhibited excellent linearity, as described by the equation y = 0.00797x - 0.0241, where y represented the peak area ratio and x denoted the Pterostilbene concentration. The high coefficient of determination (r2 = 0.9925) confirmed the robustness of the linear relationship across the tested range. The LOQ was reliably determined to be 20 ng/mL, demonstrating the assay's high sensitivity for accurate quantification of Pterostilbene in rat plasma. Precision and accuracy The precision and accuracy of the analytical method were evaluated using quality control (QC) samples (n = 6) over three consecutive validation days. As summarized in Table 1, the intra-day precision, expressed as RSD, ranged from 1.83% to 11.72%, while the inter-day precision varied between 3.96% and 7.34%. The accuracy, represented by RE, demonstrated values ranging from −1.56% to 13.56% across all QC levels. These results confirmed the method's reliability for quantifying Pterostilbene in rat plasma with excellent reproducibility and minimal systematic error.Table 1Precision and accuracy for Pterostilbene determination in rat plasma by LC-MS/MS method (n = 6).Table 1AnalyteAdded (ng/mL)Intra-dayInter-dayFoundeda (ng/mL)RSDc (%)REd (%)Foundedb (ng/mL)RSD (%)RE (%)Pterostilbene3031.1 ± 0.571.833.6732.03 ± 1.564.876.77400454.3 ± 53.2211.7213.56437.06 ± 17.333.969.2716001575 ± 43.702.77−1.561662.22 ± 122.027.343.89aMean of 6 replicates at each concentration.bMean of 6 replicates for three precision and accuracy batches.cRSD (%) = (SD/mean) × 100.dRE (%) = (mean determined concentration - nominal concentration)/nominal concentration × 100. Table 1Precision and accuracy for Pterostilbene determination in rat plasma by LC-MS/MS method (n = 6).Table 1AnalyteAdded (ng/mL)Intra-dayInter-dayFoundeda (ng/mL)RSDc (%)REd (%)Foundedb (ng/mL)RSD (%)RE (%)Pterostilbene3031.1 ± 0.571.833.6732.03 ± 1.564.876.77400454.3 ± 53.2211.7213.56437.06 ± 17.333.969.2716001575 ± 43.702.77−1.561662.22 ± 122.027.343.89aMean of 6 replicates at each concentration.bMean of 6 replicates for three precision and accuracy batches.cRSD (%) = (SD/mean) × 100.dRE (%) = (mean determined concentration - nominal concentration)/nominal concentration × 100. Precision and accuracy for Pterostilbene determination in rat plasma by LC-MS/MS method (n = 6). Mean of 6 replicates at each concentration. Mean of 6 replicates for three precision and accuracy batches. RSD (%) = (SD/mean) × 100. RE (%) = (mean determined concentration - nominal concentration)/nominal concentration × 100. Extraction recovery and matrix effect As detailed in Table 2, the extraction recovery rates for Pterostilbene demonstrated good performance across the tested concentration range. The mean recovery values were 86.33 ± 1.00% (30 ng/mL), 115.00 ± 13.11% (400 ng/mL), and 107.40 ± 12.35% (1600 ng/mL), with RSD values of 1.16%, 11.37%, and 11.50%, respectively. These results showed consistent and reproducible extraction efficiency across all QC levels, meeting the requirements of ICH M10 guidelines for bioanalytical method validation. Matrix effect evaluation revealed minimal ionization interference, with normalized matrix factors ranging from 92.68% to 103.90% across the three QC levels and corresponding RSD values of 1.32%, 7.23%, and 6.13%, respectively. These results confirmed that plasma matrix components did not significantly interfere with Pterostilbene quantification. The consistent performance across concentrations indicated the method's reliability for pharmacokinetic studies.Table 2Extraction recovery and matrix effect for Pterostilbene determination by LC-MS/MS method (n = 6).Table 2AnalyteConcentration (ng/mL)Extraction recoverya (%)Matrix effectb (%)Pterostilbene3086.33 ± 1.00103.90 ± 1.37400115.00 ± 13.1195.49 ± 6.911600107.40 ± 12.3592.68 ± 5.68aCalculated by comparing the mean area response of pre-spiked samples (spiked before extraction) to that of extracts with post-spiked samples (spiked after extraction) at each QC level.bCalculated was assessed by comparing the mean area response of unextracted samples (spiked after extraction) with mean area of neat standard solutions (in mobile phase). Table 2Extraction recovery and matrix effect for Pterostilbene determination by LC-MS/MS method (n = 6).Table 2AnalyteConcentration (ng/mL)Extraction recoverya (%)Matrix effectb (%)Pterostilbene3086.33 ± 1.00103.90 ± 1.37400115.00 ± 13.1195.49 ± 6.911600107.40 ± 12.3592.68 ± 5.68aCalculated by comparing the mean area response of pre-spiked samples (spiked before extraction) to that of extracts with post-spiked samples (spiked after extraction) at each QC level.bCalculated was assessed by comparing the mean area response of unextracted samples (spiked after extraction) with mean area of neat standard solutions (in mobile phase). Extraction recovery and matrix effect for Pterostilbene determination by LC-MS/MS method (n = 6). Calculated by comparing the mean area response of pre-spiked samples (spiked before extraction) to that of extracts with post-spiked samples (spiked after extraction) at each QC level. Calculated was assessed by comparing the mean area response of unextracted samples (spiked after extraction) with mean area of neat standard solutions (in mobile phase). Stability The stability evaluation of Pterostilbene in rat plasma demonstrated excellent performance under various storage conditions (Table 3). Short-term stability (8 h at room temperature) and long-term stability (15 days at −70 °C) results showed RSD <13.09% and RE <14.21%, well within acceptable limits for bioanalytical methods. Notably, the method demonstrated robust freeze-thaw stability with maximum RSD (11.20%) and RE (10.2%) values observed after three complete freeze-thaw cycles. These comprehensive stability results confirmed that Pterostilbene remained chemically stable in rat plasma under typical laboratory handling and storage conditions. To maintain optimal sample integrity, all plasma samples were processed within 1 week of collection and analyzed immediately after a single thaw cycle.Table 3Stability of Pterostilbene in rat plasma under various storage conditions (n = 3).Table 3Stability testConcentration (ng/mL)RSD (%)RE (%)AddedFounded8-h Exposure at room temperature in plasma3032.69 ± 0.210.658.97400453.67 ± 30.836.8013.4216001662.54 ± 121.717.323.91After three freeze–thaw cycles3033.06 ± 0.210.6510.20400412.29 ± 46.1611.203.0716001599.73 ± 129.458.09−0.02Storage at −70 °C for 15 days3034.26 ± 1.714.9914.21400451.97 ± 59.1413.0912.9916001635.68 ± 119.047.282.23 Table 3Stability of Pterostilbene in rat plasma under various storage conditions (n = 3).Table 3Stability testConcentration (ng/mL)RSD (%)RE (%)AddedFounded8-h Exposure at room temperature in plasma3032.69 ± 0.210.658.97400453.67 ± 30.836.8013.4216001662.54 ± 121.717.323.91After three freeze–thaw cycles3033.06 ± 0.210.6510.20400412.29 ± 46.1611.203.0716001599.73 ± 129.458.09−0.02Storage at −70 °C for 15 days3034.26 ± 1.714.9914.21400451.97 ± 59.1413.0912.9916001635.68 ± 119.047.282.23 Stability of Pterostilbene in rat plasma under various storage conditions (n = 3). In conclusion, we have successfully developed and fully validated a sensitive, selective, and reproducible LC-MS/MS method for the quantification of Pterostilbene in rat plasma. The validated method met all criteria in current bioanalytical guidelines. Subsequently, this method was applied to support pharmacokinetic studies in rats. Pharmacokinetic study Pharmacokinetic characterization plays a pivotal role in drug discovery and development. Despite the therapeutic potential of Pterostilbene, its pharmacokinetic profile remains poorly understood [38]. To address this, we conducted a comprehensive pharmacokinetic study in SD rats following intravenous administration of both free Pterostilbene solution and F–Pt/M formulation, using our validated LC-MS/MS analytical method. Key pharmacokinetic parameters analyzed using DAS software (v2.0, China), revealed distinct behavioral patterns between the two formulations (Fig. 2A, Table 4). As depicted in Fig. 2A, after intravenous administration, the F–Pt/M formulation exhibited a concentration-time profile consistent with a two-compartment model (weighted coefficient = 1/C2). Following an initial rapid ascending phase, with a short Tmax (time to reach maximal concentration) of 10 min, the plasma concentration decreased in a biphasic manner [39]: a distribution phase followed by an elimination phase with a terminal elimination half-life (t1/2β) of 66.6 min. The initial ascending phase may be attributable to transient pulmonary sequestration of the micelles upon first passage through the pulmonary capillary bed, followed by gradual desorption and re-entry into the systemic circulation—a phenomenon previously reported for intravenously administered nanoparticles [40]. When free Pterostilbene solution was given, it displayed a biphasic decay concentration-time profile, which was also best simulated by the two-compartment model for intravenous administration. It exhibited a rapid initial distribution phase (t1/2α = 17.0 min), followed by a relatively slow elimination phase with a t1/2β of 69.3 min. These results indicated a significant difference in the pharmacokinetic behavior between the free Pterostilbene solution and F–Pt/M in rats (p < 0.05). Several studies have demonstrated that compared to free drug solutions, micelles have the ability to significantly prolong the circulation time of loaded drugs by increasing the half-life, mean residence time (MRT), and area under the concentration-time curve (AUC) [41]. For instance, in one study, the mean residence time of indomethacin-loaded PEG-PCL micelles increased from 11.99 ± 0.83 to 21.36 ± 4.81 h compared to free drug, with a corresponding increase in AUC [42]. However, our results (Table 4) showed that the AUC(0-∞) of the free Pterostilbene group and the F–Pt/M group were 11616.01 ± 1078.43 and 4893.13 ± 637.00 μg/L min, respectively, leading to a relative systemic exposure (AUC ratio) of 42.12% for Pterostilbene loaded F–Pt/M compared to intravenous administration of free Pterostilbene. The peak concentration (Cmax) was 220.98 ± 21.55 and 71.57 ± 6.43 μg/L, respectively, and the MRT(0-∞) was 122.82 ± 17.91 and 76.98 ± 7.12 min, respectively. Interestingly, compared to free Pterostilbene, the F–Pt/M formulation resulted in significantly lower AUC, Cmax, and MRT (p < 0.05, Table 4). This observation is likely attributable to a combination of factors. First, as evidenced by the ex vivo biodistribution data (Fig. 2G), the micelles undergo opsonization and subsequent uptake by the reticuloendothelial system (RES), particularly in the liver and spleen, which actively reduces their systemic retention. Second, and more importantly, the folate ligand facilitates rapid and extensive redistribution of Pterostilbene from the bloodstream to the brain, as demonstrated by the significantly faster brain Tmax (10 min vs. 35 min for free drug) and approximately 2-fold higher brain AUC0-t of F–Pt/M (Supplementary Fig. S3 and Supplementary Table S1). This “drug sink” effect in the brain, together with the concurrent RES clearance, collectively contributes to reducing plasma concentrations, thereby shortening the systemic AUC and MRT. Notably, the relatively low critical aggregation concentration (CAC) of the micelles (2.34 μg/mL) suggests that the formulation retains adequate thermodynamic stability upon dilution in the bloodstream, supporting that rapid dilution-induced disassembly is unlikely to be the primary cause of the reduced plasma exposure. These results provide important insights for the continued development of Pterostilbene formulations for central nervous system applications, highlighting the unique advantages of receptor-targeted nanocarriers in achieving accelerated and selective brain delivery while potentially minimizing off-target toxicity [43].Fig. 2In vivo drug concentration-time profiles after i. v. administration of (A) Free Pt and F–Pt/M at a dose of 3 mg/kg in rats (n = 3). In vitro cytotoxicity of Free Pt, Pt/M, F–Pt/M and BM against (B) U251 and (C) RG-2 cell lines after treatment for 48 h (n = 5, compared to Free Pt, ∗p < 0.05). Cellular uptake studies on (D) U251 and (E) RG-2 cells using different C6 formulations. Fluorescence microscope images were captured after incubating cells with Free C6, C6/M, and F–C6/M in media with or without free folate. In the Folate + F–C6/M group, cells were initially treated with an excess of free folate (10 μg/mL) for 30 min, followed by incubation with F–C6/M. The concentration of C6 in all formulations was maintained at 100 ng/mL. C6 fluorescence was represented by the green color, while DAPI fluorescence was represented by the blue color. The intensity profile of C6 fluorescence was determined using spectrophotometer measurements (n = 3, ∗p < 0.05, #p < 0.01). In vivo and ex vivo imaging tests. (F) In vivo fluorescent and X-ray images of BALB/c nude mice at 1, 4 and 24 h after administration and (G) in vitro fluorescent imaging of brain and internal organs separated from BALB/c nude mice at 24 h after administration with Free DiR, DiR/M and F-DiR/M at a DiR dose of 500 μg/kg via tail vein. (Free Pt, free Pterostilbene; Pt/M, Pterostilbene loaded micelles; F–Pt/M, folate modified Pterostilbene loaded micelles; BM, blank micelles; C6, coumarin-6; C6/M, coumarin-6 loaded micelles; F–C6/M, folate modified coumarin-6 loaded micelles; DiR/M, DiR loaded micelles; F-DiR/M, folate modified DiR loaded micelles).Table 4Pharmacokinetic parameters after i.v. Administration of free Pterostilbene solution (Free Pt) and folate modified Pterostilbene loaded micelles (F–Pt/M) at a dose of 3 mg/kg in rats (n = 3, mean ± SD).Table 4Pharmacokinetic parametersUnitFree PtF–Pt/Mt1/2αmin17.04 ± 1.45865.19 ± 5.84at1/2βmin69.3266.56 ± 4.77V1L/kg14.72 ± 0.6660.42 ± 32.93aCLL/min/kg0.24 ± 0.020.67 ± 0.21aAUC(0-t)μg/L·min10107.8 ± 976.654181.75 ± 142.20aAUC(0-∞)μg/L·min11616.01 ± 1078.434893.13 ± 637.00aMRT(0-t)min66.75 ± 2.0552.10 ± 11.14aMRT(0-∞)min122.82 ± 17.9176.98 ± 7.12aTmaxmin510aCmaxμg/L220.98 ± 21.5571.57 ± 6.43aap < 0.05, compared to the Free Pterostilbene group. In vivo drug concentration-time profiles after i. v. administration of (A) Free Pt and F–Pt/M at a dose of 3 mg/kg in rats (n = 3). In vitro cytotoxicity of Free Pt, Pt/M, F–Pt/M and BM against (B) U251 and (C) RG-2 cell lines after treatment for 48 h (n = 5, compared to Free Pt, ∗p < 0.05). Cellular uptake studies on (D) U251 and (E) RG-2 cells using different C6 formulations. Fluorescence microscope images were captured after incubating cells with Free C6, C6/M, and F–C6/M in media with or without free folate. In the Folate + F–C6/M group, cells were initially treated with an excess of free folate (10 μg/mL) for 30 min, followed by incubation with F–C6/M. The concentration of C6 in all formulations was maintained at 100 ng/mL. C6 fluorescence was represented by the green color, while DAPI fluorescence was represented by the blue color. The intensity profile of C6 fluorescence was determined using spectrophotometer measurements (n = 3, ∗p < 0.05, #p < 0.01). In vivo and ex vivo imaging tests. (F) In vivo fluorescent and X-ray images of BALB/c nude mice at 1, 4 and 24 h after administration and (G) in vitro fluorescent imaging of brain and internal organs separated from BALB/c nude mice at 24 h after administration with Free DiR, DiR/M and F-DiR/M at a DiR dose of 500 μg/kg via tail vein. (Free Pt, free Pterostilbene; Pt/M, Pterostilbene loaded micelles; F–Pt/M, folate modified Pterostilbene loaded micelles; BM, blank micelles; C6, coumarin-6; C6/M, coumarin-6 loaded micelles; F–C6/M, folate modified coumarin-6 loaded micelles; DiR/M, DiR loaded micelles; F-DiR/M, folate modified DiR loaded micelles). Table 4Pharmacokinetic parameters after i.v. Administration of free Pterostilbene solution (Free Pt) and folate modified Pterostilbene loaded micelles (F–Pt/M) at a dose of 3 mg/kg in rats (n = 3, mean ± SD).Table 4Pharmacokinetic parametersUnitFree PtF–Pt/Mt1/2αmin17.04 ± 1.45865.19 ± 5.84at1/2βmin69.3266.56 ± 4.77V1L/kg14.72 ± 0.6660.42 ± 32.93aCLL/min/kg0.24 ± 0.020.67 ± 0.21aAUC(0-t)μg/L·min10107.8 ± 976.654181.75 ± 142.20aAUC(0-∞)μg/L·min11616.01 ± 1078.434893.13 ± 637.00aMRT(0-t)min66.75 ± 2.0552.10 ± 11.14aMRT(0-∞)min122.82 ± 17.9176.98 ± 7.12aTmaxmin510aCmaxμg/L220.98 ± 21.5571.57 ± 6.43aap < 0.05, compared to the Free Pterostilbene group. Pharmacokinetic parameters after i.v. Administration of free Pterostilbene solution (Free Pt) and folate modified Pterostilbene loaded micelles (F–Pt/M) at a dose of 3 mg/kg in rats (n = 3, mean ± SD). p < 0.05, compared to the Free Pterostilbene group. Cytotoxicity assay The in vitro cytotoxic effects of free Pterostilbene, Pt/M, F–Pt/M, and corresponding blank micelles (BM) against human U251 and rat RG-2 glioma cell lines were assessed using the CCK-8 assay (Fig. 2B and C). BM demonstrated excellent biocompatibility, maintaining cell viabilities near 100% across all tested concentrations, confirming their non-toxic nature. All drug-containing formulations exhibited dose-dependent cytotoxicity profiles in both cell lines, with increasing drug concentrations correlating with decreased cell viability, indicating that higher drug concentrations were necessary for effective inhibition of tumor cells. Quantitative analysis revealed IC50 values of 27.88 μM (free Pterostilbene), 57.10 μM (Pt/M), and 29.15 μM (F–Pt/M) in U251 cells, and 4.38 μM (free Pterostilbene), 5.68 μM (Pt/M), and 4.84 μM (F–Pt/M) in RG-2 cells. Notably, the folate-modified F–Pt/M formulation showed significantly enhanced cytotoxicity (approximately 2-fold lower IC50) compared to non-targeted Pt/M in U251 cells after 48 h incubation, which correlated well with the cellular uptake observations (Fig. 2D and E). This improved performance can be directly attributed to the folate receptor-mediated endocytosis pathway, which facilitated more efficient intracellular delivery of the therapeutic payload. Importantly, the comparable IC50 values between F–Pt/M and free drug confirmed that the encapsulation process preserved the inherent antitumor activity of Pterostilbene while providing the advantages of targeted delivery. These results collectively demonstrated that the F–Pt/M formulation combined the therapeutic efficacy of free Pterostilbene with the benefits of receptor-mediated targeting, resulting in superior cytotoxic effects against glioma cells compared to non-targeted micelles. In vitro cellular uptake assay To evaluate the cellular penetration efficiency and targeting capability of folate-modified micelles, cellular uptake studies were conducted using coumarin-6 (C6)-labeled F–C6/M in human U251 (Fig. 2D) and rat RG-2 (Fig. 2E) glioma cell lines. The experimental design capitalized on the well-documented overexpression of folate receptors in malignant cells [44], which facilitates receptor-mediated endocytosis of folate-conjugated nanocarriers. As shown in Fig. 2D and E, fluorescence microscopy revealed distinct uptake patterns: the F–C6/M groups exhibited brighter intracellular green fluorescence compared to the non-targeted C6/M groups, implying that folate significantly facilitated and enhanced the entry of micelles into glioma cells, while free C6 displayed diffuse cytoplasmic distribution potentially due to its highly hydrophobic nature. Competitive inhibition studies demonstrated that the pretreatment of cells with free folate (10 μg/mL, 30 min) resulted in a notable inhibition of the uptake of F–C6/M in both U251 and RG-2 cells. Particularly in U251 cells, F–C6/M achieved 2.13-fold (1 h) and 2.97-fold (4 h) greater cellular accumulation compared to free folate pre-saturated F–C6/M (Fig. 2D), confirming the specificity of folate receptor-mediated uptake. These results provided conclusive evidence that folate modification enhanced cellular internalization through folate receptor-mediated endocytosis, consistent with established mechanisms of folate-decorated nanocarriers [45]. This targeted delivery platform demonstrated significant potential for improving therapeutic index in folate receptor-positive malignancies while minimizing off-target effects. Brain targeting To explore the brain-targeting efficiency and biodistribution profile of polymeric micelles, the in vivo distribution of free DiR, DiR/M, and F-DiR/M in mice was monitored using a non-invasive imaging system (IVIS). At 1 h after intravenous injection, F-DiR/M exhibited a detectable fluorescence signal in the brain. When the observation period was extended to 4 h, the fluorescence intensity of F-DiR/M in the brain increased (Fig. 2F). Notably, the fluorescence signal of F-DiR/M in the brain was significantly stronger than that of DiR/M, confirming the superior targeting ability of F-DiR/M and its potential to enhance micelle distribution in the brain (Fig. 2F). In our previous study on Pterostilbene distribution, we found that F–Pt/M increased the brain penetration of Pterostilbene by 2-fold compared to free Pterostilbene (Supplementary Fig. S3 and Supplementary Table S1). Additionally, the tissue targeting index for F–Pt/M in the brain was calculated to be 4.89, further demonstrating that folate modification significantly improved brain targeting efficiency [43]. These findings suggested that folate-modified nanocarriers could serve as a viable strategy for delivering bioactive compounds across biological barriers, potentially enhancing their therapeutic effects. Throughout the study, mice administered free DiR showed almost no brain fluorescence. To further investigate the distribution of F-DiR/M in major organs, ex vivo imaging was performed 24 h post-administration (Fig. 2G). As expected, F-DiR/M exhibited the highest fluorescence signal in the brain compared to free DiR and DiR/M. Quantitative fluorescence analysis revealed that the brain-to-liver ratio for F-DiR/M was 0.013 ± 0.002, which was significantly higher than that of DiR/M (0.008 ± 0.001) (p < 0.05). Similarly, the brain-to-spleen ratio for F-DiR/M was 0.081 ± 0.014, compared to 0.035 ± 0.008 for DiR/M (p < 0.05). The significantly higher brain-to-liver and brain-to-spleen ratios for F-DiR/M compared to the non-targeted formulation demonstrate the enhanced brain-targeting efficiency achieved through folate modification. Ex vivo imaging of major organs also revealed that both DiR/M and F-DiR/M primarily accumulated in the liver and spleen (Fig. 2G), suggesting that these organs play a key role in their metabolism and clearance. Collectively, the in vitro cellular uptake and in vivo imaging studies demonstrated that F–Pt/M facilitated folate receptor-mediated endocytosis and promoted effective brain accumulation. In vivo anti-glioma activity of F–Pt/M in orthotopic RG-2 glioma-bearing rat models Building upon the demonstrated in vitro efficacy of F–Pt/M against U251 and RG-2 glioma cells, we evaluated its therapeutic potential in vivo using orthotopic RG-2 glioma-bearing rats. Animals were divided into four treatment groups receiving: normal saline (Control and Sham groups), free Pterostilbene solution (Pterostilbene group), or F–Pt/M (F–Pt/M group) via daily intravenous tail vein injection (1 mg/kg Pterostilbene equivalent) beginning on postoperative day 3. Within one week of tumor implantation, control animals developed characteristic glioma symptoms including lethargy, anorexia, mobility impairment, and signs of intracranial hypertension (ocular/nasal hemorrhage, Fig. 3A). A conventional MRI scan conducted 7 days after transplantation showed evidence of tumor progression at the inoculation site (marked with a red circle in Fig. 3B). Terminal dissection revealed extensive tumor infiltration in the left hemisphere, evidenced by darkened tumor tissue, midline shift (Fig. 3C), and histopathological features including cellular atypia, nuclear hyperchromasia, and disorganized architecture in H&E-stained sections (Fig. 3D), confirming successful model establishment.Fig. 3Orthotopic RG-2 glioblastoma model in rats and the anti-tumor performance of various formulations. (A) One of the representative signs (eye and nasal mucosal bleeding) of rats with orthotopic RG-2 glioma. (B) Conventional MRI on day 7 post-transplantation revealed a growing tumor (indicated by red circle) in the inoculation site. (C) Visual representation of the glioma in the rat brain. (D) H&E staining of the RG-2 glioma-bearing rat brain (Bar = 500 μm). (E) Body weight variation after intravenous injection of saline (Sham group and Control group), free Pterostilbene (Pt group), and folate modified Pterostilbene loaded micelles (F–Pt/M group) in RG-2 glioma-bearing rats. (F) Kaplan-Meier survival analysis. (n = 5, statistical significance indicated by ∗p < 0.05 when compared to the Control group). Orthotopic RG-2 glioblastoma model in rats and the anti-tumor performance of various formulations. (A) One of the representative signs (eye and nasal mucosal bleeding) of rats with orthotopic RG-2 glioma. (B) Conventional MRI on day 7 post-transplantation revealed a growing tumor (indicated by red circle) in the inoculation site. (C) Visual representation of the glioma in the rat brain. (D) H&E staining of the RG-2 glioma-bearing rat brain (Bar = 500 μm). (E) Body weight variation after intravenous injection of saline (Sham group and Control group), free Pterostilbene (Pt group), and folate modified Pterostilbene loaded micelles (F–Pt/M group) in RG-2 glioma-bearing rats. (F) Kaplan-Meier survival analysis. (n = 5, statistical significance indicated by ∗p < 0.05 when compared to the Control group). Body weight changes were closely monitored in all four rat groups (Fig. 3E). Before tumor induction, initial weights showed no significant differences between groups. After implantation, all rats displayed temporary weight gain due to normal growth and early glioma development. By day 5 post-operation, the Control group began showing weight loss, indicating accelerated tumor growth. In contrast, the Pterostilbene and F–Pt/M groups exhibited delayed weight reduction starting at days 9 and 11, respectively. Notably, F–Pt/M showed better body mass preservation than free Pterostilbene. Meanwhile, Sham group rats maintained consistent weight gain throughout the experimental period, confirming the tumor implantation surgery itself minimally affected body weight. These findings collectively demonstrated that intravenous F–Pt/M administration effectively delayed weight loss in glioma-bearing rats, underscoring its potential to mitigate tumor progression. Survival analysis was conducted (Fig. 3F). Kaplan-Meier analysis revealed that both Pterostilbene and F–Pt/M treatments significantly prolonged survival compared to the Control group (p < 0.05). Notably, F–Pt/M treatment achieved a median survival of 30 days, representing a 76.5% increase over Control animals (17 days; p < 0.01) and a 20% improvement versus free Pterostilbene (25 days; p < 0.05). These results demonstrated that folate-mediated active targeting in F–Pt/M significantly enhanced survival in glioma-bearing rats. The extended survival and reduced tumor progression observed with F–Pt/M underscore the potential of receptor-targeted nano-delivery systems to enhance therapeutic benefit, which may be attributed to the improved brain accumulation of F–Pt/M. In our orthotopic glioma model, F–Pt/M micelles are most likely to enter the brain parenchyma via passive permeation through the compromised vasculature associated with the tumor, consistent with size-dependent extravasation through the disrupted blood-brain tumor barrier (BBTB). In addition to passive extravasation, alternative folate-mediated transport pathways may also contribute to brain accumulation. As summarized by Aboulhassane et al. [23], folate transport in the brain is mediated by three major pathways: folate receptor alpha (FRα), proton-coupled folate transporter (PCFT), and reduced folate carrier (RFC). Folate uptake primarily occurs at the blood-cerebrospinal fluid barrier (BCSFB) through the concerted actions of FRα and PCFT, suggesting that our folate-decorated F–Pt/M could potentially enter the CSF via the choroid plexus. Furthermore, Mazrad et al. [24] recently demonstrated that folic acid-conjugated brush polymers specifically target RFC expressed on brain endothelial cells, exhibiting efficient receptor-mediated transcytosis across the BBB and enhanced accumulation in glioma cells, with the transport pathway confirmed to rely on RFC-dependent mechanisms. Collectively, while passive extravasation through the disrupted BBTB remains one reasonable explanation for the initial brain accumulation of F–Pt/M in our glioma model, the potential contribution of RFC- and BCSFB-mediated folate transport pathways cannot be excluded and warrants further investigation in our future work. To evaluate treatment-induced apoptosis in glioma tissues, TUNEL staining was performed on brain paraffin sections (Fig. 4A). Sham group tissues showed minimal apoptosis with normal nuclear morphology, whereas Control group gliomas exhibited characteristic densely packed nuclei with negligible apoptosis. Both therapeutic interventions significantly increased TUNEL-positive apoptotic cells, with F–Pt/M demonstrating superior pro-apoptotic effects compared to free Pterostilbene. Western blotting analysis (Fig. 4B) showed that the anti-apoptotic protein Bcl-2 was highly expressed in Controls but significantly reduced in treatment groups, particularly in F–Pt/M treated animals, whereas the pro-apoptotic protein Bax showed markedly increased expression following both treatments, resulting in a dramatically decreased Bcl-2/Bax ratio. These molecular changes were further confirmed by immunohistochemistry (Fig. 4C), which demonstrated consistent Bax up-regulation and Bcl-2 down-regulation in treated tumors. Collectively, these results indicated that intravenous administration of F–Pt/M effectively induced apoptosis in RG-2 glioma-bearing rats. Once delivered into the tumor microenvironment, the folate moieties on F–Pt/M are expected to enhance cellular uptake and retention within glioma cells via folate receptor-mediated endocytosis [46]. This is consistent with the overexpression of folate receptors on glioma cells [22] and our in vitro observations of enhanced cellular uptake and cytotoxicity in folate receptor-positive U251 and RG-2 cells.Fig. 4In vivo effect of F–Pt/M on apoptosis in glioma cells. (A) TUNEL staining of RG-2 glioma-bearing brain (Bar = 100 μm). (B) Western blotting analysis of apoptosis-related proteins, Bax and Bcl-2 expression, in RG-2 glioma-bearing rat brain obtained after different treatments. (C) IHC staining of Bax and Bcl-2 proteins in RG-2 glioma-bearing rat brain (200 × ). In vivo effect of F–Pt/M on the activation of STAT3 in glioma cells. (D) Western blotting analysis of STAT3 signaling pathway-related proteins STAT3 and p-STAT3 expression. (E) IHC staining of STAT3 and p-STAT3 proteins in RG-2 glioma-bearing rat brains following various treatments. (Sham, Sham operation group; Control, RG-2 glioma model group; Pt, free Pterostilbene treatment group; F–Pt/M, folate modified Pterostilbene loaded micelles treatment group. Compared with the Control group, ∗p < 0.05, n = 3). In vivo effect of F–Pt/M on apoptosis in glioma cells. (A) TUNEL staining of RG-2 glioma-bearing brain (Bar = 100 μm). (B) Western blotting analysis of apoptosis-related proteins, Bax and Bcl-2 expression, in RG-2 glioma-bearing rat brain obtained after different treatments. (C) IHC staining of Bax and Bcl-2 proteins in RG-2 glioma-bearing rat brain (200 × ). In vivo effect of F–Pt/M on the activation of STAT3 in glioma cells. (D) Western blotting analysis of STAT3 signaling pathway-related proteins STAT3 and p-STAT3 expression. (E) IHC staining of STAT3 and p-STAT3 proteins in RG-2 glioma-bearing rat brains following various treatments. (Sham, Sham operation group; Control, RG-2 glioma model group; Pt, free Pterostilbene treatment group; F–Pt/M, folate modified Pterostilbene loaded micelles treatment group. Compared with the Control group, ∗p < 0.05, n = 3). Signal Transducer and Activator of Transcription 3 (STAT3) is a key regulator in multiple oncogenic signaling pathways, and its constitutive activation is detected in nearly 70% of human cancers [47], including gliomas [48]. The persistent activation of STAT3 is critical for driving glioma cell proliferation, migration, and epithelial-mesenchymal transition (EMT). Our previous research showed that Resveratrol (a structural analog of Pterostilbene) induced distinct apoptosis in bladder cancer cells by suppressing STAT3 activation. Furthermore, another study reported that Resveratrol potentiated Temozolomide's antitumor effects and overcame Temozolomide resistance by inhibiting STAT3 signaling [49]. However, the interaction between Pterostilbene and the STAT3 pathway remains unexplored. Computational chemistry analysis revealed that Pterostilbene influenced STAT3 phosphorylation by modulating the dynamic properties of the STAT3 SH2 domain. Molecular docking demonstrated that Pterostilbene bound to STAT3, with its phenolic hydroxyl group forming a 2.2 Å hydrogen bond with GLN-644 residue on the STAT3 protein (Fig. 5A). Subsequent 100 ns MD simulations of the Pterostilbene-STAT3 complex confirmed binding stability. RMSD analysis (Fig. 5B) showed the complex maintained structural stability with minimal fluctuations, while Rg analysis (Fig. 5C) and SASA analysis (Fig. 5D) indicated consistent compactness without major conformational changes. These results collectively demonstrated that Pterostilbene formed a stable complex with STAT3, likely affecting its phosphorylation through SH2 domain modulation.Fig. 5Molecular docking and MD simulation analyses of Pterostilbene-STAT3 interactions. (A) Molecular docking revealed hydrogen bonding between Pterostilbene's phenolic hydroxyl group and STAT3's GLN-644 residue. (B–D) Time-dependent evolution of (B) root-mean-square deviation (RMSD), (C) radius of gyration (Rg), and (D) solvent-accessible surface area (SASA) for the Pterostilbene-STAT3 complex during 100-ns MD simulations. (E) Comparative root-mean-square fluctuation (RMSF) profiles of STAT3's SH2 domain (residues 584–687) with/without Pterostilbene binding. Molecular docking and MD simulation analyses of Pterostilbene-STAT3 interactions. (A) Molecular docking revealed hydrogen bonding between Pterostilbene's phenolic hydroxyl group and STAT3's GLN-644 residue. (B–D) Time-dependent evolution of (B) root-mean-square deviation (RMSD), (C) radius of gyration (Rg), and (D) solvent-accessible surface area (SASA) for the Pterostilbene-STAT3 complex during 100-ns MD simulations. (E) Comparative root-mean-square fluctuation (RMSF) profiles of STAT3's SH2 domain (residues 584–687) with/without Pterostilbene binding. To further investigate Pterostilbene's effects on the STAT3 SH2 domain (residues 584–687), we conducted parallel 100 ns MD simulations of both the STAT3 monomer and the Pterostilbene-STAT3 complex, with comparative analysis of SH2 domain flexibility through RMSF calculations. As shown in Fig. 5E, the RMSF profile of the SH2 domain exhibited significant alterations upon Pterostilbene binding, demonstrating that ligand interaction directly modulated local structural dynamics. Given the established correlation between SH2 domain conformational stability, phosphorylation efficiency, and subsequent dimerization [50,51], these observed RMSF variations strongly suggested that Pterostilbene could mechanistically inhibit STAT3 phosphorylation by specifically modifying the dynamic properties of the SH2 domain. To determine whether F–Pt/M suppresses tumor growth in rats by modulating STAT3 activation, we performed Western blotting. Quantitative analysis of band intensities revealed that total STAT3 expression was comparable across all four brain tissue samples, with no significant differences. In contrast, the expression of phosphorylated STAT3 (p-STAT3), the active form of STAT3, was negligible in normal brain tissue (Sham group), highly elevated in untreated tumor tissue (Control group), and markedly reduced following Pterostilbene and F–Pt/M treatment (Fig. 4D). Immunohistochemical (IHC) analysis further supported these findings, showing strong p-STAT3 expression and prominent nuclear translocation in the Control group. However, F–Pt/M treatment significantly suppressed p-STAT3 levels and reduced its nuclear accumulation (Fig. 4E), corroborating the Western blot results. These data suggested that the antitumor effects of F–Pt/M were mediated in part through modulation of the STAT3 pathway. These in vivo findings illustrated that F–Pt/M exerted potent anti-glioma activity in orthotopic glioma models, delaying weight loss, extending median survival, and mechanistically inducing apoptosis while suppressing the STAT3 signaling pathway. Conclusion In conclusion, this study demonstrates that folate-decorated polymeric micelles (F–Pt/M) effectively overcome the solubility and brain delivery limitations of Pterostilbene, a naturally occurring bioactive compound with neurotherapeutic potential. The F–Pt/M formulation induces apoptosis and inhibits the STAT3 signaling pathway, leading to significantly enhanced anti-glioma efficacy in an orthotopic rat model. These findings establish F–Pt/M not only as a promising strategy to potentiate Pterostilbene's bioactivity but also as a versatile nanoplatform for increasing brain accumulation of poorly soluble neuroactive agents. Future efforts will focus on assessing long-term safety, optimizing scalable manufacturing, and advancing preclinical validation to support the clinical translation of this targeted micelle platform for central nervous system disorders. CRediT Statement Conceptualization, X.S. and B·I.; Methodology, Y·W., X.Y. and L.L.; Software, Y·S., H.Z. and L.L.; Validation, H.J., C.L., W.S. and M.L.; Formal analysis, C·W., Y.Y. and W·Y.; Investigation, Y·W., Y.S. and X.Y.; Resources, X.S. and J.N.; Data curation, M.W. and Q.W.; Writing—original draft preparation, Y·W.; Writing—review and editing, X.Y. and X.S.; Visualization, J.W. and Z.Z.; Supervision, B.I. and X.S.; Project administration, X.S.; Funding acquisition, X.Y., C.W. and X.S. All authors have read and agreed to the published version of the manuscript. Data availability The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. Funding This work was supported by the National Natural Science Foundation of China (81672945, 81502992), the Foundation of Liaoning Province Education Administration (LJ212410161009 and JYTQN2023153), and Dalian Medical University Interdisciplinary Research Cooperation Project Team Funding (JCHZ2023011). Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.